ISPE_Water and steam systems_Vol4.pdf

GE Power & Water Water & Process Technologies Analytical Instruments

Simplied On-line TOC Measurement Measurement Ideal for PAT, QbD, and Other Quality Optimization Initiatives With the widest range of pharmaceutical Total Organic Carbon (TOC) analyzers and sensors, ready-to-use Sievers Standards, validation documentation, Quality System Optimization (QSO) services, technical support, and more than 20 years’ leadership in measurement technologies for pharmaceutical water, GE Analytical Instruments oers a complete solution for: • Compliance with US, European, Chinese, Indian, and the Japanese (JP 16) Pharmacopeia TOC monitoring requirements • Continuous on-line quality assurance • Real-time pharmaceutical water testing and release • On-line cleaning validation and verication • Optimizing and rationalizing cost-eective sampling plans with lower risk

The featured Sievers 500 RL TOC Analyzer oers: • • • • •

Unsurpassed accuracy and precision Broad range of detection (0.03 ppb to 2,500 ppb as C) Four-port Super iOS that automates all validation protocols Easy operation, no reagents, reagents, and low maintenance The most reliable reliable TOC system suitability measurement available

50 0 0 RL O NLINE

Learn more. Contact us at +800 255 6964, geai@ge.com, or visit www.geinstruments.com.

Sievers 500 RL On-Line CheckPoint Pharma On0.03 ppb to 2,500 ppb Line/Portable TOC Sensor 0.21 ppb to 1,000 ppb

GE imagination at work

Sievers 900 Laboratory and GE Autosampler 0.03 ppb to 50 ppm

Id e ea l f or Re al -T i im e T e es t i in g and C l le ani ng V al e i i d da t i io n

Sievers 900 On-Line

Sievers 900 Portable

0.03 ppb to 50 ppm

0.03 ppb to 50 ppm

Volume 4

Water and Steam Systems Second Edition / December 2011

Disclaimer: This Guide is meant to assist pharmaceutical manufacturers in the design and construction of new and renovated facilities that are required to comply with the requirements of the US Food and Drug Administration (FDA). The International Society for Pharmaceutical Engineering (ISPE) cannot ensure, and does not warrant, that a facility built in accordance with this Guide will be acceptable to the FDA. Limitation of Liability In no event shall ISPE or any of its afliates, or the ofcers, directors, employees, members, or agents of each of them, be liable for any damages of any kind, including without limitation any special, incidental, indirect, or consequential damages, whether or not advised of the possibility of such damages, and on any theory of liability whatsoever,, arising out of or in connection with the use of this information. whatsoever © Copyright ISPE 2011 2011.. All rights reserved. All rights reserved. No part of this document may be reproduced or copied in any form or by any means – graphic, electronic, or mechanical, including photocopying, taping, or information storage and retrieval systems – without written permission of ISPE. All trademarks used are acknowledged. ISBN 978-1-936379-29-3

For individual use only. © Copyright ISPE 2011. All rights reserved.

ISPE Baseline® Guide: Water and Steam Systems

Page 2

Foreword The global pharmaceutical industry and regulators are responding to the challenge of signicantly improving the way drug development and manufacturing is managed. New concepts are being developed and applied including sciencebased risk management approaches, a focus on product and process understanding, and modern Quality Systems. Uncertainty about the requirements for regulatory compliance may discourage innovation and technological advancement, and can drive up costs. ISPE Guides aim to describe current good practices that can help a company to develop an approach that is effective, cost-efcient, cost-efcient, and in compliance with existing regulations and related guidance. We thank the FDA for their review and comments to this Guide. ISPE seeks close involvement of international regulators, including the US FDA, in the development of the ISPE Guides, which cover many important aspects of pharmaceutical development and manufacturing. These Guides are excellent examples of how the ISPE, regulators, and industry can work co-operatively for public benet. The Guides are solely created and owned by ISPE. They are not regulations, standards, or regulatory guideline documents, and facilities built in conformance with the Guides may or may not meet FDA or other regulatory requirements. A continued working relationship between ISPE and international regulators will be fruitful for regulators, industry industry,, and most importantly for public health.

Special Dedication to Carl Roe The revision and update of this Guide is dedicated to the memory of Mr. Carl C. Roe. Carl undertook the extensive effort to update this Guide out of his passion to help humanity through improvements in the biopharmaceutical industry. His persistence and fortitude as the chairman during the initial efforts of revision laid the groundwork for the success and exceptional quality of the updated Guide. It is with sincere thanks and fond memories that the contributors wish to recognize the everlasting memory of Carl’s efforts and friendship. We all hope and pray that one day advances in the biopharmaceutical industry, industry, which we in some small way may assist, can discover a cure for cancer and other deadly and debilitating diseases.



Page 3

Acknowledgements The team of authors and contributors for this revision of the ISPE Baseline Pharmaceutical Engineering Guide for Water and Steam Systems would like to recognize the signicant efforts by the original team of authors and contributors and applaud their efforts on the rst edition of the guide. The authors and contributors for the Second Edition of the Water and Steam Systems Guide were:

Chair: Cameron Sipe

Pzer

USA

Chapter 1: Introduction Cameron Sipe (Chapter Lead) Nissan Cohen Gary Zoccolante

Pzer Start-up Business Development Siemens Industry, Inc. Water Technolog echnologies ies Unit

USA USA USA

Chapter 2: Key Design Philosophies Cameron Sipe (Chapter Lead) Pzer Andrew W. Collentro Water Consulting Specialists, Inc. Joseph J. Manfredi GMP Systems, Inc.

USA USA USA

Chapter 3: Water Options and System Planning Andrew W. Collentro (Chapter Lead) Consulting Specialists, Inc.

USA

Chapter 4: Pretreatment Options Andrew W. Collentro (Chapter Lead) Michael E. Holland Robert Vecc Vecchione hione Gary Zoccolante

USA USA USA USA

Consulting Specialists, Inc. GE Water and Process Technolog echnologies ies Christ Aqua Pharma & Biotech Siemens Industry, Inc. Water Technolog echnologies ies Unit

Chapter 5: Final Treatmen Treatmentt Options: Non-Compendial Waters, Compendial Puried Water and Compendial Highly Puried Water Gary Zoccolante (Chapte (Chapterr Lead) Siemens Industry, Inc. Water Technolog echnologies ies Unit USA Michael E. Holland GE Water and Process Technolo echnologies gies USA Bruno Rossi Millipore France Anders Widov, MSc Chem Eng FR PHARMA Sweden Chapter 6: Final Treatmen Treatmentt Options: Water for Injection Peter T. Vishton (Chapter Lead) Pzer Global Engineering (Retired) Joseph J. Manfredi GMP Systems, Inc. Bill Alkier formally with Paul Mueller Company Sharif Disi MECO Timo Heino STERIS Finn-Aqua Brian McClellan Aqua-Chem, Inc.

USA USA USA USA Finland USA

Chapter 7: Pharmaceutical Steam Sharif Disi (Chapter Lead) Cameron Sipe

USA USA

MECO Pzer

Chapter 8: Storage and Distribution Systems Cameron Sipe (Chapter Lead) Pzer Joseph J. Manfredi GMP Systems, Inc. Anders Widov, MSc Chem Eng FR PHARMA Robert M. Augustine Eli Lilly and Company T.C. Soli, PhD Soli Pharma Solutions Solutions,, Inc. Thomas Beck Recipharm Stockholm AB Robert Vecc Vecchione hione Christ Aqua Pharma & Biotech


USA USA Sweden USA USA Sweden USA


Page 4

Chapter 9: Laboratory Water Philip E. Sumner (Chapter Lead) Michelle M. Gonzalez Bruno Rossi T.C. Soli, PhD Paul Whitehead

Pzer Global Engineering Amgen Inc. (Retired) Millipore Soli Pharma Solutions Solutions,, Inc. ELGA LabWater Global Operations

USA USA France USA United Kingdom

Chapter 10: Rouge and Stainless Steel Michelle M. Gonzalez (Chapter Lead) Amgen, Inc. (Retired) Patrick H. Banes Astro Pak Corporation Sunniva Collins Swagelok Swagelo k Technolog echnology y Services Co. Bill Huitt W.M. Huitt Company Ken Kimbrel UltraClean Electropolish, Inc. Joseph J. Manfredi GMP Systems, Inc. Andreas Marjoram Bayer Healthcare AG Peter Petrillo Millennium Facilities Resources, Inc. Daryl Roll Astro Pak Corporation Peter T. Vishton Pzer Global Engineering (Retired) James “Jim” Vogel Process Facilitie Facilities s Services, Inc.

USA USA USA Germany USA USA USA USA

Chapter 11: Control and Instrumentation Nissan Cohen (Chapter Lead) Start-up Business Development Dr. Anthony Bevilacqu Bevilacqua a Mettler Toledo Thornton Joseph J. Manfredi GMP Systems, Inc. Bruno Rossi Millipore

USA USA USA USA

Chapter 12: Commissioni Commissioning ng and Qualication Alex J. Konopka (Chapter Lead) Engineering Consultant (Retired) T.C. Soli, PhD Soli Pharma Solutions Solutions,, Inc. Graham C. Wrigley, PhD Pzer Inc. Joseph J. Manfredi GMP Systems, Inc. Jay C. Buf Pzer Inc. Rostyslaw Slabicky Boehringer Ingelheim Pharmaceuticals Inc.

USA USA USA USA USA USA

Chapter 13: Microbiological Considerations for Pharmaceutical Water Systems T.C. Soli, PhD (Chapter Lead) Soli Pharma Solutions, Inc.

USA

Chapter 15: Glossary Michelle M. Gonzalez (Chapter Lead)

USA

Amgen Inc. (Retired)

USA USA

FDA Reviewers We would like to thank the following FDA representatives for providing comments on this Guide: Tara Gooen, Team Leader, CDER/DMPQ CDER/DMPQ/FDA, /FDA, USA Richard Friedman, Director Director,, CDER/FDA, USA Brian Hasselbalch, Consumer Safety Ofcer Ofcer,, CDER/FDA, USA The Water Revision Task Task Team Team would like to recognize the signicant efforts by the original team of authors and contributors and applaud their efforts on the rst edition of the Guide. The Team also would like to thank ISPE for technical writing and editing support by Gail Evans (ISPE Guidance Documents Writer/Editor) and Sion Wyn (ISPE Technical Advisor).

Cover photo and diagram: Courtesy of Pzer Pzer,, www.pz www.pzer.com. er.com.



Page 5

Table of Contents 1

Introduction ............ ......................... .......................... .......................... .......................... .......................... .......................... ......................... ......................... ................... ...... 9 1.1 1.2 1.3 1.4

2

Key Design Philosoph Philosophies ies ............. .......................... .......................... .......................... .......................... .......................... .......................... ................... ...... 13 2.1 2.2 2.3 2.4 2.5 2.6

3

Introduction ................. ............................ ...................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ....................31 .........31 Process Design of Pretreatment .......... ..................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ................. ......31 31 Feed Water to Pretreatment Pretreatment Quality: Testing Testing and Documentation Documentation ......... .................... ...................... ..................... ..................... ....................33 .........33 Output Water Water from Pretreatment: Pretreatment: Quality of Feed Water to Final Treatment Treatment........... ...................... ..................... ..................... ............. .. 34 Control of Fouling: Removal of Turbidity and Particulates Particulates ........... ...................... ..................... ..................... ...................... ..................... ...................35 .........35 Control of Scaling: Scaling: Removal of Hardness Hardness and Metals .......... ..................... ..................... ..................... ...................... ..................... ..................... ................ ..... 37 Organic Material and Removal Removal ............... .......................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ............. ... 41 System Design for for Control of Microbial Growth ......... .................... ...................... ..................... ..................... ...................... ...................... ..................... ............... ..... 44 Removal of Microbial Control Agents.......... .................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... .....................46 ..........46 Changes in Anion Composition/Concentration ..................... ................................ ..................... ..................... ...................... ..................... ..................... ................ ..... 49 The Importance of pH in Pretreatment........................ Pretreatment.................................. ..................... ...................... ..................... ..................... ...................... ..................... ............... ..... 51 Materials of Construction and Construction Practices .......... ..................... ..................... ..................... ...................... ..................... ..................... ................ ..... 51 Water Conservation .......... .................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ............... .... 52 Summary...................... Summary........... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ....................52 .........52

Final Treatment Options: Non-Compendi Non-Compendial al Waters, Compendial Puried Water, and Compendial Highly Puried Water ............ Water ......................... .......................... .......................... .......................... ........................ ........... 53 5.1 5.2 5.3 5.4 5.5

6

Introduction ................. ............................ ...................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ....................19 .........19 Water Quality Options ............ ....................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ....................19 .........19 System Planning ............. ........................ ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ................ ......25 25 System Design .............. ......................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................30 .......30

Pretreatment Pretreatm ent Options ............. .......................... .......................... ......................... ......................... .......................... .......................... .......................... ............. 31 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.1 1 4.12 4.13 4.14

5

Introduction ................. ............................ ...................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ....................13 .........13 Pharmacopeial Water Water and Steam .......... .................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ............... .... 13 Specication of Pharmaceutical Water Quality .......... ..................... ...................... ..................... ..................... ...................... ...................... ..................... ............... ..... 14 Critical Process Parameters Parameters ................ ........................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ................. ......15 15 CGMP Compliance Issues .......... ..................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ............... .... 16 Design Range versus versus Operating Range .......... ..................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ............... ..... 16

Waterr Options and System Planning ............. Wate .......................... .......................... ......................... ......................... .......................... ............... 19 3.1 3.2 3.3 3.4

4

Background ............ ....................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ................ ......9 9 Scope of this Guide....................... Guide.................................. ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ............... .... 9 Key Topics include included d in this Guide .............. ............................ ........................... ........................... ............................ ............................ ........................... ........................... .............. 10 Guide Structure.......... .................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ......................12 ...........12

Introduction ................. ............................ ...................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ....................53 .........53 Ion Exchange ........... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ............. .. 55 Reverse Osmosis Osmosis.......... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................61 .......61 Continuous Electrodeionization Electrodeionization (CEDI) ........... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ............... ..... 66 Polishing and Removal of Specic Contaminations Contaminations.......... ..................... ...................... ..................... ..................... ...................... ...................... ...................69 ........69

Final Tre Treatment atment Options: Water for Injection (WFI) ............. .......................... .......................... .......................... ................ ... 77 6.1 6.2 6.3 6.4 6.5 6.6

Introduction ................. ............................ ...................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ....................77 .........77 Pharmacopeial Issues............ Issues....................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ....................77 .........77 General Technolog echnology y Discussion ............ .......................... ............................ ............................ ........................... ........................... ............................ ........................... .................. ..... 79 Process and Systems Description Description .......... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ............. ... 83 Final Treatment – General System-Wide Controls and Instrumentation Instrumentation........... ...................... ..................... ..................... ....................90 .........90 Summary and Technology Technology Comparison ........... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ............... ..... 90



Page 6

7

Pharmaceutical Pharmac eutical Steam ............. .......................... .......................... .......................... .......................... ......................... ......................... ........................ ........... 93 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 7.10

8

Storage and Distribution Systems ............. .......................... ......................... ......................... .......................... .......................... ................ ... 107 8.1 8.2 8.3 8.4 8.5 8.6

9

Introduction .......... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... .............. .... 165 Regulatory Stance .......... ..................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... .............. .... 166 Surface Conditions and Treatments Treatments.......... ..................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ....................167 ..........167 Rouge Formation .......... ..................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ...................... ................ .....171 171 Rouge Detection (Methodology) .................... ............................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ............... .... 178 Risk Analysis – Rouge and its Remediation ................ ........................... ...................... ..................... ..................... ...................... ..................... ..................... ............. 183 Rouge Remediation (Methodology) ................... .............................. ..................... ..................... ...................... ..................... ..................... ...................... ..................... ............ 185 Conclusions ........... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ...................... ..................... ..................... ...................... ..................... ............ .. 190

Control and Instrumenta Instrumentation........ tion..................... .......................... .......................... ......................... ......................... .......................... .................. ..... 191 11.1 11.2 11.3 11 .3 11.4 11 .4 11.5 11 .5 11.6

12

Introduction ................ .......................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ....................139 .........139 System Design Considerations.......... ..................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ................. ......139 139 Determining User Needs............. Needs........................ ..................... ..................... ...................... ...................... ..................... ..................... ...................... ..................... ..................... ............. .. 140 Water Puricati Purication on Technolog echnologies ies ........... ..................... ..................... ...................... ...................... ..................... ..................... ...................... ..................... ..................... ............... .... 152 Laboratory Water Supply Options .......... .................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ............. .. 152 Maintenance ................... .............................. ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... .............. .... 161 Instruments and Calibration .......... .................... ..................... ...................... ...................... ..................... ..................... ...................... ..................... ..................... ......................161 ...........161 Commissioning and Qualication Qualication........... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ............. .. 162

Rouge and and Stainless Stainless Steel Steel ............. .......................... .......................... .......................... .......................... .......................... .......................... ............... .. 165 10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8

11

Introduction ................ .......................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ....................107 .........107 Purpose........... Purpose ...................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ...................107 .........107 System Components............ Components....................... ..................... ..................... ...................... ...................... ..................... ..................... ...................... ..................... ..................... ....................107 .........107 Materials of Construction/Finishes.......... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ............ 118 Microbial Control Considerations ........... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ............. .. 123 System Designs ............ ....................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ................ .....124 124

Laboratory Wate Waterr ........... ........................ .......................... .......................... .......................... .......................... .......................... .......................... ................... ...... 139 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8

10

Introduction ................. ............................ ...................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ....................93 .........93 Common Steam Terms and Denition Denitions s ........... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ............... ..... 93 Types of Steam .............. ............................ ........................... ........................... ............................ ........................... ........................... ............................ ............................ ........................... ............... 94 Regulatory and Industry Guidance Guidance ............. ........................ ..................... ..................... ...................... ..................... ..................... ...................... ..................... ....................95 ..........95 Background and Industry Practices Practices .......... ..................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ............. 96 System Planning ............. ........................ ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ................ ......97 97 Steam Generation ........... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... .............. .... 100 Steam Attributes and Condensate Condensate Sampling .......... ..................... ..................... ..................... ...................... ..................... ..................... ...................... ................. ......102 102 Materials of Construction .......... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ............... .... 103 Distribution .......... .................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ................ .....104 104

Introduct ion ............. Introduction ........................... ........................... ........................... ............................ ........................... ........................... ............................ ............................ ........................... ................... ...... 191 Principles Principle s .............. ........................... ........................... ............................ ........................... ........................... ............................ ............................ ........................... ........................... ...................... ........ 192 General Instrumentation Requirements ........................ .................................. ..................... ...................... ..................... ..................... ...................... ..................... ............ 193 Design Conditions versus Operating Range........... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ................. ......200 200 Instrumentation Spikes .......... ..................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................201 .......201 Control Systems Systems.......... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................201 .......201

Commissioning and Qualication ........... ........................ .......................... .......................... .......................... .......................... .................. ..... 205 12.1 12.2 12.3 12.4 12.5

Introduction .......... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... .............. .... 205 Sampling for Water Systems .......... .................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ....................205 .........205 Sampling for Steam Systems.. Systems............. ...................... ..................... ..................... ...................... ..................... ..................... ...................... ...................... ..................... ................ ......206 206 Acceptance Criteria......... Criteria.................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... .............. .... 206 Change Control and Maintaining Maintaining the Qualied Qualied State of the System .......... ..................... ..................... ..................... ...................... .............. ... 208



13

Page 7

Microbiological Microbiolo gical Consideratio Considerations ns for Pharmac Pharmaceutical eutical Wate Waterr Systems Systems ............. ........................ ........... 209 13.1 13.2 13.3 13.4 13.5 13.6 13.7 13.8

Introduction .......... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... .............. .... 209 The Microbial Growth Process in High Purity Water Systems .......... ..................... ...................... ..................... ..................... ...................... ............. 209 Detrimental Effects of Biolm Biolm .......... .................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ....................212 .........212 Microbial Control Strategies ............. ........................ ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................214 ........214 Sanitizer Choices.............................. ........................................ ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................220 .......220 Assessing Microbial Control Success ........... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ................. ......235 235 Functional Microbiological Pharmacopeial Compliance........................ Compliance................................... ..................... ..................... ...................... ...................239 ........239 Microbial and Endotoxin Control in Pure Steam Steam Systems Systems ........... ...................... ..................... ..................... ...................... ..................... ................. ....... 241

14

Appendix 1 – References References ............. .......................... .......................... .......................... .......................... ......................... ......................... .................. ..... 243

15

Appendix 2 – Glossary Glossary ............. .......................... .......................... .......................... ......................... ......................... .......................... ...................... ......... 247 15.1 Acronyms and Abbreviations ........................... ...................................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ............. .. 247 15.2 Denitions ........... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ..................... ..................... ...................... ................ .....249 249


ISPE Headquarters 600 N. Westshore Blvd., Suite 900, Tampa, Tampa, Florida 33609 USA Tel: +1-813-960-2105, Fax: +1-813-264-2816 ISPE Asia Pacic Ofce 20 Bendemeer Road, #04-02 Cyberhub, Singapore 339914 Tel: +65-6496-5502, Fax: +65-6496-5599 ISPE China Ofce Suite 2302, Wise Logic International Center No. 66 North Shan Xi Road, Shanghai, China 200041 Tel +86-21-5108-1512, Fax +86-21-51 +86-21-5116-0260 16-0260 ISPE European Ofce Avenue de Tervueren, Tervueren, 300, B-1150 B-1150 Brussels, Belgium Tel: +32-2-743-4422, Fax: +32-2-743-1550 www.ISPE.org


1

Introduction




Page 9 Introduction

1

Introduction

1.1

Background The design, construction, and verication (commissioni (commissioning ng and qualication) of water and steam systems for the pharmaceutical industry represent signicant challenges as these systems need to meet current Good Manufacturing Practice (cGMP) regulations, while remaining in compliance with all other governing codes, laws, and regulations. The design, complexity complexity,, and cost of these systems are highly variable, compounded by interpretation of regulatory requirements and corresponding design approaches. This Guide is intended to offer a practical and industry accepted interpretation of regulatory requirements, while providing a broad spectrum of innovative and proven approaches to water and steam system design, construction, commissioning commissioning and qualication. This Guide was prepared by ISPE and leading industry experts, with representative feedback from all areas and disciplines of the industry, industry, and comments provided by FDA. It reects ISPE’s current thinking related to new water and steam systems and takes into account the FDA’s guidelines for Pharmaceutical cGMPs for the 21st Century – A Risk-Based Approach (Reference Approach (Reference 1, Appendix 1) and other worldwide regulatory guidance documents. It is recognized that industry standards evolve, and this document reects the understanding of them as of the publication date.

1.2

Scope of this Guide This Guide is intended to assist with the design, construction, operation, and maintenance of new water and steam systems. It is neither a standard nor a detailed design guide. The validation of water and steam systems, which comprises commissioning commissioning and qualication activities, are not discussed in-depth in this Guide, but is covered in the ISPE Baseline ® Guide on Commissioning and Qualication (Volume (Volume 5) (Reference 2, Appendix 1). This Guide focuses on engineering issues and provides innovative alternatives for water and steam systems. It is not intended to replace governing laws, codes, guidelines, standards, or regulations that apply to pharmaceutical water and steam systems. Where non-engineering issues (e.g., microbiological topics) are covered, the information is included to stress the importance of such topics and their impact on water and steam system design. The use of this document for new or existing water and steam systems is at the discretion of the designer, owner, or operator. The information provided in this Guide does not supersede regulatory guidance documents which focus on high purity water or steam systems, such as the “Guide to Inspections of High Purity Water Systems” published by the FDA (Reference 3, Appendix 1). This Guide is intended primarily for interpretation of regulatory compliance involving the US FDA and the US Pharmacopeia (USP) (Reference 4, Appendix 1). To recognize the status of worldwide harmonization, observations applicable to the European Union, Japan, and other markets, including their representative pharmacopeias are included, along with applicable references to the World Health Organization (WHO), American Society for Testing and Materials (ASTM), American American Society of Mechanical Engineers (ASME), International Organization for Standardization (ISO), and others.



Page 10 Introduction Introduct ion

1.3

Key Topics included in this Guide The following key topics are addressed within this Guide:

1.3.1

•

specifying water quality and system planning

•

project management

•

pretreatment, nal treatment, and storage/distribution options

•

instrumentation and control

•

pharmaceutical steam

•

rouge

•

microbiological considerations

Specifying Water Quality and System Planning The selection or specication of the quality required for the water or steam is potentially the most critical step in planning a new or renovated pharmaceutical water or steam system, from a regulatory, technical, and nancial perspective. This selection selection likely is to have a more signicant impact on performance and reliability of the system than subsequent design decisions; the risk of noncompliance and system failures should be considered. The designer should have knowledge of the applicable regulations and the technologies capable of consistently meeting those regulations. Once process water or steam requirements are determined, system design options can be evaluated. This Guide presents alternative baseline water treatment technologies and water system design options with their respective advantages and disadvantages. These baseline alternatives are evaluated relative to issues such as: •

capital costs

•

product water quality

•

chemical handling

•

water consumption

•

energy consumption

•

maintenance requirements

•

chemical/microbial/endotoxin chemical/microb ial/endotoxin removal performance

The Guide emphasizes evaluating system design alternatives based on: •

feed water quality

•

pretreatment and nal treatment technologies

•

storage and distribution system design alternatives

•

operation/maintenance requirements



Page 11 Introduction

The goal of this selection is a system design that meets user requirements and is compliant with regulatory requirements.

1.3.2

Project Management Project management as it applies to pharmaceutical water and steam systems, recognizes that these systems require adequate commissioning and qualication. Planning should ensure appropriate documentation, inspection, and eld testing. Good project management capitalizes upon this practice suggesting that manufacturers engage stakeholders (engineers, operators, Quality Assurance, etc.) very early in the planning, design, construction, and commissioni commissioning/ ng/ qualication phases to ensure that systems are appropriately documented for regulatory qualication.

1.3.3

Water System Design Options This Guide emphasizes that water systems can be designed in many different ways, and still meet the overall requirements of users. A carefully carefully planned approach to the design, with input from appropriate areas of an organization (e.g., engineering, operations, manufacturing, Quality Assurance), should be implemented.

1.3.4

Control and Instrumentat Instrumentation ion This Guide includes general guidance on dening and monitoring critical process parameters for water and steam systems (see Chapter 11 of this Guide). Critical parameters are dened as those parameters that directly affect the product quality. For example, since microbial quality cannot currently be directly monitored in real time, the parameters relied upon to control microbial growth normally are considered critical. These may include: •

temperature

•

UV intensity

•

ozone concentration

•

circulating water under positive pressure

Quality attributes (properties of water produced) may be monitored at or after each process step for chemical purity purity,, and the correct performance of that operation conrmed directly. directly. For a system producing compendial water or steam, properties mandated in the applicable pharmacopeia normally constitute critical parameters.

1.3.5

Pharmaceutical Steam Industry standard practices for pharmaceutical steam used in various applications that are required to meet regulatory requirements are included (see Chapter 7 of this Guide). Denitions of the steam types are included. In addition, system planning and alternative design practices are discussed in detail to help facilitate development of the most appropriate overall system design to meet the user requirements.

1.3.6

Rouge The latest theories and industry practices addressing the formation of rouge in high purity water and pure steam systems are presented (see Chapter 10 of this Guide). Detailed information is provided on the types or classications of rouge typically occurring in these systems, as well as methodology that may be used to address the issue. Typical examples are given and analysis provided into the science of rouge formation with the intention to allow informed decisions about the presence of rouge.



Page 12 Introduction Introduct ion

1.3.7

Microbiologicall Considerations Microbiologica Controlling microbial proliferation requires thorough consideration throughout the conception, design, construction, qualication, operation, maintenance, and monitoring of a water system. Waters may be sufciently rich in nutrients to support the growth of some types of microorganisms, as a result of remnants of very low levels of inorganic and organic contaminants. The microbial growth in water purication systems as well as distribution systems needs to be controlled, if not completely prevented. The nished water must be suitable and completely safe for use in pharmaceutical, biopharmaceutical, biopharmaceutical, and medical device applications and for use by associated patients and consumers (see Chapter 13 of this Guide).

1.4

Guide Structure Figure 1.1 shows the structure of this Guide. The chapters have been organized to assist the decision process by initially determining the type of water required and then determining the system design needed from pretreatment to nal treatment. Guidance through options for storage and distribution, selecting the proper instrumentation and control, microbiological control methods, and recommendations for commissionin commissioning g and qualication is provided. Additionally,, the Guide contains Additionally contains detailed information information on related topics, topics, such as laboratory water, water, rouge, and pure steam. Figure 1.1: Structure of the ISPE Baseline ® Guide: Water and Steam Systems


2

Key Design Philosophies




Page 13 Key Design Philosophies

2

Key Design Philosophies

2.1

Introduction Pharmaceutical water is the most widely used utility and ingredient in drug manufacturing and the main component in equipment/system cleaning; therefore, systems for the production, storage, and distribution of pharmaceutical water and steam constitute essential elements in most manufacturing facilities. The control of potential sources of contamination (chemical and microbiological), while delivering the required quantity of water, is the primary goal of pharmaceutical water system design. This Guide identies technologies that can assist in achieving this goal and industry methods (current at this of publication) available to engineers for design of systems which minimize the risk. The quality of pharmaceutical water and steam is critical both from a regulatory point of view and from a nancial perspective. The pharmaceutical pharmaceutical water and steam specication has a signicant impact on the life cycle cost of the system. It must be demonstrated that pharmaceutical steam and waters (non-compendial and compendial) can be produced and distributed consistently to meet stated specications. Establishing the level of microbial control needed in a pharmaceutical water and steam system used in the manufacture of a biopharmaceutical product requires an understanding of both the use of the product and the manufacturing process. Pharmaceutical water and steam users should dene the appropriate water purity based upon sound process understanding and system equipment capability. They should determine the: •

specic purication capability for each processing step

•

limitations of the unit operation

•

critical parameters, which affect the specied water/steam quality (chemically (chemically,, physically physically,, and/or microbiologically)

The major compendia (USP (USP,, European Pharmacopoeia (EP), and Japanese Pharmacopoeia (JP)) (References 4, 5, and 6, Appendix 1) describe two bulk compendial waters (Puried Water (PW) and Water for Injection (WFI)). Additional bulk waters, such as USP -water for for hemodialysis and and EP - highly puried water, water, are identied in this Guide. This Guide primarily supports PW, WFI, and pre steam, plus additional non-compendial waters, including “laboratory water.” It is common practice to name non-compendial waters (exclusive of “drinking water”) used in pharmaceutical manufacturing by the nal treatment step (i.e., Reverse Osmosis (RO) water, Deionized (DI) water, etc.). Guidance on establishing specications for monographed water is provided in the major compendia (References 4, 5, and 6, Appendix 1).

2.2

Pharmacopeial Water and Steam The major compendia specify standards of quality quality,, purity purity,, generation, packaging, and labeling for a number of waters, including two bulk waters, “water for injection” and “puried water” and “pure steam” used in the preparation of compendial dosage forms. This Guide is concerned with the production, storage, and distribution of compendial bulk waters and steam, and does not address the other “packaged waters” monographed by these compendia.




2.2.1

Puried Water and Water for Injection Ofcial requirements for PW and WFI are provided in the monographs included in the compendia. These monographs provide the minimum requirements for production methods, source water quality, and quality attributes. The major compendia have been largely harmonized (with some exceptions). The water user should review the monographs for specic attribute requirements to ensure compliance for each specic case.

2.2.2

Pure Steam Ofcial requirements for “pure steam” (also referred to as “clean steam”) are provided in the monographs included in USP. USP. This monograph provides the minimum requirements for production methods, source water quality quality,, and quality attributes. The pure steam user should review the monographs for specic attribute requirements to ensure compliance for each specic case.

2.2.3

Testing Requirements

2.2.3.1

Conductivity and Total Organic Carbon (TOC) The compendia provide chapters that describe the test method and instrumentation requirements for the testing of conductivity and TOC. Conductivity measures non-specic conductive ions in the water. TOC is an indirect measure, as carbon, of organic molecules present in high purity water. Instruments are available for measuring conductivity and TOC at-line from slipstreams, and off-line from grab samples manually removed from the water system. Automatic Automatic off-line sample introduction systems are available for processing large numbers of grab samples. For further information, see Chapter 11 of this Guide.

2.2.4

Microbial and Endotoxin Testing Microbial contaminants and endotoxins are traditionally sampled at the points of use in a water system. Additional sampling points are recommended at the outlet of the generation system and periodically before and after any unit operation design for total viable bacteria or endotoxin reduction (e.g., reverse osmosis). For further information, see Chapter 13 of this Guide.

2.2.5

Validated (Veried) Backup Instrumenta Instrumentation tion Failure of a monitoring instrument should not be precluded when making decisions concerning type, location, and the extent of validation (verication). On-line installations should be supplemented with a calibrated laboratory instrument as backup. Validation Validation (verication) should include the operation in off-line mode as a supplement or alternate to online instrumentation. Off-line laboratory testing also can include a backup instrument to be maintained calibrated in case of failure of the primary unit. Backup also may be performed by using a qualied third party laboratory. laboratory.

2.3

Specication of Pharmaceutical Water Quality

2.3.1

Establishing Establishin g Acceptance Criteria It is the responsibility of the water user to conrm the quality of water supplied in any pharmaceutical process is consistent with the quality required for the nal product. It may not be sufcient to specify a water quality that meets the specication of the grades of bulk water outlined in the compendia. These grades, described in the PW, WFI, and EP highly puried water monographs, are minimum standards. A more stringent specication could be required depending on the intended use of the product and on the process used to manufacture that product.




Pharmaceutical water uses typically include: •

an ingredient in a dosage form manufacturing process

•

an ingredient in an Active Pharmaceutical Ingredient (API) process (the term API is used interchangeably with Bulk Pharmaceutical Chemical (BPC))

•

a nal rinse during Clean in Place (CIP), clean out of place, or manual cleaning

•

a solvent or diluent for research or laboratory processes

Water intended for use as a dosage form ingredient must be compendial water and must be produced consistently to meet monograph requirements. Evidence of control is required for all critical process parameters that may affect the nal drug characterization. WFI must be used for parenteral manufacture. Waters with endotoxin control are expected for some ophthalmic and some inhalation products. For further information, see Chapter 3 of this Guide. The monographs for PW and WFI provide the baseline requirements for water used in production, processing, or formulation for pharmaceutical activities. For some applications where there are no requirements for compendial waters, the user may establish quality specications equivalent equivalent to PW or WFI, depending on the specic application, or equivalent to these waters with additional/fewer requirements. Note: the use of compendial monographs when not required should be closely evaluated due to cost and maintenance impacts. Specications for water used as an ingredient (exclusive of sterile bulks) in the manufacture of APIs or as solvent in the wash or rinse cycles should be determined by the user. In some cases, “drinking water” may be acceptable, or certain chemical or microbial or endotoxin quality specications may be established, or one of the compendial waters may be used. The specication should be based on the potential for alteration of the nal drug product. With the appropriate justication, non-compendial waters (including “drinking waters”) may be utilized throughout pharmaceutical operations, including production equipment washing/cleaning as well as rinsing, laboratory usage, and as an ingredient in the manufacture or formulation of bulk active pharmaceutical ingredients. However, However, compendial water must be used in the preparation of compendial dosage forms. In both compendial and noncompendial waters, the user should establish an appropriate microbial quality specication. specication. The signicance of microorganisms in non-sterile pharmaceutical products should be evaluated in terms of the use of the product and the nature of the product and the potential harm to the user. Manufacturers are expected to establish appropriate microbial alert and action levels for microbial counts associated with the types of pharmaceutical waters utilized. These levels should be based on process requirements and the historical records of the system in question. Acceptance criteria for pharmaceutical waters and steam are determined by the manufacturing process, end product, and applicable regulatory and compendia requirements. Pretreatment and nal treatment subsystems, storage and distribution systems, systems, and operator/maintenance procedures are then designed based on, among other criteria, quality of the feed water to meet the acceptance criteria.

2.4

Critical Process Parameters Critical Quality Attributes (CQA) and Critical Process Parameters (CPP) are dened in ICH Q8 (Reference 7, Appendix 1) and ASTM E2500 (Reference 8, Appendix 1) and should should be referenced where applicable applicable to a water system. For further information on CQAs and CPPs, see to Chapter 12 of this Guide.




2.5

CGMP Compliance Issues Satisfying regulatory concerns is primarily a matter of establishing proper specications specications,, and using effective and appropriate methods to verify and record that those specications are satised. Issues such as quality of installation, sampling and testing procedures, operating and maintenance procedures, training, record keeping, etc., may be considered as signicant as the particular technologies selected to purify and distribute the water. Each pharmaceutical steam and water system should be viewed in its entirety entirety,, as design and operational factors affecting any unit operation within the system can affect the whole system. It is useful to identify both the quality parameters of water entering the system and the quality parameters of the water or steam to be produced. Water quality should be enhanced with each successive step. It does not necessarily follow that measures enhancing one quality attribute (such as conductivity, conductivity, particulate level, or color) will always enhance another (such as microbial population).

2.6

Design Range versus Operating Range This Guide provides detail on recognizing the distinction between design range, allowable operating range, and normal operating range and the impact this distinction has upon qualication and facility system operation (see Chapter 11 of this Guide). Details on identifying the use of action and alert levels also are provided. •

Alert levels are based on normal operating experience and are used to initiate corrective measures, such as sanitizations, but may not initiate a formal corrective action plan.

•

Action levels are based on operating experience considering yearly uctuations or worst case and upset conditions.

While a water or steam system should meet all stated design conditions, the acceptability of the system for operation from a CGMP standpoint depends on operating within the allowable operating range. Example For example, the design range for a pharmaceutical water system may require a nal product water quality conductivity of 0.5 µS/cm (2 Mohm-cm) or lower as a specication limit. The However, the allowable operating range for this pharmaceutical water may allow for generation of water quality with a conductivity of 1.3 µS/cm (0.77 Mohm-cm) or lower. The normal operating range for generating water may, in the end, be set by the manufacturer at conductivity value approaching 1.0 µS/cm (1.0 Mohm-cm) or lower to provide a comfortable environment for the operation. The normal operating range generally does not exceed the allowable operating range for the product water. Figure 2.1 denotes normal operation. Testing Testing of a water system may exceed action levels.




Figure 2.1: Val Values ues of Critical Parameters for Product Water

Note: These Note: These are general representations provided for example.



3

Water Options and System Planning



ISPE Baseline ® Guide: Water and Steam Systems

Page 19 Water Options and System Planning

3

Water Options and System Planning

3.1

Introduction This chapter outlines basic water system design criteria, and along with subsequent chapters, aims to provide a better understanding of pharmaceutical waters, how they are used, and their quality requirements. The primary goal of this chapter is to provide the user with a methodology for: •

understanding, evaluating, and selecting among water quality options

•

evaluating various system congurations able to meet selected quality requirements

•

understanding system planning and programming steps

Information regarding unit operations, system design, maintenance, and relative costs is addressed in subsequent chapters. This chapter also outlines the system planning effort for pharmaceutical water systems, beginning with the selection of water quality, based upon product requirements, processing operations, and end use. A decision tree is included to assist in selection of compendial and non-compendial waters for use in production, cleaning, and support. Subsequent information guides the user through use-point and system analysis resulting in an overall water system distribution strategy. strategy. Evaluation points are provided to assess various system congurations.

3.2

Water Quality Options Quality requirements for water used in pharmaceutical manufacturing and product development are driven by the product characteristics, manufacturing processes, and route of administration of the product. To To aid in the water selection process, compendial monographs dene minimum requirements for general types of pharmaceutical water used in most applications. However, there is also the opportunity for a manufacturer to establish water quality requirements, different from those dened in the monographs, based on specic product characteristics characteristics and processing operations. Ultimately, Ultimately, the product manufacturer is responsible for assuring that water used to manufacture the product is appropriate, and can be proven to reliably produce safe product. Although water quality requirements are product specic, it is impractical to produce special water that is specic to each situation. Based on cost and other limitations, manufacturing operations typically generate and distribute only a few, or perhaps just one, quality of water; therefore, products and operations requiring similar water qualities are commonly grouped. The waters dened by monographs are generally viewed as adequate for production of safe product; however, more or less stringent water quality specications may be appropriate for some products and processes as determined by that products manufacturer. More stringent requirements may apply to some processing operations involving signicant concentration concentration steps or products with high water content and which may be applied in large volume doses. Likewise, processes involving reliable purication and/or sterilization steps that remove impurities may, in some cases, not require water qualities as strict as those dened in the monographs. Other process characteristics characteristics can affect water quality requirements as well. In manufacturing operations, where the generation of only one quality of water is practical, the water system should be designed to meet the most stringent requirements of the most demanding product or process. When more than one quality of water is available, products and processes are often categorized and fed by the most appropriate system. The number and types of water generated is most often a function of the volume of water consumed, the variation in quality required and the cost differential. Larger consumers may nd it economical to generate and distribute multiple grades of water, while smaller users will often generate only one quality of water.




Waters used in the development, manufacture, or preparation of drugs and drug substances can be classied broadly into two categories; compendial and non-compendial waters. Simply, compendial compendial waters are those that are dened by monographs in recognized pharmacopeias. The three pharmacopeias discussed most frequently in this Guide include: •

United States Pharmacopeia (USP) (Reference 4, Appendix 1)

•

European Pharmacopoeia (EP) (Reference 5, Appendix 1)

•

Japanese Pharmacopoeia (JP) (Reference 6, Appendix 1)

On a worldwide basis, additional pharmacopeias and compendia that prescribe quality requirements for pharmaceutical waters may be applicable. Ultimately, Ultimately, the locale or region where the drug product or substance is sold or used dictates the inuence of a specic pharmacopeias. As As the requirements for specic water qualities are not completely harmonized, it is prudent to consult with the appropriate quality unit to ensure all applicable regulations are met. •

Compendial waters meet waters meet the r equirements for specic types of water in the applicable monographs. Examples of compendial waters include: -

Puried Water (PW) – dened by USP USP,, EP EP,, and JP monograp monographs. hs.

-

Water for Injection (WFI) – dened by USP USP,, EP EP,, and JP monogra monographs. phs.

-

Highly Puried Water (HPW) – dened by EP monograph.

-

Water for Hemodialysis (WFH) – dened by USP monograph.

The specications outlined in the monographs are the minimum requirements for compendial waters. Additional specications may be added or combined based on process requirements. For instance, an endotoxin specication could be added to PW if needed for the process or product application. •

Non-compendial waters meet waters meet the requirements of potable water at a minimum and are often additionally treated to meet process requirements. They may contain added substances for microbial control and do not have to meet all compendial requirements. Non-compendial waters are sometimes described by the nal or critical purication process used, e.g., RO water. In other cases, non-compendial waters are described by a specic quality attribute of the water, e.g., low endotoxin water. It is important to note that non-compendial waters are not necessarily of lesser quality than compendial waters. Non-compendial waters may, in fact, be of much higher quality than compendial waters if required by the application. Common non-compendial waters include: -

Potable water – water that meets US Environmental Protection Agency (EP (EPA) A) National Primary Drinking Water Regulations (Reference 9, Appendix 1) or comparable regulations of the European Union, Japan, or the World Health Organization (WHO). This is the minimum quality of water used in pharmaceutical processing and the starting ingredient for preparation of any compendial waters.

-

Softened water – potable water that has been additionally treated and includes a water softening process to remove hardness generally associated with calcium and magnesium contaminants, as either the nal or most important unit operation

-

RO water – potable water that has been additionally treated and includes reverse osmosis as either the nal or most important unit operation




-

UF water – potable water that has been additionally treated and includes ultraltration as either the nal or most important unit operation

-

DI water (or EDI water) – potable water that has been additionally treated and includes a deionization process such as ion-exchange or electrodeionization as either the nal or most important unit operation. May be classied as EDI water if the deionization process is specically electrodeionization electrodeionization

-

Distilled water – potable water that has been additionally treated and includes distillation as either the nal or most important unit operation

-

Laboratory water – potable water that has been additionally treated and meets the requirements for water used in laboratory applications. For further information, see Chapter 9 of this Guide.

Non-compendial water is not necessarily less critical, less costly to produce, or more difcult to qualify qualify,, than compendial water. It can enable the manufacturer to set product specic quality and/or test criteria that are appropriate for the specic product and processes. In addition, non-compendial water systems may or may not be validated (see Chapter 12 of this Guide). Generally, more highly puried water is more expensive than less puried water; however, the specics of each Generally, operation are different. For example, a plant with existing excess capacity of WFI might elect to use WFI rather than a suitable lower grade of water. In the example case, documentation should identify the quality required for the product and why the WFI was used instead. An assessment also should be conducted to ensure that the water is not too pure for any application. Figure 3.1 shows a decision tree that can be used by manufacturer organizations to determine appropriate requirements for water used throughout their pharmaceutical manufacturing processes. The completed diagram should be accompanied by documentation supporting the options chosen with review and approval by the appropriate Quality Unit. Decisions should be based on product and process specic requirements as water supplied to any process must consistently meet or exceed the requirements, dened by the manufacturer, for safe and reliable manufacture of that product. Figure 3.1 includes the most common water types and selections; however, it is impractical to provide a single decision tree capable of covering the full breadth of diverse choices possible.




Figure 3.1: Pharmaceutical Water Quality Decision Tree

Table 3.1 provides baseline requirements for most product contact water applications. Water quality criteria for pharmaceutical manufacturing and product development are driven by the product characteristics, manufacturing process, and the intended use of the product. Specic product and process characteristics may dictate that more or less stringent criteria than shown are appropriate. Figure 3.1 gives engineers some general guidance on selection of pharmaceutical water quality. quality. Quality Units should be consulted to give further advice on this critical selection of pharmaceutical water.




Once water quality requirements have been determined, Table 3.1 identies common unit operations that may be applicable. The arrangement of components varies widely throughout the industry industry.. With the exception of WFI production, in accordance with the EP and JP, the method of manufacture of compendial waters is not dened. The primary criteria in evaluating process options is whether the selection will assist in consistently producing water of the required quality This table should not be viewed in absolute terms. Rather, it should provide a general recognition that systems can be created using a variety of components when suitably arranged. This is not to say that every system should have each of the components listed or that components can only be used as indicated. Table 3.1 Typical Unit Operations Based on Water Quality Requirements Turbidity and Particulate Reduction

n o i t a r t l i F e s r u o C / y r a m i r P

g n i n e t f o S

PW

X

X

WFI

X

Highly Puried

Hardness and Metal Reduction

n o b r a C d e t a v i t c A

Chlorine Removal and Organic Reduction

Organic, Inorganic, Bacteria and Bacterial Endotoxin Reduction

n o i t a l l i t s i D t c e f f e i t l u M

n o i t a l l i t s i D n o i s s e r p m o C r o p a V

Ionic Reduction

Bacteria or Endotoxin Reduction

r o , n o i t a r t l  o r c n i o i M t , a t r i t n l  U a r t V l U U

n o i t c e j n I e t  l u S

O R s s a P e l g n i S

Note 1

X

X

N ot ot e 3

No te te 2

No te te 2

X

Note 1

X

Note 6

Not e 6

Note 4

Note 4

X

X

Note 1

X

X

Note 3

X

Softened Water

X

X

DI Water

X

Note 1

X

X

Note 3

X

Note 5

Note 7

RO Water

X

X

Note 1

X

X

Note 3

Distilled Water

X

X

Note 1

X

Note 6

Not e 6

Laboratory Water

X

X

X

X

Note 3

X

Note 5

Note 7

O R s s a P o w T

I D E

X

e g n a h c x e n o I

Note 5

g n i h s i l o P p o o L

Note 7

Note 7

Potable Water

Note 4

Note 4 X

Notes: X Commonly employed or applied process. 1. Activated carbon is commonly used in in lieu of sulte injection on surface feed water water supplies or feed waters high in organic organic content. 2. Distillation may be used to produce PW. PW. 3. An RO second pass may be used on feed feed waters high in TDS or for specic contaminant contaminant reduction. 4. Single or multi-effect is commonly used used for lower capacity systems where vapor compression compression is common for higher capacity systems. For further information, see Chapter 6 of this Guide. 5. Service exchange (non-regenerable) (non-regenerable) DI is most commonly used for low low ow or intermittent intermittent demand systems. 6. RO is used for vapor compression pretreatment for specic specic contaminant reduction, such as silica silica or endotoxin. RO (single or two pass) is more more commonly employed for multiple effect distillation feed. 7. UV units are used throughout water treatment systems for the reduction of bacteria, bacteria, chlorine, organic material. Filters are often employed employed for reduction of suspended solids, colloidal material, bacteria. For further information, see Chapters 4 and 5 of this Guide.



3.2.1


Life Cycle Costs Estimating the cost of pharmaceutical water production is not complicated although it may not always be easy to properly identify all of the associated costs. Costs are actually quite predictable; however, they vary signicantly depending on scale of operation, system design, actual usage, utility costs, maintenance philosophy, philosophy, and complexity complexity.. Items that can have a signicant impact on overall costs include: •

labor for operation

•

monitoring

•

preventive maintenance and service

•

waste disposal

•

sampling costs

The total operating cost to produce pharmaceutical waters is obtained by adding the cost of feed water to the operating costs of the entire system (including pretreatment and nal treatment processes), along with any ancillary items identied. The system operating cost includes common expenses, such as utility costs, etc., but also will likely include costs somewhat unique to water systems, such as periodic de-rouging and passivation of stainless steel systems. Other signicant costs costs should be anticipated for validation, ongoing QA/QC, and calibration, as well as waste treatment and sewerage. In addition, regulated industries must consider the risks (cost) of noncompliance and water system failures. Storage and distribution systems also should be considered when estimating system operating, maintenance, and calibration costs. To correctly compare various technologies or to determine the total cost of generating pharmaceutical water, the overall life cycle cost, including capital and operating costs over the expected lifetime of the system, should be investigated. This relatively simple simple exercise is a critical process to determine the actual costs that a pharmaceutical water system will incur. A system with a large service component may be the least expensive to purchase, but require a far greater operating expense. If one can reasonably predict the lifetime of a system, one may be able to justify a greater greater capital investment investment if a lower operating operating or overall life cycle cost is realized. Conversely Conversely,, when the water requirement is temporary or of unknown duration, a system requiring minimal capital investment, such as exchangeable deionization, may be an attractive alternative. System scalability also should be considered in life cycle cost analysis. The designer must be aware of the impact on the overall design if future production quantities need to be increased or decreased. Certain designs may not be easily scaled up or down and the storage and distribution systems may not be able to handle capacity changes. The generation technology of choice and its associated capital cost are utilized to determine the total pharmaceutical water system Net Present Value (NPV). The technology choice is based on any applicable regulatory requirements for method of manufacture, capacity requirements, feed water, total dissolved solids and hardness levels, organic and colloidal content, as well as anticipated water system utility costs (acid, caustic, salt, power, power, and source water). Consideration also should be given to maintenance requirements and available resources to maintain continuous operation. Although water treatment systems for generating either compendial or non-compendial pharmaceutical process waters signicantly vary in system operational costs, NPV for each of these various types of process waters are quite similar. The only exception is DI process water generated through the use of a non-regenerable exchange system, typically regenerated off site. The NPV analysis is usually based on the water sys tem capital cost and a system operating cost for an estimated period (e.g., 5 to 10 years). The period chosen has to be long enough to allow operating cost to be a signicant factor, factor, but short enough for reasonable analysis of operating cost returns versus increased capital expenditures.




Due to varying feed and product water conditions, site-specic utility costs and availability availability,, and end-user driven requirements, an economic evaluation should be performed for every proposed system. One cannot assume that the process design based on the optimum life cycle costs will be equivalent for different capacities and applications. This exercise is critical to the basis of design development and may result in signicant savings over the expected life of a water system. While it may be common to base decisions on capital cost and return on investment, neglecting operational and maintenance costs can lead to erroneous conclusions and result in a higher cost of products. Ensuring the most suitable water quality is selected and the life cycle cost of the system is evaluated can typically result in overall costs nearing optimization; however, additional opportunities for cost savings may be found in reusing system wastewater for suitable applications (e.g., lawn irrigation, humidication, boiler feed or make-up, cooling tower make-up). Materials of construction (including nishes), instrumentation and controls, and redundancy/reserve capacity are also extremely inuential to the capital cost associated with a technology choice. Subsequent chapters in this Guide also address cost savings issues associated with the basis of design, unit operation selection, and overall system conguration.

3.3

System Planning Water and steam are often the most widely used raw materials or utilities in a pharmaceutical facility facility.. They also may be a considerable production expense. Improper sizing or selection of a system or its components could dramatically impact facility operations by limiting production, if under sized, compromising quality and reproducibility and/or increasing capital cost, if oversized. It is most important to recognize that system design or sizing is not the starting point in design. Dening water quality requirements, acceptance criteria, and usage should be the rst step and if properly determined they will optimize capital, while minimizing construction and operational costs. Figure 3.2 provides a graphic representation of the system boundaries, limitations, and restrictions the designer must address when planning a pharmaceutical water system. During initial planning, the limits of each boundary should be established. The arrows encircling each boundary represent inputs that establish more specic operating strategies and ranges. When documenting these requirements, the designer should, whenever possible indicate ranges of acceptability, rather than a specic value or position. Ranges allow more exibility in nal planning and when making detailed design decisions. Certain restrictions can necessitate a specic strategy; however, as long as the decision leads to a result that is within the limits of the established system boundaries, it will most likely be acceptable. An An example might be a facility that requires a new use-point to deliver non-compendial water with microbial control. Within the facility, there happens to be an existing oversized WFI system in an adjacent area that the designer decides to utilize for the new use-point since the quality required will be exceeded by WFI. In this example case, documentation should identify the quality required for the product and why WFI was used instead. Of primary importance is the systematic approach to planning a pharmaceutical water or steam system. Planning should begin with the determination determination of water quality quality.. Then, Points Points of Use (POU) delivery criteria are evaluated, possibly with multiple POU arrangements, followed by an initial system planning exercise. Often, these sequential steps are repeated as information in the design process iterates, and further criteria about the overall system boundaries are identied.




Figure 3.2: Pharmaceutical Water System Planning

3.3.1

Establish Water Quality The rst step in the evaluation of water systems is the selection of water quality required for the specic product and process operation. Selection is based primarily on the dosage and form, the microbiological and chemical purity criteria for the product(s) for which the water is used (production and cleaning), and any applicable regulatory requirements. The selection must consider underlying factors that will impact quality control; installed and operating cost; maintenance and practicality practicality..




System design constraints may provide the motivation to challenge water quality or other criteria, particularly when it can be demonstrated that the change does not affect product quality or manufacturing controls.

3.3.2

Characterize Use Point Once the initial selection of water quality has been established, the operational criteria should be characterized for each use point. When evaluating POU requirements, it is often prudent to create a spread sheet or design database that summarizes all pertinent data, allows adequate space for notes and is updated to serve as a reference throughout the design process. Each POU should be identied with a unique tag and annotated with the proper values for pressure, ow, and the temperature range of water delivered for use. Establishing a range, when possible, rather than a xed value, increases opportunities for system optimization optimization by allowing a more exible approach to nal design. Classication of each use-point should include the following: •

unique identier (tag)

•

purpose or user (e.g., cleaning, batch vessel)

•

maximum instantaneous ow rate

•

periodic consumption requirements (e.g., daily daily,, weekly) and duration

•

pressure requirement

•

temperature requirement

•

utilization schedule

•

method of delivery (automatic or manual)

•

type of connector (xed or hose)

•

notes or other special requirements (if applicable)

This data can be organized in many ways, but a suitable spreadsheet can simplify the planning process and indicate decision pathways for future detailed design activities. Table Table 3.2 provides an example of a spreadsheet that might be used for system planning. In general terms, ow rate is primarily used for sizing distribution lines, whereas the total daily consumption (considering diversity for peak days), divided by the operating hours for the process gives the minimum rate for the generation plant. Both can be useful in determining storage requirements. A diversity factor is one way to normalize anticipated usage, assuming that not all loads happen every day or at the same time. Diversied usage can be applied if activities will happen either at the same time or if they can be staggered so that a smaller pump, equipment, and piping network might be used. Table 3.2 indicates a CIP system and stopper-washer that are both likely to be used on the same day day,, but not at the same time. Therefore, only the higher ow rate is relevant to loop sizing as shown in the Flow Rate column. Demand ow rates are eventually used for user connection sizing.




Table 3.2: Point of Use Criteria ID (Tag)

Temp. °F (°C)

Press. PSIG (BARG)

Type A = Auto

Equipment Name (Purpose)

M= Manual

Flow Rate

Daily Use

Comments

Dema De mand nd

Dive Di vers rsit ity y

Desi De sign gn

Dema De mand nd

Dive Di vers rsit ity y

Desi De sign gn

GPM (LPM)

Factor

GPM (LPM)

GPD (LPD)

Factor

GPD (LPD)

GPM (LPM)

Factor

GPM (LPM)

GPD (LPD)

Factor

GPD (LPD)

WFIWF I-1 1

181-18 181185 5 (83-85)

60-65 (4-4.5)

M

CIP Wash Tank

10.6 (40)

1

10.6 (40)

317 (1200)

1

317 (1200)

Assume a recirculating cycle in 4 steps for total of 23 minutes

WFI-2

68-74 (20-23)

35-40 (2.42.75)

A

Stopper Washer

5.3 (20)

0

0

122 (460)

1

122 (460)

Assume one cycle per day. 100liters/ rinse, 3 rinse/ cycle @ 2lpm for 60mins.

After location and qualities are nalized, the various properties can be charted on a requirements analysis histogram. This can be created with the aid of a computer and either simulation or spreadsheet software for larger s ystems, or manually for systems with minimal users. At this point, basic process ow diagrams also provide a good pictorial view of the water qualities, locations, and the point-of-use properties. The key to establishing the usage prole and generating a water usage chart is to break down the water usage to specic time intervals (e.g., hourly periods) for each user and sum all the requirements over that time interval. For a specic generation rate and storage tank volume, the tank volume can be plotted based on the consumption specic for each time interval. The capacity of the tank and the generation rate can be optimized considering other design requirements such as space and economics. Figure 3.3: Water Usage Chart



3.3.3


Establish System Criteria Based on the established water quality and POU criteria, the system criteria can be established with the following having a signicant inuence: •

generation rate

•

storage capacity

•

peak and average demand

•

system demand limit based on maximum loop draw-off

•

effect on existing system(s) (if applicable)

Histogram analysis is benecial for determining overall system peak demand(s), average demand, and the relationships between peak demand time periods and their ow rates. Figure 3.4 shows a hypothetical storage tank prole using the 24-hour demand prole from Figure 3.3. There is no guideline for minimum or optimum water level prior to start of relling the vessel. Constant storage tank level may not be a requirement, particularly if cycling of the generation system on and off may be undesirable depending on the technology and water ush requirements with system start-up. Generating these charts as shown in Figure 3.3 and Figure 3.4 provide the tools for creating various scenarios to simulate; recovery times from a failure, future expansion or reduced capabilities, and to analyze other factors that assist design of a properly sized water generation, storage, and distribution system. System planning and analysis also reveals other requirements that inuence design, and often lead the designer to re-evaluate the primary boundaries as discussed earlier. These issues might include: •

The system must be available at all times.

•

Shutdowns must be limited in number and duration.

•

Plant and personnel are not equipped to handle chemicals properly properly.. No permits in place.

•

Production is batched versus continuous.

•

Products dedicated or multiple product groups?

•

Product campaigns dictate unique or restrictive operating requirements.

•

Limited time is available for sanitization.

These issues should be addressed adequately and may dictate redundancy be provided to allow adequate time for water generation (tank ll), sanitization, maintenance, service, etc.




Figure 3.4: Storage Tank Level Chart

3.3.4

Revisit Water Quality With all use points characterized, the quality of water may be revisited. A thorough review may reveal a wide range of acceptable delivery conditions. Since it is often not practical to operate multiple water systems to provide the exact water conditions desired for each end product, compromises are typically required. These compromises may include providing water of a higher quality than required to simplify or reduce cost; however, water should continue to be delivered at conditions within the boundary limits, unless these requirements are changed.

3.4

System Design Once all ancillary functions are dened, detailed design of the system can begin. The process requirements determine the Points of Use (POU) in the distribution system. User locations determine how to distribute the water (e.g., central storage, multiple loops/sub-loops, etc.). One or several of the use points may require different properties that necessitate a change in distribution philosophy from a simpler design. In this case, alternatives to the water system criteria are considered, such as increasing the number of loops, changing the loop conguration (e.g., changing from a cold loop to a hot loop with POU coolers where necessary) or creating sub-loops. Plant operating schedule also must be considered since an inability to perform regular tasks, such as sanitization may render the system inoperable. The boundaries, limitations, and restrictions that were identied in the initial planning stages must be integrated into the design approach. Additional considerations might include the physical area required. The space needed could be simply an allocation within an equipment area or depending on the system complexity, complexity, may require additional support space, space at a remote site for satellite equipment or even multiple locations throughout a single building or campus setting. Plans should be in place to manage both scheduled and unscheduled maintenance. When practical or necessary equipment redundancy allows for work to be performed with only a partial loss of capacity, backup equipment for critical functions should be evaluated during the design process and is usually driven by production or related requirements. Specic details of activities required to complete a design are provided in the following chapters and include various alternatives for unit operations and system concepts. Rationale is provided for any recommendations included; however, any design team must recognize that the requirements unique to their products and facilities will ultimately dictate the nal outcome.


4

Pretreatment Options




Page 31 Pretreatment Options

4

Pretreatment Options

4.1

Introduction Pretreatment is often referred to as those process steps upstream of the primary ion reduction step(s). Pretreatment generally involves a series of unit operations to condition the feed water quality so that it will be of adequate quality to optimize the performance of the nal treatment step. Final treatment may be: •

RO

•

Ultraltration

•

Deionization

•

Distillation

RO is unique since it also can be a pretreatment step. For further information, see Chapters 5 and 6 of this Guide. This chapter discusses the process design for pretreatment including feed water quality and pretreatment efuent quality,, followed by a discussion of the selection of treatment options for the primary groups of impurities: quality

4.2

•

Control of fouling – removal of turbidity and particulates (suspended materials).

•

Control of scaling – removal of hardness and metals (dissolved materials).

•

Control of organics including microbiological impurities.

•

Removal of microbial control agents (sanitization agents).

Process Design of Pretreatment Process design of the pretreatment is the specication of unit operations or process steps required to properly treat the feed water in order to optimize primary treatment functionality. functionality. Typical information includes ow rates, temperatures, pressure, and composition of all streams. Detailed mechanical design of the equipment for individual unit operations or process steps is beyond the scope of this Guide. The process design issues for pretreatment may include: •

Required quantity and quality of water needed to be produced by the nal treatment process.

•

Temperature constraints for the water used in a pharmaceutical process and the approach to microbial control.

•

The nal treatment option that has been chosen, (this affects the required water quality leaving pretreatment).

•

Characteristics of the feed water that is supplied to the pretreatment (details typically are evaluated over a period of one year).

•

Evaluation of the difference between supply water quality and desired product water quality quality,, which effectively denes impurities to be removed by the pretreatment, and is the result of performing a material balance. Consideration should be given to other inuences affecting overall water quality. quality.




•

Pretreatment options able to provide the desired impurity removal considering factors, such as: -

skill level of the labor force

-

economics

-

waste disposal

-

environmental considerations

-

validation

-

required space

-

available utilities

In addition to dening the options for removal of non-viable impurities, the approach taken for microbial control is an integral part of the pretreatment process design and consideration should be given to: •

The feed water to the pretreatment must comply with suitable local codes for drinking water, such as US EP EPAANPDWR (Reference 9, Appendix 1) or WHO standards for drinking water (Reference 10, Appendix 1), and typically will contain chlorine or chloramines as a microbial control agent. In Europe, ozone may be a more common microbial control agent. The residual disinfectant concentration should be evaluated and may be sufcient to protect the initial steps of the pretreatment.

•

If the concentration and application of microbial control agents are insufcient, additional microbial control functions may be added, including provision to periodically sanitize the initial equipment in the pretreatment system. This is more likely if supply water comes from an unregulated source, such as a private well. Increased monitoring of feed water and pretreatment also may be warranted.

•

In most cases, the microbial control agent must be removed prior to nal treatment. A means of either continuous or periodic sanitization must be implemented for subsequent activities following removal of the microbial control agent.

The regulatory requirement that compendial waters should contain “no added substances” restricts any addition of chemicals to PW or WFI. The addition of chemical agents is not prohibited during treatment or for sanitization, provided they are removed from the nal product water; therefore, various substances are often added and removed during treatment to optimize the overall treatment regime. Examples include: •

chlorine, chloramine, chlorine dioxide, or ozone (to control microbial growth) removed in later stages of treatment

•

sodium ions (in softener with exchange for multivalent ions) removed during an ion removal process

•

acid (for degasication to remove carbon dioxide, counter ions (ions of opposite charge)) typically removed in a subsequent ion removal process

•

sulte (to reduce chlorine to chloride or chloramines to ammonium and chloride, while forming sulfate) removed by softening or during an ion removal process

•

sequestrants (to prevent scaling in nal treatment) removed during nal treatment

•

pH control agents (to convert CO2 to weak acid) removed by an ion removal process




Where added substances result in increased microbial growth or higher levels of endotoxin, they should be evaluated. It may be necessary to verify the absence of any added substances in the nal product water, in particular those that are not ubiquitous in drinking water supplies. An important consideration is the relationship between investment and operating costs as they relate to pretreatment, and its impact on the performance and operating cost of the nal treatment process. The following are generally true: •

Operation of nal treatment will be adversely affected if operation of the pretreatment system is inconsistent, unreliable, or inappropriate.

•

Inadequate pretreatment function (e.g., breakthrough of particulates, hardness, or chlorine) may result in product water quality failures. Most likely, likely, however, it will not affect water quality from nal treatment, but will rather be reected in long term maintenance costs and operating reliability. reliability.

•

Investment in pretreatment often returns many times its value in nal treatment maintenance costs savings.

•

Pharmaceutical water systems are expected to produce water that meets exacting regulatory or product specic requirements. Systems should be designed to control impurity variability in the feed water, as well as seasonal impurity prole changes. A robust pretreatment design minimizes the impact of these variations on downstream unit operations and nal treatment.

There is no single correct answer to the process design for pretreatment, but rather there are a series of choices and options, each with advantages and disadvantages that must be evaluated with decisions appropriate for the specic application.

4.3

Feed Water to Pretreatment Quality: Testing and Documentation Compendial water systems typically must use feed water complying with drinking water standards, such as that from a local municipal source which is commonly treated with a microbial control agent such as chlorine or chloramine. Based on seasonality, seasonality, both feed water composition and microbial control agent concentration may be subject to variations. In addition, events such as droughts, oods, and other anomalies may periodically occur. These variations may impact water quality, and can be detected only by extensive sampling. In addition, water arriving at a site may not be equivalent to that supplied by a municipal treatment facility, due to potential for contamination or loss of microbial control agent in the distribution system. Ongoing evaluation of feed water quality is recommended using data provided by the supplier (if applicable) supplemented by an appropriate level of testing at the site to ensure the equipment will be able to address any adversity without inappropriate risk to product and/or patient. If a feed water source does not meet local drinking water requirements, separate conditioning techniques upstream of the water treatment system are required to assure these requirements are met. Further, if a feed water tank (buffer tank) is utilized, the system must be designed to ensure that the quality of the feed water is maintained within drinking water specications. Typical Typi cal contaminants in feed water include: •

Turbidity and particulates: silt, dust, pollen, pipe scale, iron and silica, undissolved minerals, and organic compounds.

•

Inorganics: calcium and magnesium salts, heavy metals (iron, aluminum, and silica) with their corresponding anions.

•

Organics: naturally occurring byproducts of vegetative decay decay,, i.e., humic and fulvic acids and “man-made organics” such as pesticides, herbicides, and automotive pollution (oils).

•

Bacteria: bacterial contamination and its byproducts, endotoxins, and pyrogen.




Documentation that feed water meets the necessary quality may be based on municipal data alone, if proven to be reliable and representative, or supplemented/replac supplemented/replaced ed by other suitable testing. The frequency of in-process testing may be affected by the reliability of municipal data, importance of monitored variables, and an organization’s philosophy. Microbial control agent monitoring should begin at or before water enters the system, most often at the start of a pretreatment system. A total chlorine level of 0.2ppm to 0.5ppm is generally considered adequate to control microbial growth and usually has negligible effects on pretreatment equipment and its performance. Chlorine levels are affected by pH, and as a result, pH monitoring may be required if chlorine levels are erratic or difcult to maintain. Controlling chlorine excursions is critical due to the potential effect on nal treatment processes (see Table Table 4.1). Specic testing for contaminants known or suspected of being present in the feed water should be performed, particularly if they appear at variable levels, times, or if slight changes can impact system operation. This additional testing may be necessary if data from the source is inadequate or if unpredictable events can signicantly alter feed water quality, quality, such as when surface waters are contaminated by run-off of pesticides used in agriculture. Periodic monitoring of these parameters to monitor seasonal uctuations may be required.

4.4

Output Water from Pretreatment: Quality of Feed Water to Final Treatment Pretreatment should provide water quality that optimizes the operation and maintenance of subsequent treatment equipment and permit the nal treatment step to produce water meeting the desired specications, reliably and routinely. Impurities that must be removed by pretreatment, to enable reliable operation of the nal treatment, depend on the nal treatment design and its tolerance for those impurities. If pretreatment is inadequate, resulting problems can be signicant, and are categorized in Table 4.1. Table 4.1: Problems in Final Treatment Caused by Impurities Impurity Magnitude of Problems in Final Treatment Caused by Type of Impurity

FOULING: caused by particulates

SCALING: caused by hardness and minerals

CORROSION: caused by chlorides

DEGRADATION: caused by chlorine

Reverse Osmosis

Moderate

Moderate

None

Large

Continuous Electrodeionization

Moderate

Moderate

None

Large

Single Effect Distillation

Moderate

Moderate

Moderate

Large

Multi-Effect Distillation

Large

Large

Moderate

Large

Vapor Compression Distillation

Moderate

Large

Small

Large

As previously noted, unit operations selected for use as pretreatment are typically based on the needs of the primary or nal treatment that follow.




Examples •

Membranes used in either RO or ultraltration can be fouled by suspended solids (particulates) and scale (precipitating solids) as water is removed. A typical goal for particulate control in pretreatment could be represented by a Silt Density Index (SDI) of less than 5 and hardness of less than 1ppm for precipitate analysis. Most membranes are sparingly tolerant of chlorides, but only certain membranes are able to withstand exposure to chlorine or chlorine compounds. For further information, see Chapter 5 of this Guide.

•

Distillation can be adversely affected by scale formation, due to hardness and corrosion from chlorides. This may require additional treatment beyond pretreatment, i.e., RO or ultraltration. Distillation has little tolerance for chlorine, due to corrosion and carryover into the product; however, it can have some tolerance for particulates. For further information, see Chapter 6 of this Guide.

Selecting pretreatment to reliably provide suitable feed water for nal treatment, regardless of anticipated spikes in feed water quality, will reduce operating and maintenance costs associated with nal treatment.

4.5

Control of Fouling: Removal of Turbidity and Particulates Particulates typically are insoluble suspended materials present in the water. Concentrations should be measured in milligrams per liter (mg/l). Sources of particulates include: •

dust

•

pollen

•

silica

•

insoluble minerals

•

corrosion products

Turbidity is a cloudy appearance caused by the presence of suspended and colloidal materials. Rather than a physical property, property, it is an optical property based on the amount of light reected by the suspended particles and is measured in Nephelometric Turbidity Units (NTU). For example, the US EPA limit for turbidity in drinking water is 1 NTU. Turbidity cannot be directly related to particulate counts since it is affected more by particle size, shape, and color rather than concentration. Light colored particles reect more light than dark colored particles and small particles may reect more light than larger particles of equivalent concentration. Removal of particulates and turbidity is required to prevent fouling/plugging of nal treatment processes, especially those using a membrane (e.g., RO). Common methods for removal of turbidity and particulates include: •

clarication and the accompanying operations of occulation, coagulation, and sedimentation

•

media ltration, including single and multi-media ltration

•

barrier ltration, including pre-coat lter, surface and depth media, such as cartridges and ner barriers, such as nanoltration or ultraltration

Factors affecting the removal of turbidity and particulates include: •

particle size and shape relative to the ltration media

•

coagulation or the tendency of particles to adhere to each other or the media, which may be enhanced by addition of a occulating agent or alum




•

surface effects, including surface tension, hydrogen bonding, and electrostatics

Addition of alum, lime, ferric chloride, or other occulating agents, as well as pH adjustment aids the sedimentation and clarication to remove particles larger than 25 µm. Flow rates, generally, are large and cost per unit volume is low. This process is typically not utilized for the production of PW as it would be redundant to treatment typically performed by a municipality; however, it may be employed if the feed water source is an unregulated source (e.g., private well). Clarication, typically, typically, is not 100% efcient, and may require additional ltration to prevent particulates from causing blockage in subsequent pretreatment operations. Media Filtration Depth ltration is an effective and common method of removing particulates from the water with some minor effect on reducing turbidity. Design can incorporate either a single size media or multi-sized media in a housing that has the means to support the media. With multi-sized media, the larger media is typically at the top with the ow directed downward through progressively ner layers of media. The overall porosity of the bed, based largely on packing, permits removal of particles in the range of 10 to 40 µm, relative to the media selected. Particulates accumulated during operation are removed by a back ush (back wash) operation based either on pressure drop or time. This back ush in the upward (reverse) direction also decompresses the lter bed and is followed by a down-ow rinse to resettle the media and remove nes. Wastewater from the back ush is generally not considered chemical waste and is typically 3 to 10x the design forward ow rate. Back ush durations vary from 15 to 30 minutes. Sand is the most common ltration media, based on cost and availability in a wide range of sizes and purities. Sand lters are, generally, not considered depth type lters that contain media with different densities. The ltration media used in a depth lter may include anthracite, carbon, or manganese. Anthracite Anthracite may be used when leaching of the silica from sand is a problem due to high temperatures or alkalinity. Depth lters using anthracite may allow higher ow rates and require less back washing (regeneration) because of the sharply angular particles; unlike silica particles that have a more rounded shape. A depth depth lter using carbon might be selected if the water has a high loading of organics or if there is a particular reason to combine removal or particulates, organics, and chlorine. A layer of an activated granular carbon, such as coconut, lignite, or anthracite may be added. A depth lter using media coated with potassium permanganate or manganese zeolite may be selected for water having high concentrations of iron or manganese. Generally, an oxidant, such as potassium permanganate or chlorine and permanganate, is added prior to the lter to convert metals to higher oxidation states that are insoluble, so that reduction of particulates with nominal size ranges as low as 0.03 mg/l of iron and 0.05 mg/l of manganese are possible. Microbial growth is a key consideration in a lter and particularly in a depth lter, because of the large surface area and relatively low velocity. velocity. In the case of a carbon lter, the media itself is a source of nutrient; therefore, appropriate consideration should be given to sanitation, such as the presence of a disinfectant (e.g., chlorine or chloramines) in the feed water, an added disinfectant, or the ability to sanitize routinely chemically or with heat. The lter bed also may be designed with constant recirculation to minimize stagnation and growth. Advantages: large capacity per unit cost, low cost of operation, and maintenance if properly applied, well suited to Advantages: large chlorinated water supplies, excellent for feed water pretreatment. Disadvantages: particle reduction limited to 10 µm (approximately); can become a source of microbiological Disadvantages: particle contamination if improperly maintained, designed, or operated. Barrier Filtration Barrier ltration includes: •

cartridge or membrane ltration

•

pre-coat ltration




•

ultraltration

•

nanoltration

This type of ltration relies on a barrier through which the water must ow. The barrier retains particulates that are removed allowing for their elimination by either replacing the barrier (cartridge and pre-coat) or via a purge stream (ultra and nanoltration). Barrier ltration normally is not used as the primary method of particulate removal. Relatively high particulate loading in typical feed water requires frequent barrier replacement, which results in higher operating costs associated with the purchase and disposal of the media, and corresponding labor. Barrier ltration is frequently used to polish the water after another pretreatment process, such as ion exchange or carbon ltration, and before processes such as RO. When used for polishing, the barrier typically will be rated nominally for 1 to 10 µm for simple removal of any particulate carry-over from the previous operation or to protect against upset. Alternatively Alternatively,, if the objective is to remove insoluble forms of silica and iron to achieve an SDI of less than 5 for suitable feed to an RO, a rating of < 1 µm is more appropriate, and available at the lower end of microltration or the upper end of ultraltration. An absolute rated cartridge lter or a pre-coat lter also may be suitable. The latter is not used usually, unless the high particulate loading offsets the associated labor cost. Ultraltration (0.1 to 0.001 µm) and nanoltration (0.005 to 0.0005 µm) should not be used for primary removal of particulates; associated associated cost and rapid blinding of the media should be considered. Advantages: lower Advantages: lower capital cost, particle removal (size) based on cartridge selection. Disadvantages: ongoing Disadvantages: ongoing cost of operation and maintenance, potential for microbial growth if not suitably maintained.

4.6

Control of Scaling: Removal of Hardness and Metals When water is separated from its impurities in the nal treatment process, those compounds with low solubility are concentrated to the point where they precipitate. This precipitation or scaling is the result of exceeding the solubility of the divalent and trivalent cations, usually as a sparingly soluble salt, such as carbonate or sulfate. Methods of control include: •

Ion exchange: primarily removal of calcium and magnesium, as well as divalent and trivalent cations, such as iron, aluminum, and silica (e.g., water softening ion exchange, which removes divalent and trivalent ions and replaces them with sodium, is a very common process used in pretreatment of pharmaceutical water).

•

Chemical injection: can range from simple: acidication to convert carbonate to carbon dioxide (which may be removed by degasication), to complex: using proprietary sequestrants.

•

Barrier ltration: such as nanoltration, where undesirable compounds are retained by the membrane and removed via a ush stream.

Water softening via ion-exchange is applicable for all ow rates and hardness levels, and is well understood and easy to operate. It involves only the handling of salt, and produces a non-hazardous waste stream; however, the high Total Dissolved Solids (TDS) in the waste stream may limit disposal options. Water softening also is easily controlled manually or automatically. Chemical injection is one method of controlling the ions or compounds that contribute to scaling. Degasication, typically after acidication, may be considered, including including between the stages of a two pass RO. The principal advantage of degasication is that carbon dioxide is released to the atmosphere prior to nal treatment. Issues relating to the handling of acid and base solutions, as well as the initial cost and maintenance of additional equipment should be evaluated against the benets.




Alternatively, injection of a sequestering compound (usually a proprietary organic compound) to the nal treatment Alternatively, feed water acts to bind and complex the offending ions or compounds to form a complex or compound that is more soluble and will not precipitate in the nal treatment process. The ion and sequestrant have a large molecular weight and are removed in the nal treatment by the purge stream. Sequestrants are commonly proprietary compounds that require testing to verify applicability, applicability, dosage level for a particular feed water, and analysis to verify removal in the nal treatment process. Nanoltration is a barrier membrane process that may be applicable with certain feed waters and specic situations. The ltration methodology is usually cross-ow and involves a signicant purge stream. It is much like RO, the differences being pore size in the membrane and the corresponding effect on ion removal. Removal of divalent ions can be greater than 98%.

4.6.1

Water Softening by Ion-Exchange Hardness in a water supply can result in scale formation, which is a deposit of minerals left over after the water has been removed or evaporated. This can be found in boilers, cooling towers, RO machines, clean steam generators, and distillation systems. The function of an ion exchange water softener is removal of scale forming calcium and magnesium ions from hard water. In many cases, other multivalent ions, such as soluble iron (ferrous) and ionized silica are removed with softeners. An ion exchange system consists of a tank and a cartridge or cylinder containing small beads of synthetic resin. The beads are treated to selectively attract either cations or anions and exchange these ions based upon their relative activity compared to the resin. This process of ion exchange will continue until all available exchange sites are lled, at which point the resin is exhausted and should be regenerated by the appropriate chemicals. For removal of hardness and metals, cation exchange will remove positively charged ions (metals) and exchange them for sodium ions. For further information, see Chapters 5 and 6 of this Guide. A standard water softener has four major components: •

resin tank

•

resin

•

brine tank

•

valve or controller

The softener resin tank contains the treated ion exchange resin, usually small beads of polystyrene. Capacity depends on volume of the resin bed. The resin beads initially adsorb sodium ions during brine regeneration. The resin has a greater afnity for the multi-valence ions, such as calcium and magnesium than it does for sodium. As a result, when hard water is passed through the resin bed, calcium, magnesium, and other multivalent ions, such as iron and silica, adhere to the resin, releasing the sodium ions until equilibrium is reached. The water softener has exchanged its sodium ions for the calcium, magnesium, and iron ions in the water. Regeneration is achieved by introducing an aqueous sodium chloride (NaCl) solution to the resin bed, exchanging the hardness ions for sodium ions. The resin’s afnity for the hardness ions is overcome by using a highly concentrated solution of NaCl (brine). The spent brine solution plus the associated water back-ushes and rinses are waste streams and might typically approximate the nominal throughput for one hour for each regeneration cycle. Advantages: well known and recognized technology, Advantages: well technology, low cost and effective. Works Works well in a chlorinated environment for microbial control with chlorine having only minor effect on resin life and efciency efciency.. Regeneration should not be relied upon for microbial control.




Disadvantages: salt handling for brine regeneration and disposal of spent brine solution. Resins may be Disadvantages: salt incompatible with ozone, certain sanitization agents, or elevated temperatures. A key decision in the process design of the pretreatment is location of the softener softener.. The primary options are either before or after removal of the microbial control agent (often chlorine) that is in the feed water or which may have been added for control of microbial growth. Softener located prior to removal of microbial control agent: the principal advantage is protection of the softener from microbial growth by the microbial control agent present in the feed water. If the microbial control agent is chlorine, it will have only a minor effect on resin life and efciency at the chlorine levels typically encountered in chlorinated municipal feed waters (< 1ppm). Softener located after removal of microbial control agent: the advantage is somewhat increased resin life and capacity (due to absence of chlorine, if used as a microbial control agent). However, this must be balanced by the need to protect the softener from microbial growth.

4.6.2

Demineralization/Deioni Demineralizati on/Deionization zation for Specic Contaminan Contaminantt Removal If present in high concentrations, certain impurities such as iron, silica, and aluminum present unusual removal problems. a.

Iron

Iron is a common water contaminant. It is one of the more difcult contaminants to remove because it may change valence states; that is, change from the water-soluble ferrous state to the insoluble ferric state. In solution, ferrous iron behaves like calcium and magnesium; however, when oxygen or an oxidizing agent is introduced, ferrous iron becomes ferric and precipitates, leading to a rusty (red brown) appearance in water. Certain bacteria can further complicate iron problems. Organisms, such as Crenothrix , Sphaerotilus, and Gallionella, use iron as an energy source, eventually forming a rusty, gelatinous sludge that can clog piping and equipment, particularly barrier processes; such as nanoltration and reverse osmosis. One removal method for iron in the oxidized state is a replaceable barrier ltration such as a cartridge lter with a rating of < 1 µm. b.

Silica

Like iron, silica may be present in more than one form and is a major problem in some parts of the world. It may be a soluble ionized species or an insoluble material, sometimes as a colloidal mixture with organics and other metals. The concentration of ionized silica will be reduced by a water softener and insoluble silica forms can be removed by a replaceable barrier ltration with a rating in the range of 0.5 µm. The insoluble silica also can be removed by nanoltration, reverse osmosis or strong base ion exchange. c.

Aluminum

Like iron and silica, aluminum can exist in multiple valences and its chemistry is complex. It also can be a component of colloidal complexes. Its solubility, particularly as hydrated oxide compounds, is a function of pH. Aluminum may be present in the water either naturally or as a result of an alum treatment used by a municipality for coagulation. Aluminum that is present as a colloidal component component can be removed by ne barrier ltration. ltration. Softening or deionization removes aluminum in an ionized form. Aluminum also can be removed by reverse osmosis if the pH is less than 6 or greater than 8. If aluminum is a potential concern, softening or deionization followed by pH adjustment and then RO may be required for removal.




4.6.3

Acidication Acidicatio n and Degasication a.

Carbon Dioxide

The use of a pH adjustment step is common in the process design to favor or inhibit the formation of CO 2. If high levels of CO 2 are present in the water, it can be removed down to a concentration of about 5 ppm to 10 ppm with an atmospheric degasier. degasier. An atmospheric degasier has the potential of increasing bacterial burden and should be located where bacterial control measures are available. One example is to locate the degasier between the stages of a two pass RO system. b.

The Acidication/De Acidication/Degasication gasication Process

The process is well known and accepted in water purication systems. It is usually used where there is a high ow rate (> 50 gpm or 0.18 m 3/min) or high hardness (> 50 ppm). The incoming water is acidied before the RO unit and a degasier is used to remove residual CO 2 prior to moving on to a second pass RO or a mixed bed Deionization (DI) unit. In this pretreatment process, the incoming water is adjusted to a pH in the range of 3.8 to 4.2 with sulfuric acid. The acidied water is sent to a packed column degasier for removal of free CO 2 by air. Removal efciency of CO 2 is better than 98% (typical commercial degasiers are designed to reduce outlet CO 2 to less than 5ppm). This residual CO2 should not pose a problem for downstream single and mixed bed deionization units or electro-deionization. The residual CO2 also can be removed by addition of a base to increase the pH to 8.0 to 8.5 that converts it to CO 3= which is removed in the second stage of the RO. Commercial degasiers range typically from 14 inches in diameter for 50 gpm to 72 in. diameter for 680 gpm (0.46 m for 0.18 m3/min to 1.83 m for 2.57 m 3/min). Fan power requirements will range from 1/2 HP to 10 HP for the preceding sizes. Smaller and larger units are possible to meet exact needs. Standard packed tower design methods are used. The pH after removal of CO2 will be in the range of 6.5 to 7.0. For RO feed waters, the pH should be adjusted to approximately 8.0 to 8.5 in order to minimize the amount of free CO 2 still remaining in the water and enhance removal of remaining carbonate in the second pass of the RO. The acidication/degas acidication/degasication ication process has some associated problems. Air borne bacteria, if a problem or concern, can be removed by a HEP HEPA A lter in the inlet-air line. The air also may oxidize any iron present to form solids. Water from the degasication column is usually collected in a holding tank. Further treatment in this tank is possible for TOC removal and microbial control. A multimedia multimedia lter usually follows the degasier for removal of initial incoming solids and any solids generated in the degasication step. Advantages: replaces softener and the handling of large amounts of salt for softener regeneration; the CO 2 is Advantages: replaces released to the atmosphere rather than being purged as an ion in a waste steam; the added sulfate ion from acidication is easier to remove in RO than added sodium ion from softening. Disadvantages: handling acid for acidication; instrumentation, Disadvantages: instrumentation, chemical handling for two pH adjustments, and air contamination.

4.6.4

Nanoltration Nanoltration is a pressure driven membrane process with performance characteristics between RO and ultraltration. The theoretical pore size of the membrane is one nanometer (10-9 meter). These membranes are sometimes referred to as “softening membranes” and will remove anions and cations. The removal of the larger anions (e.g., sulfate) is easier than the removal of a smaller anion (chloride) as discussed earlier.




The nanoltration membrane offers high rejection of salts of divalent anions as well as organics with molecular weights above 200. This includes color bodies, trihalomethanes (THM) precursors, and sulfates. The rejection is lower, but effective effective for salts with monovalent anions or non-ionized organics with a molecular weight above 150. Final product conductivity will range from 40 to 200 µS/cm depending on the inlet water total solids and mineral species make-up. A single pass RO unit will produce conductivity of 5 to 20 µS/cm. The investment cost and size of a nanoltration system is about the same as for a RO system. Energy use is lower because they operate at 70 PSIG to 150 PSIG (4.76 BARG to 10.2 BARG) as opposed to 150 PSIG to 350 PSIG (10.2 BARG to 23.8 BARG) for reverse osmosis membranes. Operating pressures are always a function of temperature, feed water salinity, and recovery. Nanoltration membranes, like other membranes, are to a large extent application dependent. Key factors are the quality of the feed water and the quality of the product water required. The feed water should be processed through a multi-media ltration system prior to going to the membranes. Potential applications are: •

removal of color

•

removal of THM precursors and organic carbon compounds from surface waters

•

removal of hardness, radium, and TDS from well water

•

removal of high silica

Industrial uses are where moderate water quality is required.

4.7

Organic Material and Removal Organic removal may be required for nal water quality obtainment or to minimize the potential of organic fouling of downstream components. Common methods for removal of organics are:

4.7.1

•

ozone

•

strong-base ion exchange

•

barrier ltration (microltration, ultraltration, or RO)

•

ultraviolet light

•

carbon

Introduction Organic and microbiological contaminants need to be addressed in water treatment systems. The concerns are twofold: contaminants entering the system and contaminants created/growing in the system. Organics usually enter with the feed water, but also may leach from some non-metallic materials of construction. Microbiological contaminants may enter with the feed water or grow in the system. The rst issue to consider is water source since it affects organic loading. If the water is drawn from a well, organic loading is usually not very great. Surface water (lake, river, or reservoir) will probably contain relatively high levels of organics and the composition and quantity may show seasonal variation.




Water from a municipal system is usually chlorinated, sometimes with ammonia added to form chloramines. Microbiological content of the feed water will be low and will generally be inhibited until the chlorine/chloramine is removed. The second issue to address is biological growth occurring within the water pretreatment system. Most pretreatment systems are designed to keep an oxidant in the water for as long as possible to minimize the potential for growth. Special design and maintenance requirements need to be addressed in all equipment that operates without a microbial control agent, chlorine, or chloramine present. These include materials of construction and piping layout (set up and ttings for sampling and periodic sanitization and instrumentation for monitoring) compatible with the sanitization method selected.

4.7.2

Organic Contaminants The organic contaminants found in many water sources include: Bacterial Contamination: this Contamination: this is usually expressed as total viable microbial counts per ml or as Colony Forming Units (CFU) per unit volume. CFUs are determined by counting the growth resulting from incubating samples. Each colony is assumed to form from one bacterium. Pyrogenic Contamination: Contamination: pyrogens are substances that can produce a fever in mammals. Pyrogens are often endotoxins, organic compounds (lipopolysaccharide) (lipopolysaccharide) that are shed by bacterial cells during growth or are the residue of dead cells. They are chemically and physically stable and are not necessarily destroyed by conditions that kill bacteria. Their molecular weight may vary, generally 12,000 to 320,000. Pyrogen levels are quantied in Endotoxin Units (EU) per milliliter milliliter.. Pyrogens are of great concern to the pharmaceutical industry, industry, since high concentrations may cause responses in humans ranging from fever to shock, or death. Total Organic Carbon (TOC): this (TOC): this is a measure of organic materials contaminating the water and is specied in ppm or ppb. TOC is a direct measure of the organic material that is oxidizable. TOC is a very ne measurement used in sophisticated water treatment systems where any organic contamination can adversely affect product quality. TOC is not a good measure of microbial contamination. Dissolved Organic Compounds: Compounds: these these occur both as the product of the decomposition of natural materials and as synthetic compounds, such as oils or pesticides. Naturally occurring organics include tannin, and humic and fulvic acids. They detract from the aesthetics of water (i.e., color), but unless they come in contact with certain halogens, they have no known health consequences in normal concentrations. Under conditions of free halogen compounds (principally chlorine and bromine), they form chlorinated hydrocarbons and THMs, which are suspected carcinogens.

4.7.3

Removal of Organics Technologies available to remove organic materials have different benets and drawbacks. The use of chlorine and chloramines to remove bacterial contamination are the most common (see Section 4.8 of this Guide). Treatment devices used to remove one or more of other types of organic material include: a.

Ozone

Ozone is an extremely powerful oxidant that prevents microbial growth, as well as reducing the concentration of organics. Ozone is not used frequently in pretreatment systems, due to the preference for chlorine and materials of construction that are readily degraded by ozone. For further information, see Chapter 8 of this Guide. b.

Strong-Base Ion Exchange

Organic scavengers or traps are ion exchange resins that contain strong-base anion resins and are regenerated with sodium chloride brine. Most naturally occurring organics have a slightly negative charge and are absorbed by the anion resin. After the resin is loaded, the organics can be displaced by high concentrations of chlorides during regeneration.




Advantages: removes Advantages: removes most natural organics; can be regenerated. Disadvantages: disposal of brine and organic solution; requires chemicals for regeneration; brine carryover may Disadvantages: disposal result after regeneration. c.

Carbon

Carbon is considered one of the most common method of reducing organics. It provides multiple functions; including removal of organics, as well as removal or reduction in the amount of chlorine and chloramines (if these are present and the carbon lter is appropriately designed). Activated carbon reduces organic material concentration by adsorption onto the carbon particles in the bed. Removal efciency depends on bed depth, cross-ow velocity, velocity, and adsorptive capacity of the carbon. Designs of carbon columns based on organic removal are generally based on bed depths of 4 feet (1.2 m) and hydraulic rates of 0.5 to 1.5 gpm/ft3 (67.5 to 202.5 l/min/m 3) of empty bed volume depending on the sanitant to be removed and the organic concentration in the feed water. Carbon bed volume is a balance between total adsorptive capacity and the frequency of replacement of the carbon bed. Reduction of feed water organic concentrations typically typically ranges from 30 to 60 percent. Additional Additional downstream process steps, e.g., RO, may be required to meet compendial limits for TOC or oxidizable substances. The capacity for organic removal is nite and will vary for each particular feed water source. Periodically the carbon should be replaced when its capacity to adsorb diminishes. For further information, see Section 4.9 of this Guide. Advantages: reduces organic concentration; removes color; removes chlorine effectively; technically not complex; Advantages: reduces relatively low cost. Disadvantages: high potential for increase in bioburden and biolm development; medium to high capital cost for Disadvantages: high thermally sanitized units; shedding of nes requires downstream ltration; periodic replacement of the spent carbon is required. d.

Microltration

Microltration includes the use of depth cartridge lters, pleated lters, and cross-ow ltration membrane elements. These lters can remove particles ranging in size from 100 µm down to 0.1 µm, thus capturing bacteria, cysts, and large molecular weight organics. Depth and pleated lters allow water to ow through a wall of bers perpendicular to the water direction. The particles are trapped on the outside wall of these lters or within the lter walls (for depth lters), because of the pore size of the lter. When the lter lls up with these particles, it should be replaced with a new lter. Cross ow microltration forces the water to ow parallel to the ltering media, and the particles which are too large to pass through the lter are expelled from the system in a concentrate stream to drain (typically 5% to 10% of the feed ow). This allows the lters to be self-cleaning and eliminates the need to replace these lters frequently. frequently. For further information, see Section 5.5 of this Guide. e.

Ultraltration

Ultraltration can be used to remove organics and bacteria, as well as viruses and pyrogens from a water source. Filtration is typically from 0.1 µm down to 0.001 µm. Cross ow ultraltration forces the water to ow parallel to the lter media, and particles which are too large to pass through the membrane elements are expelled from the system in a concentrate stream to drain (typically 5% to 10% of the feed ow). This allows the lters to be self-cleaning and eliminates the need to replace these membrane elements frequently. frequently. The UF membrane elements will need to have any suspended solids removed from the feed stream prior to the UF system. Advantages: effective Advantages: effective ltering barrier; no by-products; works with chlorine. Disadvantages: medium Disadvantages: medium to high capital cost; 10% constant concentrate stream; can be source of microbial growth.




f.

Reverse Osmosis (RO)

RO, if included in a pretreatment system to remove anions or cations, also will remove organics and microbiological impurities. Like ultraltration, a purge stream removes impurities that are too large to pass through the RO membrane. Advantages Advantages and disadvantages are similar to ultraltration. Tolerance Tolerance to chlorine depends on membrane selection. For further information, see Chapter 5 of this Guide. g.

Ultraviolet (UV) Light

Low pressure 185 nm UV lamps can be used for organic reduction medium pressure UV units are also commercially available. Successful Successful organic reduction is a function of the UV dosage, average irradiance, and contact time in a UV chamber.. Consideration of this process is based on the concentration of organics in the feed water and the purity of chamber the water (interfering compounds). In addition, the reduction will depend on the composition of the organics in the feed water stream, as some organic molecules may be more difcult to oxidize than others. The UV dose needed for of a particular water stream also may depend on: •

the type of residual chlorines present; free chlorine versus chloramines

•

the background of natural organic concentration in feed water source

•

turbidity, turbidi ty, color, and suspended solids

•

the ratio of target efuent chlorine concentration to inuent chlorine concentration

Advantages: no harmful chemicals are added to the water stream; low maintenance, and can be hot water sanitized Advantages: no or ozonated. Disadvantages: systems can be large in size and take up oor space based on inlet ow rates. The effectiveness Disadvantages: systems of the process depends on many variables of the feed water quality quality.. The initial capital cost can be higher than other methods.

4.8

System Design for Control of Microbial Growth Common methods used in pretreatment to control microbial growth include: •

microbial control agents, such as chlorine or chloramines

•

periodic sanitization (heat or chemical)

•

UV light

•

avoiding dead legs and avoiding water stagnation

A common strategy in the design of the pretreatment system is to leave the microbial control agent provided by the municipality in the water through as many pretreatment steps as possible, in order to protect these steps from microbial growth. The microbial control agent (chlorine or chloramine) must be removed at a specied point in the process, since it, typically, is not compatible with the nal treatment processes. At this point, the only option is periodic sanitization, either with heat or a chemical disinfectant. This must be included in the design of the pretreatment system, along with the provisions for validation and monitoring its effectiveness via sampling and testing. If a chemical disinfectant is used, provisions to remove it and monitor its removal are also required. For further information, see Chapter 13 of this Guide.



4.8.1


Periodic Sanitization Periodic sanitization methods, employed on a scheduled or as needed basis, include: •

heat

•

chemical sanitization

•

regeneration or replacement of media

•

ushing

•

drainage

With heat, USP indicator organisms are killed above 60°C and the majority of pathogenic organisms will not proliferate. Temperatures Temperatures above 80°C result in complete kill of all non-resistant bacteria. Sanitization times might be one to two hours at the specied temperature. Total Total cycle time, including heat-up and cool down, may be 4 hours to 8 hours. Heat is commonly used in carbon beds, lters, and distribution systems. For further information, see Chapter 13 of this Guide. Chemical sanitization agents (when chlorine cannot be used) include hydrogen peroxide, iodine, ammonium compounds, and organic or inorganic per-oxygen compounds. Sanitization times may be 0.5 hours to 4 hours with additional time for set up to feed the sanitization agent and to ush it from the system. Total Total cycle time may be eight hours. Controlling temperature to minimize microbial growth allows an increase in the period between sanitizations. Temperatures below 15°C slow microbial growth, but may be more expensive to operate than systems held at ambient temperature. Avoiding Avoiding stagnation and dead legs also minimizes microbial growth. Recycle loops around various unit operations can be used during shutdown periods, i.e., recycle around depth lter and softener while sanitizing the carbon bed or while cleaning and sanitizing RO systems. Times for specialized periodic sanitization methods, such as regeneration, replacement of media, and drainage will depend on the equipment piece and specic design. Sanitization methods (frequency and length of sanitization) are system and sanitizing agent dependent, and should be veried.

4.8.2

Ultraviolett (UV) Light Ultraviole Treatment with UV light at 254 nm wavelength is a popular form of microbial control and disinfecting, based on the ease of use. Water is exposed, at a controlled rate, between UV light tubes. The UV light deactivates DNA in the microorganisms, preventing preventing duplication and leading to a reduction in bacteria. In pretreatment systems, UV normally is used when chlorine/chloramine and heat are not available or feasible. The feed water to a UV system needs to be free of suspended solids, which can “shadow” bacteria, preventing adequate UV contact. UV is typically used in controlling feed water to an RO unit that cannot accept chlorine or heat, and in controlling non-chlorinated water re-circulation during system idle time. The UV system does not leave a residual in the treated water and, therefore, is effective only if there is direct UV light contact with microorganisms. UV lights should be continually monitored to ensure the design intensity is within the specied range to ensure adequate microbial control. Note: UV lights may not adequately perform when solely utilized as the primary means of microbial or organic contaminant control, but perform well as a part of a total microbial control plan.




4.8.3

Chlorine Municipalities frequently use chlorine, often introduced as sodium hypochlorite to disinfect water before and during distribution. Chlorine is fed into the system to kill bacteria at typical dosage levels of 0.2 ppm to 2.0 ppm. In Europe and elsewhere in the world, residual disinfectant levels may be much lower or altogether absent. In the US, the chlorine level at outlying distribution points is often targeted at about 0.2 ppm to 0.5 ppm. If the water supply is heavily contaminated with organics; however, however, the chlorine may react and form chlorinated hydrocarbons (THMs). In other cases, chlorine can dissipate and no residual level is maintained at outlying points in a municipal distribution system. Chlorine concentration should be monitored in the feed water and in parts of the pretreatment system prior to its removal. Re-chlorination in a water treatment system should consider adequate contact time to be effective. This may require the use of feed water break tanks for certain systems. Molecular chlorine can have adverse effects on the components in a water purication system. It will cause oxidative deterioration of the membranes, particularly polyamides, used in ultraltration and RO. It also will cause degradation, embrittlement, and loss of capacity in deionization resins (oxidation rate varies with resin type) although the amount is low to moderate at chlorine concentrations usually found in drinking water. It also will cause corrosion of stainless steel, particularly at elevated temperatures and may carry over into the product in a distillation system. system. Therefore, in most systems making PW, the chlorine is removed during pretreatment. Advantages: low capital cost; common treatment; complements municipal water treatment; maintains a residual; Advantages: low easy to test and maintain levels. Disadvantages: can create THMs; does not affect all organics; residual chlorine may cause degradation to many Disadvantages: can nal treatment systems. The two principal methods of chlorine removal are activated carbon and reduction, often with sulte.

4.9

Removal of Microbial Control Agents During pretreatment, microbial control agents should be removed because of their detrimental effect on nal treatment equipment and performance. Chlorine causes deterioration of most RO membranes and is corrosive in distillation. Chloramines can pass through pretreatment and decompose in the distillation process with an adverse effect on water quality. For chlorine removal, activated carbon is a simple process for the absorption of chlorine. The carbon will reduce some of the chlorine to chloride ion, which is then removed in the nal treatment ion removal process. Sulte reduction is also simple with sulte being oxidized to sulfate and chlorine being reduced to chloride ion. Chloramine removal can be complex. Chloramine adsorption on carbon occurs at a much slower rate than chlorine, requiring longer contact times, and lower hydraulic ow rates. The potential for dissociation of the absorbed chloramines into ammonium ion and ammonia is problematic. Ammonium is removed by RO, but decomposes to ammonia in a distillation process. Ammonia passes through both RO and distillation processes during nal treatment. Sulte reduction for chloramines results in ammonium and chloride ions. These can be removed by RO. The ammonium ion partially decomposes to ammonia in the higher temperature distillation process, resulting in carryover and effects on the water quality. Removal of ammonia (from chloramine) and carbon dioxide requires appropriate pH control to maintain these species as ions for removal in an RO. The equilibrium of carbonate, bicarbonate, and carbon dioxide is pH dependent with alkaline conditions required to maintain the ionic species. The equilibrium between ammonium and ammonia is pH and temperature dependent with acidic conditions required to maintain the ionic species. At no single pH point are these species all carbonate and ammonium ions. Thus two pH adjustment steps followed by the appropriate removal technologies are required to remove both chloramines and carbon dioxide.



4.9.1


Activated Carbon for Removal of Microbial Control Agents Activated carbon removes the chlorine by adsorbing it onto the carbon particles in a carbon bed. There is also some reduction of chlorine to chloride. Removal efciency depends on bed depth, face velocity, velocity, and adsorptive capacity of the carbon. Design is based on the rate of adsorption with adsorption rates typically being more rapid for chlorine than organics if this is done in the same operation. Designs based on chlorine removal will occur with bed depths of as little as 2 to 4 feet (0.61 to 1.2 m) and hydraulic rates of 2 to 4 gpm/ft 3 (270 to 540 l/min/m 3) of empty bed volume. Carbon bed volume is a balance between total adsorptive capacity and the frequency of replacement of the carbon bed. Use of carbon to remove chlorine provides the perfect conditions for microbiological growth: low ow rates in a warm media with lots of nutrient present; therefore, a program to periodically sanitize the carbon bed is required. Heat (either pure steam, process steam, or chemical-free steam or hot water at 65°C or above) is effective with sanitization frequency varying from daily to a couple of times a week or less. With a proper sanitization program, microbial growth in carbon beds can be controlled. Although it may be difcult to constantly control activated carbon efuent streams to less than 500 cfu/ml (drinking water specication) under all circumstanc circumstances, es, this is generally considered a target value. Sanitization frequency varies based on this target. Following the sanitization, the carbon bed is usually rinsed to remove nes before being returned to service. Frequent backwashing of the carbon beds may assist microbial control within these units. Advantages: reduces organic concentration; removes color; removes chlorine effectively effectively;; technically not complex; relatively low cost. Disadvantages: high potential for increase in bioburden; medium to high capital cost for thermally sanitized units; shedding of nes requires downstream ltration; periodic replacement of the spent carbon.

4.9.2

Reduction of Microbial Control Agents The addition of a reducing agent will reduce the chlorine to chloride. Sulte, usually as sodium bisulte, is generally the reducing agent of choice. The chemistry is: SO3- + Cl2 + H2O ----> 2Cl- + 2H+ + SO4= The addition of sulte also may require an accompanying pH adjustment step. The chloride and sulfate that are formed may be removed by a subsequent deionization step or RO. Advantages: effective removal of chlorine; lower capital cost than carbon lters that can be heat sanitized; no regeneration or replacement required; low operating cost. Disadvantages: technically more complex, chemical handling including sodium bisulte and acid/base for pH adjustment; potential for microbial growth in sulte feed tank requires frequent (< 5 days) preparation of sulte solution; higher capital cost for feed systems and monitors; higher cost than disposable carbon. Additional Additional concern regarding build-up of sulte in recirculation loops.

4.9.3

Removal by UV Irradiation Ultraviolet light also can be used in the reduction of chlorine. In this process (UV-photo (UV-photolysis), lysis), where 100% photodecomposition photodecompositio n of free chlorine is achieved, UV light photolysis decomposes free chlorine to form approximately 80% chloride ions and 20% chlorate ions. The typical UV dose for photodecomposition of free chlorine is about 20 times higher than the standard disinfection UV dose. Generally low pressure 185 nm UV lamps are incorporated as the radiation wavelength. Successful chlorine removal is a function of the UV dosage of average irradiance and contact time in a UV chamber chamber.. Consideration of this process is based on the ppm level of chlorine concentration in the feed water. The UV Transmittance Transmittance (UVT) in a typical municipal feed water source is about 85% to 95%. The UV dose required for dechlorination of a particular water stream also depends on the following factors.




•

the type of residual chlorines present, free chlorine versus chloramines

•

the background of natural organic concentration in feed water source

•

turbidity, turbidi ty, color, and suspended solids

•

the ratio of target efuent chlorine concentration to inuent chlorine concentration

Advantages: no harmful chemicals are added to the water stream, low maintenance, and can be hot water sanitized Advantages: no or ozonated. Additionally, Additionally, it provides signicant TOC reduction. Disadvantages: systems can be large in size and take up oor space based on inlet ow rates. The initial capital Disadvantages: systems cost can be higher than the other methods.

4.9.4

Chloramines Chloramin es and Chloramine Removal Chloramines are formed by the reaction of chlorine and ammonia. Municipalities add ammonia to form a longer acting disinfectant than chlorine and to reduce the formation of THMs during the chlorination of municipal water. Chloramines consist of three compounds: monochloramine (NH2Cl), dichloramine (NHCl2), and trichloramine (NCl3). Dichloramine is a particularly strong biocide. Chloramines present problems since their removal typically is not a single step. The methods for chloramine removal are: •

activated carbon

•

reduction (sulte injection)

•

UV irradiation

4.9.4.1 Activated Carbon

Chloramines, like chlorine, can be removed by carbon; however, the absorption is much slower than for chlorine or organics. Chloramine adsorption will require hydraulic rates as much as 3 to 6 times less than chlorine and empty bed contact times of 3 to 6 times those required for chlorine. The removal of chloramines by activated carbon results in dissociation of some of the chloramines to ammonium ion and ammonia. The ratio is dependent on pH and temperature. The ammonium ion can be removed by cation exchange (water softening). If chloramines are present in the feed water, it may be desirable, therefore, to locate the carbon bed for removal of microbial control agent prior to the water softening operation in the pretreatment system design. The advantages and disadvantages of carbon are similar to those for chlorine. The potential dissociation of chloramines to form ammonia is a disadvantage and can cause problems in nal treatment. 4.9.4.2 Reduction

Reduction with sulte will convert chloramines to ammonium ion and chloride ion. These are removed by an ion exchange operation or the ion removal process in nal treatment. Again, if chloramine is present, it may be desirable to locate the microbial agent removal prior to the water softening operation in the pretreatment design. The advantages and disadvantages of sulte reduction are similar to those for chlorine.




4.9.4.3 UV Irradiation

Chrloramines, like chlorine also can be removed by UV irradiation. All the similar properties and reactions of chlorine apply to the reaction and removal of chloramines. UV photolysis reduces chloramines to chlorides and nitrates, which are easily removed by RO. A major design consideration for UV dechloramination is the sizing of the UV unit regarding the number of UV lamps required. The typical UV dose for photodecomposition of chloramines can range from 20 to 30 times higher than the standard disinfection UV dose. Feed water parameters, such as turbidity, dissolved solids, and color, should be considered for proper sizing. Assistance from UV suppliers or pilot testing of UV irradiation units for both chlorine and chloramines according to a supplied water analysis should be considered. The advantages and disadvantages of UV irradiation are similar to those listed for chlorine.

4.10

Changes in Anion Composition Composition/Concentrati /Concentration on Pretreatment systems typically remove non-ionic impurities and cations. Thus, any change in anionic composition or concentration is usually secondary. Distillation Distillation processes in nal treatment may be affected by chlorides, which can be removed by an RO prior to the nal treatment step. The pretreatment processes that affect anionic composition are: •

deionization

•

degasication

•

carbon bed ltration for removal of chlorine and chloramines

•

reduction to remove chlorine and chloramines

•

barrier ltration (nanoltration, ultraltration, and RO)

Ion exchange resins are designed to remove either cations or anions. An ion exchange resin that is designed to remove anions (anionic resin) typically will exchange the anions: •

chloride

•

sulfate

•

nitrate

•

carbonate

•

bicarbonate (if the pH is appropriate for the hydroxyl ion)

The ion exchange may be in a single bed, mixed beds, or twin beds and will affect anionic composition if an anionic resin is present. For further information, see Chapter 5 of this Guide. Carbon bed ltration adsorbs chlorine and chloramines from feed water. A quantity of the chlorine is reduced to chloride and is removed in a subsequent ion removal process, usually in nal treatment. The removal of chlorine and chloramines by reduction often with bisulte, changes ionic composition and concentration, as the bisulte is oxidized to sulfate and the chlorine or chloramines are reduced to chloride and ammonium.




Some barrier ltrations (particularly nanoltration) remove some of the larger anions. Reverse osmosis may be used to remove chloride ion prior to some distillation processes.

4.10.1 pH and Carbon Dioxide pH, the negative log of the hydrogen ion concentration, is a measure of the concentration of hydrogen ions (H+) in a water-based solution. The more hydrogen ions that are present, the lower the pH and the more acidic the solution. The concentration of H+ ions (pH) is especially important because it affects the chemistry of the water. For instance, the pH of the water, along with other parameters, signies whether the water will corrode piping or if certain contaminants (carbonates) are likely to precipitate and cause scaling. In water or aqueous solutions, a certain ratio of water molecules, H2O, separates (or dissociates) into ions, H + and OH-. H2O <----> H+ + OHBecause of the properties of water, when the concentrations of hydrogen (H+) and hydroxyl (OH-) present in any water-based solution are multiplied together, the value is always the same. This number is the “equilibrium ion product,” Kw, which has been determined to have the value shown below: Kw =[H+] × [OH -] = 1.01 × 10-14 at 25°C Where

[H+] = Concentration of H + (moles/liter) [OH-] = Concentration of OH- (moles/liter)

Free carbon dioxide in water is produced by the decay of organic matter, dissolution of carbon dioxide from underground sources, and solution from the atmosphere. Since the carbon dioxide content of the atmosphere is quite low (less than 0.04%), this is not a major source of carbon dioxide in the water and surface waters normally are relatively low in free carbon dioxide; however, well waters usually contain an appreciable quantity of free carbon dioxide. Free carbon dioxide is the term used to designate carbon dioxide gas dissolved in water. The designation free carbon dioxide differentiates a solution of carbon dioxide gas from combined carbon dioxide present in the form of bicarbonate and carbonate ions. In the case of high purity water, low levels of carbon dioxide from the atmosphere can cause the pH to drop from 7.0 to 5.5 and the conductivity to increase from 0.1 µS/cm µS/cm to 1 µS/cm. Low levels of CO2 also can prevent a water purication system such as a two-pass RO from producing water with a conductivity of ≤ 1 µS/cm. The pH of the water causes the equilibrium between free carbon dioxide (gas) and bicarbonate alkalinity (dissolved ion) to shift, more or less, to carbon dioxide. As the pH is lowered, the equilibrium is shifted toward carbon dioxide, which is a neutral species dissolved in the water with the ionic charge being maintained with anions from the added acid and the net formation of water. As the pH is increased, the equilibrium is shifted toward bicarbonate and then carbonate with the ionic charge being maintained with cations from the added base and the net formation of water. The determination of the level of CO 2 present in the water as it proceeds through the treatment process is important to understand because it can: •

affect the nal water quality

•

cause premature exhaustion of ion exchange systems



4.11


The Importance of pH in Pretreatment Typical drinking water standards require that the pH of the water be within a range of approximately 6.5 to 8.5; Typical however, the range of most water is much narrower due to the corrosive nature of water with an acidic pH and a scaling potential at a high pH. The feed water pH is very important when designing the pretreatment for a PW system. In addition, the pH is an important parameter in designing a RO system or an ion exchange system. The lower the pH from a pH of 8.0 to 8.5, the higher the potential capacity for dissolved CO 2. The CO2 will directly pass through pretreatment and an RO membrane and depress the conductivity and pH, making it difcult to meet the compendial conductivity requirements. requirements. If the system has an ion exchange system following the RO, high levels of CO 2 will produce a high ionic loading on the system. High CO 2 may require the use of a degasier to remove the CO 2. For further information, see Chapter 11 of this Guide. If the feed water pH is between 6 and 10, the RO system has the potential to incur hardness scaling. Adding acid to the feed water controls the deposition of scale, but this converts carbonates to CO 2 that will pass through both the RO and distillation nal treatment processes. Conversely, Conversely, the addition of base converts the bicarbonate to carbonate, and carbon dioxide (CO2) to bicarbonate. These ions will be removed by an RO unit, but also will cause scaling. Most pharmaceutical organizations incorporate the use of ion exchange softening in order to prevent scaling from occurring in the RO membrane. Chloramine in the feed water can result in ammonia or ammonium. The pH, as well as temperature, affects the equilibrium between ammonia and ammonium. Acidic conditions are required to maintain the ionic species for removal in an RO unit. In nal treatment, ammonia will pass through an RO unit. In distillation, the higher temperatures shift the equilibrium from ammonium toward ammonia. Ammonia will affect conductivity and pH, making it difcult to meet the USP conductivity requirements. Water systems may include the capability to add either acid or a base to the water in order to optimize the performance of the system. The most common acid used for pH adjustment is sulfuric acid, because it is readily available and less corrosive than hydrochloric acid. The most common base used for pH adjustment is sodium hydroxide (caustic soda) because it is readily available, and the nal treatment process will remove sodium ions.

4.12

Materials of Constructio Construction n and Constructi Construction on Practices Piping to the pretreatment system may be copper, galvanized steel, or a suitable thermoplastic. Piping in the pretreatment system, where high temperatures are not encountered, is usually plastic (PVC, CPVC, polypropylene, or other material) based upon cost and corrosion resistance. Leaching from some plastics such as PVC and CPVC may make these materials undesirable. Vessels Vessels may be Fiberglass Reinforced Plastic (FRP), lined carbon steel, or stainless steel. The piping and equipment in a portion of the pretreatment system may encounter high temperature (periodic heat sanitization) or high pressure (RO plus degasication). In these portions, piping typically is stainless steel or a plastic that can be heat sanitized, such as PVDF. PVDF. Equipment designed for high pressure may be carbon steel, lined carbon steel, stainless steel, or a high-strength plastic. Mill nish is satisfactory for these materials; electropolishing is unnecessary. The cost of sanitary construction practices, such as orbital welding and sanitary ttings, may not be warranted in the pretreatment system. Use of plastic pipe that is solvent cemented or heat fused, stainless steel pipe that is welded or anged with mill nish, or tubing with compression ttings is common. Ball or diaphragm valves predominate for ow diversion with globe and needle valves for ow control. Selecting the minimum cost piping components that will not degrade water quality is considered an area for major cost savings. Sample points should be provided upstream and downstream of each piece of equipment for monitoring and for troubleshooting. Points for eld measurement of pressure and temperature are also useful for troubleshooting.




4.13

Water Conservation Many of the unit operations used for pretreatment systems employ packed bed columns. These are common for media, activated carbon, and water softening processes. These processes require periodic backwash or regeneration maintenance steps that use high instantaneous ow rates. The frequency of these maintenance steps is a function of the incoming water quality. For certain feed waters, the frequency, and subsequently the water consumption, can be excessive. Traditional Traditional water softening requires a brine regeneration step and rinse step in addition to a backwash. Water consumption may be minimized by optimizing the regeneration or backwash frequency of these units. This may include performing these maintenance steps based on system performance and operation, rather than a preset time interval. Unnecessary or excessive backwash episodes can be particularly wasteful for media lters. The volume of water discarded for some systems can be as much as twice the normal feed water ow rate. Additional novel or innovative technologies also may be available to minimize water consumption and recycle. These may include backwash steps modied with air scouring techniques, high-efciency softening systems, alternate hardness removal techniques, or organic reduction techniques other than traditional packed bed activated carbon adsorption.

4.14

Summary The design of pretreatment components can have a major impact both on investment and on continuous operation and overall life cycle cost. Reliable operation and control of pretreatment can signicantly reduce operating and maintenance costs associated with the nal treatment technology. technology. Important process steps in pretreatment include: •

removal of turbidity and particulates to minimize membrane and equipment fouling

•

removal of hardness and metals to prevent scale formation in nal treatment

•

removal of organics and microbiological impurities

•

control of microbial growth and removal of microbial control agents to prevent degradation of nal treatment

These process steps are important because of their potential effect on water quality from nal treatment or their longterm effect on nal treatment equipment performance and possibly, their effect on water quality from nal treatment.


5

Final Treatment Options: Non-Compendial Waters, Compendial Purified Water, and Compendial Highly Purified Water




Page 53 Final Treatment Options: Non-Compendial Waters, Compendial Puried Water Water,, and Compendial Highly Puried Water

5

Final Treatment Options: Non-Compendial Waters, Compendial Purified Water, and Compendial Highly Purified Water

5.1

Introduction This chapter discusses the nal treatment technologies and basic system congurations related to the generation process of compendial PW (e.g., USP, EP, and JP), EP highly puried water and non-compendial waters. Various basic system congurations are presented with prevailing RO, ion exchange, and distillation nal treatment. System congurations are intentionally exible as many varied congurations can produce water meeting specications.. Equipment and system materials, surface nish, and other design factors are discussed for appropriate specications selection of components, piping, instrumentation, and controls. PW, highly puried water, and non-compendial water can be produced by an extensive combination of unit processes in various congurations. The most common pretreatment and nal treatment technologies used in these systems are shown in Figure 5.3. Consideration is given to the nal treatment unit processes currently utilized, including: •

ion exchange

•

reverse osmosis

•

continuous electrodeionization

•

membrane degasication

•

ultraltration

•

microltration

•

ultraviolet light

These technologies, as well as distillation are used for the production of compendial PW, Highly Puried Water Water,, and non-compendial waters. For further information, see Chapter 6 of this Guide. The most common PW system designs implement a reverse osmosis membrane based process with nal polishing by continuous electrodeionization, ion exchange, or a second reverse osmosis stage as the nal primary treatment process. Membrane based system usage has increased due to chemical consumption reduction, contaminant rejection (ionized solids, organics, colloids, microbes, endotoxins, and suspended solids), reduced maintenance, consistent operation, and effective life cycle cost. The various membrane based system congurations are compared with ion exchange and distillation in Table 5.3. Water and energy conservation has become an important design factor in a signicant percentage of systems. Equipment construction is discussed for each unit process section to promote proper selection of materials, surface nishes, and other design factors. The total system capital cost is inuenced more by equipment design details than by process selection. Many aspects of equipment can be “over-designed” and hence, become unnecessarily costly.. Appropriate consideration should be given to an individual component’s function, location, required microbial costly performance, sanitization, and other factors to optimize design. In most cases, it is not necessary to construct every




makeup system component with the same level of surface nish and detail as the distribution system for successful operation. Care should be taken so that all components of the system are still capable of producing the high quality outputs required. Many material selections are made erroneously to conform to perceived requirements that do not actually dictate the details of construction for most nal treatment components. Material selection should be made to conform to actual requirements. A design review should be performed to optimize the system for consistent operation to specications and life cycle cost optimization. The requirement to replace system components (e.g., lters, RO membranes) at a frequency such that the appropriate water quality and system output are maintained should be considered. This chapter does not differentiate between compendial and non-compendial water system equipment. Noncompendial water often is manufactured and validated in a manner consistent with compendial water. Noncompendial water for pharmaceutical manufacturing is simply water that is dened by factors other than the compendial requirements of various pharmacopeial groups. The water quality parameters may be less stringent or more stringent than compendial waters. Non-compendial waters may include several compendial water quality attributes, as well as additional attributes. The lowest quality non-compendial water for pharmaceutical use is generally water meeting drinking water standards, such as US EPA National National Primary Drinking Water Standards (Reference 9, Appendix 1) or equivalent. Non-compendial waters can exceed the quality standards of PW or Water for Injection where product or process requires extremely pure water. The systems described are most commonly implemented in a makeup mode to a storage tank based upon a level control signal. Systems may not employ a storage tank and feed distribution loops directly. directly. Systems may recirculate a quantity of water back from distribution to the makeup system when no makeup demand exists. This design helps to maintain minimum conductivity in storage and distribution and can be particularly useful in systems where ozonation or other factors may increase conductivity in storage.

5.1.1

Highly Puried Water

5.1.1.1 Description Most pharmacopeias dene PW and WFI. European Pharmacopoeia 7 and EMEA Note for Guidance on Quality of Water for Pharmaceutical Use (2002) (References 5 and 11, Appendix 1) also dene Highly Puried Water (HPW). HPW is not required to be produced by distillation differentiating differentiating HPW from WFI as dened in the European Pharmacopoeia 7 by EP and EMEA References 5 and 11, Appendix 1). In the European Pharmacopoeia 7, highly puried water is dened as “Water intended for use in the preparation of medicinal products where water of high biological quality is needed, except where water for injections is required.” EMEA Note for Guidance on Quality of Water for Pharmaceutical Use (2002) Reference 11, Appendix Appendix 1 offers additional insight into the use of HPW and possible production methods. 5.1.1.2 Application The denition of HPW makes it an alternative to WFI in those processes where the quality of WFI is desirable, but not required. The quality attributes of HPW are the same as the quality attributes of EP WFI, but distillation is not a required nal treatment process. Typical Typical production methods include one more unit process compared to production of PW, such as double-pass reverse osmosis plus ultraltration or deionization, or reverse osmosis followed by continuous electrodeionization followed by ultraltration. 5.1.1.3 Pretreatment Requirements The pretreatment for HPW is the same as for PW. HPW is produced in much the same way as PW, often with one extra purication pass to assure removal of endotoxins.




5.1.1.4 Cost Savings Factors HPW should be compared to WFI and may be less expensive to produce. 5.1.1.5 Advantages and Disadvantages

Advantages: •

generally less expensive than WFI

•

generally chemical free method

•

generally low operating costs

Disadvantages: •

cannot replace WFI in all applications

5.2

Ion Exchange

5.2.1

Description Ion exchange is a process utilizing organic polymer based resins for removal of ionized contaminants. Water passes through porous ion exchange resin beads and ionized contaminants (cations and anions from the salts in the feed water and charged organic compounds) are exchanged for hydrogen and hydroxyl ions. Cation and anion exchange resins are manufactured from organic polymers which can be made to function with a xed positive or negative chemical charge. The xed charge site has a mobile counter ion attached to be in electrical equilibrium. Cation resins have negative xed charge and remove cations (positively charged) from water. Anion resins have xed positive charges and remove anions (negatively charged) from water. As water passes through the ion exchange resins, the exchange of ions in the water stream for the hydrogen and hydroxide ions, held by the resin, occurs readily and is driven by chemical equivalent weight. Higher equivalent weight ions, such as sodium, calcium, magnesium, chloride, sulfate, bicarbonate, etc., readily displace the hydrogen and hydroxyl ions from the exchange sites. The product stream has signicantly lower conductivity as the conductive elements have been almost entirely removed. Typical ion exchange systems are easily capable of producing water meeting Stage 1 PW conductivity requirements. The ion exchange resin reaches an exhausted state when most of ion exchange sites have exchanged H+ and OH- for ionized water contaminants. The resins must be regenerated when ionic leakage in the efuent produces unacceptably high water conductivity. conductivity. The ion exchange sites prefer to remain in the exhausted state with a higher chemically charged ion than H+ or OH-. Thus, the regeneration process is driven by excess chemical concentrations concentrations of a regenerant acid or base solution. Ion exchange systems are available in tank or cartridge congurations using virgin resin or resins regenerated on site or off site. Ion exchange systems are available in two basic physical congurations commonly referred to as two bed (or separate bed) demineralizers or mixed bed demineralizers. A two-bed two-bed ion exchange system includes both cation and anion resin tanks. Two-bed ion exchange systems often times function as the workhorse of a strictly Deionization (DI) water system in terms of salt removal. Two bed demineralizers utilize a simple regeneration process, but typically produce a higher conductivity efuent than a mixed bed demineralizer. Among two bed systems, there are two types: co-current and counter-current regeneration units. In co-current regeneration systems, the regeneration uid ows in the same direction of the process water stream. In a counter-




current system, these uids ow in opposite directions. The practical results of counter-current regeneration are higher quality product water and possible noticeable reduction in chemical usage. Mixed-bed demineralizers consist of a single tank with a mixture of cation and anion removal resin. The resins are thoroughly mixed in the service cycle and must be separated into two distinct layers for regeneration. The regeneration cycle is more complex and hence possibly less consistent, but extremely low conductivity water can be produced. 5.2.1.1 Component Description A typical ion exchange system is comprised of: •

one or more tanks

•

ion exchange resin

•

piping and valve system

•

water/chemical distributors internal to the tanks

•

regeneration system

•

conductivity or resistivity meter and cell

Other instrumentation, typically typically,, may include a ow meter and pressure gauges. Ion exchange systems are available in both on-site regenerable and off-site regenerable (rechargeable) versions. In both versions, tanks may be constructed from berglass, stainless stainless steel, or carbon steel with an inert interior lining, such as vulcanized rubber, Polyvinylidene Fluoride (PVDF), Polypropylene (PP), Polyethylene (PE), or Polyvinyl Chloride (PVC). In selecting the material, care should be taken to select a material which will not leach high amounts of unwanted ions. Offsite regenerated or rechargeable systems are typically transported off site to a facility that is equipped to either regenerate or replace the resin. For this reason, these units are typically supplied with berglass or light gauge stainless steel tanks in sizes typically ranging from less than 1 ft 3 (0.03 m3) to 50 ft3 (1.4 m3) per tank. Larger, off-site regenerated systems are recharged with new or regenerated resins on-site, and the exhausted resin is returned to an off-site regeneration facility. facility. New resin provides greater capacity and some possible quality control advantages, but at a higher cost. Regenerated resin produces a lower operating cost, but may raise quality control issues, such as resin segregation, regeneration quality, and consistency. On-site regenerated units are designed with more complicated valve and piping systems to accommodate on-site chemical regeneration and rinsing. These systems should be selected when larger volumes of water are required on a continuous basis, thus justifying the higher capital investment by a lower operating cost. On-site regeneration requires chemical handling and disposal, but allows for internal process control and microbial control of the resin bed and sections of piping in contact with the regeneration chemicals. 5.2.1.2 Tanks For pharmaceutical applications, given the typical quantity of water utilized, ion exchange tanks rarely need to be more than 3 to 4 feet (0.91 to 1.22 m) in diameter diameter.. Tank Tank shell straight sides, typically, typically, are 3 to 8 feet (0.91 to 2.44 m). Steel tanks are welded and typically manufactured and designed in accordance with the ASME Code for operating pressures between 100 PSIG and 150 PSIG (7 and 10.5 kg/cm 2 gauge). ASME Code stamping or equivalent is not necessarily required for this type of equipment; however, local regulations and end user safety concerns should govern this decision. (References 12, Appendix 1).




5.2.1.3 Distributors Each ion exchange tank includes distributors at all pipe to tank interfaces. Distributors are required to ensure that resin does not escape from the tank while water is owing through the system and to provide adequate distribution of ow through the vessel. Distributors typically typically are supplied in stainless steel, PVC, Chlorinated Polyvinyl Chloride (CPVC), Polypropylene (PP), or PVDF. PVDF. Structural integrity of a distributor system is a key element in any design since a ruptured distributor can cause a signicant loss of resin and may require signicant time for repair. 5.2.1.4 Piping and Valves The selection of a piping and valve system depends upon several factors including budget, product water quality (in terms of chemical analysis), and preferred methods of sterilization. Most ion exchange systems are provided with PVC or CPVC piping and valves. The advantages of these materials include low cost, ease of assembly assembly,, and high corrosion resistance. Specialty plastics, plastics, such as polypropylene and PVDF also have been utilized in DI systems to a great extent. These materials are more expensive than either PVC or CPVC; however, these materials are superior in terms of the lower level of organic leachables into the process water. Furthermore, these materials are available in a piping design that more closely resembles the orbital welding in sanitary stainless piping systems. Stainless steel systems may offer greater structural integrity than plastic piping systems and may require less support and smaller expansion loops than thermoplastic piping in hot water or steam sanitization applications. Stainless steel is more vulnerable to corrosion than thermoplastic piping. Thermoplastic Thermoplastic piping in PVDF and PP can be hot water sanitized and PVDF also with steam with proper attention to piping support and thermal expansion. Selection generally is based on: •

total cost of ownership considering installed cost

•

ongoing maintenance requirements

•

sanitization method

•

corrosion resistance

•

service life requirements

5.2.1.5 Regeneration System On-site regenerated ion exchange systems require a regeneration system that includes chemical pumps and/ or eductors, chemical tanks, piping and valves, and related instrumentation. The cost critical component in a regeneration system is the chemical pump. Pumps better suited for this application require inert materials of construction and the capability to closely meter or regulate the chemical dosage. Positive displacement pumps driven by either electric motors or compressed air are ideally suited for this application. Chemical eductors offer another option to deliver chemicals to an ion exchange system; however, chemical dosing may be inconsistent based on variable dilution water pressure and ow.

5.2.2

Application The major purpose of ion exchange equipment in water systems is to satisfy the conductivity requirements of the water quality specication. Deionization (DI) systems may be used in isolation or in conjunction with reverse osmosis to produce compendial PW or various types of non-compendial waters. Ion exchange systems can effectively reduce organics in many applications with proper ion exchange resin selection and maintenance. Ion exchange systems may not meet compendial TOC requirements without additional membrane processes in certain applications where high feed water TOC levels exist.




Many small volume systems for both compendial and non-compendial waters and for both production and laboratory water use off site regenerated ion exchange units as the primary treatment. When using off site regeneration, the segregation of resin may be desired. For further information, see Chapter 4 and Chapter 9 of this Guide. The systems may include ultraviolet light units or microbially retentive lters to meet specied microbial limits. Capital cost and chemical handling are minimized, but operating costs can vary signicantly signicantly.. Maintenance requirements typically are low and outside services normally are used to replace the resin tanks and other consumable items. These systems can produce extremely low conductivity water. water. Larger volume systems may use on site regenerated demineralizers to minimize operating cost, but these are uncommon in newer installations. These systems can be low in capital cost, but can require signicant regenerant chemical usage. Membrane based systems have replaced a signicant percentage of these systems as many organizations minimize chemical handling and discharge.

5.2.3

Pretreatment Requirement Requirements s Ion exchange systems require pretreatment to remove undissolved solids from the water stream and to avoid resin fouling or degradation. Although dechlorination also is recommended to avoid resin degradation by oxidation, the low levels of chlorine normally found in potable water supplies usually demonstrate only long-term effects on most ion exchange resins.

5.2.4

Cost Savings Factors Most of the cost savings opportunities for these systems revolve around the correct choices in materials of construction, pretreatment options, instrumentation, and sizing of the DI system. Acceptable Acceptable piping materials of construction can vary from PVC to 316L Stainless Steel. A correctly designed system will minimize the equipment size, considering microbial control, and maintenance. Choosing to monitor only the critical parameters such as conductivity (resistivity), (resistivity), ow, pressure, etc., can minimize instrumentation. Cost savings choices should be made with respect to capital purchase and on-going operating costs with options including DI off-site regenerable bottles, on-site regenerable DI vessels (with automatic or manual controls), or another water treatment unit operation.

5.2.5

Microbiologicall Concerns, Cleaning, and Sanitizati Microbiologica Sanitization on Although ion exchange resins beds, due to the hydrogen ion and hydroxide ion exchange sites, have pH values at the extreme ends of the range, microbiological activity activity remains a concern. The regeneration of both the cation and anion exchange resin beds effectively sanitizes the system; however, as the system processes water, the resin becomes exhausted and the pH approaches neutral. Organic matter, which may be deposited on or be absorbed by the resins, particularly an anion resin, and the laminar ow of water through the bed foster bacteria growth in ion exchange beds. For this reason, regeneration frequency is more important to ion exchange systems that are not designed with auxiliary microbiological control components, such as UV lights. Polishing ion exchange systems typically are positioned in a system with bacterial control elements, such as sub-micron lters and ultraviolet sterilizers, and may operate for several weeks without requiring r egeneration. Ion exchange resins can be sanitized chemically with a variety of agents. The degree of resin attrition is a function of resin type and the chemical agent. Chemical cleaners include peracetic acid and sodium hypochlorite. Some resins are capable of hot water sanitizations at temperatures between 65°C to 85°C. Ion exchange resins suitable for limited thermal sanitizations include: strong acid cation resin, and standard polystyrene cross-linked with divinylbenzene Type 1 strong base resin.



5.2.6


Advantages and Disadvantages Major advantages are the degree of exibility in ow rate of ion exchange systems, lack of sophisticated maintenance requirements, consistent production production of Stage 1 conductivity conductivity,, and the ability to use the chemical regeneration of ion exchange resins as a means of microbial control. The major disadvantages include the requirement to store and handle acid and caustic, the requirements to neutralize waste chemicals (for on-site regenerated systems only), and the reduced ability of ion exchange resins to remove dissolved organics relative to membrane based systems.

Advantages: •

simple design and maintenance

•

exible in water ow production

•

good upset recovery

•

low capital cost for single train DI systems

•

removes ionizable substances (ammonia, carbon dioxide, and some organics)

Disadvantages: •

high cost of operations on high Total Dissolved Solids (TDS) in-feed-water

•

requires chemical handling for on-site regenerable DI (safety and environmental issues)

•

full on-site DI system can take signicantly more oor space due to primary vessels, chemical storage, and neutralization system

•

off-site DI systems will require outside service and signicant costs for regeneration services

•

off-site regeneration involves consequent loss of control over the use, handling, and care of DI vessels

•

DI vessels are excellent places for microbial growth to occur between regenerations




Table 5.1: Comparison for Ion Exchange Unit Operations Off-Site Regenerated

On-Site Regenerated

Chemical Use:

N/A

Extensive

Sanitization Method:

Change Out or Hot Water

Regeneration

Capital Cost:

Minimal

Extensive

Operation Cost:

High to Medium

Low

Water Consumption:

None

Medium

Energy Consumption:

Minimal

Minimal

Maintenance Requirements:

Minimal

Medium

Outside Service Used:

Extensive

Low

Reliability:

Good*

Good

Upset Recovery Operations:

Good, Replacement

Good

Material Control*:

Low to Medium

High

*Note: Note: having having the DI bottles regenerated by an outside service does not relieve the manufacturer of the responsibility to have quality control of their ion exchange system.

Table 5.2: Limits of Operation and Expected Performance Feed Quality: Total Suspended Solids (turbidity):

Filtration of 10 micron is recommended.

Chlorine Tolerance:

Varies with type of resin, generally at 0.5 ppm, Varies Some resins are rated up to 1 ppm.

Total Dissolved Solids (TDS):

< 200 ppm, operation at higher TDS levels is possible, but operating costs can be high.

Temperature:

Most cation resin up to 121°C; most anion resin 40°C to 70°C; some anion resin up to 100°C.

Conductivity:

Can achieve conductivity below 1.0 microsiemen/ cm depending on the system pretreatment and regeneration schedule.

Regeneration and Chemical Efciency:

Linear variation is inverse to the feed water total dissolved solids – best below 200 ppm.

Feed TOC

Ability to avoid avoid fouling varies with type of resin. resin.

Product TOC:

May increase or decrease incoming TOC levels depending on resin type and feed water – difcult to predict.




5.3

Reverse Osmosis

5.3.1

Description Reverse Osmosis (RO) is a pressure driven process utilizing a semi-permeable membrane capable of removing dissolved organic and inorganic contaminants from water. A semi-permeable semi-permeable membrane is permeable to some substance, such as water, while being impermeable to other substances, such as many salts, acids, bases, colloids, bacteria, and endotoxins. RO membranes are produced commercially in a spiral wound conguration for pharmaceutical water production. Membranes are commonly available in two materials; cellulose acetate and thin lm composite (polyamide). Cellulose acetate is the oldest commercially produced membrane. Cellulose acetate has the advantages of typically being the lowest initial cost membrane and is chlorine tolerant. This type of membrane also is more resistant to some types of fouling compared to other membranes. The primary disadvantages of cellulose acetate membranes are the fastest loss of rejection among membrane types and the potential requirement to operate at a low pH range to reduce hydrolysis. Cellulose acetate membranes are relatively intensive in energy considerations since the membranes normally operate at a high pressure (300 PSIG to 500 PSIG or 21 BARG to 35 BARG) and commonly operate at elevated feed water temperatures to prevent excessively high pressure operation. Thin lm composite RO membranes offer the highest rejection of contaminants of all membrane types in production (at time of going to press). Thin lm composite membranes operate effectively at low water temperatures (40°F (4°C) and higher) and low pressures (100 PSIG (7.0 BARG) and higher). The initial capital cost of thin lm composite membranes may be higher compared to the cost of cellulose acetate membranes, but thin lm composite membranes have a longer life expectancy. expectancy. The primary negative aspect of thin lm composite membranes is a low tolerance for chlorine in the feed water. Thin lm composite membranes are degraded rapidly in the presence of chlorine at municipal drinking water levels. The dechlorination of the feed water does allow the opportunity for some bacterial growth to occur and more sanitization efforts are required with thin lm composite membranes than with chlorine tolerant types. Both thin lm composite and cellulose membrane elements can be appropriate for use in pharmaceutical water systems. The optimum membrane selection selection is based upon an analysis of capital cost, operating cost, membrane life, rejection, and bacterial control. The majority of pharmaceutical systems utilize thin lm composite membranes. Some spiral wound RO membranes incorporate a brine seal at the leading end of the RO element. This seal is designed to expand between the membrane and the pressure vessel that contains the RO membrane. The purpose of the seal is to prevent the passage of water between the membrane and the pressure vessel and to divert the ow of water across the RO membrane surface. The brine seal can cause bacterial problems, as stagnant water conditions are created by the presence of the seal and chlorine has been removed upstream of the RO elements using thin lm composite membrane material. Other spiral wound membranes are manufactured without a brine seal on the leading end. These membranes are commonly referred to as loose wrap or full t, and are congured in several different ways with the same goal of allowing modest controlled ow between the RO membrane and the pressure vessel. This type of membrane is, generally,, superior for use in pharmaceutical RO units, as bacterial contamination is minimized. Some elements may generally incorporate a front seal with some allowable bypass ow to minimize microbial growth in a similar fashion.

5.3.2

Application RO units can be successful successfully ly implemented in pharmaceutical systems in several ways. RO may be used without post-treatment to produce non-compendial water for still feed, rinse water, or other applications where some quality parameters may be less stringent than compendial waters. RO units also can produce compendial waters meeting required quality parameters when properly implemented on appropriate feed waters. RO systems have been used in non-compendial and compendial applications meeting or exceeding the CQAs for WFI.




RO units can be used upstream of deionizers to reduce regenerant acid and caustic consumption or to minimize resin replacement costs. Two-pass Two-pass RO units (two RO units arranged in series, product staged) with proper pH control can be capable of producing water that meets the regulatory requirements for TOC and conductivity. RO systems can vary in conguration. They are designed in arrays such that turbulence is reasonably maintained to minimize scale precipitation and fouling. Greater turbulence also decreases the boundary layer and reduces the salt level at the membrane surface improving permeate quality. As permeate (product) is formed, the feed water stream concentrates in contaminants. Typical Typical arrays frequently reduce the number of membrane pressure vessels in parallel, as the feed water ow is reduced. Recycling of the reject concentrate may be employed to maintain the higher turbulence without using more feed water.

5.3.3

Limitations RO quantitatively reduces bacteria, endotoxins, colloids, and high molecular weight organics from water; however, RO cannot remove 100% of contaminants from water and has either very low or no capacity for the removal of several extremely low molecular weight dissolved organics. A concentrate ow is essential to remove the contaminants that are rejected by the membrane. The waste stream from an RO unit may be used for cooling tower make-up water or compressor cooling water, etc. Recovery is dened as the percentage of feed water that becomes puried product water. Recovery can range from 20% (or less) up to 90%, depending on factors such as: •

feed water quality provided to the RO unit

•

system capacity/con capacity/conguration guration

•

life cycle costs requirements

•

expected water quality

•

wastewater limitations

•

maintenance factors

A recovery unit with a high percentage output may have less waste to achieve the desired output rate, but will tend to have higher maintenance costs due to the fouling/scaling effects effects of the concentrate. Water quality produced by an RO system is dependent upon a number of factors, including, but not limited to: •

membrane type

•

operating pressure

•

feed water quality

As a result, it is conceivable that the feed water quality could change sufciently so that the product quality from the RO system may no longer meet water quality requirements. Carbon dioxide, ammonia, and other dissolved gasses pass directly through an RO membrane and will be in the RO product stream at the same level as that in the feed water stream. Carbon dioxide or ammonia in the RO product stream may increase the product conductivity beyond the PW Stage 1 conductivity limit. Carbon dioxide and ammonia also contribute to the loading of ion exchange resin, which may be present downstream of the RO unit. To alleviate these potential issues a degasication unit operation could be added before or after the RO process.




Another option would be to adjust the pH of the feed water water prior to the RO to minimize dissolved dissolved gas levels. levels. Carbon dioxide and ammonia are minimized at different pH values, so it can be difcult to minimize both using this method. Carbon dioxide is minimized by raising the pH to 8.3, converting the carbon dioxide to bicarbonate, which is more readily rejected by the RO membrane. Ammonia Ammonia can be minimized by lowering the pH near 7.0 converting ammonia to ammonium allowing rejection by the RO membrane. Most RO systems operate between 5°C and 28°C in production mode. Most membrane elements are limited to high pressure production operation at 45°C as damage can occur at higher temperatures. The capability for RO systems to continuously operate at high temperature (65°C to 80°C) should alleviate this; however, to date, this design is rather limited in use. RO operation at temperatures below 65°C can allow some microbial growth, and therefore, most ambient operation RO units are periodically sanitized with chemicals or hot water at 80°C.

5.3.4

Pretreatment Requirement Requirements s RO membranes must be protected from scale formation, membrane fouling, and membrane degradation.

5.3.4.1 Scaling Minimization Scaling is possible since the contaminants present in the feed water stream are being concentrated into the waste stream, which is an average of 25% of the feed stream. Scale control is normally prevented by the use of water softening upstream of the membranes, the injection of acids to lower the pH of the feed water stream or an antiscalant compound to prevent precipitation. Scale control through acidication of feed stream to lower the pH is considered effective. The principal negative aspect of this method of scale control is in the formation of free carbon dioxide from bicarbonate that is present in the feed water as the pH is lowered. Anti-scalant chemicals also are available for injection into the RO feed water stream. The anti-scalant agents are considered effective in minimizing scale formation through a sequestering action that increases the time necessary for crystal formation of the precipitate. 5.3.4.2 Fouling Minimization RO membrane fouling is reduced through the use of back-washable multi-media lters, cartridge lters, or ultraltration for suspended solids, greens and ltration or softening for colloidal iron removal, and various microbial control pretreatment methods to reduce biological fouling. 5.3.4.3 Oxidation Minimization The main causes of membrane degradation are oxidation of certain membrane materials and heat degradation. Thinlm composite membrane systems, which have an extremely low tolerance for chlorine, incorporate activated carbon or injection of various sodium sulte compounds for dechlorination. Protection against high temperature is normally incorporated where the feed water is preheated and the membrane material cannot tolerate high temperature. For further information on RO pretreatment unit operations, see Chapter 4 of this Guide.

5.3.5

Performance A single stage of RO elements typically reduces the level of raw water salts, colloids, organics, bacteria, and endotoxin by 90% to 99%. Single stage RO product water does not normally meet the conductivity requirements of most compendial PWs without further purication steps. Some two-pass units (two sets of RO membranes in series) produce water that can pass the Stage 1 conductivity requirements. Membrane selection should be based upon: •

pretreatment requirements



•

operating performance characteristics

•

sanitization options

•

warranties

•

capital and operating costs

•

feed water source


RO units should incorporate sufcient membrane area for reliable operation. Membrane manufacturers offer recommendations for membrane area or ux (gallons/day/square foot of membrane) required as a function of the feed water quality and operating temperature and pressure. One of the most important factors for optimization of membrane area is the understanding of fouling and scaling potential. The Silt Density Index (SDI) reading of fers fers an indication of the tendency of the feed water to foul membranes as a result of lterable contaminants. Lower SDI values tend to reduce membrane cleaning frequency. Many membrane manufacturers and system suppliers recommend a feed water SDI reading of three or less for reliable membrane operation. In general, the higher the membrane area for a xed product water ow rate, the lower the rate of membrane fouling. This may not be the case in every instance, as differences in the percent recovery may be a factor. A high high percentage recovery for an equivalent ow and membrane area will tend to foul more. An increase in membrane area usually will cause an increase in the capital cost of the equipment, due to the increased requirement for membranes and pressure vessels. Optimization Optimization of RO systems generally involves selection selection of membrane area to control capital cost without resulting excessive membrane cleaning and replacement. RO systems typically are designed at a specic baseline operating temperature and operating pressure to predict the ux, or permeate ow rate per membrane area. Any deviations from the specied baseline conditions could result in changes in the amount or quality of permeate water produced. For example, an increase in temperature above the design temperature could lead to increased permeate ow, but potentially decreased product water quality (assuming constant baseline feed pressure). RO operating cost optimization involves the analysis of pumping costs and water heating costs. Raising the feed water temperature reduces the required pump pressure to produce the desired product ow. The cost of energy to raise the feed water temperature may be much higher than the cost of pump energy required to produce the desired product ow at a lower feed water temperature. A lower feed water temperature generally increases salt rejection and lowers microbial fouling rates.

5.3.6

Advantages and Disadvantages

Advantages:

•

RO units eliminate or signicantly reduce chemical handling and disposal, relative to regenerable ion exchange systems.

•

Generally,, RO has more effective microbial control than ion exchange systems. Generally

•

Integrity testing can be accomplished by salt challenge and measurement of differential conductivity conductivity..

•

RO removes a wide variety of contaminants, including ionized solids and non-ionic materials (e.g., colloids, bacteria, endotoxin, and some dissolved organics).

Disadvantages: •

Water consumption can be signicantly higher than ion exchange systems, unless the wastewater is reused.



5.3.7


•

Energy consumption is generally higher than ion exchange and less than distillation.

•

No removal of dissolved gases (e.g., carbon dioxide and ammonia).

Control and Instrumentat Instrumentation ion Proper instrumentation for monitoring and controlling critical and non-critical parameters of operation is essential to any RO system design. Small changes in feed pH can have a very dramatic effect on nal RO system conductivity conductivity.. This parameter should be monitored and controlled (if pH control is a part of the system) utilizing an accurate pH meter with a feedback loop for any chemical feed pump. Using a simple on/off signal for chemical feed pump control is not recommended. The product water quality from the RO system is directly related to the feed water quality quality.. As a result, monitoring the feed water quality can provide a method for notifying an operator of an impending issue before the product water quality from the RO system degrades. The feed water and permeate conductivity can be directly compared to provide on-line monitoring of RO ionic rejection. Product conductivity and temperature may be used for water quality compliance measurement when the RO unit is the nal conductivity reduction process. Permeate ow, waste ow, and feed temperature typically are monitored, as well as pressure for feed, concentrate, and permeate using properly calibrated instruments. RO units also may monitor additional parameters, such as membrane inter-stage pressure and conductivity conductivity.. Scale will tend to form on downstream membranes and pressure drop can be an indicator. Chlorine, Oxidation Reduction Potential (ORP), or sulte level monitoring may be used to protect chlorine intolerant membranes against oxidation. RO units may incorporate some level of automation. Protective devices typically are included to protect: •

pumps against low suction pressure

•

membranes against high pressure and high temperature

Valves on RO units normally are manually adjusted. Automatic valves are used to accomplish product side ushing and system sanitization. Membrane cleaning is performed manually in most systems, but also can be automated.

5.3.8

Reverse Osmosis Concentrate Reuse RO concentrate is frequently used as cooling tower make-up (with or without some level of purication), or for noncontact cooling for compressors, or other heat loads. Concentrate may be further treated for reintroduction as system feed water. Concentrate from the second pass of a two pass RO normally is returned to the feed water stream of the rst pass RO.

5.3.9

Cleaning and Sanitization RO units usually need periodic cleaning. Acid based cleaners are used to remove accumulated metals and salts from the membranes. Alkaline Alkaline detergent cleaners remove silt and organic foulants from the membranes. Cleaning need is based upon a loss of rejection, an increase in the feed to waste membrane pressure drop or a loss of product ow. RO normalization of product ow corrected for the operating temperature and trans-membrane pressure is a useful tool to determine the requirement to clean the RO membrane elements. RO membranes can be sanitized with chemical agents dependent on membrane selection. Specially constructed membranes are available for hot water sanitization at 60° C to 85°C. Hot water sanitization may be implemented for critical microbial control.



5.4

Continuous Electrodeioniz Electrodeionization ation (CEDI)

5.4.1

Description


Continuous Electrodeionizat Electrodeionization ion (CEDI) is a technology combining ion exchange resins, ion selective membranes, and the use of an electrical eld to continuously remove ionized species and regenerate the resins. The CEDI processes are distinguished from other electrodeionization (such as capacitive deionization) and ion exchange processes in that the processes are continuous rather than batch or intermittent, and that the ionic transport properties of the ion exchange media are a primary sizing parameter, parameter, as opposed to ionic capacity. capacity. CEDI units typically have a number of successiv successively ely functioning ion depleting (purifying) and ion concentrating cells which can be fed from the same water or different water sources. Water is puried in CEDI devices through ion transfer. Ionized species are drawn from the water passing along the ion depleting cells and transferred into the concentrate water stream passing across the ion exchange membranes. The ion exchange membranes are permeable to ionized species, but not permeable to water. The ion purifying cells typically have continuously regenerated ion exchange media between a pair of ion exchange membranes. •

Units may incorporate mixed (cationic and anionic) ion exchange media between a cationic membrane and an anionic membrane to form the purifying cell.

•

Units may incorporate layers of cation and anion ion exchange media between ion exchange membranes to form the purifying cell.

•

Single purifying cells (cationic or anionic) may be created by incorporating a single ion exchange media between ion exchange membranes.

CEDI units can be designed with the cells either in a plate-and-frame or in a spiral wound conguration. The power supply creates a Direct Current (DC) electric eld between the cathode and anode of the CEDI device. It contributes to the ionic transport for: •

continuously removing ionized species

•

continuously regenerating the ion exchange resin

Continuously Removing Ionized Species Cations in the feed water stream passing across the purifying cell are directed toward the cathode. Cations, stopped by the anion permeable membrane, are transported through the cation exchange media and the cation permeable membrane to the concentrating compartment. Anions, blocked by the cation permeable membrane, are directed toward the anode. They are transferred across anion exchange media and the anion permeable membrane to the concentrating compartment. The cations and anions removed from the purifying cells and gathered in the concentrating cells are eliminated to the waste through the water ow.




Continuously Regenerating the Ion Exchange Resin As the ionic strength of the PW stream decreases the high voltage gradient at the water-ion exchange media interfaces can cause water to split into its ionic constituents (H + and OH-). The H+ and OH- ions are created continuously and regenerate the cation and anion exchange resins, respectively, respectively, at the outlet end of the purifying cells. The continuous high ion exchange resin regeneration level allows the consistent production of high purity water (0.055 to 1 µS/cm referred to 25°C) in the CEDI process. A typical CEDI process drawing is shown in Figure 5.1. Figure 5.1: Typical CEDI Process Drawing

5.4.2

Application CEDI is commonly found downstream of pretreated water from RO units in compendial PW or non-compendial water installations. This type of combination allows the consistent production of water with low conductivity and organic levels, while optimizing the CEDI unit life span. Puried Water systems. systems.

5.4.3

Limitations CEDI technology aims to remove the ionized species, but is not designed to have an effect on the uncharged contaminants from water. A concentrate stream is required to remove the contaminants from the system, and so a part of the water is constantly rejected. CEDI has temperature limitations in production operations. Most CEDI units are operated between 5°C and 40°C (41°F to 104°F).

5.4.4

Pre-Treatment Pre-Treatm ent Requirements CEDI units should be protected from scale formation, fouling, and thermal or oxidative degradation. The RO/ pretreatment equipment typically reduces hardness, organics, suspended solids, and oxidants to acceptable levels.



5.4.5


Performance CEDI unit performance is a function of feed water quality and unit design. Ionized solids reduction is generally greater than 99% allowing production of lower than 1 µS/cm (reference value of 25°C) conductivity water from RO feed water. RO/CEDI systems may produce water with 0.1 µS/cm or lower conductivity.

5.4.6

Cost Savings Factors Pharmaceutical applications normally do not require any post-treatment after electrodeionization. Systems may incorporate ultraviolet light or sub-micron ltration either to reduce sanitization requirements or to provide microbial levels considerably below those expected for PW production.

5.4.7


Advantages:

5.4.8

•

typically designed to produce low conductivity water (in compliance with compendial PW)

•

elimination of chemical handling and associated costs

•

elimination of outside service needs (off-site regenerated resin, management costs, etc.)

•

removal of ionized substances, including the weakly ionized substances (e.g., carbon dioxide, ammonia, and some ionized organics)

•

limited environmental effect technology (low energy required, low waste generated, no chemicals required for regeneration, etc.)

•

electric eld in membrane/resin module provides some bacterial control

Disadvantages: •

ionized contamination specic purication technology

•

uniqueness of designs for each manufacturer (modules often are not interchangeable)

•

UV or sub-micron ltration may be required for further bacterial reduction

•

RO generally required as a pre-treatment

•

rinse up after chemical sanitization may take hours to reach low conductivity and TOC

Sanitization CEDI units typically are chemically sanitized with a number of agents including: •

peracetic acid

•

sodium percarbonate

•

sodium hydroxide

•

hydrogen peroxide




Hot water sanitization normally is used with a number of specic CEDI modules. Modules may tolerate heat sanitization as well as, or better than, chemical sanitization, while others may experience longer life with chemical sanitization.

5.5

Polishing and Removal of Specic Contaminatio Contaminations ns

5.5.1

Ultraltration

5.5.1.1 Description Ultraltration (UF) is a cross-ow process similar to RO. A pressurized feed stream ows parallel to a porous membrane ltration surface (unless dead ended operation is selected). A pressure pressure differential forces water through the membrane. The membrane rejects particulates, organics, microbes, pyrogens, and other contaminants that are too large to pass through the membrane, while allowing most ionic contaminants to pass through. UF systems may run dead-ended. Membranes are available in both polymeric and ceramic materials. Polymeric membrane elements are available in spiral wound and hollow ber congurations. Ceramic modules are available in single channel and multiple channel congurations. 5.5.1.2 Application UF is used both for pretreatment and in several ways in PW systems. UF normally is used downstream of ion exchange processes for organic, colloidal, microbial, and endotoxin reduction. PW with low endotoxin levels (< 0.25 Eu/ml) may be used by manufacturers in ophthalmic solutions, topicals, topicals, and bulk pharmaceutical chemicals that will be utilized in parenteral manufacturing and other applications. In Europe such water can, if all other limits are fullled, be dened as Highly Puried Water (HPW). UF normally is used in still feed water systems, in combination with ion exchange, to limit the endotoxin and colloidal silica feed levels to the still. 5.5.1.3 Limitations UF cannot remove 100% of contaminants from water. No ionic rejection occurs and organic rejection varies with the various membrane materials, conguration, and porosity. Different nominal organic molecular weight rejection ratings are available. Dissolved gases are not rejected by UF UF.. Ultralters normally require a waste stream to remove the contaminants on a continuous basis. The waste stream varies, but usually is from 2% to 10%. Ultralters, when in a polishing position, may be used in a dead-ended conguration like single use cartridges without a waste stream. 5.5.1.4 Performance UF is used to remove a variety of contaminants. The appropriate UF membrane should be selected to meet the performance requirements. Organic molecules can be rejected signicantly, signicantly, but the rating of UF membranes varies in molecular weight cutoffs from 1,000 to 100,000. Reduction of typical raw water organics is not as effective as RO. Pressure drops vary with membrane selection and operating temperature. UF membranes may be capable of continuous operation at temperatures up to 90°C to provide good microbial control. UF reduction of endotoxin (pyrogens) varies from 2 log10 to 4 log 10 as a function of membrane selection. UF has been shown to be capable of consistent production of water meeting the compendial WFI endotoxin limit of 0.25 Eu/ml in typical system applications. applications. UF produces good microbial reduction with typical ratings of 3 log 10 to 4 log10 reduction. UF produces good particle reduction.




5.5.1.5 Advantages and Disadvantages

Advantages: •

UF can remove a number of contaminants, such as endotoxin and organics, better than microltration.

•

UF can have lower operating costs than microltration, e.g., in high particle loading applications.

•

UF elements may tolerate more rigorous sanitization procedures using steam or ozone, than other membrane lters (e.g., MF or RO).

•

The waste stream generally is much less volume than waste from RO units.

•

UF is generally less energy intensive than RO.

Disadvantages: •

UF cannot remove ionic contaminants, (RO can).

•

UF can require a waste stream, which can be a signicant cost factor factor..

•

UF membranes may be more difcult to integrity test than microltration cartridges.

5.5.1.6 Cost Savings Factors Capital costs can be inuenced by the optimum sizing of membrane area and membrane selection. Piping material and nish signicantly impact capital cost. Systems may incorporate a variety of plastic piping materials while others use sanitary 316L Stainless Steel. The sanitation method and the frequency of sanitation usually plays a major role in material selection and may inuence instrument costs in the monitoring parameters of operation and sanitation. 5.5.1.7 Sanitization UF membranes can be sanitized in several different ways. Polymeric membranes usually are tolerant of a wide variety of chemical sanitizing agents, such as: •

sodium hypochlorite

•

hydrogen peroxide

•

peracetic acid

•

sodium hydroxide

Polymeric membranes may be capable of withstanding hot water sanitization or steam sanitization. Ceramic UF elements usually can tolerate common chemical sanitizing agents, hot water, steam, and ozone during sanitization or sterilization procedures. 5.5.1.8 Waste Water Recovery Pharmaceutical UF units normally are fed deionized water for PW or highly puried water production or special noncompendial water applications; therefore, the wastewater is still low conductivity water that can be recycled upstream to RO units or fed directly to boilers, cooling towers, or other uses.



5.5.2


Microltration

5.5.2.1 Description Microltration is a membrane process used for the removal of ne particles and microorganisms. Generally Generally,, a waste stream is not employed in microltration processes. Microltration cartridges normally are disposable and are available in a wide range of materials and pore sizes. In nal ltration, the lters usually range from 0.45 µm down to 0.04 µm. Microlters can be used in a wide range of applications, including aseptic lling of pharmaceutical products, which are not tolerant of terminal sterilization. Microlters normally are employed in pharmaceutical water systems for microbial retention downstream of components where some microbial growth may exist. Microlters can be extremely effective in this area, but operating procedures should be in place to assure lter integrity during installation and membrane replacement to ensure proper performance. The lters should be a part of a comprehensive microbial control plan, not a solitary microbial control unit operation in the system. Minimizing Minimizing the number of locations of microltration makes appropriate maintenance easier. easier. For further information, see Chapter 13 of this Guide. 5.5.2.2 Advantages and Disadvantages

Advantages: •


•

exible in water ow production

•

no waste stream

•

cartridges are integrity testable

•

heat and chemical sanitization of microlters is possible

Disadvantages: •

can be used only as part of a total microbial control plan

•

no ion or endotoxin removal

•

shorter life, due to dead head design, so replacement is required

5.5.2.3 Performance Microltration can be as effective as ultraltration in microbial reduction and can minimize water consumption, as no waste stream is necessary; however, microltration cannot reduce dissolved organic levels or remove particles as small as those removed by ultralters, due to the difference in pore size. Heat and chemical sanitization of microlters is possible with the appropriate selection of material.

5.5.3

Ultraviolett Light Treatment Ultraviole

5.5.3.1 Description Ultraviolet (UV) light treatment aims to decrease the bacteria and organic contamination levels in water. UV wavelengths and intensity level required depend on the expected function. As the intensity penetrating the contaminants drives the reduction process efciency, lms on the UV window, particulates, and other elements which could decrease the energy applied to the contaminants, should be avoided.




Two types of lamps used in the pharmaceutical industry include: •

low pressure lamps

•

medium/high pressure lamps

Medium/high-pressure lamps emit radiation at several wavelengths, whereas low pressure lamps have smaller Medium/high-pressure spectra; therefore, the low pressure lamps emit more of the energy output at the specic wavelengths 254 nm and 185 nm. Microorganism Contamination Level Reduction In general, low pressure UV light units at 254 nm wavelength output or medium pressure UV units are used for decreasing/limiting decreasing/limit ing the bacteria contamination. UV light rays damage microorganisms (bacteria, virus, yeast, mold, or algae) and break through their outer membrane to modify the DNA. The modied DNA code prevents the replication of the microorganisms. At sufcient doses, UV radiation is an efcient bactericide. Microbial control UV light units generally are installed in generation systems downstream of units, such as activated carbon units, where microbial levels may need control. Filtration upstream of UV units may help to reduce the likelihood of microorganisms being shielded from UV light by particulate from carbon units, softeners, or other media type processes upstream. Organic Contamination Level Reduction UV low pressure units which combine 185 nm and 254 nm UV light, or medium pressure units, generally are used to effectively reduce residual organic contamination. Organic molecules are broken into ionized species under the effects of the UV light. Then the use of ion exchange resins of high purity grade (which will not re-introduce organics) normally is required for maintaining the ionic water quality. Organic reduction UV units generally are installed in generation systems downstream of primary organic reduction processes, such as RO, continuous electrodeionization, or high grade polishing ion exchange resins. Installation downstream of primary organic reduction processes minimizes the required UV energy necessary to meet the TOC specication. UV organic reduction generally is implemented only where the TOC specication is signicantly below the compendial limit of 500 ppb. The UV intensity energy level for TOC reduction usually is signicantly higher than recommended UV intensity levels for microbial control. For further information, see Chapter 13 of this Guide. 5.5.3.2


Advantages:

•


•

no waste stream

•

heat, ozone, and chemical sanitization are possible

Disadvantages: •

can be used only as part of a total microbial control plan

•

no ion or endotoxin removal

•

no disinfection residual

•

particulate can shield organisms from UV light



•


dead organisms are not removed from the water

5.5.3.3 Performance UV light is used as a nal treatment step to address microbial control and TOC reduction (where necessary), after deionization processes. It is not intended to be used as the primary means of either microbial control or TOC reduction. 5.5.3.4 Maintenance A UV-lam UV-lamp p degrades with time and a plan for frequency of exchange should be established. The frequency depends upon the application.

5.5.4

Mixed Bed

5.5.4.1 Description A mixed bed is an ion exchange process with exchange of both cations and anions. It is described previously under ion exchange. The advantages and disadvantages of using it as a polisher is described below. 5.5.4.2 Advantages and Disadvantages

Advantages: •

no waste stream

•

produces the lowest conductivity water of all polishing methods

•

in frequent use

Disadvantages: •

no microbial function

•

resin may get into the polished water

5.5.4.3 Performance A mixed bed typically is used as the nal step after ion exchangers or after RO. If appropriately designed, it should remove ions to trace levels.

5.5.5

Membrane Degasier

5.5.5.1 Description Carbon Dioxide (CO2) and ammonia (NH3) are dissolved gasses that exhaust ion exchange r esins and contribute to water conductivity. conductivity. Carbon Dioxide (CO 2), hydrogen carbonate (HCO3-), and carbonate (CO3=) forms an equilibrium in aqueous solutions, such that carbon dioxide exists only at pH below 8.4 and CO 3= only above that level. Traditionally, Traditionally, CO2 has been r emoved by dosing with sodium hydroxide, driving the equilibrium toward formation of HCO 3- and CO3=. These are ions and can be removed in an ion exchange or RO. Ammonia Ammonia is in equilibrium with ammonium (NH4-) ion. Ammonia can be converted to ammonium through pH reduction. Carbon dioxide and ammonia can be removed without addition of chemicals by using a membrane degasser. degasser.




5.5.5.2 Application The purpose of a membrane degasser in a water system is to remove the carbon dioxide or ammonia. A membrane degasser consists of a hollow ber membrane with water on one side and a strip gas on the other. The strip strip gas can be drawn by vacuum or pressurized on the inlet. The membrane is hydrophobic and the water cannot penetrate the pore. The strip gas lowers the partial pressure of the gas phase which causes the gases from the liquid phase to diffuse through the membrane into the gas phase. Figure 5.2: Membrane Degasser

5.5.5.3 Pretreatment Requirements As the hollow bers are thin bers, a cartridge lter should be installed before the membrane degasser degasser.. 5.5.5.4 Cost Savings Factors If the membrane degasser is used in an RO-CEDI-system RO-CEDI-system,, the pH decreases in the RO permeate and the most efcient position is after the RO. The degasser should be of hygienic design, including all demands on the system, after RO. Membrane degassers in polypropylene are available (at time of going to press) which can be put before RO, which can help to simplify validation of the unit. 5.5.5.5 Advantages and Disadvantages

Advantages: •


•

chemical free method

•

simple installation

•

low operating costs

Disadvantages: •

more expensive than dosing sodium hydroxide

Figure 5.3 shows common congurations for Puried Water systems.




Figure 5.3: Puried Water

Table 5.3 Puried Water Systems Comparison Chart Off-Site Regenerated Ion Exchange

RO/Off-Site Regenerated Ion Exchange

On-Site Regenerated Ion Exchange

RO/On-Site Regenerated Ion Exchange

RO/ Continuous Electrodeionization

Two-Pass Reverse Osmosis

Distillation

Capital Cost

L

M

M

M

M

M

H

Chemical Handling

N

L

H

M

L

L

L (2)

Energy Consumption

L

M

L

M

M

M

H

Water Consumption

L

H (1)

M

H (1)

H (1)

H (1)

M (3)

Outside Service Costs

H

M

M

M

M

M

L

Operational Maintenance

L

M

M

M

M

M

L

Product Conductivity (μS/CM @ 25°C)

1.0 to 0.06

1.0 to 0.06

1.0 to 0.06

1.0 to 0.06

1.0 to 0.07

2.5 to 0.5

1.0 to 0.1

Product TOC (PPB)

(4)

< 500

(4)

< 500

< 500

< 500

< 500

Microbial Performance

L

M

L

M

M

H

H

Notes: 1. High-water consumption consumption unless wastewater wastewater is reused - cooling cooling tower makeup, etc. 2. Total chemical requirement dependent dependent upon pretreatment selection. 3. Total water consumption consumption dependent upon pretreatment selection. selection. 4. USP TOC TOC requirement is met in most cases, but may not be if feed water is high high in TOC (> 2ppm). 5. High microbial performance refers to low microbial count, in relative terms. 6. N = None, None, L = Low, M = Medium, H = High



6

Final Treatment Options: Water for Injection (WFI)




Page 77 Final Treatment Options: Water for Injection (WFI)

6

Final Treatment Options: Water for Injection (WFI)

6.1

Introduction WFI is the purest grade of bulk water covered by pharmacopeial monographs (at time of going to press) based on strict microbiological expectations. expectations. Preparations made with WFI and injected directly into patients bypass the body’s primary defensive systems. These These preparations can cause irreparable harm or death if not suitably pure, based on their rapid distribution through the body and given that, once injected, they are almost irretrievable. The expectation is for WFI to be used for the manufacture of parenteral, inhalation, and some ophthalmic products, as well as for the nishing steps of parenteral grade Active Pharmaceutical Ingredients (APIs). The regulations for WFI are more stringent than for other grades of pharmaceutical water and also more prescriptive, because of its criticality. criticality. This chapter is intended to address nal treatment methods for the production of compendial WFI, the signicant variation between compendia, and the requirements that will control when world-wide distribution of products is planned. Distillation normally is used as the nal processing step for qualied WFI systems, in part due to the rejection of alternative technologies by particular compendia; however, alternative nal treatment technologies are suitable in locales where regulations allow for validation of alternate nal treatment technologies. To select appropriate technologies, the manufacturer of products should rst consider where the products will be distributed and what regulations are applicable to those markets. Using the most stringent aspects of each applicable regulation usually ensures the system design will meet necessary requirements. Technological aspects of the most commonly considered unit operations are considered in this chapter, including their function, maintenance, and associated costs. Specic feed water pretreatment may be necessary to allow nal unit operations to meet compendial requirements. These issues are discussed as they relate to WFI production. In addition, still outlet piping, including any valves, instruments, and accessories, usually is the responsibility of an installer, and should be suitably designed and installed to allow for maintenance of water quality quality,, including appropriate sanitization.

6.2

Pharmacopeial Issues Pharmacopeial publications are revised annually with additional supplements as necessary necessary.. This Guide used the versions current at time of development of this Guide. The US Pharmacopeia 34 Monograph for Water for Injection (Reference 4, Appendix 1) requires that WFI is prepared from water complying with the US Environmental Protection Agency National Primary Drinking Water Regulations (Reference 9, Appendix 1) or with the drinking water regulations of the European Union, Japan, or with the World Health Organisation’s Guidelines for Drinking Water Quality. “Water for Injection is water puried by distillation or a purication process that is equivalent or superior to distillation in the removal of chemicals and microorganisms. microorganisms.”” •

Test for Bacterial Endotoxin <85> applies less than 0.25 USP EU/ml

•

Test for Total Organic Carbon (TOC) <643> applies

•

Test for Water Conducti Conductivity vity <645> applies

Microbial alert and action levels are not part of the individual water monographs. However, the maximum action levels are suggested by the USP for WFI at 10 CFU/100 ml in Informational Chapter <1231> (or lower if technologically appropriate based on system capability) and typically are enforced by FDA.




The European Pharmacopoeia 7.0 Monograph for Water for Injections (Reference 5, Appendix 1) (Aqua ad injectabilia) requires: injectabilia) requires: “Water for injections in bulk is obtained from water that complies with the regulations on water intended for human consumption laid down by the competent authority or from puried water by distillation in an apparatus of which the parts in contact with the water are of neutral glass, quartz or suitable metal and which is tted with an effective device to prevent the entrainment of droplets.” •

“Under normal conditions, an appropriate action level level is a microbial microbial count of 10 CFU per 100 ml…..” ml…..”

•

Total organic carbon (2.2.44): Maximum 0.5 mg/l.

•

Tests for Conducti Conductivity vity apply.

•

Appearance: clear and colourless liquid.

•

Test for Nitrates applies: maximum 0.2 ppm.

•

Test for Aluminium (2.4.17) applies if intended for use in the manufacture of dialysis solutions: maximum 10ppb.

•

Test for Bacterial endotoxin (2.6.14) applies: less than 0.25 IU/ml.

The Japanese Pharmacopoeia XVI Monograph for WFI (Reference 6, Appendix 1) requires: “Water meets the quality standards of water supplies under Article 4 of the Water Supply Law (the Ministry of Health, Labor and Welfare Ministerial Ordinance No. 101, 30 May 2003), also meets the following requirement:” Purity; Ammonium <1.02> <1.02> (Ammonium no greater greater than 0.05 mg/l) mg/l) “Water for Injection is water either prepared by distillation of water or puried water, or by the reverse osmosisultraltration (a RO membrane, an ultraltration membrane, or a combined purication system using these membranes) of puried water…” •

Meets JP specications for pH, Chloride, Sulfate, Nitrate, Nitrite, Ammonium, Heavy Metals, Potassium permanganate-reducing substances and evaporation.

•

For water for injection prepared by reverse osmosis-ultralt osmosis-ultraltration, ration, substitute total organic carbon testing for Potassium permanganate-reducing substance specication, “take precaution against microbial contamination…” and assure “quality being equivalent to that of water prepared by distillation.”

•

Test for Bacterial endotoxins applies <4.01> Less than 0.25 EU/mL.

•

In Chapter 21 of the JP guidance is provided regarding performance “…by reverse osmosis (RO), ultraltration (UF) capable of removing substances having molecular weights of 6,000 and above, or a combination of RO and UF.”

Canadian GMPs (Reference 13, Appendix 1) require: •

PW is used as feed water for WFI.

•

WFI is produced by Distillation or RO.



6.3


General Technology Discussion WFI should be produced using robust technologies, which are capable of consistent and reliable operation. The disparity in compendial requirements, there is only universal agreement regarding distillation, creating additional challenges for those wishing to apply alternative technologies. Distillation is used as the bench-mark to judge other technologies. WFI using RO and UF UF,, both independently or in conjunction, may be an alternative to distillation and can be reliable, cost effective, and practical. Individual regulatory agencies may allow the use of RO and UF for WFI production, used separately or in combination. This discussion may be of use if the system design under consideration will produce product only for use in an area where distillation is not a requirement. Various design options for distillation units, as well as for RO and ultraltration are discussed. A suitable quality pharmaceutical still should purify water chemically and microbiologically through phase changes, while ensuring entrainment cannot compromise quality. quality. Through this process, water is evaporated, producing steam. The steam disengages from the water leaving behind dissolved solids, non-volatiles, and other impurities. Impurities may be carried with water mist/droplets, which are entrained in the steam; therefore, separation devices should be used to remove ne mist and entrained impurities, including endotoxins. Puried steam is condensed into WFI. Distillation systems systems typically are able to provide a minimum of 3 log10 (99.9%) reduction in endotoxin concentration. Design options to be considered include: •

Single Effect (SE)

•

Multiple-Effectt (ME) Multiple-Effec

•

Vapor Vap or Compression (VC)

It should be recognized that ME distillation uses multiple effects (columns) to reduce energy costs; however, as in the case of SE and VC distillation, an ME still distills the water just once. In an ME system, puried steam produced by each effect is used to heat water and generate more steam in the subsequent effect. It is only the rst effect that requires heat from an external source, due to the staged evaporation and condensation process. It is only the puried steam produced by the nal effect which is condensed, using an external cooling medium. VC stills produce similar quality water using a different technique. Energy imparted to the generated steam, by a mechanical compressor, compressor, results in compressed steam with increased pressure and temperature. The higher energy steam is then discharged back into the evaporator/condenser vessel to generate more steam. General areas of concern for stills include carryover of impurities, evaporator ooding, stagnant water, and pump and compressor seal design. These concerns may be addressed using mist eliminators, high water level indicators, use of sanitary pumps and compressors, proper drainage, adequate blow down control, and conductivity sensing to divert unacceptable water to drain. RO and UF unit designs may include heat sanitization and may allow operation continuously at sanitization temperatures. These unit operations have the ability to remove chemical and biologic contaminants; however, each application should be reviewed with the equipment manufacturer under consideration, as proper operation is dependent on feed water pretreatment, ux, recovery, membrane/media material selection, temperature and other critical factors.



6.3.1


Stills Incorporating Rising Film Evaporation The rising lm design incorporates a vertical shell and tube evaporator coupled at the top and bottom to a reservoir of feed water. Utility steam is applied to the outer surface of the evaporator tubes with the inner surface of the tube in contact with feed water. As the feed water is heated, boiling occurs producing a vapor/liquid mixture with steam occupying the center of the tube and a liquid layer at the tube surface. The difference in density between the liquid/ vapor mixture in the tube and the liquid in the reservoir induces a ow of feed water from the reservoir to the bottom of the evaporator thus supplying additional feed water to the evaporator. This process of evaporation and feed water replenishment occurs continually and results in a circulating action with a mixture of steam and water discharged from the top of the evaporator tubes and additional feed water supplied from the reservoir. A separation device removes entrained water and returns it to the liquid reservoir for reuse while allowing steam to be discharged as product. The rising lm evaporator evaporator,, also referred to as a natural circulation evaporator evaporator,, ensures constant wetting of tube surfaces thus minimizing scale formation and corrosion of the tube surface. The application of utility steam to the evaporator is necessary to induce the rising lm action and circulation through the evaporator tubes.

6.3.2

Stills Incorporating Falling Film Evaporation This applies to ME stills. Feed water is pumped continuously into the still where it is progressively heated as it ows through the condenser and any pre-heaters before entering the rst effect. The preheated water is then dispersed over the tube bundle of the evaporator using a diffuser plate, creating a thin falling lm on the inner surfaces of the evaporator tubes. The falling lm is further heated using plant steam on the outer tube surface causing the water to become steam. This phase transformation signicantly increases the velocity of the steam as it approaches the bottom of the column. Upon reaching the bottom of the column, the steam is forced to change direction by 180 degrees, forcing any droplets of water along with any impurities or particles to the bottom of the column, additionally reducing entrainment. This disengaged water becomes feed water for the next column with a portion removed for blowdown. The steam continues vertically up the riser space of the column around spiral shaped guides that induce uniform, centrifugal action forcing any remaining microscopic droplets that may contain impurities through vertical apertures in the separation segment of the column. These remaining droplets coalesce to form a liquid lm that ows, under gravity,, down the cylindrical space to the bottom of the column and is removed as blowdown. gravity

6.3.3

Heat Exchanger Design Heat exchangers may be used in the design of frequently used distillation units, to provide added energy using utility steam as a method to preheat the water, or provide continuous motive steam for operation. Additionally, Additionally, heat exchangers typically are used to condense the highly puried steam into liquid. It is important to provide assurance that uids of a lower quality are not allowed to enter product streams, because of the critical nature of WFI. This is particularly important in shell and tube heat exchangers where utility steam and coolant are used. In these areas, heat exchangers should be equipped with a double tube sheet arrangement. The double tube sheet provides an air gap to ensure that in the event of a leak at the location where the tube is joined to the tubesheet, the leak will be to the atmosphere rather than into the process stream; however, it should be noted that this design feature will not overcome issues related to tube failure. Where it can be demonstrated that the higher purity uid also is at a higher pressure, a single tubesheet heat exchanger may be used, e.g., in multiple effect stills for the second through the last effects. In these evaporators, the pure steam will be at the higher pressure based on the design of multiple effect stills. For further information, see Chapter 8 of this Guide.




Design of sanitary heat exchangers should consider the need for periodically cleaning, passivation, and inspection, and allow for thermal expansion. Figure 6.1: Heat Exchanger Tu Tubesheet besheet Design

6.3.4

Distillation Technology Pretreat Pretreatment ment Requirement Requirements s Distillation technologies require adequate pretreatment to prevent or minimize corrosion and scale formation in the equipment. The scope of the pretreatment will vary based on the source of the feed water, as well as the distillation technology utilized. Vapor Vapor compression stills and single effect stills generally are more tolerant of dissolved mineral content in the feed water, because of the lower operating temperatures. Conversely, multiple effect stills, because of the higher operating temperatures require more stringent control of dissolved minerals; however, these higher operating temperatures provide a more robust mechanism to ensure pyrogen destruction. Silica scale is of concern for all stills and should be evaluated as part of the feed water analysis. In many cases, ensuring that the blowdown rate is sufcient to maintain the silica level below saturation prevents silica scale formation. VC stills operating at low temperatures may be able to tolerate higher levels of silica in the feed water. Silica scaling issues should be discussed with potential equipment suppliers during evaluation.

6.3.5

Basis for Economic Comparison: Estimates of the utilities consumed per volume of WFI produced are discussed. The three primary factors affecting the operating cost of any still design are: •

steam

•

electricity

•

coolant consumption

Using these estimates and the local cost of each utility utility,, expense can be projected based on anticipated production volume.




In general, equipment which uses less energy per volume of WFI produced will be more expensive to purchase (e.g., a larger number of effects will result in a higher capital cost, but lower operating cost for the same rate of production). The appropriate selection should be based on a comparison of operating cost versus capital cost over a suitable time frame.

6.3.6

Construction Constructi on Materials The recommended material for contact with WFI and pure steam is 316L Stainless Steel (SS). Legacy systems continue to operate successfully using other construction materials, such as 304 Stainless Steel (SS) or titanium; however, these materials are relatively uncommon and rarely used for new installations. Gasket materials in contact with WFI or pure steam normally are required to be in compliance with the Class VI of the classication of plastics as determined in USP <88> Biological Reactivity Tests, Tests, (Reference 4, Appendix 1) in vivo. vivo. This requirement should be assessed based on the use of the WFI product. Common sanitary gasket and sealing materials include; blended PTFE/316LSS, PTFE, PTFE/EPDM Envelope, EPDM, and Viton. The use of solid PTFE gaskets is not recommended for high temperature and pressure applications, as it does not have elastomeric properties. Gasket materials should be suitable for the maximum temperature to which they will be exposed, and as VC and SE stills generally will be exposed to lower temperatures a broader selection of suitable materials may be indicated. A preventative maintenance program to replace elastomers should be established for reliable operation. The proper inspection and replacement interval is considered site and application dependant. As with any sanitary requirement, sealing mechanisms should be appropriate with sanitary clamp and orbitally welded joints are prevalent to allow for cleaning (refer to ASME-BPE standard) (Reference 12, Appendix 1). The use of threaded, anged, or other non-sanitary joining mechanism is not recommended where these joints may come in contact with WFI or its precursor water or steam.

6.3.7

Surface Finish Although it is common for the stainless steel components within a still to be polished, the type and level of polish varies signicantly. signicantly. Mechanical polishing and electropolishing are both used with the intention to minimize the potential for microbial growth on the surfaces and to improve cleanability. cleanability. Common industry practice has been to mechanically polish surfaces in contact with feed water to 25 microinch Ra Max (.64 μm) and to electropolish surfaces in contact with pure steam and WFI to 20 microinch Ra Max (.51 μm). The use of mechanical polished surfaces of up to 35 microinch Ra Max can be acceptable. The use of highly polished nishes can increase equipment costs signicantly and because of the relatively high operating temperatures, microbial growth within a still will probably be inhibited, regardless of nish used. Typic Typically ally,, this is not the case for RO and UF, where the production of WFI at ambient temperatures presents a greater opportunity for microbial growth. Passivation of equipment when new, and periodically thereafter, is often recommended and will improve corrosion resistance. In addition, rouge formation is common in distillation systems based on operating temperature and routine periodic inspection is recommended to monitor the presence of rouge and remove it at appropriate intervals. For further information, see Chapter 10 of this Guide. Systems, including stills and distribution, normally require derouging/re-passivation.




6.4

Process and Systems Description

6.4.1

Single Effect Distillation

6.4.1.1 Introduction The Single Effect (SE) still is the simplest distillation design concept with the primary components including an evaporator,, separator mechanism, and a condenser. The evaporator boils the feed water using an external energy evaporator source; the separator removes entrained water and associated contaminants from the steam produced; the condenser changes the pure steam into water for injection utilizing an external cooling source. SE stills may use falling or rising lm technology. technology. The SE still typically will have the lowest capital cost and smallest physical size compared to VC and ME stills of similar capacity. capacity. The primary disadvantage to single effect stills is their operating costs. In contrast to ME and VC stills, there are minimal devices incorporated into typical SE still designs to recover energy energy.. SE still applications are considered ideal when the WFI volumes required are small or usage is infrequent, such that the utility considerations are of a secondary concern. Typical Typical SE applications would be at production rates less than 150 gallons per hour (570 LPH). SE stills typically operate at atmospheric pressure and 212°F (100°C), and incorporate non-pressure rated vessels. WFI is delivered at atmospheric pressure and 176°F to 190°F (80°C to 88°C); thus a distillate transfer pump may be needed, unless the WFI receiving tank is at a lower elevation than the WFI discharge point on the still. SE stills can easily be converted to or used as a pure steam generator which can become a source of pure steam for use in manufacture or other uses appropriate to the facility facility.. If the SE still is expected also to produce pure steam, additional provisions may be required to control non-condensable gasses. In addition, when used as a pure steam generator, the operating pressure may be greater and vessel pressure rating requirements may change. It should be noted that based on size, units may not be suitable for simultaneous production of WFI and pure steam. 6.4.1.2 Utilities and Pretreatment SE stills are available in electric or steam powered versions although electric units typically are more expensive in capital and operating cost. Steam powered units typically operate using utility steam in the range of 15 PSIG to 60 PSIG (1 BARG to 4 BARG). Cooling uid is required for both steam and electric powered versions. SE still designs may require high coolant ow rates or necessitate a high temperature rise in the coolant if adequate ow is not available, because of the high WFI discharge temperature. Appropriate Appropriate commercial cooling technology can be applied providing it is non-corrosive to the contact materials, will not foul the exchanger, and has suitable capacity. Care should be taken when using a coolant source with high total dissolved solids content, as treatment for scale prevention may be required. Chlorinated water generally is not advisable because of the corrosive nature of chlorine in contact with stainless steel at elevated temperatures. Feed water pretreatment should be adequate to remove volatile compounds (such as ammonia) which may be carried through, contaminating the steam and WFI produced. In addition, pretreatment should remove chlorine and control silica and hardness levels, typically to less than 1ppm (approximately). Chlorine will damage the still and although excessive levels of silica and hardness generally will not affect the quality of the WFI produced, scale formation and fouling may result in more frequent cleaning. SE stills typically are capable of a 3 log pyrogen reduction. Feed water pretreatment for WFI should comply with local drinking water regulations; therefore, microbial control normally is anticipated. The pretreatment system should be properly designed and maintained to ensure that feed water microbiology and endotoxins are controlled. In general, a pretreatment system consisting of softening, dechlorination, and deionization could be sufcient. For further information, see Chapters 4 and 5 of this Guide.




6.4.1.3 Economics Commercially available SE stills are congured similarly and equipped with the same basic components; however, capital cost can increase with added instrumentation and controls, as well as by higher grades materials and nishes. Steam consumption is approximately 1.10 to 1.25 times the distillate produced on a mass basis depending on the feed water temperature. Lower feed water temperatures require more energy to bring the temperature of the feed water to the boiling point. For example, to produce 100 pounds (45 kg) of distillate will require 125 pounds (57 kg) of utility steam when using feed water at 70°F (21°C). When water is the coolant, approximately 9 to 10 times the WFI production rate is required based on a coolant temperature rise of 100°F (56°C). It is important to recognize that the ow and temperature rise are inversely related, such that as the temperature rise is decreased the coolant ow rate is increased. The heat capacity of alternative coolants should be taken into account, as they may be less than that of water (e.g., propylene glycol mixtures). SE stills are considered appropriate for small or intermittent applications, where operational costs are of lower importance than other factors or when pure steam and WFI are required from a common source.

6.4.2

Multi-Effect Multi-Ef fect Distillation

6.4.2.1 Introduction A Multi-Effec Multi-Effectt (ME) still is similar in design to an SE still except that multiple evaporator and separator stages are included. ME stills use a staged evaporation and condensation process to produce WFI at reduced energy consumption compared to an SE still. Utility steam is applied only to the rst evaporator (effect) with the subsequent effects using the steam produced in the previous effect as the source of energy. Similarly, Similarly, coolant is applied only at the nal effect to condense pure steam into WFI. The effects subsequent to the rst effect use feed water as the cooling source (which is evaporated) as the pure steam from the preceding effect is condensed (producing WFI). It is important to note that the ME still does not evaporate the water more than once, and does not produce water of a higher quality than an SE still. The interconnecting effects of the ME still serve only to minimize utility consumption. Typical Typi cal ME stills have 3 to 6 effects. As the additional effects are included, the energy efciency of the still is increased; however, however, the capital cost and oor space required are also increased and should be evaluated to optimize the design. It may be possible to use the rst effect of an ME still as a source of pure steam for use in manufacturing or for other uses appropriate to the facility. If the ME still is expected also to produce pure steam, additional provisions may be required to control non-condensable gasses. For further information, see Chapter 7 of this Guide. Typical designs deliver the water at atmospheric pressure to the WFI storage tank via gravity Typical gravity.. Options are often available such as; a WFI transfer pump, elevated condenser, condenser, or pressurized condenser if the WFI tank is at a higher elevation. Typical Typical ME stills are designed to discharge WFI in the temperature range of 176°F to 190° F (80°C to 88°C). 6.4.2.2 Utilities and Pretreatment ME systems typically require utility steam in the range of 80 to 120PSIG supplied to the rst effect. Electrically operated units are available; however, electrically operated stills typically are selected for applications of 100 GPH (380 LPH) or less. The capacity of an ME still is proportional to the pressure applied at the rst effect. As the available utility steam pressure is reduced, the capacity of the still will be diminished. Based on the higher operating pressures, ME stills often are specied with vessels that are pressure rated. Cooling uid is required for both steam and electric powered versions. ME still designs may use feed water as a source of coolant to conserve energy by preheating the feed water (less heating needed), while simultaneously




reducing the condenser load (less cooling needed). However, feed water alone may not be sufcient and supplemental cooling may be required. ME stills may require high coolant ow rates or necessitate a high temperature rise in the coolant if adequate ow is not available, because of the high WFI discharge temperature. Appropriate commercial commercial cooling technology can be applied providing it is non-corrosive to the contact materials, will not foul the exchanger, and has suitable capacity. Care should be taken when using a coolant source with high total dissolved solids content, as treatment for scale prevention may be required. Chlorinated water is generally not advisable because of the corrosive nature of chlorine in contact with stainless steel at elevated temperatures. Pretreatment recommendations are similar to those for SE stills. 6.4.2.3 Economics Commercially available ME stills are congured similarly and equipped with the same basic components. Increasing the number of effects increases the capital cost and can be compounded by the added instrumentation and controls, as well as by higher grades materials and nishes. The primary costs associated with ME still operation are for utilities and may vary by manufacturer manufacturer,, based on the number of effects included or based on local utility costs. Increasing the number of effects in an ME still does not result in increased output or improved WFI quality quality,, but reduces the amount of steam and cooling water required to produce an equal amount of distillate with fewer effects. Table 6.1 shows the impact that additional effects have on efciency using an SE still as the basis of comparison. These consumption estimates are typical and may vary based on application specics and manufacturer. Coolant conditions are expressed in coolant duty rather than in units of ow to ensure a common basis for comparison. Table 6.1: Utility Consumption Ratios by Effect Effects

1

3

4

5

6

Steam

1

0.35

0.28

0.23

0.21

Coolant

1

0.26

0.19

0.14

0.11

Table 6.2 shows information which is representative of applications with feed water at 70°F (21°C) and WFI discharge at 185°F (85°C). Note that electrical operating costs are negligible. Table 6.2: Approximate ME Still Utility Consumption per 1000 Gallons (3785 Liters) of WFI Produced Effects

1

3

4

5

6

Steam Lb/Hr

9,800 – 10,200

3,300 – 3,700

2,600 – 3,000

2,100 – 2,500

1,900 – 2,300

Kg/Hr

4,445 – 4,626

1,496 – 1,678

1,179 – 1,360

952 – 1,134

861 – 1,043

7.9 – 8.1

2.0 – 2.2

1.4 – 1.6

1.0 – 1.2

0.8 – 1.0

2,313 – 2,371

585 – 644

409 – 468

293 – 351

234 – 293

Coolant MBTU/Hr (KW)

Utility consumption estimates can be inuenced by feed water temperature, product water temperature requirements, plant steam pressure, and other factors including supplier equipment conguration.




An increased number of effects will decrease the operating cost; however, there is a diminishing return as the number of effects is increased. While stills with as many as 8 effects or more are available, usually the maximum number of effects that can be justied is 6. The appropriate number of effects should be determined by nancial analysis comparing the capital and annual operating costs for alternatives under consideration. The result can be inuenced by site specic utility costs, utilization, and the time frame used for the nancial analysis.

6.4.3

Vapor Compression Distillation

6.4.3.1 Introduction Vapor Compression (VC) distillation conceptually is similar in design to a heat pump or the mechanical refrigeration cycle. Major system components include: •

evaporators

•

compressors

•

heat exchangers

•

deaerators

•

pumps

The compressor typically is a single-stage centrifugal type with relatively low developed pressure. VC is intrinsically a thermally efcient distillation process, because it recycles a high percentage of the latent heat. Unlike an SE or ME still, VC stills have a primary chamber in which evaporation takes place on one side of the heat transfer surface, and condensation takes place on the other side of the same surface. Although in some industrial applications, the heat transfer surface uses plates; for WFI, only tubes are used because of the stringent quality required. Two common VC congurations are vertical tube and horizontal tube. In the horizontal tube conguration, feed water normally is circulated, using a pump, and then sprayed by nozzles on the outside of the tubes, where evaporation takes place, while condensation is on the inside of the tubes. In the vertical tube design, feed water is circulated naturally (without a pump), inside the bank of tubes, where evaporation takes place, while condensation is on the outside of the tubes. Feed water is evaporated on one side of the tubes, and the generated steam passes through a disengagement space and separation system, to remove entrained water droplets, before the pure steam is drawn through the compressor. The energy imparted by the compressor results in compressed steam with an increased temperature of 7°F to 10°F (4°C to 5.5°C) which is equivalent to increased pressure of 3 PSIG to 5 PSIG (0.2 BARG to 0.34 BARG). The higher energy steam is discharged to the condensing side of the heat transfer surface. There, the steam condenses giving up its latent heat, which is transferred through the tube wall to the feed water. Additional water is evaporated, generating more vapor, as the process is repeated. The condensate produced is WFI and the portion of the feed water that is not evaporated is recycled. A portion of the feed water referred to as blowdown, has highly concentrated impurities and is usually discharged as waste or used for other applications. Although VC stills may be congured to produce WFI at 176°F to 185°F (80°C to 85°C), applications may produce “cold WFI” typically at 10.8°F to 25°F (6°C to 13.9°C) above the feed water temperature. Intermediate temperatures also are achievable. Shell and tube heat exchangers are used to recover heat from the discharged WFI and blowdown streams, where it is used to preheat the feed water, minimizing utility steam consumption.




6.4.3.2 Utilities and Pretreatment Energy is input in two forms: •

electricity to drive the compressor and pumps

•

relatively low pressure steam as makeup heat

The amount of WFI produced is proportional to the amount of compressor work input; it is limited by the amount of heat transfer surface and heat transfer efciency. A feed water analysis should be performed and the results used in selecting and conguring a suitable VC still system. It is recognized that pretreatment requirements may be less stringent than for a ME still, because of the lower operating temperatures in a VC still. If the feed water is chlorinated or treated with chloramines, removal of the chlorine and chloramines is necessary to minimize of corrosion and rouging. Residual ammonia, if not removed, will contribute to conductivity, conductivity, such that the distillate may not meet the requirements for WFI. If a carbon lter is used for chloramine removal, it should be recognized that ammonia is a by-product of the chloramine removal inside the carbon lter. If the carbon lter is installed upstream of a softener, ammonia (a cation) will be removed in the softener. Hardness will be removed rst, then ammonia, because of the cation selectivity order. order. In this type system, there is built-in scale prevention since ammonia will “break through” before hardness causing an increase in conductivity. conductivity. Conductivity, Conductivity, which normally is monitored, usually is alarmed as well. Deaeration and Ventin Venting g in VC Stills Feed waters have dissolved gasses, the amount of which depends on water temperature, composition, and pH. Other gasses occur due to the breakdown of some of the constituents during heating, as is the case of alkalinity alkalinity.. Non-condensable gases, which include carbon dioxide and oxygen, are liberated as the temperature of the water increases. These gases, if not removed, have two detrimental effects on VC units. Since the gases are noncondensable, they can blanket the heat transfer surface and inhibit heat transfer translating into reduced output. In addition, some gases may increase the potential for corrosion. VC stills may have feed water deaerators that remove most non-condensables. Remaining gases are liberated in the evaporator,, and as the water boils inside the evaporator evaporator evaporator,, any remaining dissolved gases are liberated, make their way to the deaerator, combine with gases released there, and all non-condensables are vented to atmosphere. Such deaerators use the vent steam from the evaporator to preheat the feed water. VC stills that do not use feed water deaerators normally use vent condensers. Non-condensables are liberated from the water inside the evaporator; make their way with the steam to the condensing side of the tubes, and eventually to the vent condenser, where they are released to the atmosphere. 6.4.4

Reverse Osmosis and Ultrafltration

6.4.4.1 Introduction RO purication of water is accomplished by creating a pressure difference across a semi permeable membrane. The membrane used is permeable to water, but impermeable to many impurities. RO systems may be capable of producing water that meets the standards for WFI, and under suitable conditions, may reduce system capital and operating costs when compared to a distillation system. This is dependent upon regulatory acceptability for the application.




RO can remove many salts, acids, bases, colloids, bacteria, endotoxins, and some dissolved organics below WFI requirements; however, RO cannot remove 100% of the contaminants as well as some low molecular weight organics. UF is an extremely ne ltration although not as ne as RO. Pressure across a porous ltration material separates water being puried from particulates, organics, microbes, and pyrogens that are too large to pass through the media. UF does not reject low molecular weight organics and ionic contaminants. Unlike distillation, which relies on phase separation and entrained droplet removal, both RO and UF rely upon the integrity of the media itself and the seals that separate treated and untreated water; however, in specic global regions, these technologies are acceptable for producing WFI, either independently or in combination. Typically,, there are two primary concerns when considering RO for production of WFI. The rst is the ability to Typically meet the conductivity specication, specication, which may determine the use of double pass RO, and the second is ongoing compliance with microbial limits. For further information, see Section 6.1 and Chapters 4 and 5 of this Guide. 6.4.4.2 Pretreatment and Utilities Thorough analysis of feed water composition is considered vital for the success of an RO or UF application. Scaling and fouling are two of the concerns that should be addressed by pretreatment; which usually will include, as a minimum, preltration and softening. In addition, consideration should be given to chlorine removal, as RO and UF media may be quickly damaged by chlorine or chloramine. Another important consideration is pH adjustment, which may be used to address specic dissolved gases not rejected by RO membranes. As a pressure driven process, both RO and UF rely on pressure differential created primarily by pumps for proper operation; therefore, the primary utility required is electricity. However, However, based on their design, sanitization is a concern requiring a method for heating, such as steam, or additional electricity to drive a heat exchange process. In addition, cooling also may be required, either as feed water or from another source to bring the system back to its normal ambient operation. It should be noted that hot UF systems have been in operation. Continuously hot RO systems may be developed for future use. A unique aspect of RO that should be considered is its requirement for a continuous waste stream of water to eliminate contaminants, ranging typically from 15% to 30% of the total feed water. This can be a signicant cost factor,, unless the wastewater can be utilized, such as for cooling tower water. For further information, see Chapter 5 factor of this Guide. 6.4.4.3 Economics The costs (capital and operating) should be considered carefully and reconciled against alternative options during the system evaluation. Typical operating costs include supply water, waste water, plant utilities (electric, steam, coolant, compressed air, air, vents, etc.), replacement parts, and man hours. A ten year life is often used as the basis for evaluation; however, costs developed should be accurate and reliable, including consumable costs and realistic replacement frequency. frequency. The total life cost and cost per unit volume of WFI are often compared. Waste water discharge from ROs and UFs should be evaluated for possible reuse in cooling towers and for other functions. RO sanitization costs also should be considered relative to system evaluation based on a realistic schedule and the cost of media replacement, which can range from 3 to 5 years, must not be overlooked. 6.4.4.4 Construction Materials and Surface Finish RO and UF systems typically operate at ambient temperature allowing for the use of non-metallic construction materials in an appropriate low pressure portion of the system. Sanitary piping and valves may be optional for some systems based on the location of the RO or UF in the generation train. Often for the higher pressure section of an RO, prior to the nal purication step, may be constructed of 304SS pipe to minimize cost.




Sanitary components with either mechanical polishing or electropolishing usually are reserved for the lower pressure product side of an RO or UF, especially especially if the output will not have additional treatment, or serves as the nal product. If operation is planned for continuously hot or periodic hot sanitization, suitable materials should be employed throughout the unit. If UF is suitable and meets system, as well as regulatory requirements, its lower waste water losses and lower cost materials may prove benecial/advantageous. Systems congured with UFs may be more applicable to API production. 6.4.4.5 Instrumentation and Controls Proper instrumentation for monitoring and controlling critical and non-critical parameters of RO and UF generation systems are essential as in all WFI systems. For further information, see the ISPE Good Practice Guide for Commissioning and Qualication of Pharmaceutical Water Systems (Reference 14, Appendix 1). Parameters that require monitoring can have an impact on nal product conductivity conductivity,, including feed water conductivity and pH permeate ow, reject ow, temperature, and operating pressures. These changes can alert trained personnel of a trend in system operation that eventually may have a negative effect on nal product quality. quality. 6.4.4.6 Advantages and Disadvantages It has been estimated that WFI produced by RO or UF can offer lower life cycle cost, when compared to distillation. RO/UF systems can improve conductivity, conductivity, TOC, and microbial and endotoxin levels. When properly designed and operated, these systems are capable of reliably meeting WFI requirements. UF alone does not reduce conductivity conductivity,, so UF should be used in conjunction with other treatment technologies. UF generation systems can offer lower life cycle cost alternatives in appropriate applications and may be more applicable to API production. UFs may be able to operate at 80°C and may accomplish a 4 log reduction of endotoxins. RO systems also can operate at 80°C. Disadvantages of RO and UF UF,, when compared to conventional distillation, are that they are not accepted in Europe and they utilize limited-life media that requires routine cleaning, sanitization, and replacement, and can fail catastrophically.. In the US, there are very few RO based WFI systems in operation. catastrophically Figure 6.2: Typical RO/UF WFI Generation System



6.5


Final Treatment – General System-Wide Controls and Instrumentat Instrumentation ion Controls and instrumentation for nal treatment equipment varies widely widely,, based on the supplier of the equipment and the technology used. In general, the instrumentation should monitor the CPPs of conductivity and temperature, and protect the equipment. Parameters within the nal treatment system that are considered critical for proper operation of the system should be monitored, and alarmed where appropriate, to allow for predictive/preventive maintenance and to ensure product protection. These parameters may include: •

storage tank level

•

incoming water conditions (including quality quality,, ow, and pressure)

•

system temperatures, ows, and pressures

•

utility pressures, ows, and temperatures

Interfaces between the system controls and related equipment also should be considered. The nal treatment system receives the output of a pretreatment system, and typically typically,, supplies a WFI storage and distribution subsystem. In addition, the storage and distribution subsystem may be interfaced with a batch control system that ensures proper delivery and prevents over-draw. A wide variety of methods can be employed for suitable interface including Ethernet, relay contacts, and pneumatic signals. Typical interfaced equipment and systems may include: •

valves, with or without position indication

•

pretreatment system

•

batch controls or product tanks

•

plant wide distributed control system

For further information, see Chapter 11 of this Guide.

6.6

Summary and Technology Comparison

6.6.1

Distillation Distillatio n Applicatio Applications ns and Capacities The majority of USP WFI produced in the United States is produced by distillation at time of going to press. SE stills may be found in areas where small quantities of WFI are required. Where larger amounts of WFI are required; however, economics economics of operation may dictate the use of either ME or VC.




Table 6.3: Typical Component Summary Component

RO

UF

SE

ME

VC

Evaporator

No

No

Yes

Yes

Yes

Condenser

No

No

Yes

Yes

Yes(Note 1)

Compressor

No

No

No

No

Yes(Note 4)

Product Pump(Note 2)

No

Yes

No

No

Yes

Blowdown Pump

No

No

No

No

Yes

Feed Pump(Note 3)

No

Yes

No

Yes

No

Cooling Water Required for Operation

No

No

Yes

Yes

No

Pure Steam Generation possible

No

No

Yes

Yes

No

Periodic Sanitization Required

Yes

Yes

No

No

No

Notes: 1. Integral with evaporator. evaporator. 2. May be required depending upon delivery requirements. requirements. 3. May be required based on available water pressure or if fed from a break tank. 4. Compressors for VC stills typically require a pumped lubrication system. system.

Table 6.4: Typical Operatio Operation n Summary Parameter

RO

UF

SE

ME

VC

Required Steam Pressure, PSIG (BARG)

NA

NA

15 – 60 (2.1 – 4.1)

44 – 125 (3.0 – 8.6)

15 – 40 (2.1 – 2.8)

Normal Operating Temperature, °F (°C)

70° (21°)

70° (21°)

212° (100°)

274° – 330° (134° – 166°) In rst effect

224° (107°)

Typical Capacity Ranges GPH (LPH)

< 1 – 10 (< 3.79 – 38.0)

< 1 – 10 (< 3.79 – 38.0)

1 – 150 (3.79 – 36.63)

50 – 3,000 (189 – 11,356)

300 – 6,000 (1,136 – 22,713)

The utility costs associated with VC still operation may vary by manufacturer manufacturer,, based on the net saturated temperature across the compressor and based on local utility costs. The amount of WFI produced is proportional to the amount of compressor work input and is limited by the size of the heat transfer surface. Q = U * A *ΔT Q = Amount of compressor energy added U = Heat Transfer Coefcient A = Heat Transfer Area ΔT = Net saturated temperature across the compressor




Trade off is between increasing surface area (higher capital costs, reduced compressor energy energy,, reduced operating costs) and increasing ΔT (lower capital costs, increased compressor energy, energy, increased operating costs). Table 6.5: Utility Consumption per 1000 Gallons (3785 Liters) of WFI Produced Capacity Range 200 GPH (757 LPH) to 7200 GPH (27,255 LPH)

Minimum

Maximum

44

80

Steam Lb/Hr

900

1100

Kg/Hr

408

498

2

4

7.6

15.2

Electricity kWh

Coolant GPM LPM

The information in Table 6.5 is representative of applications with feed water at 70°F (21°C) and WFI discharge at 185°F (85°C). Utility consumption estimates can be affected by feed water temperature, product water temperature requirements, plant steam pressure, and other factors, including supplier equipment conguration.


7

Pharmaceutical Steam




Page 93 Pharmaceutical Steam

7

Pharmaceutical Steam

7.1

Introduction This chapter aims to simplify and standardize the process of selection, programming, and design of pharmaceutical steam systems. Guidelines, Guidelines, information, and options of proven practices (as of publication) and technologies are provided, along with advantages and disadvantages. There are few industry guidelines for the specication, installation, and quality assurance of pharmaceutical steam. Regulatory guidance (e.g., USP (Reference 4, Appendix 1)) for production and purity of pure steam typically is consistent with the guidelines for high purity water with added conditions for steam quality (i.e., superheat, saturation or dryness, non-condensable gases). International standards (e.g., British Standard EN 285, HTM 2010 (References 15 and 16, Appendix 1)) also provide requirements applicable to the purity and quality of pure steam used in the bioprocessing and pharmaceutical industries. These guidelines include such things as material specications, dimensions/tolerances, dimensions/tole rances, surface nish, material joining, and quality assurance. The chapter establishes standard denitions for terms commonly associated with pharmaceutical steam and provides information that helps facilitate making correct and cost effective decisions.

7.2

Common Steam Terms and Defnitions

7.2.1

Plant Steam Plant steam is non-direct impact steam (utility steam) produced by the feed of potable water or equivalent to an industrial type boiler. Corrosion control additives may be used in the maintenance of the boiler system. Typically, Typically, this steam is used for non-direct contact process heating.

7.2.2

Chemical Free Steam Chemical Free Steam (CFS) is non-direct impact steam produced from pre-treated water with no volatile boiler additives. Non-volatile boiler additives should meet the FDA Generally Recognized as Safe (GRAS) listed additives or other equivalent international standard where applicable. Typically, Typically, CFS is used for humidication and is not used for product contact operations.

7.2.3

Process Steam Process steam is direct impact steam that once condensed, meets the quality characteristics of potable water. Typically Typi cally,, this steam is used in manufacturing areas for direct injection heating and sterilization.

7.2.4

Pure Steam or Clean Steam Pure steam or clean steam is direct impact steam, which is produced by a steam generator. When condensed, the steam condensate meets requirements for USP/EP WFI. Pure steam is predominantly used for sterilization. Sterilization steam used in autoclaves for international manufacturing also should meet the requirements of British Standard EN 285 (Reference 15, Appendix 1) for non-condensable gases, degrees of super heat, and dryness.




7.3

Types of Steam Pharmaceutical steam is classied into two types based on their respective sources. These are:

7.3.1

1.

utility-boiler produced steam (plant steam)

2.

non-utility boiler produced steam (pure steam)

Plant Steam Plant steam is characterized normally by having: •

chemical additives to control scale and corrosion

•

relatively high pressure with the potential of generating superheat during expansion

•

relatively high pH

Chemical additives: plant steam normally is produced using conventional steam boilers, usually of steel construction. Such boilers are usually provided with systems that inject additives into the feed water to protect the boiler and steam distribution piping from scale and corrosion. The scale and corrosion inhibitors may include amines and other substances that may not be acceptable in steam being used in pharmaceutical processes. The user must determine which additives are used, and verify if they are acceptable in the particular application, i.e., do not add any impurities or create a reaction in the drug product. The creation of steam condensate that can meet the potable water specication may be possible on condensation. Plant steam typically is used for heating coils in HV HVAC AC applications and as the heating media in heat exchangers. Additionally,, plant steam can be used for sanitization Additionally sanitization of non-product non-product contact equipment or biological destruction of solid or liquid wastes. Plant steam can be ltered to remove particulate matter matter,, but ltration does not remove dissolved substances and volatiles such as amines. Superheat: superheated steam is produced by heating the steam beyond saturation temperature or by generating the steam at a higher pressure in a re tube boiler and then reducing the pressure through a regulating valve. When the pressure is reduced, condensate is not produced due to the higher temperature beyond saturation; therefore, superheated steam is not suitable for sterilization. Additional information can be found in ASME BPE (Reference 12, Appendix 1). Superheat Superheat may be dissipated dissipated downstream of of the regulating valve valve because of heat heat loss in the lines. This excess of sensible heat must be removed before steam will condense. This makes steam more difcult to condense, as a portion of the heat exchange surface will be used to remove the sensible heat before a phase change can occur. This is benecial to transport (less loss in the steam lines), but can be a problem in heat exchangers and sterilization processes. Control of pH: to protect carbon steel from corrosion by the plant steam, additives should be used to raise the pH to between 9.5 and 10.5.

7.3.2

Pure Steam There are many terms used in guidance documents and the pharmaceutical industry to describe pure steam. In addition to pure steam, these include: •

Clean Steam

•

Pyrogen Free Steam




•

WFI Steam

•

USP PW Steam

Guidance documents from regulatory agencies include various denitions for these terms. The most commonly used terms are “pure steam” and “clean steam”. In this Guide, the term “pure steam” is used. Pure steam is generated from treated water that meets applicable drinking water regulations and that is free of volatile additives, such as amines or hydrazines, and is used for thermal disinfection or sterilization processes. It is considered especially important to preclude such contamination from injectable drug products: Pure steam is characterized as having: •

no additives

•

limited generated superheat except when the generated pressure is signicantly higher than the use pressure of the steam

The condensate of pure steam should meet requirements for USP/EP WFI, has no buffer buffer,, and has a relatively low pH compared to that of plant steam.

7.4

Regulatory and Industry Guidance The user has the ultimate responsibility for system design and performance and for ensuring that the proper type of steam is used for a given process.

7.4.1

United States Pharmacopeia Guidelines The USP (Reference 4, Appendix 1) provides guidance as to the generation, quality attributes, and uses of pure steam. The Pure Steam USP monograph provides direction for the feed water source, added substances, and testing condensate attributes. Pure steam dryness and non-condensable gases; however, should be determined by the user, based upon the specic application. In addition, CGMPs issued in 1976 for Large Volume Parenterals (LVP (LVPs) s) indicated that feed water for boilers supplying steam that contact components, drug products, and drug product contact surfaces shall not contain volatile additives, such as amines or hydrazines. Among US Government publications, the FDA FDA’s ’s Code of Federal Regulations (CFR) (Reference 17, Appendix 1) provides culinary steam recommendations and stipulations related to heat exchanger and tank air vents design and construction. The culinary steam recommendations apply to food applications only. only. US Public Health Service/Dairy Industry Committee, 3A Sanitary Standards, Number 609-02 (Reference 18, Appendix 1), adds additional limitations to culinary steam feed water additives for food applications. It should be noted that boiler feed water additives permitted in food for human consumption may not be acceptable in drinking water or orally ingested drug products.

7.4.2

European Guidelines European Guidance for pure steam is provided in the British Standard EN 285 (Reference 15, Appendix 1) and in Health Technical Technical Memorandum (HTM) 2010 (Reference 16, Appendix 1).




British Standard EN 285 stipulates that steam for sterilization equipment meets the following steam physical quality attributes: •

Contains no more than 3.5% V/V of non condensable gases.

•

Dryness value not less than 0.95 for metal loads.

•

Dryness value not less than 0.90 for other types of load.

•

When the steam is expanded to atmospheric pressure, the superheat shall not exceed 25°C.

British Standard EN 285 and HTM 2010 provide guidance on the test methods for these attributes.

7.4.3

Industry Guidance CGMP guidance documents, such as ASME BPE (Reference 12, Appendix 1) govern the design and construction of pure steam generators and distribution systems. This This standard provides recommended practices for pure steam systems and high purity water systems in the bioprocessing and pharmaceutical industries. Comparable guidelines in Europe and Asia include DIN and JIS-G. These guidelines include such things as material specications specications,, dimensions/ tolerances, surface nish, material joining, and quality assurance.

7.5

Background and Industry Practices

7.5.1

Purity of Sterilizing Steam When steam or the resulting condensed water comes in direct or indirect contact with the drug product, the purity should be equivalent to the water purity acceptable for manufacture of the drug product. Note: a continuous supply of dry saturated steam at the Points of Use (POU) is considered necessary for efcient Note: a steam sterilization. Water Water carried by the steam in suspension may reduce heat transfer and superheated steam is considerably less effective than saturated steam when used for sterilization. Non-condensable gases, if contained in the steam, act as insulation through blanketing of heat transfer surfaces, and may prevent the attainment of sterilization conditions in parts of the sterilizer load.

7.5.2

Steam for Humidication

When steam is used for indirect humidication, such as injection into HVA HVAC C air streams prior to nal air ltration, the steam does not need to be purer than the entrainment air and suitable plant steam may be used. When humidifying process areas; however, the potential level of impurities, including amines and hydrazines should be evaluated in order to ascertain the impact on the nal drug product. This is particularly important in areas where open processing takes place, such as aseptic lling suites and formulation areas. If the diluted water vapor is found to contribute signicantly to the contamination of the drug, a purer grade of steam should be selected.

7.5.3

Industry and Baseline Practices for the Productio Production n of Steam The Table 7.1 represents typical uses for pure steam and the commonly accepted generation methods used to meet regulatory requirements in the pharmaceutical industry. industry. The table is not intended to be denitive or all-inclusive.




Table 7.1: Typ Typical ical Uses and Generation for Pure Steam Intended Use of Steam

Method of Steam Generation or Steam Type

Parenteral and Non-Parenteral Dosage form applications where steam is in direct contact with the drug.

The use of a Pure Steam Generator (PSG).

Critical step in the manufacture of API where steam is in direct contact with the API.

The use of a PSG.

Non-Critical Step in the manufacturer of an API where added impurities will be removed in a subsequent step.

A PSG PSG is commonly commonly used; however, however, chemical free free steam may be acceptable.

Sanitization or Sterilization of a High Purity Water system.

A PSG PSG is commonly commonly used; however; chemical free steam may be acceptable followed by adequate ushing.

Humidication for dosage form production where steam is in direct contact with the drug, where open processing occurs, or where chemical additives may be detrimental to the drug product. Humidication of non-critical HVAC systems such as rooms and areas where the drug product is not directly exposed to the room atmosphere.

7.6

The use of a PSG.

A PSG PSG is commonly commonly used; however; chemical free steam or plant steam may be acceptable.

Humidication of critical process cleanrooms.

The use of a PSG.

Heat source for non-critical and CGMP heat exchangers.

Chemical free steam or plant steam.

Deactivation of solid or liquid biologic process waste.

The use of PSG or chemical free steam in a dedicated deactivation vessel.

Sterilization of direct product contact production equipment, process vessels, containers, or packaged product.

The use of a PSG.

System Planning Pharmaceutical steam system planning is the process of establishing system boundaries, limitations, and restrictions. Initial system planning should reveal primary boundaries that establish the foundation for design criteria. The system boundaries include: •

steam requirements

•

system design

•

POU criteria

•

distribution system requirements

The limitations of these system boundaries establish more specic operating strategies and ranges. To allow more exibility in nal planning and detailed design, the designer should always indicate ranges of acceptability, rather than a specic value or position.




7.6.1

Steam Requirements The planning process starts with the listing of all steam requirements and applications that include:

7.6.2

•

Company standards including QA/QC requirements and published SOPs.

•

The categorization of use-points by: -

Type of application (humidication, contact with product, API, and dosage form applications).

-

Purity selection primarily based on the application and the endotoxin and chemical purity criteria set for the product with which the steam, or its condensate, will be in contact. The selection should consider underlying factors which have impacts on purity control, installed and operating cost, maintenance, and practicality.

-

Steam quality (dryness, non-condensable gas limits, and maximum superheat).

System Design Pharmaceutical steam is generated using different methods. The most appropriate method for each application should be selected. The process should continue with an evaluation of the steam system requirements (generation) that includes: the selection of the type of generation system that would satisfy each category, category, which would include:

7.6.3

•

The types of generation systems available. (If both pure steam and lower quality steam are required, the practicality and economy of producing only pure steam should be considered.)

•

The type and number of systems required based on feedback from the “Distribution System” evaluation.

•

The condensate sampling needs.

•

Safety considerations.

Point of Use Criteria The third step should dene the specic delivery requirement ranges for pure steam at the Points of Use (POU) including:

7.6.4

•

utilization, which is determined from system peak demand(s), average demand, and the relationships between peak demand periods and their ow rates

•

pressures, ow rates, and sampling requirements

•

use periods and histogram analysis, if available

Distribution Distributi on System The fourth step includes the distribution system evaluation, which includes: •

condensate, non-condensable, and moisture removal

•

pipe size and insulation requirements

•

materials of construction, sanitary design requirements, and surface nish




•

physical location of each use point

•

heat and temperature losses

•

natural drainage

Note: as the steam quality will decline, condensate formation due to heat losses, with time, the efciency of Note: as the insulation and the length of the distribution system, the quality at the POU will not be expected to reect the generation quality level. The inclusion of high quality condensate removal traps that are strategically located can greatly reduce this effect.

7.6.5

System Planning Re-evaluation These sequential steps are repeated and re-evaluated as information in the design process iterates, and further criteria about the overall system boundaries are identied. In operations with a requirement for only one grade of steam, the steam system should be designed to meet the requirements of the product or process. Where more than one purity grade of steam is required, products and processes frequently are categorized and fed by the most appropriate steam system. The number of types of steam generated is usually a function of the volume of steam consumed, economics, and variation of purity required. Figure 7.1: Pharmaceutical Steam Purity Decision Flow Chart




7.7

Steam Generation

7.7.1

Plant Steam (Utility Steam) Plant Steam is produced in conventional plant utility boilers whose typical design and construction are outside the scope of this Guide.

7.7.2

Pure Steam (PS) Pure steam is produced in specially designed non-red generators or from the rst effect of ME WFI stills, which do not use scale or corrosion inhibitor additives. The generator is fed with water pre-treated for the purpose of removing elements that contribute to scaling or corrosion, and the materials of construction are resistant to corrosion by steam that has no corrosion inhibitors. The dedicated PSG is very similar in design and construction to an SE still or the rst effect of an ME still. For further information, see Chapter 6 of this Guide.

7.7.2.1 Dedicated Pure Steam Generator

There are various designs of pure steam generators using the vaporization of feed water to produce the pure steam. Typically Typi cally,, vaporization is accomplished in a steam-to-steam evaporator, which can be of the vertical or the horizontal type, depending on the manufacturer and the overhead space available. The feed water disengagement space and the moisture separator may be housed in the same vessel as the evaporator or in a separate vessel. Sanitary construction includes 316L SS material, orbital Tungsten Inert Gas (TIG) welded wherever possible, or mechanically welded with the inner surface ground smooth after welding. Movable connections normally use in-line “sanitary” ttings. Flanges and threaded connections on the pure steam portion are not considered “sanitary” and are not recommended. For further information, see ASME BPE (Reference 12, Appendix 1). Heat exchangers, using plant steam as the heat source, including the evaporator should be of the double-tubesheet, tubular design to prevent the contamination of the pure steam by the heating medium. Most PS generators, except those with heated feed water or with lower capacity capacity,, normally are tted with feed water heaters. In addition, a blow down cooler typically is included to avoid discharge of very hot and ashing condensate. A feed pump may be required if the feed water supply pressure is inadequate. Depending on system design and the manufacturer,, a feed pressure of approximately 10 to 15 PSIG (0.7 to 1.0 BARG) above the maximum expected pure manufacturer steam pressure is required. This allows for pressure drop in piping and valves. A sample cooler tted with conductivity element is often used to monitor pure steam condensate purity purity.. This is an optional feature, the use of which should be decided based on need. Monitoring the conductivity of the condensate can alarm problematic operation and provide information regarding the suitability and applicability of the distributed steam for its nal use. Operating Principles of a Typ Typical ical Pure Steam Generator Pure steam normally is generated in a shell-and-tube heat exchanger like evaporator. Feed water is introduced on one side of the tubes, while the heating medium is introduced on the other side. Heating of the feed water to above the boiling temperature causes the water to evaporate, producing steam. The heating medium is normally plant steam, and does not come in direct contact with the feed water or with the pure steam. Clean Steam (CS) generators may be designed to use other heating medium, such as electric immersion heaters.




Pure steam pressure should be selected by the user. Typic Typical al units are designed for pharmaceutical applications at 30 PSIG to 60 PSIG (2 BARG to 4 BARG). Pure steam pressure is maintained by a standard Process and Instrument Diagram (PID) control loop, which modulates the supply steam control valve. The evaporator feed water is independently controlled using a level sensor and a control loop to start/stop a feed water pump or open/close a feed water supply valve. The feed water level is controlled to protect against ooding of the evaporator and carryover of endotoxins. A high level alarm and subsequent shutdown normally are incorporated into the design. The plant steam supplied to the generator, typically 80 PSIG to 120 PSIG (5 BARG to 8 BARG), should be at a higher pressure than the required pure steam pressure. In general, for a given size generator, the greater the differential between the plant steam pressure and the pure steam pressure, the higher the pure steam production rate. Plant steam pressure should be greater than 30 PSIG (greater than 2.0 BARG) higher than the pure steam pressure, to optimize the heat transfer and production rate. Plant steam consumption will be approximately 10% to 20% greater than the quantity of pure steam produced. Moisture entrainment separators normally are designed to function over an optimum range of steam velocity velocity.. Caution should be taken if the volume of steam increases substantially beyond design capacities. The increase in steam velocity associated associated with the higher capacity could result in carryover of endotoxins through the moisture separator. This condition also can exist if the steam pressure differential signicantly signicantly exceeds design conditions. Under these conditions, the velocity of the steam through the separator also may be excessive. Specication of the pure steam generator should be at the maximum output of the generator and at the highest possible pressure difference. An alarm and equipment shutdown is recommended and can be incorporated into the controls to protect against such conditions. 7.7.2.2 Pure Steam from Multiple Effect Stills

The rst effect of an ME still can be used to produce pure steam. The ME still may or may not produce pure steam when the still is producing WFI. A common practice is to increase the size of the rst effect to provide the required pure steam capacity for distribution to use points plus the quantity to supply subsequent effects of the still. This design is commonly used if simultaneous production of pure steam and WFI is preferred.

Advantages: •

Does not require a separate pure steam generator with the associated cost, space, installation, operation, and maintenance.

Disadvantages: •

Output may be limited to the capacity of the rst effect of the ME still.

•

May not produce WFI when using excessive quantities of pure steam. In an ME still, the steam generated in the rst effect becomes the motive (power) steam for the second effect, which in turn produces motive steam for the third effect, etc. The impact of utilizing an excessive quantity of pure steam higher than the sizing of the rst effect, therefore, drastically reduces the ability for WFI production.

The ME still manufacturer should be advised in advance if simultaneous production of WFI and pure steam is preferred.

7.7.3

Feed Water Pretreatment Feed water pretreatment is an important consideration in the design and successful operation of the pure steam generator. All generators are susceptible to scaling and corrosion if the feed water is not pretreated properly. The basic functions of the pretreatment are to minimize or prevent scale formation and to minimize or prevent corrosion. Additionally,, the pretreatment system can Additionally can remove objectionable objectionable volatiles, such such as ammonia, that are not removed removed in stills, and would carry over into the distillate and adversely affects distilled water quality. quality.




The pretreatment of feed water for a pure steam generator is similar to that of an ME still; therefore, the guidelines for feed water treatment for ME stills should be closely followed for pure steam generators. For further information, see Chapter 6 of this Guide. Additionally Additionally with some pretreatment systems, if the pure steam is to meet the requirement of British Standard EN 285 (Reference 15, Appendix 1) for non-condensable gases, then the pure generator may need to be tted with a feed water deaerator or heated break tank. Hot WFI or PW with Stage 1 conductivity or better, which has not been nitrogen blanketed, should not contain sufcient non-condensable gases to result in failure of the British Standard EN 285 test. 7.7.3.1

Common Practices for Feed Water

It is common practice to generate pure steam from water that has been treated to remove impurities, which may be detrimental to the operability of Pure Steam Generator (PSG). Many manufacturers recommend the removal of particulates, chlorine/chloramines, chlorine/chloramines, and hardness ions as a minimum; however, additional levels of water treatment (e.g., reverse osmosis, deionization) usually are utilized to aid in the removal of impurities and extend the mechanical reliability of the pure steam generator. A common common practice for supplying the treated water is to use the same water treatment system as the facility’s high purity water system. Water is supplied from the treatment system following the treatment step required by the PSG manufacturer or desired by the owner. Another common practice for supplying water to a pure steam generator is to use the facility’s high purity water system (e.g., PW, PW, WFI). (Note: the pure steam generator does not require feed water meeting PW or WFI). This practice, however, however, ignores the ability of the pure steam generator to remove impurities. A common application for this practice is when the steam quantity is small and the cost and maintenance of a dedicated feed water system exceeds the cost of using high purity water.

7.8

Steam Attributes and Condensate Sampling

7.8.1

Treatment of Plant Steam It may be necessary to lter or condition plant steam. In specic applications, it also may be necessary to change the steam boiler treatment and substitute additives that do not contain amines or hydrazine. Given that the type and degree of conditioning are dependent on the application, as well as on the quality of the plant steam and additives present, this Guide all possible scenarios are not addressed. Prior to the elimination of amines and hydrazines, by the substitution for standard boiler pretreatment additives, the plant steam boiler manufacturer should be consulted regarding the impact on equipment warranty, performance, and expected life. Substitute additives may not be as effective as those normally used.

7.8.2

Pure Steam Purity Sampling Purity requirements for steam used in pharmaceutical manufacturing and product development normally are driven by the product characteristics, manufacturing process, and the intended use of the product. The drug product manufacturer is responsible for ensuring that steam used to process the product is appropriate. When required by the process, the steam purity should be monitored through acceptable sampling techniques. A slipstream slipstream of the steam may be passed through a sample condenser/cooler, condenser/cooler, tted with a sampling valve. Steam should not contribute to drug product contamination; sampling should be included during commissioning and qualication per established regular/periodic procedure, and prior to each occasion on which the steam is used. If the sampling requirement is for endotoxin or pyrogen testing, the sample cooler, tubing, and valve should be of sanitary construction.




Sample coolers can be tted to the PS generator, or located in the distribution line, or at the POU (recommended location), or a combination thereof. It is normal practice to t a sample cooler with conductivity monitors and alarms on the outlet of the pure steam generator. Additionally Additionally,, sample locations typically are located at sterilization equipment and at the farthest points in the distribution system. These sample locations can have permanently installed sample coolers or a portable cooler can be utilized. Endotoxin removal: the removal: the condensate sample from a pure steam generator with separator is expected to show 3-4 log10 level reduction in pyrogens compared to the level in the feed water (similar to reduction in WFI stills). Though steam purity requirements are product specic, it may be impractical to produce special steam for each situation. Manufacturing operations typically typically generate and distribute only one or two steam purity grades. Although, for a given application, the condensate of pure steam may not be required to meet WFI attributes, it is important to note that, as a rule, if the condensate does not meet the attributes, the generator design/operation or distribution system should be evaluated.

7.8.3

Pure Steam “Quality” Sampling The term “quality” when referring to steam indicates the level of steam saturation or dryness (ratio of the vapor mass to the mass of the steam mixture) and the amount of non-condensable gases in the pure steam. The quality of the pure steam should be established by the application and as required by the applicable regulatory guidelines. Dry saturated steam with minimum superheat is necessary for efcient steam sterilization. Regulations for pure steam quality are commonly applied to sterilization applications due to the international nature of the industry. British Standard EN 285 (Reference 15, Appendix 1) and other international regulatory requirements commonly are adopted for pure steam systems globally. Water can be generated and carried by steam within distribution systems in two ways: 1.

in suspension as moisture when the steam quality is less than 100%

2.

as condensate separated from the steam

Water vapor carried in suspension may be reduced by: •

reducing the pressure

•

reducing the velocity

•

adding a steam entrainment separator

Steam traps may reduce water moisture and condensate. The installation of a properly designed condensate trap is recommended immediately up stream of steam quality testing locations, such as prior to sterilization equipment. Steam “quality” sampling may be employed to determine the level of saturation, superheat, and non-condensable gasses. This can be determined by applying a steam calorimeter and measuring the dryness or saturation level. A steam calorimeter measures the percentage by weight of steam in a mixture of steam and entrained water. British Standard EN 285 (Reference 15, Appendix 1) provides guidance on steam quality test methods.

7.9

Materials of Constructio Construction n

7.9.1

Materials of Construction for Plant Steam Conditioning and Distribution Chemical compatibility with the plant boiler generated steam and the carried over feed water chemicals normally are required for all materials used to condition the contaminated steam.




Based on the particulate levels in the steam and the required steam purity purity,, more than one ltration stage may be utilized. Distribution of plant steam following ltration follows similar practices as PS to control condensate build up, noncondensable gases, and saturation levels as required for the application. Acceptable materials should be relatively inert and may include SS or tin-coated copper.

7.9.2

Materials of Constructi Construction on for Pure Steam Generators Structural integrity and chemical compatibility with the contact uid and its constituents are two of the more practical issues that drive construction material selection for PS systems. The intrinsic corrosion potential forces pure steam generator manufacturers to consider relatively inert metals, including stainless steel or titanium. Sanitary piping and valves, considered unnecessary for plant steam boilers, often are standard features for PS systems based on the specic manufacturer and model. The materials chosen should not contribute to contamination of the drug product. The most commonly used material of construction is 316 and 316L SS, except for the plant steam supply piping normally is carbon steel, and so are the skid and structural. Chlorine and/or chlorides combined with water will damage the stainless steel generator and distribution piping regardless of the nish; appropriate selection of chloride free insulation is recommended. Insulation also should be rated for the appropriate temperature to prevent material breakdown, which may contribute to chloride-induced corrosion of stainless steels.

7.9.3

Surface Finish for Pure Steam Generation Systems Mechanical polishing, electropolishing, and passivation processes are commonly implemented in stainless steel pure steam systems. The operating temperatures of pure steam systems are sufcient for inhibiting microbiological growth. Typical Typi cal mechanical polishes (35 to 20 Ra) usually are acceptable for nal nishing of generators and distribution piping. Additionally, Additionally, the internal surfaces, including mechanical welds, are treated with nal passivation to improve the corrosion resistant chromium oxide barrier. Electropolishing Electropolishing of the internal surface also can be used, but may not be necessary because of operating temperatures and the inherent formation of rouge. For further information, see Chapter 10 of this Guide.

7.10

Distribution Distribution systems for pure steam follow the same engineering standards commonly used for plant steam; however, contact materials should be inert to the aggressive aggressive nature of pure steam. Corrosion-resistant Corrosion-resistant 304, 316, or 316L grade Stainless Steel “tubing” or solid-drawn “pipe” normally are used. Surface nish is not critical because of the self-sanitizing nature of the pure steam. Mill nish or mechanically polished pipe or tubing usually is sufcient. Orbital welding and post-installation passivation is considered appropriate for this application. Piping should be designed to allow for thermal expansion and to drain condensate. Piping should include minimum slope in the direction of steam ow to a low point condensate trap for adequate removal of condensate. Note: the sloping of distribution piping in opposition to the direction of ow is not recommended because of the possibility of condensate collection and moisture entrainment. Sanitary high pressure clamps or pipe anges are frequently used where the piping should be joined, but welded connections are used where possible to eliminate safety concerns, maintenance costs, and the potential for leaks. Threaded connections may be suitable for instrumentation if positioned to drain condensate and remain hot; however, sanitary clamp clamp connections are normally preferred. Ball valves are commonly used for isolation. The use of diaphragm valves should be closely considered because of the limited ability of the elastomeric diaphragms to hold up well in this service. Where diaphragm valves are used, PTFE-encapsulated EPDM or Viton diaphragms usually provide the best long-term performance.




Steam quality sampling should be included in the distribution design, and sample locations should be designed with condensate traps, entrainment separators, and vents. (Condensate trap maintenance cannot be over emphasized because of the small orices required in the separation of gas and liquid.)

7.10.1 Line Sizing Steam distribution headers are normally sized for a maximum velocity of 120 ft per sec. (37 m per sec.) to limit pressure drop and extend the life expectancy of the piping. Condensate line sizing should follow industry standards for utility condensate. A pure steam system including the pure steam generator normally is not designed to provide the connected load requirement. A diversity diversity factor should be established for the pure steam system consistent with the usage prole; however, consideration should be given to changes in the usage prole, otherwise increased usage may cause difculties. A common problem in established pure steam systems is the addition of equipment without consideration of maximum supply capacity of the distribution piping. The size of the distribution system should be evaluated along with usage diversity when new equipment or points of use are added. Designing pure steam main headers and main branches of a distribution system at lower velocities (e.g., 100 fps) can compensate for future connected loads or decreasing diversity.

7.10.2 Point of Use Design Designs for a pure steam POU normally include an accessible isolation ball valve and an appropriately oriented condensate trap. The supply piping to the POU valve typically is designed as a branch of piping extending from the top of the distribution main to the condensate trap. The condensate trap usually includes an isolation valve, for trap maintenance, and outlet piping to a drain. The condensate drain piping should include an air gap separation to the oor drain. Consideration should be given to the temperature and the aggressive nature of condensate verses the drain piping material. The POU valve should extend from the top of a horizontal run of the branch pipe prior to the trap in order to prevent moisture entrainment at the POU valve.

7.10.3 Instrumentation Monitoring the pressure of the pure steam system is considered essential for proper operation and quality quality.. Pressure monitoring is recommended at the pure steam generator as well as critical process locations on the distribution system. The pressure at the extents of the distribution system tends to drop if the system is poorly designed or inadequately sized. The drop in pressure corresponds to periods of high usage. Suitable pressure monitoring can provide information for establishing usage diversity for a system. It is considered unnecessary to use a feed water conductivity monitor when the feed is from a USP puried or WFI source.

7.10.4 Moisture Removal Water vapor forms in steam systems because of heat loss, causing a change in the liquid/vapor ratio or steam “quality.” Steam may be dried by reducing the pressure just prior to the point of use to coincide with the steam temperature of saturation at the reduced pressure. Moisture entrained in the steam also can be removed by installing an in-line separator just prior to, or just after, the pressure control valve. If the moisture separator is located upstream of the pressure control valve, the piping should be designed to protect the valve from condensate damage. A sudden sudden line enlargement combined with a change in ow direction and a condensate trap also can be used for moisture removal.




In-line separators remove moisture with a series of bafes on which the suspended water droplets impinge and fall out by gravity to the outlet trap, which should be piped to drain. Separators commonly have a separation efciency of better than 99% in the removal of condensate.

7.10.5 Condensate Removal Condensate is the water that has separated from the liquid vapor mixture and forms in steam systems because of heat losses and natural separation effects. effects. Lines should be slopped in the direction of ow to a trap to prevent the buildup of condensate. The proper location of condensate traps can prevent dangerous water hammer and eliminate branches where condensate is allowed to collect. Collected condensate has the potential to cool and could cause microbial issues. Any Any untrapped vertical length of pipe can quickly ll with condensate. If this condensate is allowed to stand for sufcient time, it can cool and become a breeding ground for bacteria. This bacteria could possibly be entrained back into the main distribution header and contaminate use points downstream. The following practices are commonly employed to minimize these concerns: •

Each line is adequately slopped/supported in the ow direction to avoid sagging and subsequent condensate accumulation.

•

Steam traps are installed at points where condensate can collect (e.g., at least every 100 feet (30 meters), upstream of control valves, at the bottom of vertical risers). Steam traps used for pure steam service should be sanitary design and self-draining.

•

If the main distribution header is above the use points, the branches to the users should be routed from the top of the header to avoid excessive condensate loads at the branch. Each branch should be trapped to avoid condensate buildup.

•

An alternative is to run the main distribution header below the use points. The branches can drain back to the main distribution header, avoiding the need for additional traps; however, this design can result in condensate being discharged at the POU due to entrainment in the high velocity steam.

7.10.6 Non-Con Non-Condensable densable Gas Removal Air and other non-condensable gases should be minimized in pharmaceutical steam systems. Since air acts as an insulator,, incomplete sterilization can occur in the process. Air in a system offers a highly effective barrier to the insulator heat transfer, which may lead to a reduced temperature at the surface of a tube, system component, or process equipment. Air is heavier than the steam in a distribution system; therefore, air can be discharged using steam traps; however, excessive levels may slow down the discharge of condensate. The sub-cooled condensate can cause insufcient sterilization temperatures because of the excess water. The removal of air can be achieved by placing thermostatic steam traps at appropriate locations throughout the distribution system. These These should be placed in positions where air is prone to collect, such as the terminal points of the main and large branches of the steam header. In the case of air and condensate discharge at the bottom of large vessels, the air and condensate should be separated by suitable piping practices.


8

Storage and Distribution Systems




Page 107 Storage and Distribution Systems

8

Storage and Distribution Systems

8.1

Introduction Appropriate design and operation of both the storage and distribution systems is critical to the success of a biopharmaceutical or life science water system. There are considerable numbers of designs for storage and distribution systems. systems. The differences in these designs are directly correlated to the user requirements and the manufacturing requirements. This chapter provides an overview of common storage and distribution congurations, noting advantages and disadvantages of each alternative. Comparisons are provided for the alternatives and a decision ow chart is provided to help decide which alternative system best suits the operating requirements. Microbial considerations are discussed, as they apply to alternatives for continuous and periodic sanitization of storage and distribution. Typical Typical system components also are discussed, along with alternative materials of construction, and common installation practices for storage and distribution systems.

8.2

Purpose The purpose of a storage system is to accommodate demand peaks without the necessity of over sizing the purication system. A distribution system should transport the water to use points at the required ow, temperature, and pressure. Additionally, Additionally, the storage and distribution system should not allow the water quality to degrade below the appropriate quality for its designated end use. Storage allows a smaller, less costly water purication system to meet peak usage demand while enabling operation closer to the ideal of continuous, dynamic ow. Large manufacturing sites, with systems serving different different buildings, may use storage tanks as a means to separate sections of the system and to minimize potential cross contamination. For sites that require continuous operation, an appropriately sized storage system will allow the distribution system to remain in service and supply the users, while maintenance activities are performed on the purication equipment.

8.3

System Components This section is intended to review many of the primary components that might be found in a typical storage and distribution system. It is not meant to be limiting or all-inclusive, but rather to provide general information regarding the most common components. Storage and distribution system design is discussed; however, the trend in laboratory systems is to use non-metallic materials of construction. For further information, see Chapter 9 of this Guide. As technology has improved over the years, many design features, such as storage at elevated temperature, constant circulation, or use of sanitary components, including polished tubing and diaphragm valves, along with orbital welding and frequent sanitization, have become commonplace. To To incorporate all of these features into each new design typically leads to ever escalating costs with possibly minimal reduction in risk of contamination. Although Although each of these features provides improvements to the system, it may be unnecessary to use all of them in every system. A reasonable approach is to use design features that provide the greatest reduction in risk to operational and microbial issues at the most reasonable cost. The idea of selecting design features based on “return on investment,” where “return” is dened as reduction in risk, can be very helpful in controlling system cost and in evaluating different alternatives. The ability of a system to deliver water of the required quality may be used to determine the success of each design application. The challenge for the design engineer is to determine which features to include, achieving the required water quality with the lowest life cycle cost.



8.3.1


Tanks When properly designed, storage vessels can offer a number of advantages over tankless systems, including reserve capacity during a purication outage, atmospheric air break for loop return, or service of the upstream water purication equipment, as well as minimizing purication system instantaneous demand capacity. Careful consideration should be given to sizing, based on various factors including associated costs. The storage vessel also may be used as a contact tank for sanitization using ozone. Potential disadvantages of storage tanks are the capital cost and the cost of associated pumps, vent lters, and instrumentation; however, however, this is usually less than the increased cost of water purication equipment sized to handle the peak demand in the facility. Storage tanks may introduce regions of slow moving water inside the tank, which can promote loosely attached biolm; therefore, sanitization and continuous circulation of the water in the storage tank is important. The formation of loose biolm can be further reduced with the use of a tank spray ball to allow for continuous water turnover and wetting of the top interior surface of the tank. Generally, tanks are located near the purication equipment for ease of maintenance and to minimize cost; however, Generally, alternate locations also may be suitable. Utility areas are acceptable for this purpose, if maintenance accessibility is provided, the area is suitably maintained, and personnel access is controlled.

8.3.1.1

Tanks – Sizing and Capacity Criteria affecting storage capacity include the user’s existing and future demand prole or the amount of use, duration, timing, diversity (in the case of more than one user), and whether the system is circulating or noncirculating. Careful consideration of these criteria is necessary to optimize cost and water quality quality.. The storage tank should provide an adequate reserve to minimize cycling of the purication equipment and to offset maximum water usage. Average usage over time should be used to approximate the purication system capacity. capacity. The storage tank also should enable routine maintenance and orderly system shutdown in the event of an emergency, which can vary from a few to many hours, depending on the size and conguration of the system and maintenance procedures. In heated systems, the use of a spray ball serves to keep the top of the tank wetted and near the same temperature as the water. The use of spray balls for high purity water storage tanks may not require the stringent cleaning requirements used for CIP applications, especially in hot systems. Tank Tank spray balls may be placed on: •

the return of the distribution loop

•

a slipstream from the discharge of the distribution pump

•

a separately pumped recirculation loop

Connections on the top head (relief devices, instrument connections, etc.) should be kept as close to the head as possible to simplify the spray ball design and obtain the benet of the spray action. An exception is the vent lter, which should be removed far enough from the storage tank to avoid direct contact from the water spray spray,, which could blind the lter. If dip tubes or instruments project down from the head, multiple spray balls may be needed to avoid a “shadow” being created in the spray pattern, and vacuum-break holes may be required. For ozonated systems, spray balls typically are not used, as the agitation increases off gassing. As an alternative, return ow should enter the tank below the liquid level, in an appropriate location to maximize ozone contact. The tanks should be vented to allow lling, and emptying (especially for non-pressure or vacuum rated vessels), and a lter should be used at the vent to avoid airborne particulate and microbial contamination.




If the change in conductivity resulting from absorption of CO 2 into the water proves problematic, a number of options can be considered, including polishing deionization, blanketing of the tank head space with an inert gas (e.g., nitrogen), using a CO2 absorber with the vent lter, or ushing/replenishing with fresh water. Gasses Gasses added to storage tanks should be appropriately ltered to avoid objectionable contamination and typically, typically, require an uninterruptible source. 8.3.1.2 Tanks – Design Considerations Vessels should be designed using appropriate standards and installed such that they do not compromise the water Vessels quality or system operation. Additional information information can be found in the ASME BPE (Reference 12, Appendix 1). Horizontal tanks may be used to address space issues; however, vertical orientation is the most common based on the following advantages: •

lower cost

•

less dead volume

•

simpler spray ball design

•

less oor space required

The turnover rate may be important for systems using chemical sanitization or polishing equipment. Storage tank turnover is required to minimize low ow areas. The turnover rate is less important when storage is under continuous sanitizing conditions, including hot storage or ozone. It also may be less important under conditions that limit microbial growth, such as cold storage (4°C to 10°C). For circulating systems, tank design may include an internal spray ball to ensure that all interior surfaces are wetted and as part of the microbial control strategy. strategy. Horizontal tanks may be necessary if overhead space is limited. With a horizontal tank, there should be no unwetted surfaces during operation and multiple spray balls may be required. Tank jacketing or an external heat exchanger may be necessary in hot systems to maintain water temperature over long periods without makeup. Alternatively, Alternatively, these features may be used to cool high inuent temperatures to preclude excessive rouging or pump cavitation. The maximum size of a single storage vessel often is limited by the space available in the facility or by structural loading. Multiple tanks may be required to obtain the desired surge capacity or the desired maintenance redundancy. redundancy. In this case, interconnecting piping should be designed carefully to ensure adequate ow through all supply and return branches. In addition, if elevations require lower than desirable pump suction head, vortex breakers may be required.

8.3.2

Pumps Sanitary centrifugal pumps are commonly employed in high purity water distribution systems. Performance curves and suction head requirements should be reviewed to preclude cavitation, which could lead to particulate contamination and corrosion of pump casing and impeller impeller.. The generation of pump heat over extended periods of low or no draw also should be considered, since signicant temperature rise in cold systems or cavitation caused by vapor pressure in hot systems could occur. Where the pumps are at the low point of the distribution, casing drains allow for full system drainage. The installation of dual pumps, for standby purposes, may be considered if properly designed to avoid dead legs and suitable pump switching procedures are implemented. Additional information can be found in ASME BPE (Reference 12, Appendix 1).




8.3.2.1 Pumps – Sizing and Capacity High purity water distribution pumps should be designed to deliver the required capacity and pressure to support usage. Typically Typically,, this includes a usage diversity factor, i.e., only a percentage of the use points may require ow at any one time. The ow and pressure capacity of a distribution system should meet adequately the diversied usage with capacity to meet minimum return velocity and pressure during maximum usage and maintain fully ooded piping. Centrifugal pumps usually are selected for operation at less than the maximum pressure and ow capacity; using variable frequency electrical supply or back pressure control valves for operation at system requirements. A variable frequency electrical supply controlled by system pressure or ow can be used to modulate closely against variable usage and to optimize energy consumption. Similarly, a pressure control valve modulating on system pressure can impose pressure on a constant speed pump, which will allow similar performance at slightly higher energy costs. Distribution systems often are hydraulically modeled to closely predict performance under various usage conditions. Hydraulic modeling of distribution systems can be used during the design process to evaluate the design against user requirements. Additionally Additionally,, in existing distribution systems, a hydraulic model can be used to evaluate modications or additions that may affect pump performance. 8.3.2.2

Pumps – Design Considerations Common practice is to design circulating loops based on: •

ow

•

pressure

•

temperature

•

velocity

Pump design should include considerations for excessive pressure surges (i.e., water hammer) and adequate suction pressure to avoid cavitation. Pump seal materials should be suitable for the application with consideration given to wear and particle generation. Pumps equipped with double mechanical seals, using product water for ushing uid, as well as external single mechanical seals usually are used for high purity water applications and may minimize the possibility of contamination. The pump casing and impeller generally is recognized as a predominant area for the formation of rouge. For further information on rouge, see Chapter 10 of this Guide. This is speculated to be caused by the highly turbulent conditions inside the pump head. 316L Stainless Steel or higher alloy liquid end components usually are used and may reduce rouge formation. Inclusion of a casing drain along with a pump discharge orientation at 45° is common for draining and venting a pump. Alternately Alternately,, a casing vent can be included along with the casing drain for pumps requiring alternate discharge. Suction issues should be reviewed based on height requirements necessitated by the casing drain. Given that distribution pumps require routine maintenance, the design of pharmaceutical water systems may include redundant or stand-by pumps. Designs may include these secondary pumps in a primary and stand-by conguration with one pump always running and the second pump r eady for operation in stand-by service. The stand-by pumps can require ushing or sanitization prior to service or can be included in the continuously circulating system for on demand operation without sanitization. Conversely, Conversely, if it is acceptable for the system to be sanitized following pump change out, use of an uninstalled spare may be considered.




Variable frequency drives controlling the pump motors speed may be used to provide exibility in the operating conditions of the pump. Variable frequency drives may be used to control pump motors to allow for the pump speed to vary based on the operating conditions. This feature may result in slightly increased capital cost, but reduces operating and maintenance cost because the pump will only operate at the speed required to meet the system demands.

8.3.3

Vent Filters Vent lters are used on high purity water storage vessels to serve as a particulate and microbial barrier between the surrounding environment and the water. The headspace of the storage tank takes in and expels air proportional to the rise and fall of the water level. Vent lters suitable for microbial retentive gas ltration, typically are constructed of compatible materials, such as hydrophobic PTFE, PP, PP, or PVDF to prevent wetting. Filters should be capable of withstanding sanitization (i.e., heat, ozone, chemicals) and sized based on vessel design characteristics (pressure and vacuum rating), and system maximum ll or draw down rates. Areas of concern for gas ltration include: •

improper sizing

•

blockage of the lter caused by condensed water vapors

•

leakage caused by improper lter installation

•

materials incompatible with water system sanitization methods

•

defective lter cartridges

The lter cartridges should be specied consistent with the lter housing. Vent lters for compendial water storage tanks should be integrity tested following installation in the housing, but may not require to be validated as sterile lters. Testing Testing and visual inspections should be performed, including integrity testing in the housing at the end of the service life. Integrity testing testing is used to verify that the lter is not plugged or is not leaking at the time the lter is removed from service and to provide assurance that the preventative maintenance schedule is appropriate. Common lter integrity testing methods include bubble point, diffusion, and water intrusion, which vary based on the lter manufacturer, application, application, and type of cartridge. Adequate procedures should be developed for possible failure of lter testing following removal from service. Additional information can be found in ASME BPE (Reference 12, Appendix 1). 8.3.3.1 Vent Filters – Sizing Sizing and Capacity Capacity Microbial retentive hydrophobic vent lters typically have a 0.2 to 0.22 µm rating. When sizing vent lters, the lter should allow gas to enter or escape at a rate comparable to the water entry or removal to avoid creating a vacuum or positive pressure which could exceed the vessel rating or may impact operation of other devices such as rupture disks, level controls, or ozone injectors. For high ow rates, multiple vent lters may be required to avoid exceeding the recommended maximum ow rate or pressure drop across the lter. The expected life of the vent lter also will vary based on the operating conditions. The ow, differential differential pressure, and life expectancy values of lter cartridges typically vary depending on the air temperature, materials of construction, and lter manufacturer. manufacturer. For example, a lter cartridge that is periodically heat sanitized could last longer than one operated continuously at elevated temperatures.




8.3.3.2 Vent Filters – Design Considerations The design of vent lters for high purity water storage tanks can be as simple or as complicated as required to satisfy system operational requirements. This can range from a single disposable cartridge/housing combination to parallel heat-traced housings with in-situ sanitization and integrity testing capability. capability. Consideration may be given to installing redundant parallel lters or valves that will allow for lter replacement and integrity testing without exposing the contents of the tank to the environment or reagents. The design of the vent lter(s) depends signicantly upon the operating parameters of the associated storage tank and the exibility for periodic maintenance. Pressure safety devices (i.e., rupture discs, etc.) should be used for vessel over/under-pressurizat over/under-pressurization ion protection, because vent lters can plug during operation or valves installed on vent lters could be inadvertently closed. In addition, tanks that are rated for appropriate pressure and vacuum offer the safest and most trouble free operation. Typically Typi cally,, only over pressure protection is used for tanks rated for full vacuum. Where hydrophobic lters are used, care should be taken to prevent wetting of the lter. If the tank contains a spray ball, the vent lter should be located on a tank nozzle, such that the spray will not wet the lter, and where it is easily accessible for maintenance. Consideration also should be given as to the source of air for the tank. It is desirable to avoid locating the vent lter in a location where organic vapors or other undesirable vapor sources could enter the tank causing contamination or damaging the lter element. Control measures to avoid accumulation of condensed vapors in the vent lter housing include electrical or steam heat tracing and a self-draining orientation. If heat tracing is used, the temperature should be set above the dew point temperature of the vapor space and below the manufacturer’s maximum temperature rating of the cartridges. An alternate practice is to blanket the tank with ltered air or nitrogen. If a lter is used only for the supply of air or nitrogen to the storage tank, a condensation control measure may not be required.

8.3.4

Heat Exchangers Heat exchange equipment is used to heat or cool high purity water to desirable levels or to maintain temperatures. Design congurations often include: •

shell and tube

•

plate (also called plate and frame)

•

jackets applied to tanks/vesse tanks/vessels ls

Within each group there are design variations that may may,, when properly applied, improve reliability or reduce risk. Tube and shell heat exchangers typically are constructed of an outer cylindrical shell, with one or more smaller tubes installed inside. High purity water ows through the inner tube(s), while heat transfer media contacts the exterior, contained between the product tube exterior(s) and the outer shell. Plate heat exchangers are constructed of specially formed “plates” that when assembled create a “stack” with precise gaps between each plate. The stacks usually are mounted within a support frame, hence the alternate name “plate and frame.” High purity water and heat transfer media ow through these gaps in an alternating arrangement. Tank jackets typically are constructed by adding a secondary layer to the outer wall of the vessel. The gap that is created between the outer wall of the vessel and the jacketing is the space through which heat transfer media travels. Jackets can be applied to the sidewall and heads of a vessel and can be congured in a variety of ways. Additional information can be found in ASME BPE (Reference 12, Appendix 1).




8.3.4.1 Heat Exchangers – Sizing Sizing and Capacity Capacity Heat exchangers are available in a large number of standard and custom sizes. The capacity of tube and shell, or plate, heat exchangers is determined primarily by the size and number of tubes or plates. The capacity of a tank jacket is limited limited by the tank tank dimensions. Tube and shell, shell, or plate, heat exchangers often can be congured by the manufacturer to suit an application, such that the size and number of tubes or plates can be selected for a best “t.” The details of the heat transfer media will have a signicant impact on the heat exchanger sizing, along with the thickness of the materials of construction and resulting heat transfer coefcient. Sizing can be based on a oncethrough design or on a batch basis, as needed for the application. A once-through design will typically result in a larger unit with higher utility demand requirements. Utility systems that provide heat transfer media including overall capacity,, demand cycles, supply and return temperatures, pressures and ow rates should be considered. capacity 8.3.4.2 Heat Exchangers – Design Considerations Heat exchange equipment that will be in contact with high purity water should be specied with appropriate materials, nishes, and connections, etc. The principal concern with properly sized heat exchangers is the possibility of contamination from the heat transfer media into the high purity water. This possibility is compounded by the thermal expansion typically endured by heat exchangers used in systems with signicant temperature variations, or if used for both heating and cooling applications in the same exchanger. Thermal Thermal impacts on related utilities (e.g., the coolant in a cooling exchanger during sanitization) also should be considered. A variety of schemes to minimize this risk can be employed. One method is to ensure there is a higher pressure on the product side than on the media side, which requires the inclusion of pressure monitoring devices, as well as procedures for ensuring proper operation; however, this technique may not be able to overcome venturi action if a leak develops in an area of high velocity. Another method is the use of double tube-sheets for tube and shell exchangers. Double tube-sheets with an air gap between reduce the risk of contamination if the leak occurs at the tube-to-tube-sheet seal; however, this feature does not eliminate the risk if a tube ruptures. Conductivity sensors, installed at the outlet of exchangers, commonly are used as a method of detecting a tube rupture and the entrainment of media in the high purity water. Alternatively, double wall heat exchangers can further minimize the risk of contamination, especially if other scenarios Alternatively, described in this section of the Guide are employed. Double wall heat exchangers employ an added layer of material to further isolate the media from the product. These designs can be effective, but usually at a signicant loss of efciency.. For example, in the simple tube and shell heat exchanger with a single tube within the shell, the product efciency is contained within the inner tube while the media is contained between the shell and the inner tube. Employing a double wall design would require a third concentric tube, where the product remains within the inner tube, but now the media is contained between the shell and the intermediate tube. The gap between the intermediate tube and the inner tube serves as a barrier in the event of a leak and is vented to provide a method of detection. For plate heat exchangers, the plates can be doubled to accomplish the same goal. Tube and shell heat exchangers can be made completely drainable, either using slope or for “U” tube (multi-pass) units by adding weep holes at the low point of each chamber. Plate heat exchangers typically are more compact and also can be less expensive and more efcient; however, they are usually not fully drainable making them less desirable for many applications. (Gasketed units can be drained with full disassembly disassembly.) .) Appropriate construction is important including proper welding; for double tube-sheet units, the inner tube-sheet should be torque-rolled while the outer is normally welded and polished. Appropriate space for service is required. It should be noted that inclusion of all design features would be unlikely to remove all potential scenarios for contamination. An evaluation of the overall installation of any exchanger to ensure safe and proper operation, including the use of pressure relief devices as necessary, should be conducted. Sanitary heat exchangers also can become part of an “economizer” circuit to reclaim energy where appropriate and justiable, provided proper design standards are applied.



8.3.5


Piping/Tubing/Fittings/Valves Piping and Tubing: extruded Tubing: extruded seamless and longitudinally welded tubing (ASTM A270) (Reference 20, Appendix 1), commonly is used. Longitudinally welded tubing is similar to seamless in appearance and performance and can be signicantly lower in cost. Thermoplastics such as PVDF PVDF,, ABS, and Polypropylene (PP) have been shown to be viable alternatives with proper design. System temperature, extractables, and chemical compatibility are a concern when using thermoplastics thermoplastics.. Fittings: detailed information on sanitary ttings can be found in ASME BPE (Reference 12, Appendix 1). Sanitary Fittings: detailed clamp connections in a range of designs are available in a large number of sizes and are used frequently, but usually are minimized because of microbial and maintenance concerns. Valves: diaphragm valves predominantly are used in high purity systems because of their enhanced ability to be Valves: diaphragm sanitized. Other types of valves, such as ball valves and needle valves that have internal areas which harbor water, typically are not used. Additional information can be found in ASME BPE (Reference 12, Appendix 1).

8.3.5.1 Pipe/T Pipe/Tube/Fittings ube/Fittings – Sizing /Capacity The ow of water in piping should be turbulent to control biolm development on the piping wall and to allow for thorough mixing throughout all components of the distribution system (See Chapter 13 of this Guide). Turbulent ow can be achieved at relatively low velocities and is dependent on water temperature and diameter of piping. The capacity of the system should be such that the desired amount of water can be used at all times without risking the loss of return pressure. The size of the piping is then decided based on maximum pressure drop at maximum ow in the system. Designing just to obtain a velocity in the piping of at least of 0.9 m/s or 3 feet/s for a pharmaceutical water system is considered outdated. 8.3.5.2 Pipe/T Pipe/Tube/Fittings ube/Fittings – Design Considerations The installation of a piping distribution system that is fully drainable is a common industry practice, but not a GMP requirement. Systems that will be steam sanitized should be fully drainable to assure complete condensate removal. Systems that are not steam sanitized sanitized may not need to be fully drainable, as long as water is not allowed to stagnate in the system. These systems should be continuously circulating or water should be evacuated during maintenance and shutdown by mechanical means. It is considered common practice to allow for the complete draining of equipment and associated piping, facilitating system maintenance or system ushing. Drain points in non-controlled areas should be evaluated based on a possible risk of contamination from uncontrolled air entering the system. Vent Vent valves also may be needed at the high points in complex or larger systems to allow ltered air to enter the piping system to aid in complete drainage. In addition, it may be acceptable to include partial disassembly of the system to allow complete draining. Deadlegs should be minimized or eliminated where possible. A turbulent condition may be maintained in short dead ended pipe stubs, if the length of the stubs is limited. Thorough mixing is desired at these locations to facilitate sanitization. This limited length varies with the pipe stub diameter and to a lesser degree with the main pipe diameter diameter.. A specied specied minimum deadleg may be difcult difcult to achieve achieve in large mains with small branches, branches, and may result result in unacceptably long deadlegs in large branches. Rather than universally applying “deadleg rules,” it is important to recognize deadlegs as areas of concern and take appropriate steps to prevent them in the original design or implement special provisions to address them if unavoidable. Factors to consider include: •

operating temperature




•

velocity in the main

•

frequency of use each outlet

There are various regulatory guidance documents that limit deadlegs to less than 6, 3, 2, or even 1.5 branch pipe diameters. Originally, Originally, the 6D rule contained in the FDA Guide to Inspections of High Purity Water Systems of 1993 (Reference 3, Appendix 1) describes describes the distance from the center line of the pipe to the end of the deadleg. More recently,, some industry guidance documents suggested using a guideline of 2D or less. Depending on where recently measurements are taken; however, the discrepancies in the two guidelines can cause confusion. This Guide recommends that the length of the dead legs be minimized; a deadleg of any distance is problematic, if it results in a microbial issue. Operations to control microbial concerns, such as periodic ushing for known sanitizing durations or other remediation activities, may be considered. However, it should be recognized that a one-way pipe may not constitute a deadleg if it is continuously used, frequently ushed, or frequently sanitized.

8.3.6

Instruments Appropriate instrumentation should ensure proper operation of a high purity water system and to provide a means for obtaining suitable data necessary necessary to document operation. Components Components may range from local visual visual indicators to devices capable of integration with electronic systems offering control, alarm, trending, and more. Typical Typical operational parameters that should commonly be monitored by instrumentation in a pharmaceutical water system include temperature, pressure, tank level, and conductivity conductivity.. Additional parameters can include ow, pump speed, and TOC, depending upon the level of sophistication and monitoring desired. Critical instrumentation that is installed should be calibrated and included in a preventive maintenance program to ensure the reliability of the data that is provided. Instrumentation should be installed such that it is accessible for data gathering or maintenance. Instrumentation used for quality release of the water may require additional quality assurance oversight. Additional information can be found in ASME BPE (Reference 12, Appendix 1).

8.3.6.1 Instruments – Sizing and Capacity Monitoring components should be appropriately sized; accuracy at normal operating ranges, as well as extremes of operation caused by sanitization activities that may result in excessive pressures, temperatures, and/or chemical/ ozone concentrations should be considered. Concerns include material compatibility (i.e., temperature and chemical/ ozone) and physical compatibility (i.e., steam sanitizing of a pressure-based level transmitter). Pressure monitoring devices should be selected with consideration given to water hammer effects which may occur in a system as a result of closing valves or abrupt starting of a distribution pump. 8.3.6.2 Instruments – Design Considerations Design considerations should include instrument location to assure proper performance and contamination avoidance, as well as accessibility for removal, service, and calibration. Pressure monitoring devices should be of sanitary design and should include a diaphragm barrier to isolate the internal components of the device. The recommended installation conguration for temperature and conductivity elements is opposing the water ow. This assures continuous contact with the water and proper ushing of the area around the element. Additionally, Additionally, instrument connections can be oriented horizontal to the main pipe to avoid trapping air. This is particularly important for conductivity elements, which can give false reading in the presence of air. High purity water storage tanks typically are designed to include differential pressure as a means to monitor level; however, there are several commonly used sanitary methods of monitoring tank level. Additionally, Additionally, storage tanks should be equipped with a pressure safety rupture disk. A rupture disk burst indicator is strongly recommended to alarm a failure of the rupture disk and exposure to the atmosphere. For further information, see Chapter 11 of this Guide.




8.3.7

System Componen Components ts Comparison Table 8.1 shows a summary of water system components, listing common industry practices and advantages and disadvantages. Table Table 8.1 is not intended to be all-inclusive. Components not listed should be evaluated for appropriate use. Table Tab le 8.1: System Components Comparison Item

Type

Industry Practice

Advantages

Disadvantes

Connection Methods

Sanitary ( Vario Various us Types)

Common at sanit ar ar y equipment

Minimal cr ev evice Ease of inspection Ease of disassembly

Cost. Pressure limit in sizes above 4” OD.

Compression (Industrial)

Infr In freq eque uent ntly ly us used ed in pr pret etre reat atme ment nt

Skil Sk ille led d lab labor or no nott req requi uire red d

Cost, crevices

Flanged (Industrial)

Common in pretreatment

Cost Co st,, di disa sass ssem embl bly y

Alig Al ignm nmen ent, t, cr crev evic ices es

Threaded (Industrial)


Cost

Crevices

Welded-Automated (Sanitary)

Common in sanitary distribution

Highest reliability

Cost

Welded-Manual (Sanitary or Industrial)


Cost

BUNA (Nitrile )

Used for ambient temperature applications (i.e., pretreatment). (Rated by some vendors to 250°F).

Suitable for non-pure water and steam applications, temperature range, resilience, cost

Compatibility (Not Class VI), extractables, ozone incompatible

EPDM

Used for both higher temperature and ozonated applications (i.e., heat or ozone sanitized components and distribution. (Rated by some vendors to 300°F).

Suitable for pure water and steam applications, temperature range, good ozone compatibility compatibility,, resilience, cost

Several formulations are available. Some formulations may have limited steam or ozone service.

Silicone (Peroxide Cured, usually for Compression Molded Applications)

Used for ambient and higher temperature applications (i.e., heat sanitized components and distribution. (Rated by some vendors to > 400°F).

Suitable for pure water and steam applications, temperature range, resilience, cost

Ozone incompatible

Silicone (Platinum Cured, usually for Injection Molded Applications)

Used for ambient and higher temperature applications (i.e., heat sanitized components and distribution. (Rated by some vendors to > 400°F).

Suitable for pure water and steam applications, temperature range, resilience

Ozone incompatible

PTFE (solid)

Used for higher temperature applications. (Rated by some vendors to > 400°F).

Inertness, suitable for pure water and steam applications, temperature range, excellent ozone compatibility

Cold ows, non-elastic, difcult for cyclic temperatures, requires higher sealing forces

PTFE (envelope)

Used for ambient and higher temperature applications (i.e., ambient or hot systems typically when not heat sanitized. (Rated by some vendors to > 400°F).

Suitable for pure water and steam applications, temperature range, good ozone compatibility

Cost, creasing, difculty sealing

PTFE/SS Composite

Used for higher temperature applications (i.e., heat sanitized components and distribution. (Rated by some vendors to > 400°F).

Suitable for pure water and steam applications, extended life, exibility exibility,, high temperature

Cost, limited number of suppliers, higher sealing force required

Viton

Used for both higher temperature and ozonated applications (i.e., heat or ozone sanitized components and distribution. (Rated by some vendors to > 350°F).

Suitable for pure water and steam applications, temperature range, good ozone compatibility compatibility,, resilience

Cost

Gaskets


Quality/repeatability



Table 8.1: System Components Compariso Comparison n (continued) Item

Type

Industry Practice

Advantages

Disadvantages

Heat Exchangers

Concent riric Tu Tube

Common in in sa sanitary ap applications

Simple design, low pressure drop

Limited surface area

Plate and Frame (Single Wall)


Cost, large surfac surface e area

Drainage, Drai nage, not “rst in rst out,” gasket

Plate and Frame (double Wall)

Common in sanitary applications

Improved integrity

Cost

Tube and Shell-Single Tubesheet


Cost

Contamination potential

Tube and Shell-Double Tubesheet

Common in sanitary applications

Highest integrity

Cost

Centrifugal wi with Si Single Seal

Comm Co mmon on fo forr pu puri rie ed d an and d WF WFII

Cost Co st,, si simp mpli lici city ty

Seal integrity

Centrifugal with Double Seal

Common for WFI

Seal integrity

Cost, maintenance

Positive Displacement-Rotary

Can be used for high press. 200 psi

High pressure

Cost, pulsation

Positive DisplacementDiaphragm

Can be used for suction lift

Sucti Su ction on li lift ft/p /pri rimi ming ng

Limi Li mite ted d o ow w, pu puls lsat atio ion n

Single Shell

Common for ambient operation No heat sanitization

Cost, Cos t, eas easy y ins inspec pectio tion n

Sweat Sw eating ing,, ex exibi ibilit lity y

Insu In sula late ted d with with She Sheat ath h

Comm Co mmon on for for col cold d or for for amb ambie ient nt Operation with heat sanitization

E ne ne rg rg y, y, sa fe fe ty ty

Co st st , D ifif c c ul ult To I ns ns pe pe ct ct and Repair (DTI&R)

Jacketed 1/2 Pipe (and alternative)

Common for hot applications and/or heat sanitization

Ther Th erma mall ef efc cie ienc ncy y

Weld We ldin ing, g, co cost st,, (D (DTI TI&R &R))

Jacketed-Dimple Wall

Common for hot applications and/or heat sanitizations

Cost Co st,, le less ss we weld ldin ing g

Ther Th erma mall ef efc cie ienc ncy y (DTI&R)

Rupture Disc

Common for sanitary applications

Monitorable, integrity, suitable for pressure and/ or vacuum service

Cost, failure causes shutdown

Relief Valve

Common for sanit ar ary applications

Cost

Not monitorable, can result in potential contamination. Suitable for vacuum rated tanks only unless used with vacuum relief.

Diaphragm

Common for sanitary systems

Integrity

Cost, not suited for steam.

Buttery


Cost

Stem, crevices

Ball

Common in pretreatment and CS

Cost, CS service

Stem, crevices

Plug, Gate, Globe


Cost

Stem, crevices

Rising Stem Sanitary

Common in sanitary throttling

Integrity, leak detection, performance

Angle design

Pumps

Tanks

Relief Devices

Valves




Table 8.1: System Components Compariso Comparison n (continued) Item

Type

Industry Practice

Advantages

Disadvantes

Vent Devices

Vent Filter – Unjacketed

Common for ambient systems and ozonated (with proper materials)

Cost Co st,, ea easi sier er se serv rvic ice e

Cost Co stly ly if oz ozon one e co comp mpat atib ible le..

Vent Filter – Steam Jacketed

Common for hot application and/or heat sanitary

Eliminates blinding.

Disconnect. St ea eam t o chg. Energy

Vent Filter – Electric Traced

Common for hot application and/or heat sanitary

Eliminates bl blinding.

Disconnect. El Elect . To ch chg. Energy

Ozone Destruct Vent – Thermal

Common Com mon for ozo ozonat nated ed sys system tems s

Elimin Eli minate ates s cat cataly alyst, st, non non-sanitary.

Energy, high heat

Ozone Destruct Vent – Catalytic

Comm Co mmon on fo forr oz ozon onat ated ed sy syst stem ems s

Lowe Lo werr co cost st an and d en ener ergy gy,, non-sanitary

Energy catalyst can solidify/and block vent.

Elastomer Elas tomer (Sani (Sanitary) tary)

Used for lower temperature and pressure

Cost

Cost, used in conjunction with sanitary vent lter.

Elastomer (Industrial)


Cost

Durability,, used in Durability conjunction with sanitary vent lter.

Plastic (Sanitary)

Used for moderate temperature and pressure

Cost

Cost, used in conjunction with sanitary vent lter.

Plastic (Industrial)


Cost

Durability,, used in Durability conjunction with sanitary vent lter.

Metallic (Sanitary)

Used for higher temperature and pressure

Strength

Cost, exibility exibility,, used in conjunction with sanitary vent lter.

Metallic (Industrial)


Cost , st re rength

Flexibility, used in conjunction with sanitary vent lter.

Flexible Flex ible Hose Hoses s

8.4

Materials of Constructio Construction/Finishes n/Finishes Pharmaceutical equipment and piping systems rely extensively on stainless steels (typically 316L Stainless Steel) to provide the non-reactive, corrosion-resistant construction needed to meet operating conditions and sanitization methods (See ASME BPE (Reference 12, Appendix Appendix 1)). 304 SS also has been used, but is not recommended for new installations because it is more susceptible to the corrosive nature of high purity water. However, However, suitable thermoplastics (e.g., Polypropylene (PP) and Polyvinylidene Fluoride (PVDF) are available that may offer alternative benets. At elevated temperatures, structural integrity, integrity, support, and thermal expansion of piping systems are a concern. These concerns increase when using thermoplastics and may need to be addressed. Thermoplastics Thermoplastics also may be susceptible to degradation by UV irradiation; hence, it is common to use stainless piping immediately adjacent to a UV light. Material selection should be appropriate throughout the distribution, storage, and processing systems. Materials of construction “shall not be reactive, additive or absorptive so as to alter the safety, identity, identity, strength, quality, or purity of the drug product beyond the ofcial or other established requirements” (21 (21 CFR 211. 211.65) 65) (Reference 21, Appendix 1). Sanitization procedures should be considered when selecting materials. Sanitization with heat, UV UV,, chemicals, or ozone should be carefully managed with regard to concentration, pH, pressure, and temperature to avoid corrosive effects or damage to distribution systems.




High purity water distribution systems, using the material and nishes specied by the design, should be joined using acceptable welding or other sanitary techniques. For high purity water systems using a thermal sanitization method, 316L Stainless Steel normally is used. Automated Automated welding is the preferred method for joining high purity piping/tubing systems, because of the greater control over critical weld parameters and the smooth weld bead characteristics of the process. However, However, manual welding may still be required and used in specic situations, and should have increased quality inspection of the welds. The distribution and storage systems should be installed in accordance with cGMPs and fabricated, manufactured, procured, and installed in strict accordance with explicit procedures (e.g., in-house specications or other industry standards). While the use of higher-grade materials and nishes may yield benets in specic applications, applications, the additional expense should be evaluated against the advantages. There are numerous industry guidelines for the specication, installation, and quality assurance of pharmaceutical grade piping and components. Guidelines for high purity water systems include the ASME BPE (Reference 12, Appendix 1), which which provides the requirements requirements applicable to to the design of piping, piping, equipment, and systems used used in the bioprocessing and pharmaceutical industries. Comparable guidelines in Europe and Asia include DIN and JIS-G. These guidelines include topics such as: •

material specications

•

dimensions/tolerances dimensions/toleranc es

•

surface nish

•

material joining

•

quality assurance

Gaskets and seals used in high purity water systems should be reviewed for compatibility with the sanitization methods and chemicals. A variety of materials and designs can be used, including solid one-piece molded elastomers, machined gaskets, and multipart envelope gaskets. A range of grades of elastic and inelastic polymers may be applicable, but should be selected based on their sealing and compatibility properties; however, however, care should be exercised to avoid extractables/leachables (e.g., USP Class VI), and to ensure suitability with sanitization methods, including heat, chemicals, and ozone.

8.4.1

Sanitary Tubing and Piping Piping/tubing fabrication and installation should meet appropriate specications and/or standards (e.g., ASTM, ASME BPE, ISO, DIN, SMS, BS, JIS-G), as applicable. All welds should be documented and inspected and an isometric or other suitable piping drawing should be marked, identifying each weld with a unique number that corresponds to the weld inspection log and including data, such as the welder ID number, material test reports, and date. The piping support type and its spacing should be designed and installed in accordance with supplier recommendations or piping specications and applicable drawings. If necessary, the pipe slope should be checked and documented. The installed piping system should be pressure tested according to the requirements of the piping fabrication specication. Pressure tests can be performed pneumatically with proper safety precautions, particularly on thermoplastic systems. Pressure tests performed hydraulically should be evaluated based on possible residual microbial concerns. In both cases, the media should be of a quality to avoid contamination concerns. If hydraulic pressure tests are to be performed they should coincide with system start-up to avoid microbial issues.




8.4.1.1 Stainless Steel Distribution Piping Stainless steel tube, commonly used for high purity water distribution, is available in seamless drawn or welded construction, with the latter being more commonly used. Material quality is critical, such that levels of both carbon and sulfur should be precisely controlled. Levels that are either too high or too low can result in unsuitable welds and increased probability of weld failure. Tubing recommended for a high purity water distribution system should conform to ASTM A270 (Reference 20, Appendix 1). Stainless steel is susceptible to chloride attack from sources, such as chloride-containing insulation or prolonged contact with chlorinated or chloride containing water or some sanitizing agents. Temperature Temperature cycling can exacerbate the problem. Supports for the stainless piping, which incorporate isolators, should be used to preclude galvanic corrosion. The use of stainless steel tubing for high purity water distribution requires close attention to maintaining a passive layer on the water contact surface to minimize concerns with rouge development. For further information, see Chapter 10 of this Guide. 8.4.1.2

Types of Non-Meta Non-Metallic llic Materials Few compatible non-metallic piping materials are available that will withstand the rigors of a pharmaceutical high purity water system, such as: •

65°C to 90°C operation or periodic sanitization

•

121°C steam sanitizing temperatures

•

ozone contact

One such material that will support the above requirements and limit inorganic extractables to a minimum level comparable with stainless steel is Polyvinylidene Fluoride (PVDF). This material is available in a compatible range of pipe diameters, surface nishes, and fusion welding capability. PVDF is inert and will not exhibit surface corrosion when in contact with 90°C high purity water or commonly used oxidizers. The surface nish of PVDF is comparable or better to polished stainless steel and the fusion welding equipment and capabilities are similar to stainless steel orbital welding. Weldable ttings, elbows, tees, reducers, adapters, diaphragm valves, zero static valves, ow meters, regulators, etc., are available for PVDF pipe. Maximum operating pressures for PVDF piping typically are less than those for stainless steel tubing; therefore, the pressures should be evaluated and associated with pipe diameter and operating temperatures. Caution should be observed when using PVDF in combination with stainless steel, because of differing thermal expansion characteristics characteristics,, structural strengths, connection types, and applied stresses. Continuous support is recommended for systems that are operated at 65°C or above, because of softening at elevated temperatures. Polypropylene is an additional non-metallic piping material typically used in laboratory systems. This material typically is joined using heat fusion; however, sanitary anged joints also are readily available. These systems typically typically use chemical sanitization. For further information, see Chapter 13 of this Guide. The use of ABS, PVC, and CPVC for distribution systems has decreased; conversely conversely,, where non-metallic materials are used, the use of polypropylene and PVFD has increased.

8.4.2

Materials Comparison Table 8.2 is a comparison of the relative values of key factors in the use of various piping materials for the design and installation of water systems.




Table 8.2: Materials Comparison PVDF

ABS

Polypro

PVC

CPVC

316L SS

304L SS

Material Cost

H

M

M

L

L

H

M

Installation Cost(Note 1)

H

M

M

L

L

M

M

Y(Note 2)

N

N

N

N

Y

Y

Hot Water Sanitizable

Y

N

N

N

N(Note 3)

Y

Y

Ozone Sanitizable

Y

N

Y(Note 4)

N(Note 4)

N(Note 4)

Y

Y

Chemical Sanitizable

Y

Y

Y

Y

Y

Y

Y

Rouging Susceptibility

N

N

N

N

N

Y

Y

Corrosion Resistance

H

M

H

M

M

H

M

Availability

M

L

M

H

H

H

M

Extractables

L

M

L

H

H

L

L

Degree of Thermal Expansion

H

H

H

H

H

L

L

Joining Method • Sanitary Clamp • Solvent/Glue(Note 5) • Thermal Fusion/Weld

Y N Y

N Y N

Y N Y

N Y N

N Y N

Y N Y

Y N Y

Support Requirements

H

H

H

H

H

L

L

Typical Usage

M

L

H

L

L

H

L

Steam Sanitizable

(Note 3)

Legend: Y = Yes, Yes, N = No, H = High, M = Medium, L = Low Notes: 1. Based on skilled labor requirements, requirements, ease of welding, ease of visual inspection, inspection, shop fabrication requirements, requirements, etc. 2. The steam pressure and steam temperature control is critical critical to keep both below the manufacturer’s ratings. 3. Sanitization can be performed only at at low temperatures (e.g., 60°C) and with appropriate support. support. 4. Limited tolerance. May be beyond beyond Manufacturing Recommendations. 5. Solvents and glues glues may result in higher higher TOC levels.

8.4.3

Stainless Steel Polishes and Improved Finishes Historically, various terminologies have been used to refer to nish qualities. Terminology has included numeric Historically, systems, as well as systems that attempt to measure the average grit particles per inch for the polishing abrasives. These measurements have been superseded by the use of Roughness average (Ra) expressed in appropriate units (e.g., micro-inch and/or micrometer) as the industry standard. The Roughness average (Ra) reects the arithmetic mean of the surface deviations. Industry guidance documents (e.g., ASME BPE, ISO (References 12 and 22, Appendix 1)) dene grades of surface nish with associated associated Ra values and provide recommendations recommendations for application. application. Mechanical fnishing: is fnishing: is frequent and provides a suitable surface for most applications; however, abrasive polishing has particular inherent deciencies, including the tendency to enlarge the exposed surface area, mask surface imperfections, and it requires multiple steps to apply properly properly.. Electro-chemical polishing or electropolishing electropolishing:: is a reverse plating process, used to improve the surface nish of mechanically polished stainless steel components. Electropolishing is able to improve the mechanical nish by removing sharp peaks created by the abrasives, along with other advantages. Electropolishing provides a smoother surface, and reduces surface aws and contaminants resulting from mechanical polishing; however, welding of stainless components is detrimental to the electropolished surface nish. For further information, see Chapter 10 of this Guide.




The advantages of electropolishing may include: •

reduction of the height variations of surface

•

cleaning the surface

•

increasing the chromium to iron ratio at the product contact surface of stainless steel

•

revealing defects that may have been hidden by mechanical polishing

•

removing impurities trapped below folded layers of mechanically formed ridges

Electropolishing typically is applied over a mechanically polished surface that has been prepared by progressive uniform abrasive polishing applications.

8.4.4

Stainless Steel Finishes Stainless steel (300 series) typically is produced by cold rolling which imparts a granular nish to the material based on the rolling mill used, hence the term “mill” nish. This nish is not precisely controlled and as a result, typically it is not acceptable for high purity water applications and usually is subjected to subsequent enhanced nishing processes. Industry guidelines such as ASME ASME BPE (Reference 12, Appendix 1), as well as comparable guidelines in Europe and Asia (e.g., ISO and JIS-G)) should be used, as applicable. It is common industry practice to use sanitary tube OD sized materials for high purity water applications with interior nishes ranging from improved nish of 30 microinch (0.8 micrometer) Ra to mirror-like surface nishes of better than 10 microinch (0.3 micrometer) Ra. Two primary methods of interior polishing are used; mechanical polishing (using abrasives) and electropolishing. After mechanically polishing or electropolishing, the polishing compounds should be conrmed to have been completely removed, so as not to accelerate corrosion, or contaminate the water. The benets of any specied nish should be weighed against the application and the risks associated with using a lower nish. Though the value of high quality nishes is undecided, nishes in the range of 25 microinch (0.6 micrometer) Ra are most common. Systems operating at ambient temperature or with infrequent sanitization may require a smoother surface nish. The interior surfaces of stainless piping systems, in high purity water service, typically are polished to achieve a smooth surface with minimal roughness to enhance sanitization efcacy. efcacy.

8.4.5

Fabrication/Installation The distribution piping and storage systems should be installed in accordance with cGMPs and should be fabricated, manufactured, procured, and installed in strict accordance with explicit operating procedures. Fabrication of the distribution system requires extreme care and precision to ensure a smooth internal nish that will not allow any crevices that will support or promote bacterial growth, corrosion, or particulate generation. Fabrication should be performed by certied welders in a controlled environment to preclude contamination of equipment and material surfaces. Facilities dedicated to the fabrication of stainless steel (or higher grade alloys) are preferred to avoid contamination by carbon steel. Fabrication should follow an approved quality assurance plan. There should be adequate documentation in the design and construction of the system, including up to date Process and Instrument Diagrams (P&IDs), system isometrics, weld test reports, etc. A traceability matrix for validated systems, including all material certications, also should be included. Bending of stainless steel tubing should be closely evaluated because of possible damage to the surface nish (See ASME BPE (Reference 12, Appendix 1)). Additionally Additionally,, bending of stainless stainless tubing introduces introduces areas of thinner thinner wall thickness, which may result in low-pressure capability or possible stress cracking.




Tubing and piping welds, whether orbital or manual, should have a smooth internal diameter contour without excessive concavity or convexity, convexity, bead wandering, misalignment, porosity, or discoloration. One hundred percent photographic or radiographic analysis, while used to an increasing extent, may not be cost effective or justiable for a high purity water system. Appropriate Appropriate sampling is strongly recommended. A dened dened weld inspection program should be established to ensure the quality of welding. This program should include the process for reworking or replacing rejected welds. In the EU, the welding procedures should comply with BS EN 288 and the welder should be qualied according to BS EN 287 and BS EN 1418 for orbital welding (References 23, 24, and 25, Appendix 1). Thermoplastic piping should be installed in accordance with the manufacturer’s recommendations and with ASME BPE (Reference 12, Appendix 1) Part PM, following the guidelines for inspection and quality control.

8.5

Microbial Control Consideration Considerations s Given that microorganisms grow almost exclusively on surfaces, every wet surface associated with a water system is at risk of biolm growth. For further information, see Chapter 13 of this Guide. In the storage and distribution system, the impact of biolm growth is to the nal water quality from contamination with the bacteria and their cellular components shed by that biolm; therefore, consideration should be given to microbial control in all aspects of the storage and distribution, including the following elements:

8.5.1

1.

The compatibility of the materials of construction with the various planned or even unplanned sanitization approaches.

2.

Mechanisms for minimizing inux of planktonic organisms from upstream.

3.

How it is designed and operated to minimize locations that facilitate biolm development.

4.

How it is designed and operated to effectively apply apply,, distribute, and remove the sanitizing physical or chemical conditions that periodically kill and remove the developed biolm or keep it from developing.

Design and Operational Controls External contamination of a water system may be avoided by use of design and maintenance features such as: •

air breaks at drains

•

functioning vent lters

•

integral rupture disks on tanks

•

maintaining a high positive pressure on the distribution system to prevent the inux of contaminants from heat exchangers or from pressurized process vessels hard piped to the water system, and similar external sources

Distribution system components should be designed and operated to maintain the chemical purity of the water. The water ow should be fully turbulent and well mixed to assist in maintaining system-wide uniformity in temperature and chemical content during sanitization. The resulting high ow rate also helps the development of only a tenacious type of biolm that is minimally released or shed into the turbulent water when sanitizing conditions are not present. Microbial control may be achieved by a comprehensive program involving multiple design features, routine operational and maintenance approaches, and sanitization activities that work together. Design and operational elements should combine to make water of an acceptable microbiological quality. If a program feature is decient, microbial issues could be a recurring issue. The capital cost of appropriate design features and operational cost of routine maintenance and sanitization typically is less than the cost of repeated remediation and investigation (as well as potential product loss and regulatory scrutiny) because of microbial issues. For further information, see Chapter 13 of this Guide.



8.5.2


Sanitization Designs A distribution system should be designed with the capability of being sanitized using more than one type of approach. For further information, see Chapter 13 of this Guide. The rationale supporting a multiple sanitization approach capability is to have alternative approaches possible in case one approach proves to be ineffective. Materials should be compatible with the sanitizing agent or condition. See Table Table 8.1 for various distribution system material compatibilities with common sanitizing agents and conditions. Removal of the sanitizing agent or condition should be accounted for in the sanitization design. With the exception of heat and ozone which can be neutralized or removed in situ, all the other sanitants should be removed by ushing from the distribution system. A sloped and fully drainable system could facilitate a more rapid removal of the sanitant from the system, possibly using less rinsing water. The rinsing water should be available in an amount suitable for complete sanitant purging and of a chemical and microbiological purity that will not re-contaminate or re-inoculate the newly sanitized water system. This water can be provided from a reserved quantity of high quality distribution system water stored prior to the sanitization procedure or it may be freshly generated by the water purication system if available at a sufcient rate. The valving should be designed such that the sanitizing agents or conditions are able to contact all system surfaces, including bypass piping and valves and associated components, and if needed, the internal surfaces of POU valves exposed only while the valves are open. When post-sanitization system rinsing occurs, the ow path should allow sanitant purging from the system without the inux of contaminated air, as well as efcient rinsing to drain (or to a neutralization tank, depending on the sanitant employed and local plumbing codes), such that no piping section or valve is allowed to retain any unushed sanitant.

8.5.3

System Sanitization Microbial control usually can be achieved through a combination of distribution system design features, as well as effective periodic or continuous sanitization. For further information, see Chapter 13 of this Guide.

8.5.4

Monitoring for Sanitization Effectivenes Effectiveness s and Ongoing Microbial Control Microbial monitoring should be used to conrm the effectivenes effectiveness s of the sanitization process, including where the process has been ‘validated’ previously as effective. Over time, biolm development in a water system can affect a validated process unpredictably. unpredictably. The effectiveness of sanitization should be monitoring on an ongoing basis after initial sanitization. If a subsequent large rebound in microbial counts is experienced relatively soon after sanitization (within about a week), this may be an indication that sanitization was not effective and a more stringent or more frequently applied sanitization approach should be considered, even if microbial counts are undetectable immediately following sanitization. For further information, see Chapter 13 of this Guide.

8.6

System Designs This section provides information that may be useful in evaluating the advantages, disadvantages, and cost effectiveness effectiven ess of different designs commonly used to store and deliver water to use points. In addition, a method of selecting/optimizing selecting/optim izing system storage and distribution design is discussed. System designs should meet dened user requirements; better designs also provide greater reliability or reduced life cycle cost. Examples of common storage and distribution design approaches are presented in this Guide, to help demonstrate the concept of optimal system design. Alternatives provided are intended to demonstrate key concepts, which when applied and properly operated, can result in an acceptable storage and distribution system. They are not intended to indicate that these are the only designs which are considered acceptable.



8.6.1


General Considerati Considerations ons Numerous criteria should be considered when evaluating alternative designs for storage and distribution systems. The design advantages/disadvantages should conform to the user requirements for the water. The optimal design of a pharmaceutical water storage and distribution system should accomplish the following:

8.6.2

•

Maintain the chemical and microbial quality of the water within acceptable limits.

•

Minimize the conditions and locations that favor microbial growth.

•

Deliver the water to the use points at the required ow rate, pressure, and temperature.

•

Accommodate a suitable total instantaneous demand of water (i.e., diversity) to multiple use points.

•

Minimize capital and operating expenses.

•

Ensure reliability while minimizing potential disruptions to operations.

•

Account for the possibility of future expansion.

•

Maintaining the chemical and microbial quality of the water within acceptable limits is often misinterpreted. It should be noted that it is not necessary to protect the water from every form or level of degradation, only to maintain the quality within acceptable limits. Potential disruptions to water availability may be acceptable for brief periods of time (e.g., non-intrusive maintenance, utility outages) based on limited impact on microbial growth potential. However, intrusion of external contamination during these brief periods may be of concern if the system were opened. Extended disruptions may increase the risk of contamination.

Distribution Loop Velocity The primary purpose for recirculated distribution is to reduce the release of biolm organisms into the water by forcing them to tightly adhere to the interior surface of the pipe. For further information, see Chapter 13 of this Guide. Although the mechanisms mechanisms are not universally agreed upon, upon, it is thought that the velocity velocity associated associated with turbulent ow reduces the initial attachment of bacteria to interior surfaces, as well as the development of loosely attached biolm. It is generally recognized that turbulent ow is indicated by a Reynolds number greater than 4000. Distribution systems often are designed to operate with nominal ow velocities of 2 to 3 feet per second or higher, which greatly exceeds the turbulent ow threshold, and therefore, may have a greater detriment on biolm development. Selection of a design nominal velocity should be based upon the mixing effect obtained at branch connections because of the turbulence created, fully ooded piping, biolm development, and capital/operating cost. The turbulent mixing effect can assist in minimizing stagnation in the length of the branch, but may not prevent loose biolm formation. In general, velocity may drop off for short periods of time during high use without adversely affecting performance, so long as positive pressure is maintained in the system. Circulation at higher velocities also helps to maintain a uniform temperature throughout the distribution system.

8.6.3

Storage and Distribution Decision Flowcharts The decision owcharts are presented to aid in the analysis of which of the alternative designs best suits a particular application. In evaluating which conguration is optimal for a given situation, designers should consider the user requirements. There may be alternative designs that will satisfy a particular application; designers should investigate the advantages and disadvantages of each option to support a decision. The decision regarding storage alternatives can be made by evaluation of the requirement for QA release of the water prior to use. If QA release is required, a batch storage approach is considered the most suitable. If the water can be utilized while quality analyses are ongoing, a dynamic/conti dynamic/continuous nuous storage concept is considered most appropriate. Following the decision of the storage alternative, a designer can use the decision owchart to decide additional attributes of the storage system and the type of distribution system that best ts the application.



Figure 8.1: Storage Decision Flowchart

Figure 8.2: Distribution Decision Flowchart




8.6.4


Storage and Distribution Design Concepts The two basic concepts developed for storage of pharmaceutical waters are referred to as the “batch” and “dynamic/ continuous” storage concepts. The batch concept may use one or more storage tanks. With a single tank system, the vessel is lled with water, which is then quality tested. With multiple tanks, one or more may be lled and the water quality tested, while another is in service providing water to the various users. As batches are released and the in-service tank is emptied, tanks are rotated as necessary to meet demand, e.g., a two tank system. The dynamic/continuou dynamic/continuous s storage concept off-sets the peak instantaneous water demand put on the overall water system through use of a single water storage vessel. This vessel holds the generated and r ecirculated high purity water and ultimately supplies it to users via the distribution system. The advantage of the “batch” distribution concept, over the “dynamic/continu “dynamic/continuous” ous” distribution concept, is that the water is tested for compliance before use with tank QA/QC lot release (water used in each product batch lot is traced and is identiable). The advantages of the “dynamic/continuous” distribution concept include lower life cycle costs, as well as less complex piping around the storage vessel, and a signicantly more efcient operation. Distribution concepts can be divided into “branch/one-way” and “circulating” ow designs. Both design concepts mostly use continuous ow of the water as the primary method to maintain water quality. quality. The branch/one-way design consists of a single supply pipe from the storage tank to the use point. This design primarily is used when water usage is nearly continuous or stops only for short durations. The circulating design concept includes supply and return piping between the tank and the use points. This loop concept allows continuous ow of water independent of usage. The system also may include equipment designed to maintain or modify attributes of the water, including pumps for distribution, heat exchangers for heating and/or cooling, vent lters for microbial and particulate protection, and polishing equipment to maintain quality. Once a system preliminary design concept has been selected, the following additional considerations should be evaluated:

8.6.5

•

Loop conguration, including whether series or parallel loops are required, distribution loop points of use, cooling requirements (steam-able, sub-loop, or multiple branched heat exchanger assemblies), reheat requirements, secondary loop considerations, etc.

•

POU details including temperature, pressure, and ow rate (e.g., heat exchangers, pumps, ow devices).

•

Installation details (alcove, surface, direct, or cabinet mount) for POU protection/acces protection/accessibility sibility and room cleanability.

•

Sanitization method (steam, hot water, ozone, or chemical).

Examples of Design Concepts The examples describe the systems contained in the accompanying decision tree, which can be used successfully to store and distribute high purity water. The examples present simplied schematic diagrams (not meant to be P&IDs) of each conguration. System details including valves, instruments, and redundancy are not included unless specically required for describing the system. Hot storage tanks (greater than or equal to 65°C) typically are represented with steam jackets, but alternatively can use external heat exchangers in a circulating water loop. Advantages and disadvantages listed may not be all inclusive.




Example: Ambient or Reduced Temperature Storage and Distribution (Heat or Chemical Sanitization) This system is most advantageous when the water is generated at ambient temperature, will be used only at ambient temperature, and there is adequate time for sanitization. Congured in this manner, water is stored and distributed at ambient or reduced temperature with periodic sanitization accomplished accomplished by heating to sanitization temperature and circulating for an adequate amount of time; alternatively,, the system can be operated hot for extended periods and cooled for use. Heat can be supplied to alternatively the water through the use of a tank jacket, or alternatively, alternatively, by a heat exchanger in a circulating loop; alternatively, a sanitization chemical can be added to the system then ushed allowing the elimination of the sanitizing heat exchanger.. Similarly, exchanger Similarly, UV irradiation also can be included for reducing or impeding microbial growth. Cooling can be used to prevent temperature increases because of pump and UV energy, for cool down after sanitization, and as a means for reducing or impeding microbial growth. Figure 8.3: Ambient or Reduced Te Temperature mperature Storage and Distribution (Heat or Chemical Sanitization)




Example: Storage and Distribution with Continuous Polishing Continuous “polishing” (maintaining or improving the quality of the water) can be included in the design of a storage and distribution system as a water quality control mechanism. The method of “polishing” may include a technology common to purication (e.g., reverse osmosis, deionization, ltration, ultraltration, etc.). Polishing of the water may be achieved using a separate circulation loop off the storage tank that includes additional polishing equipment, or may include a POU that returns water for re-treatment by the primary generation system. The water source for the separate loop can be a separate pump or may be a branch off the main distribution loop. Flow balancing may be required to avoid over lling the storage tank or wasting of polished water. Figure 8.4: Storage and Distribution with Continuous Polishing

Example: Ozonated Storage and Distribution An ambient storage and distribution system can be operated effectivel effectively y with an ozonated storage and a periodically ozonated loop, for microbial control. Typical levels of 0.02ppm to 0.2ppm of continuous ozone protect the water from microbial contamination. Frequently, Frequently, two methods of generating ozone are employed, electrolytic and corona discharge. The corona discharge method generates ozone from oxygen in the air, while electrolytic method uses the oxygen in the water. Ozone should be completely removed from high purity water prior to usage by using UV irradiation. Ozone monitors also should be included to verify that ozone has been eliminated prior to the points of use, to maintain consistent ozone levels, and to conrm loop sanitization. In addition, atmospheric ozone monitoring should be installed to address safety concerns (See Occupational Safety and Health Administration Administration (OSHA) Standards (Reference 26, Appendix 1)). Additional Additional atmospheric monitoring of oxygen or hydrogen should be evaluated dependent upon the technology and specic installation. Ozone should be periodically circulated through the distribution loop by de-energizing the UV light, as required for sanitization and controlling biolm development. Typically Typi cally,, the level of ozone concentration during sanitization is increased.




Figure 8.5: Ozonated Storage and Distribution

Example: Hot Storage, Hot Distribution This system consists of a hot storage tank and one or more hot distribution loops. Temperature is maintained in the storage tank by heat supplied to the tank jacket or alternatively by a heat exchanger in the circulating loop. A spray ball is typically included in this design. Figure 8.6: Hot Storage, Hot Distribution




Example: Branched/One Way Distribution The branched/one way distribution system consists of one or more storage tanks with supply piping to use points, but no return piping. The system may have limited circulation of the storage tank, and typically, includes a method to periodically ush the supply piping at multiple locations. This system typically typically is used where water use is nearly continuous throughout the distribution piping. A program should be established to periodically ush and sanitize the loop to maintain microbial contamination within acceptable limits. Figure 8.7: Branched/One Way Distribution

Example: Batch Tanks, Recirculating Distribution The batch concept uses one or more storage tanks supplying a single or multiple distribution loops. The tanks and loop(s) can be hot or ambient. This system can be used where partial QA release for chemical attributes is required on a batch basis. After a tank has been lled from the water treatment system, it is isolated and the water inside is tested. Only after partial QA release is that tank put into service, but the water continues to be used at risk on a microbial basis. The water often is drained after 24 hours, but can be validated for longer periods of time. At the completion of the draining operation, the vessel and associated distribution system usually are sanitized before relling. Figure 8.8 for this example is for a two tank system. The increased use of rapid microbiological methods can result in earlier full QA release.




Figure 8.8: Batch Tan Tanks, ks, Recirculating Distribution

Example: Parallel Distribution Loops from Single Tank This system is a combination of multiple distribution loop schemes supplied by one storage tank. Figure 8.9 for this example depicts a hot storage tank with two loops; one hot and one that is cooled and reheated. An additional heat exchanger (economizer) may be included to recover the heat transferred between the two loops shown. The loops can be supplied by one pump or by separate pumps if greater reliability and separation is desired. Figure 8.9: Parallel Distribution Loops from Single Tan Tank k




Example: Hot Storage, Cooled and Reheated Distribution This design concept consists of a heated storage tank supplying a distribution loop that includes a cooling heat exchanger prior to the use points, followed by a heating heat exchanger prior to the tank return. Hot water from the storage tank is cooled through the rst heat exchanger, circulated to the use points, and then reheated in a second heat exchanger before returning to the storage tank. The purpose of this system is to maximize the time the water is at sanitization temperature and only cool the water just prior to the points of use. This design concept also allows the water to be used at hot temperatures. Sanitation of the use points is achieved by turning off the cooling medium on a periodic basis. This design is similar to “ambient or reduced temperature storage and distribution” with heat or chemical sanitization and can offer increased exibility; however, if both concepts are incorporated, increased equipment and utility sizing may be required for this option. To decrease energy and cooling media consumption, an economizing heat changer can be used which exchanges heat between the two high purity water streams (discharge and return). Figure 8.10: Hot Storage, Cooled and Reheated Distribution




Example: Hot Storage, Cooled Bypass Circulating Distribution This design concept includes hot storage with cooled distribution system that bypasses return to the storage tank and includes a cooling heat exchanger prior to the use points. Hot water from the storage tank is cooled through the heat exchanger, circulated circulated to the use points, and then returned to the pump suction. When water is drawn from a POU valve, hot water from the storage tank ows into the loop and is cooled by the heat exchanger. The loop can be sanitized on a periodic basis by turning off the coolant and opening up the return valve to the storage tank, allowing hot water to ow through the loop. The returning of cooled water to the tank can be avoided by optionally ushing the lower temperature water to drain until the loop becomes hot and then return the ow to the storage tank. An alternative is to add a heat exchanger for periodic sanitization of the distribution loop independent of the storage tank. An additional alternative alternative is to continuously continuously return a small portion of the circulating water water back to the tank. Figure 8.11: Hot Storage, Cooled Bypass Circulating Distribution




Example: Hot Distribution with Cooled Branch Use Point, Heat Sanitizable This system consists of a hot tank and hot distribution loop, but has one or more use points requiring water at lower temperature, that are equipped with POU cooling heat exchangers. Hot water is ushed to drain through the POU heat exchanger for sanitization, and then cooled before opening up the POU valve. The use of pure steam can allow for continuously sanitizing the exchanger and downstream piping when water is not called for at the use point. Alternatively,, multiple POU valves can be included on the distribution Alternatively distribution from the cooling exchanger; however, multiple POU valves can pose operational challenges. POU exchangers are most advantageous when there are both hot and lower temperature water use points off the same loop, and the number of low temperature users is small. Since they maintain the water hot until it is drawn from the loop, they provide excellent microbial control, provided they are frequently ushed or sanitized when not in use. As the number of low temperature users users increases, the capital costs costs and space requirements requirements become prohibitive, prohibitive, and one of the other conguration examples should be considered. Water consumption is high because of ushing although this is minimized by the scheme shown. Energy consumption is moderate because only water drawn out of the loop is cooled. Maintenance requirements are high because of the added exchangers and valves. Complexity is high as each exchanger should be properly ushed and sanitized. Each drop is limited in capacity by the sizing of the exchanger. Figure 8.12: Hot Distribution with Cooled Branch Use Point, Heat Sanitizable




Example: Hot Distribution with Cooled Use Point, Slip-Stream This system consists of a hot tank and hot distribution loop, but has one or more use points requiring water at lower temperature that are equipped with a cooling heat exchanger in a slip-stream off the main. The conguration continuously circulates a portion of the water through the heat exchanger and r educed temperature use points in a slip-stream that is parallel to the main loop. The water in the slip-stream returns to the main loop. An orice or control valve is included in the main loop to force water through the slip-stream. An optional ush valve may be added for cool down of the loop and to avoid returning cooled water to the hot loop; however, stopping the return ow poses operational challenges if the ow is stopped for extended periods. Sanitization is accomplished by circulating hot water from the loop, through the POU exchanger, back to the main loop. The scheme shown in Figure 8.13 results in an added pressure drop in the main loop, which leads to a larger circulation pump. Figure 8.13: Hot Distribution with Cooled Use Point, Slip-Stream

Example: Hot Distribution with Cooled Sub-Loop This system includes a pumped, cooled sub-loop off the main hot distribution loop to provide the reduced temperature water to one or more use points. The secondary pump provides the circulation through the cooling heat exchanger, to the use points, through the sanitization heat exchanger, and returning to the pump. The water in the sub-loop circulates, but does not return to the main loop.




Figure 8.14: Hot Distribution with Cooled Sub-Loop

Example: Primary/Second Primary/Secondary ary Distribution Primary/secondary distribution consists of a storage tank supplying one or more primary loops which then supply use Primary/secondary points, secondary storage tanks and/or secondary distribution loops. This design concept can be a combination of many of the example concepts. A secondary secondary pump provides ow through the secondary distribution loops. Multiple secondary loops can be used to service individual manufacturing areas from a remotely located primary storage tank and loop. The primary loop typically is larger in diameter than each secondary loop. In addition, the primary loop can be a hot loop while the secondary loop could be at a reduced temperature. A reduced reduced temperature secondary loop can be utilized to supply one or multiple use points. Figure 8.15: Primary/Second Primary/Secondary ary Distribution




Example: Use Point/Sample Valves The design of POU sample valves or sample valves in general, often poses signicant issues. A simple manual or automated POU valve that is easily accessible can be sampled representative of the way the high purity water is used, provided it is of suitable size; however, the addition of a hose to the use point, sampling a remote location, or the direct connection to a tank or other equipment can signicantly complicate sampling methodology. methodology. Hoses attached to a POU valve should be installed and maintained in a drainable position at all times, and the sample should be obtained from the outlet of the hose. Draining of the hose often is assisted through the use of suitable, pure, ltered air purging. Methodology should be established for periodic in-place or out-of-place hose sanitization or hose sanitization sanitization prior to each use. The use of a sample valve integral to the body of a POU valve that is directly connected to a process tank or other equipment is a commonly accepted practice. The integral sample valve provides a representative sample of the water supplied to the POU valve without breaching the integrity of the valve-to-equipment connection. connection. The piping between the POU valve and the equipment often undergoes a periodic CIP or SIP process for sanitization. Sample valves located in remote locations can be opened/closed using regulated air actuation to provide a sample to an accessible location through a completely draining attached hose or extended section of tubing. The extended tubing or hose should be thoroughly ushed and/or sanitized prior to sampling to minimize microbial contamination. contamination. In addition, specialized “sanitary” sampling valves can be used to sample the water from the mid-stream ow within a pipe to avoid the inuence of bio-lm within the branch piping of a sample point. For further information, see Chapter 13 of this Guide. Figure 8.16: Use Point/Samp Point/Sample le Valves


9

Laboratory Water




Page 139 Laborato ry Water Laboratory

9

Laboratory Water

9.1

Introduction Laboratory water requirements may differ from those for manufacturing. This chapter provides an overview of the different laboratory water purication and distribution approaches available, as well as a step by step method to help to determine what type of system design will best meet a user ’s needs.

9.2

System Design Consideratio Considerations ns Water quality requirements for laboratory purposes vary widely depending upon the type of analysis to be performed and the governing organization. Non-pharmacopeial related agencies, which cover water quality for laboratory purposes, such as ISO, ASTM, CLSI, typically typically are used as a source of information. The wide range of user needs leads to a variety of possible approaches. Specic laboratory user information is needed to design a cost effective and efcient solution. The design team needs to understand the related design information including, but not limited to, the issues listed: Dening User Needs 1.

What laboratory tasks will require water?

2.

What quality of water is needed for each task?

3.

What are the regulations that must be complied with?

4.

What is the location of each task?

5.

Is there a work pattern for each task?

6.

Can these tasks tasks be clustered clustered in various laboratory locations (e.g., (e.g., by water water quality needed, needed, analytes of concern)?

7.

Can one group of tasks be served by one POU outlet?

8.

How much water is needed at each POU and by task or task group?

Solution Design Based on User Needs 1.

What points of use are needed?

2.

What water characteristics (quality (quality,, etc.) are needed at each point?

3.

What purication technologies could be used to produce the water qualities required?

4.

What types of distribution system could be suitable?

5.

Other parameters to be considered, such as: a.

building characteristics

b.

laboratory architecture

c.

criticality of water in the process



Page 140 Laboratory Laborato ry Water

d.

water source options

e.

ergonomics of use points and drains

f.

economics

9.3

Determining User Needs

9.3.1

Quality Needs Laboratories often require a selection of waters with distinct purity specications specications,, dependent upon analytical applications and regulatory requirements. Compendial procedures will need to use compendial specied waters. Potential impurities in PW may be grouped into: •

inorganic ions (typically monitored as conductivity or resistivity resistivity,, or by specic chemical tests)

•

organic compounds (typically monitored as TOC or by specic tests)

•

bacteria (monitored by cultivative total microbial plate counts or other methods)

•

endotoxins (monitored by LAL test)

•

nucleases (monitored by specic enzyme assays)

•

particulates (typically managed by ltration, but not monitored usually)

•

gases (typically managed by degas/puricatio degas/purication n equipment and monitored by specic tests, if required)

Within any of these groups, particular substances also may have a specic interference in a particular test, such as components that produce over-lapping peaks in chromatography or contaminants in the water that are identical to the analytes in the test sample. Table 9.1 provides guidance regarding the types of impurities that may be important for an application when choosing a water system. Quantied impurity levels are not provided because of the wide variations in water purity needed within any one type of application. Table Table 9.1 is intended primarily for the design engineer who may not have an analytical laboratory background, and could serve as a basis of discussion with laboratory personnel regarding specic water grade/purity needs. Table 9.2 lists the purity specications for commonly used laboratory water grades. These should be considered as minimum specications. specications. Additional considerations maybe listed in source documentation for these water grades. A laboratory’s water purity needs for particular applications may exceed minimum requirements for a specic attribute of a given water grade. More stringent requirements and additional purication technologies may be applied to maintain several attributes at lower levels. Conversely, Conversely, the most suitable water grade for a particular application may exceed the purity needs of specic attributes for the application. Unless otherwise mandated by regulatory requirements, the water purity provided may be optimized with the water purity needed. Maintaining a higher water purity than needed by applications can be costly and usually is unnecessary unnecessary,, unless it is a regulatory requirement or expectation.




Table 9.1: Importance Level (Widely Observed) for Parameters/Contamin Parameters/Contaminants ants in Different Tech Techniques* niques* Technique

Application Sensitivity**

Importance Level of the Water Contaminants

Inorganic Ions

Organic Compounds

Particulates

Bacteria

Endotoxin

Nuclease

Low – High

Low

Medium

Medium

Medium

Low

Low

High

Medium

Low

High

Medium

Low

Low

Low – High

Medium – High

Medium – High

Medium

Low – High

Very Low

Very Low

Electrophoresis (Polyacrylamide Gels)

High

High

High

Medium

High

Low

Low (High for Protease)

Electrophoresis (Agarose Gels)

High

High

Medium

Medium

High

Low

High

Electrophysiology

Low – High

High

High

Medium

High

High

High

ELISA

Low – High

Low

Medium

Medium

High

Low

Low

Medium – High

Low – High

Medium – High

Medium – High

High

High – Very High

Low

Low – High

High

Low

Medium

Medium

Very Low

Very low

High

Very High

High

High

High

Low

Low

Low – High

Low – High

Medium – High

Medium

Low – High

Very Low

Very Low

High

High

High

Medium

High

Very Low

Very Low

General Wet Chemistry

Low – High

Low

Low

Medium

Low

Very Low

Very Low

Glassware Washing

Low – High

Low – High

Medium – High

Medium

Low – High

Low

Very Low – Low

Histology

Low – High

Low

Medium

Medium

Medium

Medium

Very Low

HPLC

Low – High

Low – High

Medium – High

High

Low – High

Low

Very Low – Low

LC-MS

High

High

High

High

High

Low

Low

Low – High

Low

Medium

Medium

Low

Very Low

Very Low

ICP-AES

High

High

Medium

Medium

High

Very Low

Very Low

ICP-MS

High

Very High

High

High

High

very Low

Very Low

Immunocytochemistry

High

High

High

High

High

High

Medium

Ion Chromatography

Low – High

Medium – Very High

Medium – High

High

Medium – High

Very Low

Very Low

Mammalian Cell and Tissue Culture

High

High

High

High

High

Very High

High

Microbiological Media Preparation

Low – High

Low

Medium

Medium

High

Low

Very Low

Microbiological Analysis

Low – High

Low

Medium

Medium

High

Medium

Low

Molecular Biology

High

High

High

High

High

Low – Medium

Very High

Monoclonal Antibody Research

High

High

High

High

High

Very High

Low

Bacterial Culture Clinical Biochemistry Electrochemistry

Endotoxin Determination Flame-AAS GF-AAS GC

GC-MS

Hydroponics




Table 9.1: Importance Level (Widely Observed) for Parameters/Contamin Parameters/Contaminants ants in Different Tech Techniques* niques* (continued) Technique

Application Sensitivity**

Importance Level of the Water Contaminants

Inorganic Ions

Organic Compounds

Particulates

Bacteria

Endotoxin

Nuclease

Plant Cell and Tissue Culture

High

High

High

High

High

Medium

Medium

Radioimmunoassay

Low – High

Low

Medium

Medium

High

Very Low

Very Low

High

Medium

High

Medium

Medium

Low

Low

Spectrophotometry

Low – High

Low

Medium

High

Low

Very Low

Very Low

Steam Generation

Low – High

Low

Medium

Medium

Low

Low

Very Low

TOC Determination

High

High

High

Medium

High

Medium

Low

Trace Metal Detection

High

Very High

High

Medium

High

Low

Low

Solid Phase Extraction

Notes: 1. Table 9.1 is provided as an example example of the widely observed importance level level of a selection of water quality parameters/contaminants parameters/contaminants in different laboratory techniques. The level of importance depends on the sensitivity expected for the application, the material used, the method applied, and the regulatory constraints. A range of levels is listed for a number of attributes because of a wide variety in water purities needed for t he many forms of the technique and types and levels of analytes involved. 2. Application sensitivity sensitivity refers to the level of analyte detection, quantitation, or contaminant contaminant impact expected with that application. Typically: Typically: High = ppb or higher sensitivity levels, Medium = ppm to ppb sensitivity levels, and Low = ppm or lower sensitivity levels



s e i r o t a r o b a L l a c i t u e c a m r a h P n i d e s u e b y a m t a h t s e d a r G r e t a W l a i e p o c a m r a h P n o N s u o i r a V r o f y r a m m u S n o i t a c  i c e p S : 2 . 9 e l b a T


e d i v s

S E Y

b p p 0 0 5

*

*

*

*

*

*

1 ± 2 . 8 1

S E Y

L ) b / g p µ p 0 0 2 2 (

*

*

*

*

*

*

2 . 0

S E Y

*

*

*

*

*

*

*

0 . 4

S E Y

) L / b g p µ p 0 0 0 0 2 2 (

*

*

*

*

*

*

*

0 . 1

0 . 1

S E Y

) L / b g p µ p 0 0 5 5 (

*

*

*

*

*

*

I e p y T

I , D t - l i m n o B c e i / t M F c S r l a ; m * µ u i l ) . µ o t 0 S v 2 s i . 2 i u D q 0 < ( + E

5 5 5 0 . 0

8 1

S E Y

L ) b / g p µ p 0 0 5 5 (

*

*

*

*

*

*

3 e d a r G

. t s O i R D - r o e I l g D n i r S o

5 . 7 o t 0 . 5

0 . 5

*

S E Y

*

*

4 . 0

*

2

*

*

2 e d a r G

r t o s . i t s D i + D - O e l R p r i t o l I u M D

*

0 . 1

*

S E Y

*

*

8 0 . 0

1 0 . 0

1

*

2 0 . 0

1 e d a r G

; t , t e l i s c i r F m ) µ D u - s o m 2 . e s a S µ 0 R l g 2 2 . + I r o n e 0 , i d + D ( t r l a O o i r G R F

*

1 . 0

*

S E Y

*

*

*

1 0 0 . 0

*

*

1 0 . 0

e v i s u l c n i ( C ° 5 2 t a e u ) l a e v g H n p a r

, C ° 5 2 @ m c / S µ y t i v i t c u d x n a o C m

, C ° 5 2 @ m c Ω M y t i v i t s i s i e n R m

? d t e n t a e s m n e e r p u s m a o e C M y e r t u i v t i t a r c e u p d m n e o T C

x a m , ) C s a ( C O T

) . t s b u s S e . c d n e a t s R e b t u S a n e a l b g n a z a i d m r i x e O P (

d , n t a h g x a m n e m n L , 4 s 5 h t i 2 t a n t P a l U e a e c i c c n t n a p a b b r O r o m o s s b c b A 1 A

n / o g i t a m r , o C p ° a 0 v E 1 1 r t e a t f g a i n e t x a u a e m d i H , s e n g R o K

n o i t a r , o C p ° a 5 v 0 x e 1 a r t m e a , t f g L a n m e i u t a 0 d e 0 i h 1 s / e n g R o m

x a m , L / g m ) 2 O i S s a ( a c i l i S

I d ) S E 6 L h 0 0 C t ( 4 2

e t e W  t a u ) y l i b R i c t t i L e n t r n o p C S a u t a

*

6 9 1 ) 5 6 D 0 M 0 2 T ( S A

o s r i e n t d B o a r r i a t W o a d f c e n e l a i d t d i S u p r p a G A G

e r l e b ) t a s a t i e W u ( s g S s e n ; i e k c c o r n r i r u o p D S

*

*

V I e p y T

, n , O o I i t D R a E r l o l , / i I t s D d n i D a

0 . 8 o t 0 . 5

0 . 5

I I I e p y T

t , l n , O i F o I i D R t m a E r µ l o l , / i I 5 t d 4 s . i D n D a 0 +

*

5 2 . 0

I I e p y T

r o n . o v i t i a u l q l E i t s i D

r e t a W t n ) e 1 e g t o a N ( e ) 6 R 0 r o 0 f 2 ( n 3 i o 9 t 1 1 a c D i  M c T e p S S A d r a d n a t S

e s U y r o t a ) r o 5 b 9 a 9 L 1 ( l 6 a c 9 i 6 t y 3 l O a n S I A r o f r e t a W

e c n e r e f e R / n o i t a z i n a g r O

*

*

0 1

( q

e p y T r o e d a r G r e t a W

d s n e a h e c a c r o u r o p p S A d n e o  i i c t e a c p  S i r u P

) 2 e t o N (

x a r m , e L t t / g a m M t e n l e b t a z n i o d i C x 2 O O




) d e u n i t n o c ( s e i r o t a r o b a L l a c i t u e c a m r a h P n i d e s u e b y a m t a h t s e d a r G r e t a W l a i e p o c a m r a h P n o N s u o i r a V r o f y r a m m u S n o i t a c  i c e p S : 2 . 9 e l b a T

e d i v s

*

*

*

*

*

n m d i o µ r e t i e l a 2 l t 2 p i . t 0  m i m I l y b

/ u ) f c L 0 m 0 0 1 ( 0 1 1

1 0 . 0

*

*

*

*

*

) L 1 . m 0 0 : 0 B 1 / e f u p c y T 0 1 (

0 / ) u 1 f : c L C 0 m e 0 0 p 0 0 y 1 1 T (

3 0 . 0 : A e p y T

5 2 . * : 0 C : B e * e p y p T y T

*

*

*

*

0 1

) 1 L 0 . m 0 0 : 0 A 1 e / u p f y c T 1 (

) L 1 . m 0 0 : 0 B / 1 e f u p c y T 0 1 (

0 / u ) 1 f : c L C 0 m e 0 0 p 0 0 y 1 1 T (

3 0 . 0 : A e p y T

5 2 . * : 0 C : B e * e p y p T y T

*

*

*

r y t e b l d n  e i i o m l p t a µ t m i 5 I m 4 . i l 0

5

) 1 L 0 . m 0 0 : 0 A 1 e / u p f y c T 1 (

) L 1 . m 0 0 : 0 B 1 / e f u p c y T 0 1 (

0 / u ) 1 f : c L C 0 m e 0 0 p 0 0 y 1 1 T (

3 0 . 0 : A e p y T

5 2 . * 0 : : C B e * e p y p T y T

*

*

*

*

1

1

) 1 L 0 . m 0 0 : 0 A 1 e / u p f y c T 1 (

) L 1 . m 0 0 : 0 B 1 / e f u p c y T 0 1 (

0 / u ) 1 f : c L C 0 m e 0 0 p 0 0 y 1 1 T (

3 0 . 0 : A e p y T

5 2 . * 0 : : C B e * e p y p T y T

*

*

*

y r e b t d n l  e i o l i m p t a µ m t i 2 I . m 0 i l

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

y r e b l d n t e o  i l i m p t a µ m t i 2 I . m 0 i l

I d ) S E 6 L h 0 0 C t ( 4 2


*

6 9 1 ) 5 6 D 0 M 0 2 T ( S A



V I e p y T


I I I e p y T


I I e p y T


I e p y T

I , D l - t i m n o B c e i / t M F c S r l a ; m µ u i l ) . µ 3 o t 0 S v 2 s i . 2 i u 0 < D ( q + E

3 e d a r G


2 e d a r G


1 e d a r G





*

*

*

0 1

*

/ u ) f c L 0 m 0 0 1 ( 0 1 1

/ u ) f c L 0 m 0 0 1 ( 0 1 1

0 5

) 1 L 0 . m 0 0 : 0 A 1 e / u p f y c T 1 (

( q



*

*

0 0 5

3

*

x a m , L / g µ a c i l i S l a t o T

*

0 5

0 1

5

*

x a m , L / g µ m u i d o S

*

*

x a m , L / g µ e d i r o l h C

a i r e t c x a a B m , c L i h / m p u o r f t c o t r e n u t e o H C

*

*

*

*

s n i x o L t o m d / n I E U l r a o i r L e m t c / a U B E

*

*

x a m , m p p s e t a r t i N

*

*

*

x a m , b p p m u i n i m u l A

x a m , m p p s l a t e M y v a e H

s e t u b i r t t A c i n a g r o n I r e h t O


s d i o l l o C d n a e t a l u c i t r a P


) d e u n i t n o c ( s e i r o t a r o b a L l a c i t u e c a m r a h P n i d e s u e b y a m t a h t s e d a r G r e t a W l a i e p o c a m r a h P n o N s u o i r a V r o f y r a m m u S n o i t a c  i c e p S : 2 . 9 e l b a T


e d i v s

I d ) S E 6 L h 0 0 C t ( 4 2


*

*

6 9 1 ) 5 6 D 0 M 0 2 T ( S A



s s i n a d n o a i e t t d d r e e e a t c i e c l i m n r p i o L f p a

V I e p y T


*

I I I e p y T


*

I I e p y T


*

I e p y T

I , D t - l i m n o B c e i / t M F c S r l a ; m * µ u i l ) . µ o t 0 S v 2 s i . 2 i u D q 0 < ( + E

3 e d a r G


*

2 e d a r G


*

1 e d a r G


*




( q



s e s a e t o r P , s e s a e l c u N

, r e p p o C , t l a b o C , m u i m o r h C , m u i m d a C , c i n e s r A , m u i n o m m A , . m e u d n i r i o m l u u F l A d s n a a , h e c t u a s f l . s u n n S o , o t i t i e a a t a c i l c h r p f p p o s o a L e / h h g µ P t , r 1 t e o d a f e t r e e i t N a c i x , r e e p t d o o i r n r p t o p s l a u h C s i m s r s a e n t a i o h c w c u i d n s n a s a g n o t r e o i n n a m i f d e r o n a n a , s i o c n t n o a i i r t t a n Z , c e m  i c u c n i e o n t c t a i p s l i m a T f i i t L s t o , o e e m u N i g t h i d r l o o o n o , n e S e h l h , t b c , m a t u e i c e l i l e b i s s p t a p a s s t A n o o r t e p P t t l o l , N a o e , h n k d t c i s i e r w i N i g u d i n , q e t m e c s i u t e u R d t o y s o r t e n N p i v g , e i t a d b i s M e s ,  i y e d r a c e m e a e p r i n L l S t e , : t n n o s o a i f I r e N W I

t o . . N * 1 2




s e i r o t a r o b a L l a c i t u e c a m r a h P n i d e s u e b y a m t a h t s e d a r G r e t a W l a i e p o c a m r a h P s u o i r a V r o f y r a m m u S n o i t a c  i c e p S : 3 . 9 e l b a T

/ . v s r i u s e q e t a ; e c o r W e c o r p g r n r n u i o i o o t i k S r n a e i l l r i p D t s u s i D

n o i t c e j n I r 4 o f 3 r a i e t e p ) a o 1 W c 1 a 0 m 2 r ( a r h e P t a S W U d e  i r u P

6 1 a i e o p o c a ) 1 m r 1 a 0 2 h ( P e s e n a p a J

0 . 7 a i e o p o c ) a 1 m r 1 a 0 2 h ( P n a e p o r u E


s r s e e t a ; c e r W c o P g r e n u i l o k S b a n t i r i u D S

*

*

n o i t c e j n I r o f r e t a W

F d ; U e e O  i r c r R u u r P o o r S n o r i o e t r t a e a l t l a W i t s W i D

r e t a W d e  i r u P

, n f o o i ; t a e e a r r e c i z o r t , h u n n o o i o n o S i i e t t a a r D l e l , i i n t F t a s b W U , i D m O o R c

d e  i r r u e P t a y W l h g i H

s I r + e u D e O q d t a ; R i n n a s h W e c s c F g r e a U t n u i o p - e s k S l a n 2 , i . b r a h c t . i D g e u u s s

n o i t c e j n I r o f r e t a W

r e t a ; n o i W e c t a g r l l u i n o t i k S s n i i r D D

r e t a W d e  i r u P

e r O l e R b t a s a ; r t o i d W e I u c o g r D s h * t n u r r i e e o k S o . h t m n t i r i s o D D r o


d s e n h a e c a c r o u r o p p S A d n e o  i i c t e a c p  S i r u P

) 1 e t o N (

3 . 1

) 1 e t o N (

3 . 1

*

*

) 2 e t o ) 1 N e ( t o N (

*

e v i s u l c n i ( C ° 5 2 t a e u ) l a e v g H n p a r

*

*

*

*

*

*

*

O N

0 ) 5 . d r 3 e t 0 a t o n d ( e t o n N ) m e t a b u r s s p p t L s n / 0 n o p g 0 I s m 5 ( e r

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

) d ) 3 e t e a L t o N d / ( i ) l g b a O v m p N i f 5 p . 0 , S 0 0 5 E ( Y (

*

*

*

*

*

*

*

*

) d e t a d i l O a v N f i , S E Y (

*

*

*

*

*

*

*

*

) d e t a d i l O a v N i f , S E Y (

] b u L S / g ) x m b p O 5 . p o 0 0 t l 0 t 5 [ ( A

t s ] e t C o O t e T * o v t i t a t l g A e [ N

*

*

*

*

*

, C ° 5 2 @ m c Ω M y t i v i t s i s i e n R m

? d t e n t a e s m n e e r p u s m a o e C M y e r t u i v t i t a r c e u p d m n e o T C

x a m , ) C s a ( C O T

) . t s b u s S e . c d n e a t s R e b t u S a n e a l b g n a z a i d m r i x e O P (

d , n t a h g x a m n e m n L , 4 s 5 h t i 2 t a n t P a l U e a e c i c c n t n a p a b b r O r o m o s s b c b A 1 A

n / o g i t a m r , o C p ° a 0 v E 1 1 r t e a t f g a i n e t x a u a e m d i H , s e n g R o K

n o i t a r , o C p ° a 5 v 0 x e 1 a r t m e a , t f g L a n m e i u t a 0 d e 0 i h 1 s / e n g R o m

x a m , L / g m ) 2 O i S s a ( a c i l i S

x a m , L / g µ a c i l i S l a t o T

*

*

P ] J 1 n 2 i . h * C 1 . o 2 3 f . n 1 I [

) 2 e t ) o 1 N e ( t o N (

*

O N

0 ) 5 . d r 3 e t 0 a t o n d N e t o n ( ) m e t a b u s p r s p t L 0 s n / o n 0 g p I s 5 e m ( r

P ] J 1 n 2 i . h * C 1 . o 2 3 f n . I 1 [

) 1 e t o N (

3 . 1

) 1 e t o N (

3 . 1

) 1 e t o N (

1 . 5

, C ° 5 2 @ m c / S µ y t i v i t c u d x n a o C m

) 3 e

*

t o L / ( N g ) b m 0 p p 5 . 0 0 0 5 (

) 3 e

*

L t o / N ( g ) m b 0 p p 5 . 0 0 0 5 (

) 3 e t

o L N / ) ( g b m p 5 . p 0 0 0 5 (

) 3 e t o N (

x a r m , e L t t / g a m M t e n l e b t a z n i o d i C x 2 O O




The water grades denoted in Tables 9.2 and Table 9.3 frequently do not specify limits for all the groups of impurities. This is particularly marked for the USP, EP, and JP specications for grades of pharmaceutical water; such grades may not be appropriate for all laboratory related uses. For water to conform to a particular application, its suitability for the intended use should be veried and the continuation of purity should be ensured. In addition to their purity specications specications,, waters which need to meet regulatory requirements may have additional requirements, including: •

which feed water is to be used

•

limitations on how the water is to be puried

•

how and where the purity attributes are to be monitored

•

how the monitoring equipment is to be calibrated

•

how the purication system performance is to be trended and maintained

Where a particular type of water is specied, e.g., when using an ASTM method, these requirements must be met by the purication system selected. Where laboratory water quality needs differ from those in manufacturing (e.g., microbiological limits), it may be necessary to establish laboratory water specications to be met to avoid requiring the laboratory to use water attributes and limits to manufacturing water standards that are not required.

9.3.2

Quantity Needs The water demand prole is needed to size the water generation system(s) and determine if any distribution loops are needed. The water demand/diversity information should include volume delivery information, including maximum demand, batch or continuous delivery needs as well as maximum and average ow rates, the time required to deliver the specied quantity of water, frequency of use or operation, pressure, and temperature. These values should be obtained or estimated in real time, as well as over an operating shift or day day.. Normal and worst case water usage proles should be prepared for 24-hour and 7 day periods. This information, along with diversity of use, dictates the design specication for the system(s).

9.3.3

Data Collection Table 9.4 provides an example table for data collection.




Table 9.4: Example User and Water Data Collection Worksheet User Location (Note 1)

Lab Analysis or Equipment

Water Quality Need Regulatory Requirement if Applicable (E.G., GMP GMP,, GLP,, ISO, GLP I SO, CLSI)

Purity of Water Needed as Required by Equipment or Procedure

Water Quantity Need ) 2 e t o N (

k s a T r o . p i u q E f o e n t a i o R t r e a s u U D

e s U r e p e m u l o V d e t a m i t s E

Comments

) 3 e t o N (

y a D r e p s e s U f o r e b m u N

(Note 3)

y a D r e p d e s U e m u l o V

d e d e e N . p m e T r e t a W

d e d e e N e r u s s e r P r e t a W

Notes: 1. Floor, room, and area (e.g., tech support or QA). 2. For example: glassware washer, pipette washer, washer, autoclave, or task/analysis. 3. Often several uses may be always sequenced together. together. For example, glass washer and autoclave. This must be noted in the comment section to ensure that Step 1 (glass wash cycle) is not considered without Step 2 (autoclave wash cycle) in a demand prole.

Water use information should be compiled and grouped according to the layout and common test requirements. For example, one POU may be used to supply water for several laboratory analyses.

9.3.4

Monitoring Needs Monitoring requirements for water purication systems should be assessed. It should be established that a water purication system produces water of suitable purity, purity, which is t for purpose. Testing should be performed to ensure good component operation and to minimize operational cost and quality risks; however, it may not be practical or cost-effective cost-effec tive to monitor all, or even many, many, potential impurities after each stage of purication in laboratory equipment. Tests to ensure that the water purication and distribution systems are working adequately usually are performed in addition to the measurement of specic control checks (e.g., baseline blanks) as part of test procedures. A general approach is shown in Table 9.5.




Table 9.5: Monitoring Requireme Requirements nts for Types of Impurities Impurity Type

Monitoring

Type

Location

Frequency

Ions

Conductivity or Resistivity

In-line

Pre/Post RO, post IX, post EDI, product

During operation

Organics

TOC

On-line or off-line

Product

During operation or occasionally

Particles

Particle count*

On-line

Product

Very rarely used

Bacteria

TVC, epiuorescence

Off-line

Product

Typically weekly (use dependent)

Endotoxins

LAL

Off-line

Product

Occasionally (use dependent)

Large Bio-active Molecules

Specic tests

Off-line

Product

Occasionally (use dependent)

Gases

Specic tests

On-line

Product

As required

*Normally managed by ltration.

Key parameters that should be measured or used include: •

resistivity for ionic contamination

•

TOC for organic impurities

•

total viable counts as a measure of bio-burden

Where biological contamination is critical, measurements of endotoxin and nuclease levels also can be valuable. Concentrations of weakly-ionized silicon and boron species may be controlled by suitable system design, and if required, specic monitoring. Generally, parameters are measured in the product water or as close to the dispense point as practical. This Generally, gives values for the water actually used and also avoids breaking into the water purication circuit. The exception is resistivity which is measured in-line at several points. It provides an indication of RO, EDI, and ion exchange performance and can be measured after, and on occasion, before (calculation of % ionic rejection) these technologies. The frequency of off-line measurements varies considerably considerably.. It should be based on an assessment of the effects of loss of water purity and the likelihood of a problem. Condence, obtained by regular, relatively frequent logging initially,, enables the period between analyses to be decreased with time. Particular care should be taken after initially activities that may introduce impurities, such as changing components and sanitization. Based on system design, scale differences, or regulatory requirements, the same design concepts normally are not applied for similar functionality exactly or systematically to both large (production) systems and small (laboratory) water systems. Parameters monitored in large water systems may not be systematic systematically ally measured in small systems. Comparing the rationale of monitoring to the constraints involved, e.g., the importance of making a microbiological measurement between two purication stages that are located close to one another, compared to the risks/const risks/constraints raints of installing a sampling port between the two purication elements, may be used as justication for the difference in approach. This difference in approach also could be linked to the water system design selected. A laboratory water system should not be considered as a “black box.” The control of intermediate water purication operations in the




system allows the anticipation of problems, and subsequently improves compliance with good practices; therefore, based on the regulatory requirements, an analysis of the equipment design/content may be required to ensure compliance. The extent of monitoring required depends on the size and complexity of the system(s) (see Section 9.5 of this Guide).

9.3.5

Compliance Laboratory equipment installed in pharmaceutical laboratories or used for laboratory applications may require compliance with specic regulations including: •

Quality management systems (e.g., Good Laboratory Practices, GMPs)

•

Pharmacopeia Pharmaco peia (e.g., USP USP,, JP JP,, EP (Refere (References nces 4, 5, and 6, Appendi Appendix x 1))

•

CFRs or local regulations

•

Other water standards

Additional regulations also may be requested (e.g., CE, UL marks) as a general organizational policy requirement or for local legal compliance; therefore, it is important to dene the list of compliances required to select the nal laboratory water solution.

9.3.6

Laboratory Environmen Environmental tal Needs A comprehensive review of all the laboratory work spaces requiring access to laboratory water should be performed to select suitable systems, which meet user requirements, including: •

quantity of points of use needed

•

location of the different areas needing water and the possibility to group them ,

•

location of the POU (e.g., rooms, elevation, at sinks)

•

use constraints dened by some procedures including:

•

-

maximum distance to the use location

-

lling large containers

-

large ow rates

site-location constraints including: -

limited access (e.g., cleanroom, off hours use)

-

contaminating atmospheric conditions (e.g., volatile organics, corrosive vapors)

-

noise considerations

-

utilities available (e.g., electricity electricity,, source water, drain)

-

space available




9.3.7

Costs A comprehensive cost analysis that includes the direct and indirect life cycle costs of the proposed laboratory water choice(s) should be developed by the laboratory water owner. Direct life cycle costs for the system include, e.g.: •

•

capital costs including: -

engineering

-

equipment and material procurement

-

construction/installation construction/ins tallation

-

commissioning/qualication commissioning/ qualication

operating costs including: -

consumable (e.g., lters)

-

utilities (e.g., water , energy)

-

calibration

-

maintenance

-

training

Note: more Note: more automated systems may be more expensive to purchase and less expensive to operate. Indirect costs are an estimate of the potential costs in case of problems. Such an analysis should cover cover,, as a minimum, the costs associated with water quality or quantity problems (e.g., when facing a lack of suitable water during several hours or days).

9.4

Water Purifcation Technologies There are various combinations of purication technologies with which it is possible to produce the waters that may be required in laboratories. The technologies are similar in type to those used for production purposes, usually on a smaller scale and different in detail. For further information, see Chapters 4 and 5 of this Guide.

9.5

Laboratory Water Supply Options There are a range of means of providing pure water for laboratory applications. For applications of low volume or which are localized to one area, the choice usually is between a single dedicated water purication system, typically fed from municipal water, and the use of packaged water. For larger scale installations, a water purication solution can be based on the combination of one or more approach, including: •

a purication system with permanent extensive distribution system to the entire building



9.5.1


•

one or more local purication systems with local distribution loop to a oor or laboratory

•

small purication system with local storage and loop

•

individual point-of-use units supplied with city water

•

an extension of a distribution system used for production of pharmaceuticals

•

water polishers or dispensers fed from any of these distribution systems

Water Generation System and Distribution Options When water is required at multiple locations or when large volumes of water are necessary necessary,, several different design possibilities exist: Large – Central System for Entire Building A large make-up system for which the storage is in one location, with PW distributed throughout several laboratories or oors, can provide water throughout an entire building. These systems typically are custom designed. POUs can be connected directly from the distribution loop or the loop water can be treated by local polishers (units designed to further purify pre-PW) to give enhanced purity to meet special applications. This approach also enables the distributed water in the loop to be of lower purity. Use of multiple make-up water purication systems providing PW to the same storage reservoir and distribution loop and located in the same location provides some redundancy at the primary make-up level, reducing the risk of completely shutting down a facility. facility. When one system is down for routine maintenance or service, the other system should continue to provide PW. Figure 9.1: Example of Central System




Localized – Central System for a Floor/Laboratory Several local loops (with each loop providing water oor by oor or department by department) enable each smaller system to be designed to meet specic local requirements. For example, when a high volume of water is required at a specic location (e.g., dishwashers and washing) and demand is much less in other locations, a better approach may be to dedicate a system to this high demand location rather than trying to provide for this in the complete distribution network. Medium sized generation systems typically are standard pre-packaged units with a custom designed distribution system. Figure 9.2: Example of Localized Systems

Small – One or More “Individual” Units per Building, Floor, or Laborat Laboratory ory POU systems (often referred to as bench top units), can provide a variety of different water qualities and quantities at a much smaller scale than the large production or centralized systems. In practice, they can be located in a range of locations, e.g., on the wall or under bench, and can include storage and local distribution. They also can include various monitoring and dispense options. They typically are pre-designed, pre-packaged, and compact units. The use of distribution loops may be avoided, while meeting multiple applications, with multiple individual POU units, which include the make-up purication system, storage, and additional polishing, as required. When compared with packaged waters, the advantages of POU systems include purication and monitoring of water on an on-going basis and that it is available on demand. For instruments or analyzers used for manufacturing purposes, calibration and maintenance are vital to increase the probability of ongoing good operating and performance conditions. System designs which facilitate such operations can be advantageous.




Figure 9.3: Example of Small Stand alone Units

Note: in addition to approaches shown in Figures 9.1, 9.2, and 9.3, another approach is to use large or medium loop Note: in systems for most applications and address specic needs using individual POU systems or local loops, which include the make-up purication system, storage, and additional polishing at a much smaller scale and as required. This approach should avoid the need to extend piping to all departments, potentially simplifying simplifying the design of the main water purication system. Supply from a Manufacturing Plant When PW is available from an adjacent production area, this could be distributed through a loop to the laboratory block, or locally as an extension of the loop. The water could be treated to meet specic local requirements. Note that the distribution piping, if extended to the laboratory laboratory,, may have more design, quality, and maintenance requirements (e.g., microbiological) to maintain, than a dedicated laboratory system. Contamination risks from a multipurpose system (e.g., manufacturing and laboratory) should be considered.

9.5.2

Packaged Water Packaged water is a possible alternative source of laboratory grade water. There are two general applications and associated purity levels typically used: 1.

Very high purity waters for specic analytical purposes where small volumes are needed.

2.

Less pure waters considered to be equivalent to compendial PW for general laboratory use where larger volumes typically are used. The latter types of packaged waters may be used in small scale manufacturing and process development applications that may be considered as “laboratories.”




9.5.2.1 Simplicity and Cost Advantages of Packaged Packaged Laboratory Waters The potential simplicity of using a packaged form of water may seem a more cost effective and suitable option to satisfy a relatively small volume and perhaps a high purity requirement, when considering the capital and operating costs, as well as complexity of designing, installing, qualifying, maintaining, and monitoring a water purication system, regardless of size. Packaged waters may be purchased in containers of manageable size and packaging congurations that allow their integration into laboratory operations in a similar way to other laboratory reagents, complete with expiration dating, lot numbers, and documentation certifying purity or compliance with standards. Allowed contaminants contaminants leaching from the packaging into the water may not be consistent with the implications of the name of the water and this should be considered. A risk assessment assessment may be performed to determine the potential impact of the contaminants on the uses of these waters; therefore, the potential cost and complexity of such risk-based assessments assessments may need to be included in the decision process to assure that the packaged water approach will be acceptable. 9.5.2.2 Purity Compendial monographed sterile packaged water purity requirements are dened in the USP with less stringent inorganic specications and less sensitive organic contaminant tests. The EP and JP also use less sensitive wet chemical tests for these packaged waters. These less stringent specications allow the presence of organic or inorganic packaging leachables that could exceed the specications for bulk PW, the minimal purity required by most compendia for pharmacopeial testing. A review and investigation of the purity documentation (Certicate of Analysis or Certicate of Compliance) should be performed to determine if the stated purity (or compliance with compendial specications)) relates to the source water placed into the container specications container,, or whether it relates to the guaranteed purity of the water within the container throughout its shelf life. PW packaging often is plastic or may have elastomeric closures or container entry ports, all of which may leach organic plasticizers, molding releasing agents, or glues and associated solvents and monomers into the water. The organic extractables may become evident with escalating TOC levels over the water’s shelf life. Some elastomers, as well as glass packaging materials, may be prone to leach inorganic ions into the water, degrading its conductivity and pH. Inert packaging materials which usually are used only for special purposes or more costly waters intended for specic analyses such as HPLC, may be an exception. Packaging leachables are partially controlled by the specication specications s for the waters (because of selective and broad inorganic limits and a selective organics test insensitive to many extractables). There is potential for variability in the levels of packaging leachables: •

over the shelf life of different container types

•

from supplier to supplier

•

potentially from batch to batch of a given supplier

This potential variability in purity may require risk-based assessments for containers to assure suitability suitability.. If the water’s microbial content is a concern for laboratory applications, any claimed bioburden level by the packaged water’s quality documentation is likely to be a transient attribute, and compromised as soon as the package is opened. 9.5.2.3 Verica Verication tion of Suitability of Packaged Packaged Waters Waters If a packaged form of PW is being considered for use in place of bulk PW produced by an on-site water purication system and the purity specications are not identical, the laboratory should verify tness for use in each of its applications (e.g., as per USP Chapter <1231> (Reference 4, Appendix 1)). In-house specications (e.g., TOC TOC




and conductivity) should be established for these packaged waters, tested on samples taken from representative containers, rather than basing acceptance upon the Certicate of Analysis, to help ensure that unsuitable packaged waters will not be used in susceptible applications. Suitability verication may not be required if the packaged water is being used for an analysis for which the water was specically labeled and intended, such as water for HPLC analyses; despite this, such waters may not be suitable for all HPLC analyses. A risk-based approach to assessing overall application suitability is considered appropriate. As soon as the package is opened and air is allowed to enter the container container,, the purity starts to degrade. For packaged waters, it is recommended that water from these containers is used for only a short period of time after opening unless, the purity degradation the water will experience has been shown to be inconsequential to its laboratory applications. If the water required for a given laboratory’s operations is not a compendial grade, but rather a grade specied, e.g., by ASTM, ISO, Clinical and Laboratory Standards Institute (CLSI), or similar organizations, the suitability of packaged forms of these waters for the user application should be veried, unless certied for a specic purpose by the manufacturer of the packaged water. In addition, because some of these water grades have microbial and endotoxin specications in addition to their chemical specications, consideration should be given to the potential negative impact on these attributes from how water is removed from the package and by how long the opened package is kept in use. 9.5.2.4 Balance of Cost and Suitability Packaged waters may be appropriate in laboratories, particularly where there is a need for a minimal amount for specic compatible applications. applications. Where water is more generally needed for analytical applications, the risks and cost of assuring suitability for each application, potentially for each new batch or shipment received, should be balanced against the cost of installing and maintaining a purication system to produce bulk water, the application suitability for which may be related directly to ongoing quality monitoring.

9.5.3

Related Consideratio Considerations ns

9.5.3.1 Quantity Needs Large distribution congurations usually are considered when PW is required at different locations or in large volumes. It may be practical to meet high local ow or pressure requirements by additional local pumps and reservoirs, even with a large distribution system. The distribution design should maintain water quality provided by a centralized system. The general advantages and limitations, described in Chapter 8 of this Guide, for the different distribution congurations (loop, etc.) could be directly applied for the selection of the distribution conguration used in a laboratory. 9.5.3.2 Quality Needs In general, a large system should be considered when an equivalent water quality is required in a variety of locations. Specic higher water purity requirements can be accounted for by individual polishing equipment. As the number of different water qualities required in an area increases, so does the importance of individual POU systems to account for specic needs. Several local distribution systems may be more effective in meeting diverse water purity requirements, if they can be localized. Designing a large system that will produce water based on the most stringent water purity application requirements, may be an alternative approach; however, this solution is considered unlikely to balance the advantages and limitations (costs, complexity complexity of maintaining high water quality in a distribution system). As the length of the distribution increases, the greater the difculty in maintaining the equivalent levels of quality at all points, and the greater the risk of contamination (particularly bacteriological). The distribution design, materials in contact with the water and components installed, etc., can affect differences in quality signicantly.




9.5.3.3 Monitoring and Compliance Needs The greater the degree of equipment centralization, the more monitoring controls are localized. Time spent on checking/controlling checking/cont rolling satisfactory equipment operation (when compared with individual POU systems); may be optimized by localization; however, providing monitoring information to users (quality parameters, etc.) may be important. In a centralized conguration, the installation and associated costs of remote control/alert or individual monitoring equipment at the points where such information may be important should be considered. A system that will distribute water to different locations with different regulatory requirements should align the main part of the installation requirements (design, equipment selection, maintenance program, etc.) to the most stringent regulatory requirements. 9.5.3.4 Environmental Constraints The overall laboratory architecture and that of the various departments and oors in a building should be analyzed. If a single large system is chosen, its distribution should comply with typical design practices (see Chapter 8 of this Guide) and it should be veried that successful installation is possible (wall supports, etc.). The total length and height between the lower and upper distribution points will have a signicant impact on the nal distribution design, and on whether the water quantity and quality can be delivered consistently as requested. If large-scale storage and distribution are considered for an existing building, the available architecture (door openings) should be able to accommodate integration of the solution (storage, etc.). Exposed distribution system pipework can provide needed exibility for distribution system modications. modications. The noise level should be taken into consideration when choosing the location where a water system will be installed, as equipment delivering large volumes of PW is in general noisier than standard laboratory equipment. These factors should be studied when a new construction is planned. Space available within the laboratory may be restricted. Distribution equipment and local units will use bench/wall space, or storage space if located under-bench. Placement of these alternatives should be considered. 9.5.3.5 Maintenance Needs In a single main installation, the advantage of having less systems to maintain may be offset by the complexity of the installation and the distribution system. A trained trained technician may need to be available to deal with issues in a timely fashion. The risks of problems usually increase with the complexity and length of the distribution; therefore, more frequent maintenance may be required, and depending on the conguration selected, this may affect laboratory operations. 9.5.3.6 Risk Management The consequences of a lack of water caused by preventive or corrective maintenance should be analyzed and compared to the user requirements at the laboratory, departmental, departmental, and oor levels. A sanitization or a repair of the distribution system will not have the same consequences if the installation is a centralized system and distribution is for the entire building, rather than for a department or a laboratory. A duplex (multiple) approach minimizes the risk of interruption of water supply; however, however, it adds additional costs and is limited to cover the problems linked to the make-up system. Distribution complexity, complexity, length, operating conditions (pressure, temperature, equipment, etc.) will be directly related to risk management and should be considered before making the nal conguration decision. For example, horizontal distribution (one oor) compared to vertical distribution (over several oors) will minimize pressure-drop related issues and optimize the linear velocity of water in the pipes. 9.5.3.7 Users’ Convenience The location and quantity of the PW POUs may depend upon user requirements such as: •

the time spent to get water




•

frequency at which water is required

•

the risks associated with a decrease in water quality during transportation

The nal solution proposed should take into consideration the potential water purities required and provide an optimized way of providing users with water of the appropriate quality. 9.5.3.8 Future Laboratory Expansion or Modications Modications Information on future modications of the installation should be considered when making nal design/conguratio design/conguration n selections. As As distribution complexity increases, so does the level of risk, difculty, difculty, and expense associated with changes to the laboratory’s distribution to serve future needs. The possible need to perform a revalidation could add complexity and cost. Smaller size production units also allow an easier renewal program, as the investment can be planned over several years, renewing one smaller installation at a time. 9.5.3.9 Costs The total costs equal the combination of the capital investment, operating, and risk management costs. In addition to these total costs, the nancial investment strategy of the organization (high capital – low operating costs) will have an impact on the nal decision. Initial investment will depend on site specic factors; in general, local systems are likely to be less expensive if there are few users who are widely distributed and with disparate water requirements. A large or medium distributed system is likely to be more cost-effect cost-effective ive for large volume usage of a similar water quality within one building. 9.5.3.10 Comparison The nal choice should be the result of an analysis based on the advantages and disadvantages of the various congurations in regard to the quantities and qualities of water required, geographic distribution of these requirements, budget, etc. Establishing these needs is critical. The strengths and weaknesses of the various congurations are summarized in a Table 9.6. These comments are for guidance only as other factors may drive the nal choice. Table 9.6 provides a comparison of the advantages and limitations of various system design options, which provides an overview of which factors should be considered when going through the selection process.




Table 9.6: Comparison of Laboratory Water Delivery Schemes Sche Sc heme me De Desi sign gn At Attr trib ibut ute e

Larg La rge e – Ce Cent ntra rall System for Entire Building

Localized – Central System for a Floor or Laboratory

Small – One or More “Individual” Units Per Floor or Laboratory

Supply From Manufacturing Plant

Packaged Water

Av A vailability

Custom designed purication and distribution system

Prefabricated purication unit plus custom distribution system

Prefabricated purication unit with local POU(s) or very limited distribution

Distribution system connected to or part of manufacturing system

Commercially available from scientic supply houses/catalogs

Capital Cost Per System

Large

Moderate

Low

Low – High (loop design dependant)

N/A

Capital Cost Per Unit Volume

Low – Moderate

Low – Moderate

Moderate – High

Low – Moderate

N/A

System Owner/Financial Responsibility

Site Dependant

Department

Department/Project

Manufacturing

Depar tm tment /P /Pr oj oject

Operating Costs • consumables • calibration requirements • monitoring • utilities • maintenance • revalidation

Site Dependant

Site Dependant

Moderate per unit volume

Site De Dependant

High pe per un unit volume

Failure Impact (quality or supply)

System Dependant – typically large to moderate

System Dependant – typically moderate moderate

System Dependant – typically limited

System Dependant – potentially large

Typically Limited

Maintenance

Specialized Maintenance Personnel

Experienced/ Trained Personnel

Trained Users

Manufacturing Responsibility

None

Design and Construction

Complex

Moderate

Low

Complex

None

a. Space

Large, but only 1 area

Medium in support area

Small, but could be multiple

Distribution piping only

Transfer/storage area and quantity dependant

b. Noise

Very noisy, but only 1 area

Medium in support area

Low

Low

None

Water wastage

Technology Dependant



Low

None

Poor feed water quality management

Treatment at central unit (single point)

Treatment at each unit

Treatment at each unit


N/A

Upgradability (e.g., changed usage or quality requirement)

Replace/upgrade or add – signicant engineering needed

Replace/upgrade/ add – some engineering needed

Replace/upgrade or add unit

Su pp ppl y De pe pen da dan t

B uy uy m or or e

Relocation

Difcult

Medium

Easy

N/A

N/A

Revalidation

Complex

Moderate

Easy – Moderate

Easy – Moderate

N/A

Importance of Supplier Certication Program

Variable

Variable

Variable


Critical

Quality Variability at different use points on a loop

Desi De sign gn de depe pend ndan antt

Desi De sign gn de depe pend ndan antt

N/A

Low

N/A

Location




Table 9.6: Comparison of Laboratory Water Delivery Schemes Schemes (continued) (continued) Sche Sc heme me De Desi sign gn At Attr trib ibut ute e

Larg La rge e – Ce Cent ntra rall System for Entire Building

Localized – Central System for a Floor or Laboratory

Small – One or More “Individual” Units Per Floor or Laboratory

Supply From Manufacturing Plant

Packaged Water

Management of Water System by User

Access Limited/ Remote

Generally Accessible

Local

Access Limited/ Remote

N/A

Need for Water Usage Suitability Vericatio Verication n

Low – during qualication then by meeting specs in routine sampling




High (potentially with every batch or shipment or even every container)

Assurance of Quality at time of Use

Validation, periodic sampling, and centralized monitoring

Validation, periodic sampling, and centralized monitoring

Localized monitoring and sampling

Manufacturing group responsible for validation, periodic sampling, and centralized monitoring

Refer to in-house testing and quality control process

Sanitization

Whole system shutdown

Local system shutdown

Individual unit

Whole system shutdown

N/A

Regulatory Requirements for System Management

Dened by the most stringent requirement in the whole system

Dened by the most stringent requirement in the localized system

Dened by specic local requirements

Dened by the most stringent requirement in the whole system

Dened by specic local requirements

Meeting Users Water Quality Needs

Distributed water quality established for the building, polish at POU as necessary

Distributed water quality established for the local needs, polish at POU as necessary

Water quality established for each POU

Water quality established by manufacturing, polish at POU as necessary

Water quality established by suitability verication

N/A = not applicable

9.6

Maintenance Preventive maintenance on a laboratory water purication system usually helps to anticipate problems and ensure the long-term performance and reliability of the water purication system. A maintenance program should be dened and its denition should be based on the manufacturer’s maintenance recommendations and on a risk impact analysis. A maintenance program may include the replacement schedules for consumables and spare parts caused by wear and tear, preventive verication/tests on critical components, and the denition of the monitoring program. Denition of the monitoring program could reiterate the operating ranges, the alert and actions levels that were dened for the different parameters, as well as maintenance actions to be conducted when values are out of range. When equipment is not used in a regulated environment, the actions performed on water systems used in a pharmaceutical laboratory should be documented. This traceability allows greater efciency during any future troubleshooting actions. When the equipment is used in a regulated environment, or depending on its classication in terms of criticality, criticality, the level of regulatory practices and the preventive maintenance frequency will be different.

9.7

Instruments and Calibration Calibration is required by quality management systems (pharmaceutical GMP GMP,, ISO 9001, etc.) on equipment used to measure, control, or monitor. The reliability and condence obtained from a value provided is linked to the calibration method, calibration frequency, and results obtained.




The measurement of quality parameters, e.g., conductivity conductivity,, TOC, and bacteria, after each purication step, may not be relevant in a laboratory water system when the distances between two purication stages are minimized and the risks of contamination through insertion of “sanitary” sampling instruments may be higher than the anticipated benets. The measurement of water quality and operating parameters after major water purication steps in a laboratory water system may help to anticipate operating and performance problems. Frequency of calibration should be based on the type of measurement performed (e.g., temperature, conductivity conductivity,, TOC, pressure), the type of instruments used (method, model, etc.), and the importance of the measurement for the application, in addition to the importance of the application in the entire process. The frequency should be dened following a risk impact analysis approach.

9.8

Commissioning and Qualifcation Commissioning and qualication of water systems are discussed in Chapter 12 of this Guide. The same qualication approach should be followed for laboratory water systems should be qualied using equivalent approaches to those for the manufacturing systems. If the laboratory water system is an extension of a manufacturing distribution loop into the laboratory area, the issues and testing that apply to the validation of the water system for manufacturing also apply to the POUs in the laboratory area. Where the water system is exclusively for laboratory use, the specic purity attributes required in the laboratory may differ from those required for water used for manufacturing. These laboratory purity differences (or impurity allowances) may be reected in the use of different or considerably scaled down purication unit operations, and different system materials, distribution system system designs, and POU valve types than are typically present in manufacturing water systems. Small, tightly packaged bench-top or wall-mounted water purication units may be designed to operate to the point of reduced quality (e.g., exhaustion) from one or more purication modules; the point of reduced quality usually is signaled by a built-in sensor. There is a risk of producing and using unacceptable water associated with these types of laboratory water system designs; maintenance approaches should be evaluated prior to system selection and purchase, and challenged or veried during qualication. The high number of purity attribute, operational, maintenance, and design differences between manufacturing and laboratory water systems usually requires the strategy for the qualication process for laboratory water systems to differ from that typically used for manufacturing water systems.

9.8.1

Importance of Internal Laboratory Water Specication Standards The minimum chemical purity of water normally required for analytical purposes is compendial PW. A pharmaceutical laboratory PW system may not require a microbial specication, because only the water’s chemical purity is of consequence to a number of laboratory uses. Where the water purity attributes required from a laboratory water system differ from the purity attributes needed by and described by an organization’s raw material standards or monographs for manufacturing PW or water for injection(s) systems, separate specications specications to dene laboratory water quality should be developed. These specications may be the basis for the acceptance criteria for water system qualication processes; documented standards for the quality of waters needed in the laboratory should be dened. Where these standards are not dened, the laboratory may be required to apply water attributes and limits from manufacturing water standards. The manufacturing water attributes and limits may be unnecessary or not sufciently stringent and for which the laboratory water systems may not have been designed to control these appropriately. appropriately.



9.8.2


Tailoring Laboratory Water System Qualication to Intended System Capability Large laboratory water systems may be entirely custom designed and built. These systems can be designed to include special monitoring and sampling capability between unit operations. For medium to smaller sized laboratory water systems, smaller purication units with precongured unit operations usually are obtained as a single unit. These precongured units may provide less opportunity for sampling between modules for monitoring individual unit operations. Requirements to verify the performance of individual unit operations should be determined based on the overall system design and quality needs. Retrotting sampling capabilities for qualication purposes to prefabricated systems may not be feasible, as they often are compact, “well-tuned” systems. Smaller bench-top or wall-mounted single POU systems also may have limited user access to on-board gauges and instruments for calibration and standardization purposes. Where gauges and instruments are for informational purposes only (e.g., to signal performance changes in the system), and denitive QC testing is performed either off-line (with occasional grab samples) or by portable in/at/ on-line instruments that can be calibrated/standardized, then lack of access may not be a problem. Where the onboard gauges and instruments are intended for QC purposes, then user accessibilit accessibility, y, the ability to calibrate, and pharmacopeial compliance of instrument features are considered more signicant. Therefore, such capability should be investigated during the water system selection process in order to ensure compliance with user and regulatory needs. A systematic duplication of the qualication approach used for manufacturing water system designs may not be appropriate for laboratory water systems. The duration of the qualication process, which includes the PQ, should be customized for the evaluation of the laboratory water quality attributes and the nature of the water system’s operation. For example, if there are no microbial attributes of concern for the laboratory water, the only impact of biolm development would be to reduce the longevity or efciency of chemical purication unit operations. This would be apparent in the chemical quality of the efuent water. Without microbial attributes of concern in the nal water, the period of qualication and frequency of sampling may be tailored to evaluate the consistency of the chemical purication processes. processes. However, if the unit operations are not intended to be periodically replaced because of exhaustion, such as RO modules, periodic microbial monitoring may be of operational value, rather than being a quality requirement per se. The qualication of a laboratory water system may be more limited (or more stringent) than that for a manufacturing water system, and the protocols should be developed to accommodate those differences, as long as the total qualication process reveals the operational consistency of the water system for controlling the quality attributes of concern. If a specic water system monitoring feature or a particular quality attribute is an absolute requirement for an organization’s water system validation program, a laboratory water system that contains those features and is able to control the mandated purity attributes should be purchased and installed.

9.8.3

Special Validation Considerati Considerations ons for Small Laboratory Systems

9.8.3.1 Conductivity Compliance and Consistency Most small water purication units, especially the compact, prefabricated, single POU systems, contain an on-board conductivity/resistivity conductivity /resistivity instrument with its sensor positioned within the nished water stream. If the purication unit has been in a “stand-by” mode, system manuals often instruct the user to initially recirculate the water through the unit or ush water to drain until the conductivity conductivity/resistivity /resistivity reading gives acceptable values. If that instrument can be calibrated and complies with the compendial requirements, the readings can be evaluated against those compendial specications.. If, however, the instrument cannot be calibrated or otherwise does not meet compendial requirements, specications it can still be used as an informational instrument whose performance should be periodically compared or correlated to an instrument that does meet compendial requirements through the in-line testing of the same water. The




relationship of those data correlations can then impart a level of condence to the informational instrument readings for day-to-day operations. Since the on-board instrument is only an informational instrument, periodic testing on water from the system using a compliant instrument is needed to assure ongoing compendial compliance. Whenever, the product water does not meet those conductivity specications specications after appropriate recirculation or ushing to drain, it is usually a sign that a purication module within the unit should be replaced. After After such a replacement has been executed, the water generally resumes conformance with the conductivity specications. specications. Having procedures in place to assure proper use of the water system with appropriate precautions to preclude using unacceptable water is very important. Verifying that this exhaustion/module replacement/use resumption phenomenon reproducibly occurs also can be an element of the system’s qualication. 9.8.3.2 TOC Compliance and Consistency It is not universal for such compact, prefabricated purication units to contain on/at-line TOC monitoring. For those that do not have built in TOC monitoring capability, capability, the TOC test for such systems should be performed either on grab samples from the system or by connection, often at a use point, to a TOC instrument. The TOC purication capability of such a system is an attribute that could, like conductivity, conductivity, change dramatically in a short period of time because of the way this purication equipment may be operated with ionic exhaustion usually signaling a need for reactive maintenance. Without on/at-line TOC monitoring it is impractical to determine TOC with every use, unlike what is usually possible with in-line conductivity conductivity.. The consistency of TOC concentrations is important to validate, not only during routine use up to the point of potential TOC removal exhaustion, but also immediately after any purication module replacement when TOC spikes are not uncommon. Therefore, the duration of frequent grab sample or on/atline TOC monitoring during qualication needs to at least encompass a period long enough to verify the predictability of TOC removal exhaustion, which may occur either before or after ionic exhaustion. In addition, after a specied long idle period between system uses, the point in time after pre-use ushing or polishing recirculation when the monitored conductivity signals that the water is ionically acceptable for use, also should be veried as attaining acceptable T OC quality. 9.8.3.3 Consistency of Other Attributes This same approach as used for TOC should be applied to any quality attribute dened as important and not routinely monitored with every use. This attribute performance characterization could then be incorporated into usage and maintenance SOPs for the system (which may be different from those recommended by the purication unit manufacturer) that would preclude unwittingly making and using unacceptable quality laboratory water. The qualication process could then verify that these other attributes, though infrequently routinely tested, are consistently met when the system is operated appropriately. •

the risk analysis/risk control on the remediation (de-rouging) processes

There is no absolute answer to the question on “how to deal with rouge.” Rouge may be a risk, but the alternatives offer risks, too. This chapter presents basic information on rouge and an estimation of several risks, but cannot provide the risk estimation for every process or product. Information is provided on what rouge is and what the consequences of rouge may be for the water/steam system and production equipment. Information also will be provided on de-rouging processes and their consequences. Decisions should be made based on alternatives structured around specic existing systems and production situations.


10 Rouge and Stainless Steel




Page 165 Rouge and Stainless Steel

10 Rouge and Stainless Steel 10.1

Introduction The basic corrosion resistance of stainless steel originates in its ability to form a protective or “passive” layer on its surface. The formation of this layer (not “lm”) is instantaneous in an oxidizing atmosphere such as air, water, or other uids that contain oxygen. Once the layer has formed, the metal becomes “passivated” and the oxidation rate will slow down to inconsequential limits. The passive layer can be explained as the chromium enriched oxide layer that improves the corrosion resistance of the base metal. There is no specic denition for rouging, and rouge may be confused with localized corrosion (e.g., pitting corrosion) as similarly colored products are generated. Rouge in stainless steel systems utilized in the biopharmaceutical/ life science industry, industry, is a general term used to describe a variety of discolorations on the product contact surfaces, caused by variations in hydration agents and the formation of metallic (primarily iron) oxides and/or hydroxides from either external sources, or from alteration of the chromium rich “passive” layer. Rouge is an anomaly that has been generally perceived in the industry as a nuisance or source of harm to stainless steel product contact surfaces, rather than a contaminant that may or may not adulterate the WFI, PW, pure steam, raw materials, or nal products. Concern should be focused on whether or not the presence of rouge may be detrimental to the drug products, rather than its repercussions on capital equipment protection. Rouge is not aesthetically pleasing; it is a result of the use of a material that is never chemically identical in all its forms, contains a very high proportion of iron (approximately 65% to 70%), and is exposed to a highly complex set of processes and chemical, mechanical, and electromechanical electromechanical inuences. Other than switching to the use of higher alloys to avoid the problem altogether, which could increase capital costs substantially,, the best alternative to evaluating the true impact of rouge is considered to be to provide science-based substantially results based on investigative endeavors. The presence of rouge normally is more apparent in systems operated continuously or intermittently at elevated temperatures or exposed to the presence of halogen salts (usually not present in biopharmaceutical water systems), particularly chlorides which are one of the most common elements in nature; chlorides which are soluble, forming the basis for good electrolytes, can easily compromise the integrity of the passive layer and allow corrosive attack to occur. This chapter considers: •

rouge and rouge formation

•

rouge analysis

•

the risk of rouge and the risk of remediation (de-rouging) processes

•

the risk analysis/risk control on the remediation (de-rouging) processes

There is no absolute answer to the question on “how to deal with rouge.” Rouge may be a risk, but the alternatives offer risks, too. This chapter presents basic information on rouge and an estimation of several risks, but cannot provide the risk estimation for every process or product. Information is provided on what rouge is and what the consequences of rouge may be for the water/steam system and production equipment. Information also will be provided on de-rouging processes and their consequences. Decisions should be made based on alternatives structured around specic existing systems and production situations.




Figure 10.1: Structure and Approach of Chapter 10

10.2

Regulatory Stance

10.2.1 Food and Drug Administrat Administration ion (FDA) The FDA has no written position addressing rouging, its existence, or presence in high purity water, steam, and process systems. Their Their criterion is to meet established standards of quality for those systems. 21 CFR (Code of Federal Regulations) Chapter I, Part 211, Subpart D – Equipment, Section 211.65(a) 211.65(a) – Equipment construction reads “Equipment shall be constructed so that surfaces that contact components, in-process materials, or drug products shall not be reactive, additive, or absorptive so as to alter the safety safety,, identity, identity, strength, quality, quality, or purity of the drug product beyond the ofcial or other established requirements.” requirements.” (Reference (Reference 27, Appendix 1) 21 CFR (Code of Federal Regulations) Chapter I, Part 211, Subpart D – Equipment, Section 211. 211.67(a) 67(a) – Equipment cleaning and maintenance reads “Equipment and utensils shall be cleaned, maintained, and sanitized at appropriate intervals to prevent malfunctions or contamination that would alter the safety, identity, identity, strength, quality, or purity of the drug product beyond the ofcial or other established requirements.” (Reference (Reference 28, Appendix 1) The Adulterated Drugs and Devices – Federal Food Drug & Cosmetic Act – 31 December 2004 Amendment, Section 501 (a)(2)(B) reads: “A drug or device shall be deemed to be adulterated if… the methods used in, or the facilities or controls used for, its manufacture, processing, packing, or holding do not conform to or are not operated or administered in conformity with current good manufacturing practice to assure that such drug meets the requirements of this Act as to safety and has the identity and strength, and meets the quality and purity characteristics, which it purports or is represented represented to possess”. possess”.

10.2.2 United States Pharmacopeia (USP) The USP neither identies rouge as a contaminant nor proposes alert and action limits or methods for detecting rouge product streams; the US Pharmacopeia usually does not address design or material criteria directly, but rather indirectly by dening limits for the components that ultimately will enter the human body (Reference 4, Appendix 1).




The scope of the US Pharmacopeia covers the quality of the water that is used, not the system that delivers it, and rouging is a matter that relates to material selection for the system; conversely, conversely, design criteria are intended to minimize the risk to the water. The USP requires representative sampling; therefore, inferring the sample represents the water quality for the entire system and the water quality between sampling periods; the USP gives only specications to that quality. Additionally Additionally,, design criteria are intended to ensure this quality quality,, and to keep the process under control for a longer time (even if the water is not compendial water). The owner/user should decide if the water quality obtained from a system that shows rouge is still compliant with the USP as well as internal requirements for the process. Guidelines or regulations specifying the need/requirement to de-rouge or the frequency of de-rouging do not exist. At exist. At time of publication, there are no documented cases showing that the presence of rouge in high purity water or pure steam systems has resulted in the systems being out of compliance with current Pharmacopeias requirements.

10.2.3 European Pharmacopo Pharmacopoeia eia The European monographs do not address rouge or give guidance in this matter. There is one paper describing the maximum acceptable metal residues arising from the use of catalysts in the synthesis of pharmaceutical substances (Note for Guidance on Specication Limits for Residues of Metal Catalysts – Catalysts – EMEA – December 2002 (Reference 29, Appendix 1)). This guideline applies to all dosage forms, forms, but different different limits are applied to oral and parenteral parenteral routes of administration; it can be applied as a guiding document for risk assessment on heavy metals in products and/or water systems.

10.3

Surface Conditions and Treatments

10.3.1 Oxidation Oxidation is a common form of electrochemica electrochemicall reaction where one element yields an electron, while at the same time, another substance absorbs an electron; the complete process constitutes a redox reaction, which, in this case, is the combining of oxygen with various elements and compounds in metals or alloys in interaction with their environment, such as exposure or use. It can occur regularly and slowly, as in rusting, or rapidly, as in metal pickling. It can be benecial if it is involved in forming of the passive layer, or detrimental if it is involved in the formation of corrosion.

10.3.2 Corrosion Corrosion is the chemical or electrochemical interaction between a metal and its environment, which results in undesirable changes in the properties of the metal. This metal. This interaction may lead to impairment of the function of the metal, the environment, and/or the technical system involved. Corrosion resistance is one of the main reasons why stainless steels are used in biopharmaceutical/life science systems. systems. Each form of corrosion can be identied by mere visual observation, but sometimes magnication is helpful or required. Valuable information for the solution of a corrosion problem can often be obtained through careful observation of corroded test specimens (sacricial spools) or failed equipment.

10.3.3 Corrosion and its Variables It is convenient to classify corrosion by the forms in which it manifests itself; the basis for this classicati classication on being the appearance of the corroded metal and the specic cause for its appearance which can be either a chemical dissolution of the metal or an electrically (galvanic) driven process. Additionally, Additionally, whether the corrosion is derived from an active oxide layer metal such such as iron, zinc, aluminum, and copper (anodic or least noble end in the galvanic series of metals and alloys), or a passive a passive oxide oxide layer metal such such as stainless steel, titanium, gold, and silver (cathodic or noble end in the galvanic series of metals and alloys) should be considered.




10.3.3.1 10.3.3. 1 General or Uniform Corrosion General or uniform corrosion refers to the relatively uniform reduction of thickness across the entire surface of a corroding material. It is expressed as corrosion “rate” measured in mm/year or mils/year. Uniform corrosion can occur from an overall breakdown of the passive layer, and the “rate” of corrosion is inuenced by material composition, uid concentration, temperature, velocity, velocity, and stresses in the metal surfaces subjected to attack. It is considered easy to measure, predict, and design against this type of corrosion damage. General corrosion usually is controlled by selecting suitable materials, protective coatings, cathodic protection, and corrosion inhibitors. Stainless steels are subject to general corrosion in the presence of many acids and some salt solutions. 10.3.3.2 10.3.3. 2 Galvanic Corrosion Sometimes called dissimilar metal corrosion, galvanic corrosion, galvanic corrosion is an electrically driven process by which the materials in contact with each other oxidize or corrode. There are three conditions that must exist for galvanic corrosion to occur: •

the presence of two electrochemical electrochemically ly dissimilar metals

•

an electrically conductive path between the two metals

•

a conductive path for the metal ions to move from the more anodic metal to the more cathodic metal

If any of these three conditions does not exist, galvanic corrosion will not occur occur.. When two different metals are immersed in an electrolyte and connected through a metallic path, current will ow. Oxidation occurs at the anode and reduction (normally oxygen reduction) occurs at the cathode. In general, the further apart the materials are in the galvanic series of metal and alloys, the higher the risk of galvanic corrosion, which should be prevented by design. 10.3.3.3 10.3.3. 3 Crevice Corrosion Considered a form of galvanic corrosion, crevice corrosion, crevice corrosion is a localized corrosion of a metal surface at, or immediately adjacent to, an area that is shielded from full exposure to the environment because of close proximity between the metal and the surface of another material. Areas where the oxide layer can break down also can sometimes be the result of the way systems or components are designed, for example, under gaskets, in sharp reentrant corners, and associated with incomplete weld penetration or overlapping surfaces. To To function as a corrosion site, a crevice has to be of sufcient width to permit entry of the corrodent, but sufciently narrow to ensure the corrodent remains stagnant. 10.3.3.4 10.3.3. 4 Pitting Corrosion Pitting corrosion is another form of galvanic corrosion and corrosion and is an extremely localized type leading to the creation of small pits or holes at the surface of the metal. Pitting corrosion is the most common failure mode for austenitic stainless steels. This type of corrosion is of great concern because once a pit is initiated there is a strong tendency for it to continue to grow, even though the majority of the surrounding steel is unaltered. The stimulus for this phenomenon is the lack of oxygen around a small area; this area becomes anodic while the area with excess of oxygen becomes cathodic. For specic acceptance criteria of pits in the surface of stainless steel components utilized in the life science industry industry,, refer to the ASME BPE Standard (Reference 12, Appendix 1). 10.3.3.5 10.3.3. 5 Stress Corrosion Cracking Cracking This is a type of corrosion that occurs because of sudden failure of normally ductile metals subjected to a constant tensile stress in a corrosive environment, particularly at elevated temperatures. This type of corrosion requires a susceptible material, under a susceptible stress, in a susceptible environment. A change in any of these conditions




can cause or eliminate Stress Corrosion Cracking (SCC) as a failure mode. Particular austenitic stainless steels alloys crack in the presence of chlorides, which limit their usefulness for being in contact with solutions (including water) with greater than a few ppm content of chlorides at temperature above 50°C. 10.3.3.6 10.3.3. 6 Intergranular Corrosion Intergranular corrosion is a form of relatively rapid and localized corrosion associated with a defective microstructure known as carbide precipitation. When austenitic steels have been exposed to high temperatures and allowed to cool at a relatively slow rate, such as occurs after welding, the chromium and carbon in the steel combine to form chromium carbide particles along the grain boundaries; the formation of these carbide particles depletes the surrounding metal of chromium and reduces its corrosion resistance, allowing preferential corrosion along the grain boundaries. Steel in this condition is referred to as “sensitized.”

10.3.4 Passivation The forming of a passive or inert layer on the clean surface of stainless steel alloys is a naturally occurring phenomenon when the surface is exposed to air, aerated water, or any other oxidizing atmosphere. Airborne impurities, high temperatures, lack of oxygen, and other direct contact materials can compromise the integrity of this layer causing the metal to lose its ability to ward off corrosive processes. The passive layer may be augmented by a chemical treatment that removes exogenous iron or iron compounds from the surface of stainless steel by means of a chemical dissolution, most typically by a treatment with an acid solution that will remove the surface contamination (and enhance the formation of the passive layer), but will not signicantly affect the stainless steel itself. The itself. The purpose of the passivation process is to restore and/or enhance the spontaneous formation of the chemically inert surface or protective passive layer after welding and fabrication of a new system or welding of new components in an existing system. The system. The passive layer is a product of the interaction between the stainless steel basic material and the corresponding owing solutions. The thickness and constitution often depends on the solution utilized and generally cannot be predicted or calculated. If metals or alloys that can be passivated are handled in air, a thin passive layer will have developed on the entire surface of the metal. If an active metal surface (e.g., following a pickling process that removes some material) is immersed in water, a passive layer forms spontaneously that protects the material from further attack. The passive layer changes when another solution is applied to the product contact surface: •

In a solution with no oxidant, the passive layer dissolves and the material actively dissolves (corrodes); the material is in the active range.

•

In a solution with an oxidant, in which small quantities of the metal oxides are dissolved, the passive current density increases, and the material corrodes at a slightly elevated rate. Over time, the continuous process of dissolution and formation leads to the formation of a passive layer that is typical for the respective medium; the material is in the passive range.

•

In solutions with elevated oxidation potential, thicker oxide layers form with an increased chromium oxide content; the material remains in the passive range at higher potentials.

•

In solutions with extremely high oxidation potential (E-redox > 1.2V), the material passes into the transpassive state. The Cr3+- oxides are oxidized to Cr6+ and then corrode at a high rate. The Cr- passivity of stainless steels can no longer be maintained; the material resides in the transpassive range.

The passive layer normally is just a few nanometers (< 10 nm) thick and consists primarily of chromium oxide, a mixture of iron oxides and iron hydroxides, and small quantities of nickel hydroxides. The layer displays relatively good electron conductivity, conductivity, and is formed by a reaction between metal ions on the surface with an oxidant (or redox system), such as oxygen; therefore, the oxide layer is at a lower energy level and represents a stable state. The determining factor for the further formation of the passive layer is the diffusion of the metal and oxygen ions through aws in that layer, as soon as a monolayer of oxidic compound is formed. After a specic thickness has




been reached, the passive layer is in a quasi-steady state or equilibrium, determined by the dissolution process on one hand, and the formation process on the other. In other words, the passive layer is in a process of continuous transformation depending on the nature of the (chemical/physical) environment. environment. The structure of the passive layer depends on the: •

medium (solution)

•

temperature

•

surface condition

•

duration of exposure to the environment

It is characterized by: •

its thickness

•

the chromium/iron and nickel ratios

•

the potential hydroxide ratios

•

its structural homogeneity

The chromium/iron ratio in the passive layer is often given as a measure of corrosion resistance. Raising the chromium oxide content is thought to improve corrosion resistance. This relationship is exploited for alloying metals, because increasing the chromium content in the alloy also raises its content in the passive layer, thus enhancing the corrosion resistance of the material. The higher the oxidation potential (redox potential) of a solution, the higher the chromium content in the oxide layer. Although stainless steel components may be clean and the passive layer intact prior to installation, welding disturbs the passive lm at the weld bead and in the Heat-Affected Zone (HAZ) of the weld. The distribution of elements across the weld and HAZ, including chromium, iron, and oxygen, changes when the metal is melted and the passive layer reforms so that the concentration of iron is elevated, while chromium is reduced. To prevent oxidation of the welded material during the entire weld sequence, the weld area is normally completely purged free of oxygen and covered with an inert gas, usually puried argon. Passivation may be performed to reduce the iron concentration and enhance chromium on the product contact surface. Some of the most common passivation treatments include the use of nitric acid, phosphoric acid and phosphoric acid blends, ammoniated citric acid, and mixed chelant systems. While passivation helps to build a good, thick passive layer on the product contact surface, however, it has been observed that after the system is put back on service, the passive layer will be the same that corresponds to the natural interaction between product and product contact surface, making passivation a short time effect (Reference 30, Appendix 1), 1), and showing no general benet to the corrosion resistance of the base metal. metal. In In a discussion on passivation, it should be realized that any product contact surface treatment only puts the alloy in its most corrosion resistant state for a particular environment. Passivation, by removal of free iron, is capable of dramatically increasing the chromium-to-iron (Cr/Fe) ratio on the surface of 316L Stainless Steel when properly applied. A applied. A Cr/Fe acceptance ratio, regardless of test test method, should be 1.0 or greater. However, greater. However, identical results in Cr/Fe ratio may not be the same using the different testing methods available, because of variability in their accuracy. accuracy. A measurement of the degree of enhancement of the passive layer following a chemical passivation treatment can be evaluated by the following methods of inspection: •

Gross inspections such inspections such as visual examination, wipe tests, residual pattern test, water break, water-wetting and drying, etc. (ASTM A380 (Reference 31, Appendix 1) and/or ASTM A967 (Reference 32, Appendix 1)).




•

Precision inspection such inspection such as solvent ring test, black light inspection, black light, Ferroxyl test for free iron, etc. (ASTM A 380/A 967 967 (References 31 and 32, Appendix 1)).

•

Electrochemical eld and bench tests such tests such as cyclic polarization measurements, electrochemical pen, and the Koslow test kit 2026.

•

Surface chemical analysis tests such as Auger Electron Spectroscopy (AES, or X-ray Electron Spectroscopy (XPS) also called Electron Spectroscopy for Chemical Analysis (ESCA), and Glow-Discharge Optical Emission Spectroscopy (GD-OES).

It should be noted that inspection methods do not provide an indication to the passive layer that will form after the systems are put back in service or give any information about corrosion resistance improvement. It does, however, give an indication of the corrosion resistance that the passivation solution may have imparted to the product contact surface, not more.

10.3.5 Electropolishing Rouging is assembled system specic. Irrespective specic. Irrespective of the specic rate of corrosion, rouge typically will occur in piping systems regardless of whether the materials used are unpolished, mechanically polished, or electropolished; however, the rate at which rouge occurs may differ depending upon the product contact surface nish. Additionally Additionally,, electropolishing can be utilized only for parts or components, not entire assembled systems; in these cases, the minimization of surface area and its resulting reduction of surface anomalies may show a benecial effect only in providing less room for colony forming units (cfu) to develop. The ASME BPE Standard (Reference 12, Appendix 1) is recommended for appropriate guidance in the selection of nishes, as well as for understanding which surface anomalies or indications are acceptable, and to what extent. It is considered unlikely that typical biotechnology/life science production operations, including high purity water or pure steam, will be detrimental to installed systems where electropolished components (e.g., tubing, ttings, valves, tanks) are located. Mechanical or chemical cleaning, or rouge remediation procedures, can damage electropolished surfaces, greatly reducing the effective service life of the component(s). (See Section 10.7.2 of this Guide).

10.4

Rouge Formation

10.4.1 Rouge Composition/Cla Composition/Classication ssication To help identication of rouge deposits, the various observed types are characterized following a classication classication based based on their source, i.e.: •

Class I – I – Migratory Rouge – Rouge – consist of various oxides and hydroxides derived from the source metals (iron oxide or ferrous oxide (FeO) being the most prevalent). It predominantly is orange to red-orange, is particulate in nature, and tends to migrate from its originating point on the original metal surfaces. These deposited particles can be removed from the surface leaving the composition of the stainless steel unchanged.

•

Class II – II – In-Situ Oxidation of Non-Passive Surfaces – Surfaces – localized form of active corrosion (iron oxide or ferric oxide – hematite (Fe2O3) being the most prevalent). It occurs in a spectrum of colors (orange, red, blue, purple, grey,, and black). It can, most commonly, grey commonly, be the result of chloride or other halide attack on the surface of the stainless steel. Integral with the surface, it appears more frequently on mechanically polished surfaces or where the interaction of metal and owing product may have compromised the passive layer.




•

Class III – III – Black Oxide Produced by Hot-Oxidation Hot-Oxidation – – surface oxidation condition occurring in high temperature environments, such as pure steam systems. As the rouge layer thickens, the system’s color transitions from gold to blue and to various shades of black. This surface oxidation initiates as a stable lm and is rarely particulate in nature. It is an extremely stable form of magnetite (iron sesquioxide (an oxide containing two atoms or radicals of some other substance), (Fe3O4)).

More details on these classicati classications ons can be found in the ASME BPE Standard (Reference 12, Appendix 1).

10.4.2 Hypotheses for Rouge Formation A common theory for rouge formation in pharmaceutical high purity water systems indicates that is the result of its distillation in contact with the atmosphere, and allowing it to equilibrate (degrade) with corrosive gases in that environment while in contact with 316L Stainless Steel. The mechanism supporting this theory is explained as follows: high purity water readily adsorbs carbon dioxide (CO2) which goes into solution as carbonic acid (H 2CO3). The carbonic acid fosters a chemical reducing environment in the water that attacks the passivated (chromium oxide (Cr 2O )) surface of the stainless steel. The resulting de-passivated surface surface permits iron to be exposed and oxidized, resulting in rust (rouge). The aggressive high purity water readily attacks the iron in the de-passivated stainless steel and a variety of heavy metals dissolve into solution. These metal ions will then react with oxygen and carbon dioxide in the high purity water creating various iron oxides and carbonates that compose the colored corrosion deposits recognized as rouge. This hypothesis appears to suggest that after some time the rouging process would stop, and with all carbon dioxide consumed to form rouge, there would be no carbon dioxide left to form the carbonic acid environment. As water in the distribution systems is regularly consumed and replenished by the water purication system; however, oxygen as well as carbon dioxide remain present in the water at all times; concentrations may differ due to operational conditions. Another theory asserts that the chromium oxide passive layer is dynamic in nature, continually breaking down and re-forming. This process occurs quite readily as long as there is sufcient oxygen for the reaction to occur. In air, this happens naturally; in the environment of hot compendial waters or steam; however, the higher the temperature of the system, the less dissolved air is carried by the water, and the more difcult it is to maintain a passive layer. At some point, there is insufcient air available, and rouging ferrite will be exposed to the surface. In contrast, with the two previous theories, an essay (Reference 33, Appendix 1) states that “studies have conclusively proven that the developing rouge layer, as a typical and characteristic secondary layer, comprises a multitude of poorly water-soluble (FE 2+ < 1 ppm, FE 3+ < 1 ppb, minimal contribution to water conductivity) iron-oxide dominated heavy metal particles. Analyses Analyses of these particles reveal, along with Fe, a signicant presence of the alloying metals Cr, Ni, and Mo.” It should be noted that the described theories indicate diverging science based conclusions, but do not provide a nal answer or a consensus to the source or formation of rouge.

10.4.3 Rouge and its Potential Contributors To advance the understanding of rouge formation, after performing an analysis of its potential contributors, the following parameters may be identied as possible rouge formation initiators: •

material composition

•

fabrication and installation methods

•

process environment

•

maintenance and repairs




10.4.4 Material Compositi Composition on 10.4.4.1 10.4.4. 1 Material Chemistry The preferred material for the fabrication of systems components in the manufacturing of biopharmaceutical/li biopharmaceutical/life fe science products is 316L Stainless Steel (UNS S31603) which typically has a content of approximately 65% to 70% iron (Fe). This should conform to applicable fabrication specications and standards. When this type of stainless steel is specied for automatic welding of tubing and ttings, the composition of the material may vary. Table 10.1 demonstrates that variability variability.. Table 10.1: 316L Stainles Stainless s Steel Tubing Chemical Compositio Composition n – Comparison Element

ASTM A 270

DIN 17457

BS316S12

EN DIN 1.4404

EN DIN 1.4435

C

0.035 max.

0.03 max.

0.03 max.

0.03 max.

0.03 max.

Cr

16.0 – 20.0

16.5 – 18.0

16.5 – 18.0

16.5 – 18.0

17.0 – 19.0

Mn

2.0 max.

2.0 max.

0.50 - 2.0 max.

2.0 max.

2.0 max.

Mo

2.0 – 3.0

2.5 – 3.0

2.25 – 3.00

2.0 – 2.5

2.5 – 3.0

Ni

10.0 – 14.0

12.5 – 15.0

11.0 – 14.0

10.0 – 13.0

12.5 – 15.0

P

0.045 max.

0.04 max.

0.045 max.

0.045 max.

0.045 max

Si

1.0 max.

0.75 max.

0.20 – 1.0 max.

1.0 max.

1.0 max.

S

0.005 – 0.017

0.03 max.

0.03 max.

0.015 max.

0.015 max.

0..11 max. 0

0.11 max

Balance

Balance

N Fe

Balance

Balance

Balance

In 316L Stainless Steel, the major elements and their functions are: •

Carbon – hardens all steels, but is kept to 0.03% maximum in 316L Stainless Steel to minimize carbide Carbon – precipitation on the grain boundaries during welding. These grain boundary carbides make the alloy susceptible to intergranular corrosion. Reducing the carbon level reduces the potential for carbides to occur, and if they do, they are smaller and fewer. The alloy trades off some hardness for improved intergranular corrosion resistance.

•

Chromium – gives the steel its corrosion resistance. Chromium participates in the formation of a complex Chromium – chromium oxide layer on the surface of the alloy, known as the passive layer.

•

Manganese – stabilizes the austenite, and has effects on hardenability. Manganese – hardenability. It is also present in the steel to “trap” the sulfur in manganese sulde (MnS) inclusions. Manganese has been identied as one of the elements present in weld discoloration.

•

Molybdenum – increases the ability of the alloy to resist pitting, especially to chloride solutions. Molybdenum also Molybdenum – has been shown to have a synergistic effect with nitrogen in the formation of the passivation layer.

•

Nickel – – stabilizes the austenitic structure so the alloy is nonmagnetic and ductile over a wide range of temperatures.

•

Phosphorus – like sulfur, is present to a limited degree in melts. At low levels, it contributes to corrosion Phosphorus – resistance and hardenability. hardenability. As levels increase, it has a tendency to segregate at grain boundaries and contributes to a condition called temper embrittlement, a decrease in room-temperature notch toughness. Combined levels of phosphorus and sulfur in excess of approximately 0.04% can lead to weld cracking caused by grain boundary segregation.




•

Silicon – a common deoxidizer used in steel making. It increases oxidation resistance and contributes slightly Silicon – to the hardness. When oxygen levels are low, it is dissolved in the alloy in solid solution. Silicon improves the resistance of austenitic stainless steels to stress corrosion cracking in chloride solutions.

•

Sulfur – – is present to some limited degree in melts. The sulfur content also can be directly correlated to the volume of sulde inclusions in the alloy. Sulfur enhances machinability and reduces the heat input necessary for welding, but the sulde inclusions, if sufciently large, are also preferential sites for corrosion. MnS inclusions smaller than 0.7 µm are reported not to initiate pits in type 304 Stainless Steel.

•

Nitrogen – contributes to the stability of the austenite. As a dissolved gas, it is present to some degree in all melts Nitrogen – of stainless steel. Nitrogen also can be used to increase the hardenability of the alloy. At very low levels, nitrogen content is affected strongly by processing route.

10.4.4.2 10.4.4. 2 Ferrite Content In industrial practice, austenitic stainless steels usually are formulated with a composition which will result in a microstructure predominantly of austenite with a small percentage of retained ferrite. This type of microstructure can reduce signicantly the tendency of castings to crack. The retained ferrite can be minimized or eliminated with appropriate thermal-mechanical processing. The amount of residual ferrite after thermal-mechanical processing (such as hot and/or cold work plus annealing) can be altered signicantly by the specic processing methods employed. The effect of residual (delta) δ-ferrite on the corrosion resistance of austenitic stainless steels varies depending on testing medium, material conditions, and processing history. •

Castings of 316 and 316L Stainless Steel compositions with various levels of ferrite indicate that ferrite may be benecial to the resistance to stress corrosion cracking and general corrosion, in contact with certain media. The ferrite level differences in the materials can be caused by considerable Cr and Ni content variations, particularly, particularly, if the content is out of the ASTM specied range for stainless steel castings; therefore, it is not clear how much the chemical composition variation may contribute to the corrosion resistance. Variations Variations of ferrite content in the range of 0.00% to 2.33% cause no signicant difference in corrosion resistance in austenitic stainless steel castings when characterized by the polarization behavior of these materials.

•

Welded tubing and ttings. As compared to wrought 316 Stainless Steel, welded 316 usually has a larger amount of ferrite necessary to prevent hot cracking during welding. It is recognized that the corrosion resistance of welded joints is less than the base metal. The contributing factors to this decreased corrosion resistance in welded metal are complicated by both the compositional and the variations in microstructure. In addition to the usually higher amount of ferrite, the chemical heterogeneity also can contribute to the decreased corrosion resistance in these welds. Ferrite percentage in the welds may be altered by either alloy content or heat input variations, which in turn change the chemical composition or the chemical heterogeneity of the welds; therefore, it is difcult to separate the effect of ferrite from the inuence of chemical composition variations. variations. Minor changes in the chemistries of 316L Stainless Steel can alter the way the alloy solidies during welding. The presence of δ-ferrite in welded austenitic stainless steel has been found to encourage pitting corrosion, and recent specications indicate a very low allowable δ-ferrite for use of welded components in corrosive service. It is generally understood that corrosion resistance is negatively affected in orbitally welded 316L Stainless Steel when δ-ferrite exceeds 3% in the weld.

10.4.5 Fabrication and Installation Methods 10.4.5.1 System Design/Installation Rouge in a piping system operating under a single set of uid service conditions is an anomaly anomaly,, the cause of which can be attributed to multiple factors rather than resulting from one single originating source.




The propagation of rouge is generally believed to be dependent upon three major factors: 1.

Material of Construction (MOC)

2.

system dynamics

3.

the process environment

The MOC for this discussion is 316L Stainless Steel. While this material may be the source in some situations, it may not necessarily be the entire cause. The cause of rouge can better be attributed to the inuence of the other two aspects: system design and process environment. Additional factors that may be considered include water quality and oxygen concentration. The system design characterizes the dynamics of the piping system and determines velocity velocity,, as well as high impingement regions. The process environment (temperature, pressure, etc.) affects the electrochemical properties within the tubing and components. 10.4.5.2 System Dynamics System components, conguratio conguration, n, and ow velocities contribute to the dynamics imposed on a piping system. These dynamics play a large part in the initiation and propagation of rouge in a system that may be predisposed to develop rouge. For a system to be predisposed to develop rouge, it has to be constructed of a material not entirely compatible with the uid service. A well passivated system system will show rouge after a time, as the passive layer will change to the appropriate level resulting from the equilibrium between the owing product and the product contact surfaces. Oxidation of the ferritic component in stainless steel is the root cause of rouge, but the mechanics that instigate the onset of rouge are the system dynamics coupled with a possible diminished thickness in the passive layer. The aw or imperfection in the passive layer may constitute a breach, permitting an aggressive uid to attack the chemistry of the stainless steel. This attack initiates an electrochemical reaction in which ferrous oxides are produced. At the source location, this can be structurally structurally damaging. As downstream deposits, deposits, the ferrous ferrous oxides may be unacceptable from a cosmetic or sanitary standpoint without detriment to the structural integrity of the pipe/tube or component in that downstream region.

10.4.6 Process Environmen Environment t 10.4.6.1 Water Systems Water system rouges, generally classied as Class I or Class II, are attached weakly to the product contact surface and are relatively easily removed or dissolved. 10.4.6.2 Steam Systems Steam systems generate rouge generally classied as Class III high temperature rouge. Class III rouge is much more difcult to remove compared to Class I and Class II rouge, because of structural and chemical composition differences. The high temperature originated deposits form magnetite, iron oxide with some substitution of chromium, nickel, or silica in the compound structure. Signicant amounts of carbon generally are present in these deposits, because of the reduction of organics present in the water, which sometimes produces the “smut’ or black lm that may form during de-rouging. Remediation processes for this class of rouge will etch the product contact surfaces in a greater or lesser degree based on the remediation (de-rouging) solutions used. 10.4.6.3 Process Equipment Water and steam systems, process equipment (e.g., vessels) also are likely to show rouge. This can be caused by buffers, CIP cleaning solutions, and sterilization cycles, etc.




Production equipment contacting product, particularly at the end of a line, offers a particular risk for the product; therefore, risk assessment also should include production equipment. 10.4.6.4 Gases The inuence of gases on rouge formation remains questionable. One of the many theories subscribes to the inuence of CO2 as a promoter for rouge formation, though there is little scientic evidence. Another Another theory points to the benecial inuence of oxygen, as this gas contributes to the formation of a passive layer; this would be consistent with the observation that rouge formation is more likely in hot storage systems having low O 2 solubility. Systems at lower temperatures would have improved O 2 solubility solubility,, and therefore, would be less prone to rouge formation, though this effect also remains unproven. Water systems may be operated using nitrogen for tank blanketing. This practice may be instigated as a measure for rouge prevention; however, it has not been proven that nitrogen inhibits or prevents rouge, and may not be cost effective. The decision decision to use nitrogen as an “inert” gas for tank-ventilation in a blanketing system should be carefully considered, and be recommended only when there is signicant improvement in water quality. 10.4.6.5 Flow Velocities/Force There are several ways in which the thickness of the passive layer that protects stainless steel surfaces can be compromised, such as high uid velocities and effects from high turbulence and uid impingement. Flow velocities, particularly in pump housings, tend to create microcavitation effects that may lead to the extraction of particles from the product contact surface, which can migrate throughout the system. Traditionally, ow velocities in excess of 5 feet per second have been considered the design standard for compendial Traditionally, water system tubing/piping systems. The theory was that higher velocity water ow would reduce the likelihood of bacterial adhesion to surfaces, and as a result, minimize biolm formation. It is understood that biolm formation is unlikely to be signicantly affected by velocity; therefore, system design may be evaluated more appropriately in the context of appropriate system dynamics and equipment/materials suitability (see Chapter 13 of this Guide). In addition, there may be unique areas in a system design requiring specic evaluation based on their physical and mechanical congurations These may include the effects of ow through reducers, ow orices, valves, pump impellers, sprayballs, ttings and instruments, instruments, as well as from their placement and conguration within the system. The effects of reducers and elbows close to a pump discharge, as well as convoluted piping congurations, can complicate design issues resulting in velocity effects that may weaken or breach the passive layer exposing the base metal to oxidation and corrosive attack. 10.4.6.6 Temperature Temperature is a major component in dening the type of rouge a system may encounter. The effect of temperature on the structural integrity of a piping system, if considered in isolation, may not be problematic when the system has been appropriately designed. Coupled with factors affecting the product contact surfaces, such as compromised passive layers; however, however, high ow velocities, uid impingement, and elevated temperatures can promote the onset of corrosion/erosion and change the chemical characteristics of rouge. There is no clear identication of temperatures that mark a boundary in which a system can be predicted to develop rouge. Rouge is not the result of a singular event or condition; it is the result of multiple factors acting in harmony harmony.. While temperature is a relative factor in the onset and propagation of rouge, it cannot be quantied in advance. Historical data and analytical evidence pertaining to specic piping systems with a specic set of criteria should be developed. Rouge may not be found abundantly in cold (4°C to 10°C) to ambient water systems; when discovered in these types of systems, it typically is Class I and easily removed. Conversely, rouge often is found in hot water (65°C to > 80°C) and pure steam systems. This anomaly is system specic and efforts to address rouge remediation require a wellplanned analytical approach in detecting and evaluating it on a per system basis. There is no empirical evidence




providing a causal relationship between temperature ranges and the presence of rouge although it is considered that systems operating at higher temperatures are more prone to develop rouge. Compliance and quality demands may require high temperatures to suppress threats to water quality and microbiological growth. Ambient water systems seem to exhibit less rouging than WFI or pure steam systems operating continuously at elevated temperatures. Temperature Temperature uctuations experienced during sanitization of ambient water systems do not appear to exacerbate the formation of rouge. 10.4.6.7 Spray Balls Sprayballs are used frequently in the design of both ambient and heated storage vessels for compendial water, except when ozone is used for sanitization as sprayballs increase off-gassing. off-gassing. Spray balls, typically made of 316L Stainless Steel (occasionally other materials), are static spherical devices with holes drilled to create a spray pattern that will ensure cleaning solution contact with designated components or portions of the equipment. These devices rely on water velocity and impingement to provide the desired cleaning patterns, as well as wet the surface of the dome and prevent uncontrolled microbiological growth in this area. Designs may include rotary mechanisms that may offer similar advantages at lower ows and pressures. Although these devices have been moderately successfu successfull in eliminating visible rouge within tanks, they have had little impact on rouge within the remainder of the system. It also has been observed that in some spray ball arrangements, particularly at the points of stream impact on the walls, increased rouge develops. There is a hypothesis that uid impingement compromises the surface integrity of the wall and creates a source of rouge that may be distributed throughout the system. 10.4.6.8 Pumps The detection of rouge within a pharmaceutical water system often occurs rst at the system pumps because they are routinely accessed for service. Visible rouge appears to develop at these pieces of equipment and rapidly increase, particularly in heated systems. There are many theories supporting these observations; ranging from metallurgical aws (i.e., high ferrite concentrations) to cavitation (including microcavitation), uid velocities, and temperature related issues. There are no clear explanations for the occurrence of this phenomenon and the specication and use of low ferrite pump materials may not signicantly deter the formation of rouge. Sanitary centrifugal pump designs have provided notable improvements over the recent past; however, high impeller speeds, cavitation, metallurgy, metallurgy, forming methods, etc., remain suspect as the source or initial appearance of rouge in pharmaceutical water systems. Conversely Conversely,, there may be systems that show no rouging at the pumps, though they have been operating for several years at elevated temperatures. The theory that rouge may occur particularly at pumps may be proven by common occurrences, but at time of publication this does not have a scientic basis.

10.4.7 Maintenance and Repairs 10.4.7.1 Basic Basi c Approach Approaches es The maintenance and repair of an existing hygienic process, compendial water, or pure steam system is an opportunity to either minimize the onset of rouge or conversely to set the stage for its formation. During installation and repair functions, the various product contact surfaces may be compromised by scratched markings, welding residues, etc., making the base material susceptible to corrosive attack and possibly prompting the onset of rouge. Maintenance and repair functions should be conducted based on logical and well planned dismantling and erection protocols, applied by appropriately trained personnel, and followed by stringent inspection techniques. Passivation of only the newly installed portions of an existing system, should be considered, because modication involves disturbing (cutting, welding, etc.) established circuits or loops. Design techniques and protocols for this kind of modication may consider the installation of header circuit block valves to avoid owing of passivation solutions




throughout the entire existing circuit or loops. The existing portions of the system may not substantially benet from the ow of passivating solutions and it may have a signicant cost. In addition, if rouge monitoring is used, the installation of “sacricial” sampling spools, at least at the beginning and end of the newly installed section, should be considered.

10.5

Rouge Detection (Methodology) The analytical method should be considered to help ensure that information obtained from an analysis is useful. High-end surface analysis or sophisticated water analysis may not be benecial without knowing how to use the information. The rst step should be the evaluation of the question: •

Which information is needed to make a decision (supporting a risk-based approach)?

Each analysis method provides specic information. The analysis should support the effort to estimate the risk for the specic system/process/product. system/process/product. Performing the analysis and estimating the results requires advanced knowledge and experience in the eld of material science and should be performed only by a trained expert. The presence of rouge in high purity waters and pure steam systems fabricated with 316L Stainless Steel materials and exposed to continuous or intermittent high temperatures, cannot be detected using methods involving temperature, ow, pressure, conductivity, conductivity, and TOC measurements. The presence of rouging presence can be identied either through process uid analyses of mobile constituents or solid surface analyses of surface layers composition; the latter requiring undesirable shutdowns to conduct examinations of: •

dismantled pump heads (casings and impellers)

•

valve diaphragms

•

tank interiors

•

pump discharge tubes

•

loop return tubing and spray balls at tanks

•

“sacricial” or sampling spools

Rouge monitoring methods are available (at time of publication) that use equipment to visually detect rouge by measuring the reection rate of the stainless steel surface and providing alarms if that reection rate changes, inuenced by the deposited rouge layers. In order to establish the barrier properties of the passive layer, various invasive and non-invasive analytical techniques can be employed to assess the stainless steel product contact surfaces of process and utilities uids, as well as the process uids themselves. These These analyses can help to detect rouge, determine its chemistry, chemistry, and quantify the rouge. Analytical methods can be segregated into uid (non-invasive), and surface (invasive) analytical techniques. Fluid analyses require the periodic collection of representative samples from various major locations throughout a given system; it also may require the periodic removal of a representative xed surface medium, (such as a sacricial spool or a test coupon), for visual and destructive analysis of the surface. Timing of sampling required also should be considered; samples may be taken during a variety of system conditions, e.g., after the weekend, after longer times without consumption, and in times of full production. The methods described may help to detect and analyze rouge; however, they provide information only about the rouge, and cannot help to make the remediation decisions. For example, an electron microscope picture of rouge




may show a very detailed surface with rouge on it, but does not answer questions about water quality quality.. Surface analysis may provide a better understanding of what is happening, but the liquid analysis provides more valuable information, about water quality quality.. Methods described in 10.5.1 – Process Fluid Analyses for the Identication of Mobile Constituents Constituents and and 10.5.2 – Solid Surface Analyses for the Identication of Surface Layers Composition, Composition, especially the latter, are suitable for basic research on rouge and the mechanism that triggers rouge formation. They also may be used to properly identify “rouge” (to ensure that it is not a different form of corrosion). They are not applicable in a “day to day” business.

10.5.1 Process Fluid Analyses for the Identication of Mobile Constituents Fluid analyses provide a means of identifying the mobile constituents within a subject water or steam system. They represent the current quality status of the media and the result of rouging. 10.5.1.1 Ultra-Trace U ltra-Trace Inorgan Inorganic ic Analysis (Non-Invasive) •

Inductively Coupled Plasma Mass Spectrometry (ICP/MS) – (ICP/MS) – concentrations of trace metals in ultra pure water/ steam are directly analyzed by Inductively Coupled Plasma Mass Spectrometry (ICP-MS). The analytes in the sample and standard solutions are delivered into the argon plasma using appropriate sample introduction systems. The thermal energy of the plasma dissolves, atomizes, and ionizes the elements. The analyte concentration is then calculated from the ion signal, by comparison to the signal of blank and standard solutions. Detection limits of < 0.05 – 50 ng/L (or ppt). This analysis provides highly quantitative information, providing the ability to trend data.

Ultra-trace inorganic analysis is a very elaborate and costly method. A baseline should be determined for each system analyzed, providing information about the current trace metal concentration in the water, and allowing trending of concentration changes. 10.5.1.2 10.5.1. 2 Standard Particulate Analysis Analysis via Light Obscuration (Non-Invasive) USP <788> Particulate Matter in Injections The USP Monograph (Reference 44, Appendix 1) describes the process by which a liquid sample is subjected to a laser light, which scatters upon contact with particles. The scattered light is collected, processed, segregated by channel, and displayed as a specic count for each size range analyzed. Detection limits of 0.5 µm to 50 µm size particles are used to characterize the distribution of particles; this resolution scale may not help detect rouge particles that are smaller than 0.2 µm. This analysis provides quantitative information, allowing trending of data. A baseline baseline should be determined for each system analyzed. 10.5.1.3 Ultra-Trace Particulate Analyses (Non-Invasive) Scanning Electron Microscopy (SEM) and Energy Dispersive X-Ray Spectroscopy or Analysis (EDX) Fluids are ltered via vacuum ltration and particles are collected on a ne pore lter medium. The particles are then analyzed for size and topographical features. A nely focused electron beam scans the sample surface and generates secondary electrons, backscattered electrons, and x-ray signals. These signals are collected, processed, and displayed. The technique allows for magnication from 10X to 300,000X. Particles can be further identied using Energy Dispersive X-Ray Spectroscopy or Analysis (EDX) technology. technology. This technique uses the Scanning Electron Microscopy (SEM) generated X-rays, processes them, and allows for the qualitative and quantitative identication of the elemental composition of a given particle. This analysis provides detailed physical observation and elemental composition data for mobile particulates; however, however, it is limited with r espect to organic particulate identication.




Table 10.2: Analytical Methods for the Identication of Mobile Constituents of Rouge Type of Test

Test Description

Test Criteria Pros

Cons

Process Fluid Analyses for the Identication of Mobile Constituents Ultra Trace Inorganic Analysis (ICP/MS)

Concentrations of trace metals in process solutions including pure water/steam are directly analyzed by Inductively Coupled Plasma Mass Spectrometry (ICP-MS).

Non-invasive sample acquisition. Highly quantitative information. Provides strong ability to trend data.

Limited with respect to organic particulate identication.

Standard Particulate Analysis (via Light Obscuration)

A liquid liquid sample is subjected to a laser light, which scatters upon contact with particles. The scattered light is collected, processed, segregated by channel, and displayed as a specic count for each size range analyzed.

Non-invasive sample acquisition. Highly quantitative information. Provides strong ability to trend data.

Baseline must be determined for each system analyzed.

Ultra Trace Inorganic Analysis (by SEM/EDX)

Fluids are ltered via vacuum ltration and particles are collected on a ne pore lter medium. The particles are then analyzed by Scanning Electron Microscopy for size, composition, and topographical features.

Provides highly detailed physical observation and elemental composition data for mobile particulate.

Limited with respect to organic particulate identication.

Fourier Transform Infrared Spectroscopy (FTIR)

Organic analysis of liquid samples or extracts from wipe samples. Used to identify possible organic lms or deposits.

Potentially non-invasive sample acquisition. Allows for organic identication of elastomers or alternate organic contaminants.

Organic contaminants must be proled in a specic target compound library.

Note: Table Note: T able 10.2 presents a number of the most commonly used tests for process uid analysis; however, however, being an emerging area of technology, improved techniques may be available at time of publication.

10.5.2 Solid Surface Surface Analyses Analyses for the Identicatio Identication n of Surface Surface Layers Composition Composition Surface analyses provide information on the nature, microstructure, and composition of surface layers. They may represent the future status of the media, and the possible threat of rouging to the water quality quality..




10.5.2.1 Visual Analyses With or Without Microscopy (Invasive) •

Microscopic and Human Visual Analysis: visual Analysis: visual analysis via Polarized Light Microscopy (PLM), SEM, or alternative microscopy instrumentation. Metal surface components are removed from the system and analyzed with or without a microscope to determine the gross surface physical characteristics. This is considered a good test for morphology determination, and can be coupled with EDX analysis for elemental composition information. This test requires the periodic removal of solid samples, such as coupons.

10.5.2.2 10.5.2. 2 Surface Metal Elemental Composition Analyses Analyses (Invasive and Destructive to Samples) •

Scanning Auger Microanalysis (SAM): metal (SAM): metal surface components are removed from the system and analyzed via SAM. The sample is subjected to a focused electron beam in which auger (or low energy) electrons are emitted from the surface and measured to provide elemental analysis of the surface layers. An argon beam can then be used to remove surface layers in order to perform a depth prole which shows the changes in elemental concentration with depth. This allows for a determination of the passive layer thickness and accurate chromium to iron ratios. This analysis is an extremely accurate method for positive identication and qualication of the surface metal composition; it also is used to determine the depth and composition of the passive layer. layer. This analysis requires the periodic removal of solid samples, such as sacricial spools or coupons, and is destructive to the sample surface.

•

Small Spot Electron Spectroscopy for Chemical Analysis (ESCA) or X-Ray Photoelectron Spectroscopy (XPS): metal surface components are removed from the system and analyzed via ESCA. The sample is subjected to a probe beam of x-rays of a single energy and electrons are emitted from the surface and measured to provide elemental analysis of the top surface layers. An argon beam can then be used to remove surface layers in order to perform a depth prole, which shows the changes in elemental concentration with depth, allowing the determination of the passive layer thickness. This technique allows for more detailed, yet size limited, surface compositional analysis. This analysis, destructive to the sample surface, is an extremely accurate method for positive identication and qualication of the surface metal composition. It may be used to determine the depth and composition of the passive layer. This test requires the periodic removal of solid samples, such as sacricial spools or coupons.

This method can help identify the surface composition prole of the passive layer and rouge, up to 600 nm depth (local application only). The method is considered costly and requires advanced expertise. For rouge, an expected result would show an iron rich layer with a depth of up to 600 nm. The tests described will provide information about the current system status, but will destroy the samples (e.g., sacricial spools). A replacement spool will no longer represent the system status, as it is new and shows no rouging. One available option is to insert a couple of samples and remove only one at a time for destructive analysis. This will allow data to be trended. The challenge will be to nd a sample conguration (holder for sample pieces) that fullls cGMP requirements for drainability and has no dead legs.




Table 10.3 Analytical Methods for the Identication of Surface Layers Composition Type of Test

Test Description

Test Criteria Pros

Cons

Solid Surface Analyses for the Identication of Surface Layers Composition Microscopic and Human Visual Analysis

Visual analysis via PLM, SEM, or alternative microscopy instrumentation.

Good test for morphology determination. Can be coupled with EDX analysis for elemental composition information.

Invasive test. Required the periodic removal of solid samples (e.g., coupons).

Scanning Auger Microanalysis (SAM), or (Auger)

Surface Metal Elemental Composition Analysis. Provides for detailed qualitative elemental composition data on both the surface itself and the sub-surface (or base metal).

Highly accurate method for positive identication and qualication of the surface metal composition. Utilized to determine the depth and composition of the passive layer itself.

Invasive and destructive test. Required the periodic removal of solid samples (e.g., coupons).

Small Spot Electron Spectroscopy for Chemical Analysis (ESCA) or X-ray Photoelectron Spectroscopy (XPS)

The sample is subjected to a probe beam of x-rays of a single energy. Electrons are emitted from the surface and measured to provide elemental analysis of the top surface layers.

Highly accurate method for the qualication and quantication of the surface metal composition. Utilized to determine the depth and compositional analysis of the passive layer. Provides excellent elemental analysis of the top surface layers, including which oxide(s) are present.

Invasive and destructive test. Required the periodic removal of solid samples (e.g., coupons).

Electrochemical Impedance Spectrometry

The analysis of electrochemical noise in order to quantify state of corrosion of a metallic surface.

Non-invasive, real time quantication of metallic corrosion. Provides strong ability to trend data.

Field qualication of this method is still ongoing.

Reection Grade Elipsometry

Multicolor interferometry utilizing light and it diffractive properties to assess surface conditions.

Non-destructive analysis. Known diffractive characteristics of elements could provide for qualitative analysis of surface chemistry properties

Invasive test. Required the periodic removal of solid samples (e.g., coupons). Field qualication of this method is still ongoing.

Note: Table Note: T able 10.3 presents a number of commonly used tests for surface analysis; however, as this is an emerging area of technology technology,, improved techniques may be available.



10.6


Risk Analysis – Rouge and its Remediation

10.6.1 Systems Monitorin Monitoring g A baseline level of acceptance for particulates and metal oxides should be established, based on risk assessment analyses which include potential damage to the process/product. A rouge remediation process of these systems (see Section 10.7.2 of this Guide) should then be considered, based on an observed and quantied escalating level of particulates and surface accumulation of those oxides. Table 10.4 provides an example for a risk-based approach to rouge and its remediation measures. Table 10.4: Risk Analysis/Risk Control Risk Analysis

Risk Control

Event (e.g., possible fault, potential error)

Effect of Failure

Actions: Risk Reducti tio on Strate teg gies

Particles of rouge may end up in the nal product.

Negative effects of rouge particles on patients are to be expected.

1. Calculate the amount of rouge from process process media (e.g., WFI, pure steam, cleaning media, CIP solutions) that can contaminate the nal product and compare with limits set for heavy metals like Cr, Ni, Fe, Mo (e.g., EMEA “Note for Guidance on Specication Limits for Residues of Metal Catalysts” (Reference (Reference 29, Appendix 1.)) 2. Should the result indicate that that the theoretical levels of heavy metals in the end product or water/steam are likely to exceed the limits, the nal product, i.e., water/steam is to be tested for heavy metals (Cr, Ni, Fe, Mo). 3. If test results exceed exceed the limits and an increased heavy heavy metal concentration cannot be explained by, e.g., substrates in use, a remediation (de-rouging) procedure is to be carried out. Thereafter, regular inspections of the system are to be introduced. To To protect the nal product from rouge, installation of lters in the system may be necessary.

Interaction between rouge and nal product, i.e., ingredients/media.

Negative effects of rouge particles on nal product, i.e., ingredients are to be expected.

1. Calculate the amount of rouge from process process media (e.g., WFI, pure steam, cleaning media, CIP solutions) that can contaminate the nal product and compare with limits set for heavy metals like Cr, Ni, Fe, Mo (e.g., EMEA “Note for Guidance on Specication Limits for Residues of Metal Catalysts” (Reference (Reference 29, Appendix 1.)) 2. Consider whether chemical reactions reactions between rouge components and the end products or ingredients are possible at the determined concentration. 3. If end products or ingredients react react with rouge components the following measures could be carried out. a. Filters are to be installed in appropriate locations in the system. b. A remediation procedure (de-rouging) is to be carried out.




Table 10.4: Risk Analysis/Risk Control (continued) Risk Analysis

Risk Control

Event (e.g., possible fault, potential error)

Effect of Failure

Actions: Risk Reducti tio on Strate teg gies

The system contains parts (e.g., spray balls, pumps, measuring ports, valves, heat exchangers) that may be affected by rouge.

Some parts of the system do not work properly.

Visual inspection, function tests, i.e., calibrations of the parts concerned should be carried out as part of regular maintenance procedure. This will ensure that these parts are in proper working order.

Filters may be affected by rouge particles.

Filters are blocked by rouging particles.

1. Filters should be visually visually inspected during maintenance. maintenance. 2. If lters are blocked by rouge, rouge, maintenance intervals are to be shortened. 3. If lters are damaged by rouge, differential differential pressure monitors are to be tted for a better monitoring of lters.

Detection of rouge during normal operation triggers unscheduled remediation (derouging) measures.

Shut downs interfere production.

1. Calculate the amount of rouge that can contaminate the nal product from process media (e.g., WFI, steam, cleaning media, CIP) and compare with limits set for heavy metal like Cr, Ni, Fr, Mo (in e.g., EMEA “Note for Guidance on Specication Limits for Residues of Metal Catalysts” (Reference 29, Appendix 1.)) 2. If the result indicates that the theoretical levels of heavy metals in the end product or water/steam are likely to exceed the limits, a routine visual inspection is to be introduced to allow scheduling of de-rouging measures.

Remediation (derouging) procedures alter/corrode the surface of the materials.

Changes to the roughness of the product contact surfaces, the resistance and cleanability of the materials, leakage are to be expected.

Resistance of materials to cleaning chemicals should be checked (see Section 10.7.2 of this Guide). The condition of the system after a de-rouging measure is to be checked (e.g., random roughness measurements, visual inspections, wipe tests).

Remediation (derouging) is not carried out or documented properly (e.g., use of wrong chemicals, inadequate documentation, incorrect switching of valves).

Remediation (derouging)/cleaning agents may contaminate the product.

Before proceeding with remediation (de-rouging) measures, it must be determined whether any cross reaction between the system and the de-rouging agents are likely to occur regarding leachables and/or extractables. After each remediation (de-rouging) measure, systems are to be cleaned and rinsed (see Section 10.7.2 of this Guide). After each rinse, the system (following remediation) is to be checked for chemical residues (e.g., using pH indicator strips, conductivity measurements). The success of remediation (de-rouging) measures is to be veried and documented (e.g., roughness measurements, visual inspections, wipe tests, heavy metal concentrations). Based on the success of remediation (de-rouging) measures, future measures should be (re)considered to prevent automatic reaction without improvement. Remediation (de-rouging) SOPs are to be developed.




Notes: Usually rouge is in particulate form. The particles usually are in a stable oxidation state. There are substances Notes: Usually that can loosen rouge particles.

10.7

Rouge Remediation (Methodol (Methodology) ogy)

10.7.1 Rouge Observation Rouge has not been documented to alter the water quality beyond compliance demands; therefore, water and steam systems can remain in use and in compliance with output quality requirements. Once rouge is found in a system, it is necessary to perform a risk analysis on rouge and its possible remediation (e.g., de-rouging). The risk analysis/risk control on rouge will help to dene appropriate measures to prevent compromise of product or process quality. quality. Regular analysis of water and product samples for heavy metals/particles and trend development can be a supportive action to keep track of rouge and rouge development. This may be achieved by: •

regular heavy metal analysis during routine monitoring of water system

•

visual inspection and documentation during maintenance

•

specic analysis of product and comparison with compliance and internal standards

The analyzing methods should be chosen appropriately and should identify: •

changes in rouge propagation

•

changes in rouge structure

•

changes in heavy metal concentrations

10.7.2 De-Rouging De-rouging may be a remediation procedure commonly conducted on high purity water and pure steam systems; however, as rouge typically re-occurs, current practices are to exercise different approaches; including leaving systems as they are to establishing remediation (de-rouging) practices at intervals of between one and three years or when decided on as a part of the ndings of a rouge monitoring program. De-rouging does not provide a permanent solution to the presence of rouge in a high-purity water or pure steam system; however, this procedure can minimize rouge in a system. Once a de-rouging process has been conducted, there is no existing methodology to show that surfaces exposed to the various available solutions have been thoroughly freed of all traces of rouge. Only representative system samples, e.g., “sacricial” spools, can be analyzed for de-rouging effectiven effectiveness. ess. De-rouging processes potentially may be detrimental to exposed base metal surfaces when applied with the most aggressive chemicals and the presence of variations in rouge deposits, thus increasing the chances of surface etching and erosion. It should be noted that specic systems and their components often may be found to contain more than one of the three classied types of rouge, making the process of de-rouging more challenging. A formulated formulated solution may work on one type of rouge, but not on another, increasing the possible chemical intrusion in the process/utility systems. systems. When de-rouging, a specic rouge type may be removed and r eplaced by a different type, adding to the complexity of the process and its expected results. Aggressive chemical removal of rouge often requires re-passivation procedures, which will restore the chromium oxide layer, but may benet only compromised (de-rouged) surfaces.




Available chemicals for de-rouging include: •

Phosphoric acid (H3PO4): this is useful to remove light accumulations and can be blended with other acids and compounds including citric, nitric, formic, or other organic acids and surfactants to assist in de-rouging effectiveness. effectivenes s. Citric acid (C 6H8O7) based chemicals with additional organic acids can be effective for the removal of Class I rouge, as is the use of reducing agents [sodium hydrosulte (Na 2S2O4) or equivalent]. These chemicals are processed at elevated temperatures ranging from 40°C to 80°C for between 2 hours and 12 hours. The process time and temperatures may depend upon the severity of rouge accumulation, the material of construction of system components, and chemical concentration. These concentrations generally are based upon service provider’s proprietary testing and process design criteria.

•

Class II rouge is removed with chemicals similar to those for Class I rouge with the addition of oxalic acid (H2C2O4). Oxalic acid may etch the product contact surface depending upon conditions and concentrations processed. The chemicals used to remove this type of rouge can be very aggressive and may negatively affect the product contact surface nish, if not closely monitored. Phosphoric acid (H3PO4) based de-rouging systems generally are effective only on very light accumulations of this class of rouge. The stronger organic acid blends with formic acid (CH2O2) and oxalic acid (H 2C2O4) can be effective on some of the high temperature rouges, and produce a low potential for etching of the product contact surface nish. The citric acid (C6H8O7) and nitric acid (HNO 3) blends with hydrouoric acid (HF) or ammonium biuoride (NH4HF2) often will remove the Class III rouges more quickly, but will etch the product contact surface wherever the base metal is subjected to the de-rouging uid. The uid. The amount of etching or increase in product contact surface nish roughness is dependent upon process conditions, chemical concentration, variability of the rouge thickness, and initial surface nish roughness (R ( R a). The condition of use for these processes is variable both in temperature and time required to effective effectively ly remove all of the rouge. The less aggressive organic acid chemicals usually are used at higher temperatures (60°C to 80°C) and require longer contact time (8 hours to 48 plus hours.); the nitric acid based uoride solutions often are used at lower temperatures (ambient to 40°C), while the citric acid based uoride solutions typically are used at the higher temperatures and shorter contact times (2 to 24 hrs.). These concentrations are generally based upon service provider’s proprietary testing and process design criteria.

The time and correlating temperatures given are in direct relation to the percent by weight of the base reactant(s). A change change in a formulation formulation will change the corresponding corresponding requirements. requirements. Different application application methods include include uid circulation, gelled applications for welds or surfaces, and spraying methods for vessels and equipment. Newer procedures may use pH-neutral cleaners which are much less aggressive than acid solutions. These new cleaning solutions are considered extremely effective on Class I and Class II rouge. The de-rouging time is, in most cases, much shorter than with acid solutions and this with much lower chemical concentrations. Neutral cleaning solutions are an alternative to the common acid treatments. Thorough rinsing of surfaces after processing, as well as proper waste disposal planning is critical to the de-rouging process. The waste uids generated by these processes may be classied as hazardous because of chemical constituents or heavy metal content. Rouge can be removed effective effectively ly from product contact surfaces to reduce the potential for oxide particulate generation into the process uids. These remediation (de-rouging) practices should be followed by appropriate cleaning/passivation cleaning/passiv ation processes of the product contact surfaces for restoration of the passive layer. Analytical testing of utility uids can be useful in identifying the level of particulate generation and levels of metal oxides contained in the uids, as corrosion degrades the surface. At time of publication, rouge does not inuence pharmaceutical water quality beyond compliance levels. However, that potential should be evaluated with a case to case risk analysis. The alternative, to remove rouge by the use of chemicals, if not done properly, may risk the system and by association the product quality. A thorough risk analysis should be performed, by owners/users with remediation (de-rouging) service providers, to evaluate the need of conducting such a procedure.




10.7.3 De-Rouging Procedures This section describes a suggested procedure for de-rouging. 10.7.3.1 Planning •

Assessment and evaluation of cleaning requirement:

The initial assessment should include a detailed list of all objects to be cleaned, their materials and nish quality, in order to dene the parameters of the de-rouging process. Routing should be dened for all objects to be cleaned. Dening acceptance criteria for cleaning success: •

Depending on the type type of cleaning performed, appropriate criteria criteria should should be used to to judge whether the process was successful: -

visually clean (the object cleaned has a metallic sheen and looks new)

-

shows visual cleanliness of the equipment; however, does not guarantee that the rouge layer has been entirely removed, i.e., wiping may still show rouge

-

swab clean (the object is wiped down with a white microber cloth or swab and no residues can be detected on cloth or swab)

-

shows whether there is rouge that can be removed by wiping, but does not guarantee visual cleanliness as some rouge might not easily be removed by wiping

•

Obtain quotes (if outsourcing is required).

•

Check quotes (if outsourcing is required).

When checking quotes, the method used for de-rouging should be taken into consideration. Established cleaning methods include: •

dynamic circulation through pipes and equipment

•

static cleaning by accumulation in pipes or equipment

Established cleaning types include: •

pH-neutral cleaning

•

acid based cleaning

The method should be chosen in accordance with the equipment condition, the internal evaluation of the cleaning method, the occupational health and safety requirements, and equipment safety. In addition, the resistance of the system materials and the materials structure (cast-stainles (cast-stainless-steel, s-steel, rolled stainless steel, etc.) in contact with the cleaning agent chemicals should be checked (seals, O-rings, surface materials, instrumentation, etc.). The precise composition of the cleaning agents should be known (quantitative/qualita (quantitative/qualitative). tive). It should be ensured that chemical agents will not accumulate in the equipment (e.g., insufcient solubility).




•

•

Dening the documentation: -

The type and scope of the documentation should be dened (e.g., temperature, cleaning times, cleaning chemicals used, and concentration).

-

In line with the initial appraisal and the nal evaluation, further parameters should be noted (e.g., roughness measurement, visual condition, swabbing)

creating and approving cleaning protocols (Standard Operating Procedures (SOPs)) -

For the remediation (de-rouging) procedure, an SOP should be generated detailing the steps (see Sections 10.7.3.1, 10.7.3.2, and 10.7.3.3 of this Guide).

•

scheduling of staff and resources

•

denition of occupational health and safety requirements

•

handling and storage of chemicals should be checked for safety; for a safety assessment the following questions may be considered:

•

•

-

What happens if leakage occurs (disaster plan)?

-

Does the hot cleaning process produce gasses? Are these a health threat?

-

How will the chemicals be fed into the system? Is the equipment used suitable for the purpose?

-

What protective clothing is required?

-

What rst aid equipment will be available and where will be located?

-

Which areas have to be closed off and how is this to be achieved?

-

What safe means of chemical waste collection and disposal are required?

-

Is it assured that chemical agents cannot contaminate waste water or municipal water system?

application of cleaning agents -

the plant may require modication to allow application of cleaning agents, depending on the method being used

-

rinse water analysis for chemical residues

-

suitable detection methods should be implemented (e.g., conductivity measurement and pH measurement) to ensure that all chemicals used have been removed without residue

-

all chemical agents should be completely water-soluble

neutralization and discharging of cleaning agents Depending on the cleaning method, highly corrosive chemicals may be used (cleaning agents, passivation agents, neutralizing solutions, rinse water) which must not be discharged into the normal waste water. A procedure for disposal of these should be dened. The following parameters may be used:




-

pH-levels

-

heavy metal content (Fe, Cr, Ni, Mo)

-

chemicals

-

chemical concentrations

10.7.3.2 Execution •

staff training

•

certication of occupational health and safety measures

•

assessment of current situation Before starting the cleaning procedure, the objects to be cleaned should be inspected. Equipment should be opened to allow visual inspection. In addition, roughness measurements may be performed on dened areas. Results should be documented.

•

leakage testing After the initial assessment, the system should be closed and tested for leakage under operating pressure. Water may be used as a test medium. The test should include supplier’s equipment connected to the system.

•

application of cleaning agents

•

heating of cleaning agent and cleaning until desired result is achieved As cleaning is usually carried out at high temperatures (e.g., 60°C to 85°C), 85°C), the system should have a temperature control. The maximum temperature should not be exceeded, particularly at the heat exchangers, as this will lead to corrosion damage. The cleaning time and the success of the cleaning process should be dened and documented in accordance with the method used. The following criteria may be applied: -

visual inspection

-

measurement of heavy metal concentration

-

previous experience

Depending on progress, cleaning time may be extended; the concentration of the cleaning solution increased, or the temperature adjusted. During the cleaning process, the system should be continuously checked for leaks. •

initial ushing Once the required cleaning success has been achieved, the system is ushed with water to ensure that follow up procedures (neutralizing, passivation) are not inuenced by cleaning chemicals residue.

•

neutralization (depending on protocol)




•

passivation To enhance formation of a new passive layer, passivation after de-rouging is recommended. Passivation requires the system to be wet with passivation solution for a dened time/volume ow period.

•

rinse water analysis for chemical residues

•

nal inspection Upon completion of the cleaning process, the system is to be checked and approved using the acceptance criteria for the cleaning success. Surfaces are to be checked for corrosion damage (pitting) by, e.g., using visual inspection or roughness measurements.

•

checking and approval of documentation

10.7.3.3 10.7.3. 3 Routine Cleaning of Equipment After the remediation remediation (de-rouging) process is completed, the system should undergo a routine cleaning process (e.g., CIP), or standard sanitization process.

10.8

Conclusions Rouge is an electrochemical phenomenon. Its chemistry is understood; there are diverse theories regarding its formation, and the conditions under which it is more likely to appear and progress are generally agreed. Where it originates and the specic causes for its appearance are less well understood. Rouging may occur in any pharmaceutical water or steam system, independent of how the system was designed and built. The industry utilizes an alloy (316L Stainless Steel) that offers all the best observed advantages for use in the biotechnology and pharmaceutical processes; it is not supposed to corrode, it is naturally protected by a “passive layer” that can be enriched with chemical treatments, but unfortunately can be compromised by welding, mechanical stresses, airborne impurities, chlorides, elevated temperatures, and contact with solutions that contain aggressive acids. Processes to counteract those threats include both preventive and reactive means to minimize and stabilize its presence, and in some cases, remove it from the product contact surfaces, despite knowing that it will soon reappear if exposed to similar pre-existing conditions. The basic consideration of a decision for system remediation should not be the status of the water or steam systems, but the quality of the water and steam in the systems. It is necessary to estimate if the changes in the systems may pose a threat to the water and/or steam quality and the associated production processes. On the other hand, it is necessary to discuss the available alternatives; remediation processes may remove a possible threat in the form of rouging, but on the other hand, create another risk. Remediation processes are invasive processes; the risk that arises with their application should be estimated with thorough analyses. The process should be designed in a way that there is no additional risk for other water systems, the environment, or the staff executing the process. To satisfactorily achieve these requirements, detailed information about utilization of chemicals in the remediation processes is absolutely necessary; like any other invasive process in the qualied water and/or steam systems, the remediation processes also should be evaluated. This chapter has endeavored to establish a rm foundation of knowledge based on what is currently understood about rouge. From this point forward and with the cooperation of a broad spectrum of experienced professionals, a strong structure consisting of science-based concepts, science-based risk analyses, research results, and practical experiences should be constructed. Once this structure is completed, rouge may be considered as an everyday occurrence worthy of attention, but not of unwanted or questionable reactions. The risk of rouging should be estimated; the risk of remediation also should be estimated. These two risks should be compared and a decision be made.


11 Control and Instrumentation




Page 191 Control and Instrumentation

11 Control and Instrumentation 11.1

Introduction Controls and instrumentation often are used within pharmaceutical water and steam systems to: •

control the operation of equipment and components

•

monitor and record data on the performance of equipment

•

monitor and record data on pharmaceutical water quality

The concepts and regulatory philosophy of dening critical versus non-critical parameters are discussed as it relates to controls and instrumentation. There is no regulatory requirement for the use of on-line/at-line instrumentation; however, USP Chapter <1231> (Reference 44, Appendix 1) has a statement on a monitoring program. A monitoring monitoring program may include a combination of on-line, in-line, at-line, and off-line probes, sensors, or instruments. In-line denotes the installation of the probe, instrument, or sensor in the main ow of the process water within the component or piping. On-line/in-line denotes the probe, instrument, or sensor is installed in the water system component or piping and is connected to a controller.. At-line/side controller At-line/side stream denotes the probe, instrument, or sensor is located on a side steam or take–off point from the main ow of the process water, component, or piping. Off-line denotes the probe, instrument, or sensor is unconnected to the water system. The sensor is the device that is attached to the process piping, either directly into the process stream or into tubing that diverts a small portion of the process uid (side stream). The diverted process uid usually is then directed to waste. Typical Typical in-line sensors include: •

temperature

•

pressure

•

conductivity

•

ow

Typical examples of side stream sensors are pH, ORP ORP,, TOC, and ozone. pH and other electroc electrochemical hemical sensors are normally used in side stream measurements since some types can leach electrolyte into the process uid. The instrument (or meter or transmitter) is the device that converts the electrical output of the sensor to a measurement (µS/cm, ppb, TOC, pH, etc.), and the instrument may display the result. The result also can be transmitted via analog output or digital signal to an external recording device such as a Programmable Logic Controller (PLC), chart recorder, or printer. The sensor and measurement instrument may be integrated into a single piece of equipment, such as a ow or pressure transducer. transducer. The instrument provides an operator interface to the sensor for control functions, such as calibration. If on-line/at-line instrumentation, sensors, or automated equipment are used to measure or record a critical parameter; action and alert limits may be established. A method of addressing “spikes” or anomalies should be established and can use the use of 6 Sigma (σ) methodology methodology.. The use of Standard Deviation (SD) at 1σ with limits, however, may be a better indicator of the current trends (at time of going to press) and the affect or non-affect of spikes or anomalies. If the next measurement value reverts below the set limits for the SD, the spike was an anomaly and not a trend.



Page 192 Control and Instrumentat Instrumentation ion

Systems often consist of a combination of manual, semi-automatic, and automatic features. Automatic is considered as the ability of the equipment to function without human intervention and semi-automatic is considered as having some human intervention (example: manual start of automatic process). Automation can have a signicant impact on the cost and performance of a pharmaceutical water system. There is no single optimum level of control and instrumentation for all pharmaceutical water systems. The optimum level of control and instrumentation for a system balances the benets of improved process control, improved documentation, and lower labor costs against the cost of procuring, installing, validating, and maintaining the control systems and instruments. The reliability aspects of controls and instruments should be evaluated to ensure product quality. The level of automation for a pharmaceutical water system often is consistent or exceeds the automation used for the manufacturing process it supports; however, however, this is not a requirement.

11.2

Principles To achieve GMP compliance, the manufacturer must demonstrate, through documented evidence, that the pharmaceutical water system is in control to consistently produce and deliver water of acceptable quality. quality. Although many quality attributes can be continuously monitored using on-line/at-line instrumentation, there is no compendial or regulatory requirement for on-line/at-line monitoring of pharmaceutical water quality, but it is strongly recommended in the USP Chapter <1231> (Reference 4, Appendix 1). Implementing the FDA’s guidelines for Pharmaceutical cGMPs for the 21 st Century – A Risk–Based Approach (Reference Approach (Reference 1, Appendix 1) can enhance the usage of on-line/at-line sensors, probes, and instrumentation. A monitoring monitoring program typically includes a combination of on-line, at-line instrumentation, manual or automated documentation of operational parameters, and off-line laboratory analysis of water samples. Control systems and instruments are critical and must be qualied when they are used to measure, monitor, control, or record a CPP. For example, the temperature of the nal water product may be considered critical for microbial control. In this case, the temperature controls (e.g., sensors and alarms) would be considered critical; however, however, in this case, it is not necessary to consider the temperature of the heating media (e.g., steam) as a critical parameter. parameter. Documentation should clearly indicate which instruments are critical and which are not. Identication of non-critical instruments on the eld device is recommended. All instruments and controls identied as critical by a criticality assessment require qualication. Items that should be recorded in the system documentation include: •

maintenance procedures and maintenance work performed

•

procedures for sampling and analysis

•

reporting of results

•

trend analysis of the laboratory data

The monitoring program during start-up typically denes maintenance frequency and alert and action levels for the process variables.




Off-Line Testing Water sampled for off-line testing (i.e., a grab sample) is not identical to on-line/at-line sample testing nor is the chemistry of the sampled water the same. Various conditions, introductions, and affects will change the chemistry of the off-line sample, e.g.: •

glassware and its cleaning residue

•

atmospheric contamination

•

atmospheric CO2 equilibrium

•

conductivity increases due to atmospheric gases

•

pH stability

•

physical handling

•

containment material incompatibility

These factors may cause off-line testing results to differ signicantly from on-line/at-line measurements.

11.3

General Instrumentat Instrumentation ion Requirements

11.3.1 Instrument Selection and Installati Installation on Instruments should be selected for accuracy and reliability over the entire process range and according to compendial requirements, where applicable, to meet process and product quality requirements. Instruments should be selected and installed in a way that reduces the potential for contamination. Manufacturer’s recommendations and good practices should be followed. •

For sensors, contact surfaces should be constructed of materials that are compatible with the water they contact. Materials of construction, surface nishes, and sanitary design (see Chapter 8 of this Guide) are usually specied for sensors installed in distribution systems.

•

Sensors in direct contact with waters with strict microbial limits should be of sanitary design. Non-sanitary sensors are usually used in feed water and pretreatment systems.

•

Sensors may be installed directly in the water system (in-line) or in a side stream (at-line). Side stream water may be reused, depending on suitability.

•

Installation of sensors into dead legs or creation of dead legs should be avoided (see Chapter 8 of this Guide).

When possible, sensors should be installed such that exposure to harsh process conditions, such as pH and temperature extremes, is avoided if the sensor is not designed to withstand these conditions. Chemical sanitization, such as ozone or peracetic acid can cause irreversible damage to sensors dependent on the materials of construction; therefore, sensors should be selected and protected to minimize potential damage from selected sanitant(s). Sensors that are not compatible with passivation agents, sanitization agents, or sanitization temperatures should be installed so that they may be easily removed or bypassed without creating a dead leg. Such sensors or devices may need to be identied and physically marked.




Sensors and instruments should be installed in accordance with manufacturers’ requirements to ensure proper operation. For example, ow sensors should be installed in the correct orientation and with the correct upstream and downstream straight run of pipe. The impact of process and ambient conditions on a sensor ’s accuracy and reliability should be addressed. Conductivity sensors are particularly sensitive to the presence of air or steam bubbles, which can be present where there is turbulence, cavitation, or interstitial conuence. In addition, sensors should be oriented according to manufacturers’ requirements requirements to eliminate gases or particles from collecting between the measurement electrodes. Example: locations directly before or after pumps should be avoided for conductivity measurement. Example: locations measurement. It is considered good practice to avoid installing in-line sensors in a vertical tee in order to avoid trapping air in the top of the tee. Accessibility of the sensor and instrument for maintenance should be considered, but improving control response Accessibility is usually more critical. Poor response time may be a consequence of the poor placement of a sensor, and usually can be improved by installing the sensor closer to the point of use; however, sensors and instruments should be accessible for maintenance and periodic calibration purposes. Location of sensors is dependent on the criticality of the measure of unit operations either upstream or downstream. Examples: conductivity sensors often are placed upstream and downstream of CEDI units to measure the efcacy Examples: conductivity of the treatment. Conductivity sensors often are placed on the distribution and return loop to determine nal product conductivity.. Sensors may be placed for compendial compliance or unit operations measurements. conductivity

11.3.2 Instrument Calibration The calibration of sensors and instruments should follow a regular program, which provides evidence of consistently acceptable performance. Non-critical sensors and instruments may be calibrated on a frequency deemed appropriate for the service. Calibration frequency should be determined by process parameters, criticality criticality,, and sensor or instrument performance. As-found data can can be used as calibration calibration verication. verication. Calibration should follow approved procedures and the results should be documented. Each measurement system in a control loop should be calibrated individually or the loop may be calibrated in its entirety unless otherwise specied by appropriate compendia (example: conductivity conductivity). ). All calibrations should be traceable to recognized local, national, and international standards (e.g., National Institute of Standards and Technology (NIST) (US), JIS, EP, ICH, etc.). Calibration certicates should be prepared and should reference the applicable sensor sensor,, instrument parts number, and serial numbers. Calibration stickers should be attached locally to the instrument after calibration. The impact of shipment and installation on the supplier’s calibration should be addressed in place of eld calibration. The manufacturer’s recommendations for installation, calibration, and calibration verication should be conrmed.

11.3.3 Types of Instrumentat Instrumentation ion 11.3.3.1 Pre-Treatment Instrumen Instrumentation tation These measurements, sensors, and instruments refer only to the pre-treatment section of the water purication train and normally should not be required in any other section. For further information, see Chapter 4 of this Guide.




Hardness Hardness measurements can be used to determine the Ca2+ and Mg2+ concentrations before or after softening. The bivalent cations of magnesium and calcium usually are the most prevalent elements found in untreated water, and these ions are harmful to downstream RO and distillation processes. Other multivalent multivalent cations, such as iron, aluminum, and manganese are usually at much lower concentrations, may contribute to downstream processing problems. Typically, hardnes hardness s is measured at sampling points after the softener, with off-line test kits that use chemical reagents, and normally, normally, a colorimeter to get a semi-quantitative response. The aim is to achieve a zero response after the softener, to protect downstream processes. Advanced Advanced on-line hardness measurement systems may be used although the robustness of softener systems can preclude the need for advanced hardness measurements. Hardness sensors are used to gauge the effectivenes effectiveness s of the water softener components. After water softening, conductivity is considered a better gauge for assessing water quality. Tests Tests may be conducted on-line, at-line, or off-line in a laboratory.. Manufacturers should be consulted for calibration, usage, and installation requirements. laboratory Oxidation Reduction Potential Oxidation Reduction Potential (ORP) also is known as redox and is an electrochemical measurement of the ions in the water capable of oxidizing organics. ORP can be used during pretreatment to assess required pH adjustments or to measure chlorine, chloramine, bromine, and other oxidizing agents by combining all oxidizing agents into a single parameter for measurement. ORP may be used before an RO system to ensure the elimination of harmful oxidizing agents prior to contact with RO membranes. The ORP measurement can be used to control bisulte (or other reducing agent) addition. Bisulte is used to reduce chemically harmful chlorine to harmless chloride. Measurements may be conducted on-line, at-line, or off-line in a laboratory. Manufacturers should be consulted for calibration, usage, and installation requirements. Silt Density Indicators Silt density indicators are used primarily to ensure that the turbidity of the water complies with the (US) National Primary Drinking Water Standards (Reference 9, Appendix 1), particularly in cases of supply of non-public water sources. In addition, suspended solids and colloidal material can inhibit the efcacy of RO membranes. The Silt Density Index (SDI) test is used to determine the fouling potential of water feeding a membrane ltration process such as a RO system. This test is dened by its specic procedure (ASTM D-4189 (Reference 34). The common instrument used to determine turbidity is a nephalometer nephalometer.. Nephalometers use Nephalometer Turbidity Units (NTUs) as the parametric readout. The maximum allowable NTU for RO membranes is 5. Installation is common before the sand lters. Tests Tests may be conducted on-line, at-line, or off-line in a laboratory laboratory.. Manufacturers should be consulted for calibration, usage, and installation requirements.

11.3.3.2 Conductivity Conductivity is a critical parameter for many high purity water systems. Conductivity measures non-specic conductive ions in the water. Conductivity Conductivity limits for PW, highly puried water, WFI, pure steam condensate, and many types of sterile or packaged waters are specied in the water monographs for USP USP,, EP, EP, JP (References 4, 5, and 6, Appendix 1), and other other pharmacopeias. On-line conductivity instrumentation is used to monitor and control the performance of many types of purication equipment, and to monitor continuously the quality of pharmaceutical waters. On-line conductivity instrumentation also may be used for nal quality assurance testing, thus eliminating the need for periodic laboratory analysis of water samples. This is in accordance with many of the principals espoused in the FDA’s Process Analytical Technology (PA (P AT) initiative with the opportunity for real-time response and controls.




Temperature has a profound impact upon conductivity measurement because of the temperature effect of ion mobilities. To To compensate for the temperature effect, most conductivity measurement systems include a temperature sensor embedded in the conductivity sensor, and one or more compensation algorithms incorporated into the transmitter to correct the actual conductivity measurement to a standard temperature, usually 25°C. Differences caused by a change in temperature and those caused by a change in ionic concentration can be distinguished using temperature-compensated temperature-compens ated measurements. For conductivity systems where the measurement is needed for verication of a purication process or other process control, temperature-compensated conductivity measurements are recommended. Compensated conductivity measurements are not suitable for critical quality assurance testing of PW, highly puried water, WFI, and pure steam condensate; however, because of variations in the temperature compensation algorithms of various instruments. When in-line conductivity measurements measurements are used for nal quality assurance testing of these waters, a non-temperature compensated conductivity value and the water temperature should be measured, according to pharmacopeial requirements. The limits vary with type of water and temperature. Reference should be made to applicable pharmacopeia for requirements. Compensated conductivity measurements which are used solely for process control and monitoring are normally not subject to compendial requirements. To operate properly properly,, on-line conductivity sensors should be installed such that there is continuous water ow through the sensor and air bubbles or solids cannot become trapped between the electrodes. Air bubbles will result in lowerthan-expected conductivity readings. Solids also can affect conductivity. conductivity. Pure steam should be condensed prior to conductivity measurement. Conductivity measuring instruments may be used throughout a pharmaceutical water system to monitor and control purication processes or to monitor pharmaceutical water quality; examples include: •

Feed water monitoring can detect seasonal or unanticipated quality changes that could impact pretreatment equipment operation.

•

RO feed and permeate monitoring allows calculation and trending of percent rejection. Changes in percent rejection may be a sign of membrane failure, scaling or fouling, seal failure, improper pH, inadequate feed pressure, or too high a recovery rate.

•

Differential conductivity of deionizer inuent and efuent or in-bed conductivity monitoring detects, predicts, and trends resin exhaustion initiating automatic or manual operation of regeneration cycles in resin bed systems.

•

The conductivity of pharmaceutical water may be monitored after the nal treatment step to verify acceptable quality prior to delivery. In addition, conductivity sensors often are installed in the return piping of distribution loops downstream of the nal POU. Systems may include provisions for automatic diversion to drain or recirculation back through purication equipment when water quality entering the tank is outside the acceptable range.

11.3.3.3 Total Organic Carbon TOC is a measure of the non-specic carbon dissolved in water in the form of organic compounds. It is a valuable tool for measuring the aggregate level of organic impurities in pharmaceutical water systems. A TOC test with a nominal limit of 0.5 mg/L (500 ppb) for USP PW, highly puried water, WFI, or pure steam condensate is a required test in the USP.. Similar limits are in place for similar waters in other major pharmacopeia. USP TOC sensors and instrumen instruments ts (or systems) are relative relatively ly sophistica sophisticated ted analytica analyticall instrumen instruments. ts. The USP USP,, EP EP,, JP (References 4, 5, and 6, Appendix 1), and other pharmacopeias provide guidance on how to qualify an instrument and how to interpret the instrument results.




In addition to “continuous” monitoring of equipment performance and pharmaceutical water quality quality,, on-line/at-line TOC systems may be used for nal quality assurance testing; therefore, eliminating or reducing the need for periodic laboratory analysis. When used for critical assurance testing of USP PW, highly puried water, WFI, and pure steam condensate, the instrument precision, system suitability, suitability, test methodology methodology,, and calibration procedures must meet compendial requirements. Instruments Instruments used strictly for process control and monitoring are not subject to compendial requirements. TOC may be monitored at several locations in a pharmaceutical water system; examples include: •

Feed water monitoring can detect seasonal or unanticipated quality changes that could impact pretreatment equipment operation or the potential for resin or membrane fouling.

•

Monitoring TOC downstream of carbon lters, ozone generators, organic scavengers, RO units, and UV lights installed for TOC removal, can verify appropriate equipment operation and provide advance warning of bed exhaustion, compromised membranes, or the need for lamp replacement.

•

TOC levels in pharmaceutical water may be monitored after the nal treatment step to verify acceptable quality prior to delivery. In addition, TOC measurement systems often are installed in the return piping of distribution loops downstream of the nal POU. Systems may include provision for automatic diversion to drain or recirculation back through purication equipment when water quality is outside the acceptable range.

TOC analyzers have been considered as a method to indicate endotoxin or microbial contamination. While this type of contamination will lead to higher TOC levels, there is no quantitative correlation to TOC levels; therefore, TOC results cannot substitute for microbial or endotoxin testing. 11.3.3.4 pH pH monitori monitoring ng is not require required d for nal quality assurance testing of pharmace pharmaceutical utical waters accordin according g to the USP USP,, EP EP,, or JP (References 4, 5, and 6, Appendix 1), but the JP does require an acidity/alkalinity chemical chemical test, in addition to other chemical tests. pH measurement is relatively straightforward for high conductivity water. Reliable results generally can be obtained using pH indicators or laboratory, eld, or on-line pH sensor and meters. Accurate pH measurement is difcult in many pharmaceutical waters because of its low conductivity conductivity.. Low conductivity water is susceptible to pH uctuations because of contaminants introduced from the air, sample containers, and test equipment, as well as sensor difculties associated with measuring the pH of low ionic strength solutions. pH measurement of high purity waters should be done “at-line” (sidestream) since conventional pH sensors have electrolyte which may diffuse across the glass membrane and into the uid. pH measurements made upstream (feed water, pre-RO) can be on-line or at line. Locations for on-line pH measurement and control often include: •

Upstream of cellulose acetate or Thin Film Composite (TFC) RO membranes, where acid is injected to minimize membrane hydrolysis.

•

Upstream of a degasier degasier,, where acid is injected to increase CO2 removal.

•

Upstream of an RO system where pH adjustment is used to optimize the percentage rejection efciency efciency..

On-line pH systems with PID controllers are used almost exclusively for process control applications, such as acid/ caustic injection rates.




11.3.3.5 Ozone Dissolved ozone levels should be monitored in storage and distribution systems that use periodic or continuous ozone for microbial control. Ozone levels can be determined in the laboratory using several wet chemistry methods or continuously using on-line/at-line sensors and instruments (see Chapter 8 of this Guide). For effective and safe system operation, ozone levels should be monitored at the following locations: •

at a suitable location for control of the ozone generator

•

downstream of the UV light to ensure ozone destruction prior to water delivery

•

in loop return piping to ensure proper ozone levels are maintained during sanitization of the complete loop

•

mechanical room space where ozone equipment is used

11.3.3.6 Flow A wide variety of ow sensors and meters may be used in the feed water and pretreatment portion of a pharmaceutical water system including rotary magnetic ow meters, mass ow meters, vortex shedding meters, paddlewheels, and ultrasonic sensors. In the distribution loop, sanitary ow sensors are recommended and may be installed downstream of the POU to ensure proper velocity. Sensors should be installed according to the manufacturer’s instructions to ensure proper operation. Special care should be taken to ensure that there is an appropriate length of straight pipe (or pipe diameters) before and after the ow sensor according to the supplier’s specications,, otherwise the ow sensor calibration may be compromised. specications Water ow rate (or velocity) may help to maintain temperature within hot or cold systems and may impact biolm formation. In addition, ow can be used under some conditions to control Variable Frequency Drive (VFD) pumps to optimize energy usage. Flow rate usually is veried upon startup, but may not be monitored continuously. continuously. Flow rates throughout the system may vary based on use. Flow may be monitored for information only or may be considered a critical parameter for water supply and system operation, when so noted by the operator or system’s owner. 11.3.3.7 Temperature Temperature is often monitored and controlled at various locations to ensure optimum equipment performance and for microbial control. Temperature Temperature interlocks may be used to prevent damage to membranes, resins, or equipment if water temperatures drift outside allowable ranges. In distribution systems where temperature is controlled or where heat sanitization is used, temperature may be considered critical to ensure correct system operation or effective sanitization. sanitization. Temperature Temperature criticality should be assessed based on its importance in the process. In hot distribution systems, temperature typically is monitored at the theoretical cold spot which typically is the distribution loop return. 11.3.3.8 Pressure Pressure may be monitored and controlled throughout the purication process to ensure optimum equipment performance. Monitoring differential pressure across lters indicates when backwashing or element replacement is necessary. necessary. Differential pressure measurement across resin beds is useful in detecting resin fouling and poor ow distribution. Monitoring RO feed, interstage, permeate, and concentrate pressures provides early warning of membrane fouling and scaling. Back pressure control in distribution systems may be critical, if minimum pressures are required at points of use. Distribution loop return pressure also can be used to control pump speed via a VFD. Pressure normally may not be considered a critical parameter; however, the system should maintain positive pressure at all times. It may be monitored for information only.




11.3.3.9 Tank Level Various types of tank level measurement are used in the pretreatment and distribution portion of a pharmaceutical water system, including: •

simple oat switches

•

free-space radar

•

guided radar

•

load cells

•

ultrasonic sensors

•

capacitance sensors

•

pressure sensors

Appropriate design is recommended. Sanitary tank level sensor designs are available for distribution tanks. Appropriate sanitary design is recommended recommended for the nal nal distribution tank. tank. Tank level may be monitored to control the supply of water into a tank and for control and protection of downstream pumps. The level may not be considered a critical parameter for water quality quality,, but may be considered a critical parameter for water supply and system operation. 11.3. 1.3.3.10 3.10 Liquid Particle Counter Counters s Liquid particle counters can be installed at various positions in a water system, including (but not exclusively exclusively)) after lters to measure sloughage, breakthrough, and failure; at the POU for product quality; before or after components used for particle reduction, such as pretreatment and roughing lters, RO membranes, nal lters; and as part of storage and distribution. Instrumentation should be installed at-line on a side stream from the component or section to be measured and monitored. Liquid particle counters may be ranged from < 0.1 to 5 micron sizing of particles. Ranging is user-selectable. Liquid particle counters are not a pharmacopeial mandated measurement for pharmaceutical waters, but may be a critical measurement if noted as such by the water system’s owner or operator. Particle control is required for several sterile compendial waters, such as sterile water for injection. 11.3. 1.3.3.1 3.11 1 Rapid Microbia Microbiall Measurements Measur ements Emerging technologies using Rapid Microbiological Measurements (RMM) are becoming more available and placed into current practice, usually with verication by traditional laboratory culturing methods. Key parameters to measure are the quantity and identication of the microorganisms. Examples Examples of typical RMM methodologies include: •

DNA (Genetic Fingerprinting-ID Type)

•

ATP AT P Bioluminescenc Bioluminescence e (Airborne/Direct Surface)

•

Integral Cell Membrane (Dye)

•

Enzyme-Linked Immunoabsorbent Assay (ELISA)




•

ID type- Antibody/Antigen Reaction

The integration of the new technology and its data should follow existing and accepted cGMPs, FDA guidelines, or compendial mandates, where applicable. 11.3. 1.3.3.12 3.12 New Measureme Measurement nt Technologies Technologi es Any new measurement technology technology,, whether mandated by a regulatory authority or not, can be installed on a pharmaceutical water system. All issues of commissioning and qualication, installation qualication, operation qualication, and validation, whether using the process, its attributes, and limits as the validation mechanism (as described in the PA PAT T Guideline (Reference 35, Appendix 1) or traditional validation sequencing must be adhered to for compliance. The integration of the new technology and its data will follow existing and accepted cGMPs, FDA guidelines, or compendial mandates, where applicable. Examples of new and emerging technologies include:

11.4

•

on-line speciation of bacteria

•

real-time bacteria counting

•

on-line automated endotoxin testing

•

nanotechnology instrumentation

•

on-line rouge monitoring

Design Conditions versus Operating Range The control system usage may distinguish between design conditions and operating ranges, and the impact this distinction has upon validation and facility operation. These criteria are dened as: •

Design Condition: the Condition: the specied range or accuracy of a controlled variable used by the designer as a basis to determine the performance requirements for an engineered system.

•

Allowable Operating Range: the Range: the range of parametric values within which acceptable water product can be produced and validated.

•

Normal Operating Range: a Range: a range that may be selected as the desired acceptable values for a parameter during normal operations. This range should be within the allowable operating range.

A facility should meet all stated design conditions; the acceptability of the water system for operation depends on operating within the allowable operating ranges. Normal operating ranges cannot exceed the allowable operating range for product water. Design condition selection should reect accepted common engineering practices. The concept of alert and action points action points may be applied along with normal operating range. Alert levels are based on normal operating experience and are used to initiate investigations or corrective measures, before reaching an action level. Action Action levels are dened as the level at which some corrective action must be taken to avoid jeopardizing water quality.. For further reference see USP Chapter <1231> (Reference 41, Appendix 1). quality



11.5


Instrumentation Instrumentat ion Spikes “Spikes” may be experienced in the measurement of a number of parameters. These excursions may be the result of an actual process event or they may be the result of the measurement technique, the sensor installation, electronic noise, or other artifact unrelated to the quality of the water. If a spike occurs in a system with a signicant physical size or mass, rapid changes in a parameter is impossible (depending on the frequency of the measurement), and consequently can be treated as instrumentation or electronic spikes. For example, a water system operates between 20°C and 30°C and reports temperatures every 10 seconds. A single temperature sensor in the distribution system reports a single value temperature of 85.7°C, then returns to its normal operating range. Meanwhile, no other temperature sensors in the loop report any substantive excursions. In this case, it is easy to show this was not a real process event, but an instrument anomaly. anomaly. In other cases, it may be decided to treat these spikes as alert level deviations based upon their frequency and duration even though their magnitude may exceed the action level. The use of delay timers to avoid inadvertent alarms is accepted (at time of publication). Smoothing techniques of Exponentially Weighted Process Statistics (EWPS) and statistical process control using running averages can be employed to minimize the disruption of the Out of Specication (OOS) or alert limit trigger. Additionally,, control limits, Additionally limits, also known as alert limits, limits, are initiated at levels below a specication specication limit. limit. The Standard Standard Deviation (SD) of the measurement and not the 6 sigma (σ) spread can be used to understand whether the spike is a trending function of the process or an anomaly. Using an SD with a 1σ limit above and below the traditional measurement of SD can alert an operator to a measurement anomaly or whether a trend is developing. If the spike occurs and then reverts to regular measurement within the 1σ control limits, it is considered an anomaly and the process is unaffected. If the SD continues to deviate as an excursion above the control limit, there is a possibility of affecting the process. Early intervention can avert a process problem and shutdown. A procedure for dening and handling spikes should be developed in conjunction with quality assurance based on the specic water system.

11.6

Control Systems

11.6.1 Level of Automation When selecting of a control strategy for a pharmaceutical water system, consideration should be given to: •

total cost of ownership

•

feed water quality and reliability

•

complexity of the purication and distribution system

•

labor costs

•

personnel skill levels and capabilities

•

documentation and reporting requirements

Options for control include: •

local instrumentation with manual control

•

semi-automatic control

•

automated control




•

integrated systems

11.6.1.1 Local Instrumentation with Manual Control In this option, a combination of instrumentation, periodic samples, and visual examination is used to monitor critical process parameters. Data is collected and recorded manually, and analysis and trending capabilities are limited. Excursions of critical parameters outside acceptable ranges typically trigger local alarms to reduce the risk of unacceptable water quality. Satisfactory Satisfactory manual operation requires signicant human intervention; this requires detailed operating procedures and conscientious documentation of critical quality parameters. This option has the lowest installed cost, but is very labor intensive and may be subject to human error. 11.6.1.2 Semi-Automatic Semi-Auto matic Control These systems use operator control panels, relay logic control, local chart recorders and printers, and some manual data collection to monitor and control the water system. These systems are considered less labor intensive than the manual systems, but still are considered labor intensive, based on the manual data collection and monitoring required to control the process. 11.6.1.3 Automated Automat ed Control Automated systems use central controllers or computers (PLC or DCS) to control a pharmaceutical water system. The computer system uses appropriate process monitoring instrumentation (conductivity meters, ow meters, TOC instrumentation, etc.) to gather data and make appropriate adjustments to the system automatically. automatically. Data is transmitted from the instrumentation to the controller via 4-20mA outputs, digital, or serial formats. As water generation systems become more sophisticated, relying on human intervention to control and monitor the water system becomes more difcult and labor intensive. An automated system requires less operator involvement, but requires a more highly trained maintenance and engineering support staff. Automated systems systems typically have higher initial capital expenditure. 11.6.1.4 Integrated Integr ated Systems These systems include an automated system and a Wide Area Network (WAN) connection to other computer systems in the building, site, or organization. These systems allow for central, remote, or local site monitoring, automatic electronic data collection, and centralized, remote, or local alarm monitoring with automatic recording, response, and report generation. These may need to be compliant with 21 CFR Part 11 (Reference 36, Appendix 1. Additional information on control system design is available in GAMP 5, GAMP Good Practice Guide on GxP Process Control Systems, and in various guidelines by the Instrument Society of America America (ISA) (References 37, 38, and 39, Appendix 1). Whichever level of automation is selected, the commissioning and validation effort should verify operation of the complete system, including vendor-supplied sub-systems.

11.6.2 Control System Software The software/control system may be used to measure, monitor, control, or record critical process parameters. Programming and design standards, especially concerning operator interface, control techniques, alarm handling, and interlock processing should be applied during the development, validation, and maintenance phases of the project. The control system software consists of: •

rmware, operating system, and application software

•

user congurable software




11.6.2.1 Firmware Firmware,, Operatin Operating g System, and Applicati Application on Software This is software permanently loaded into memory memory.. Users may have access to the software. While the functions performed by the control system may be divided between critical and non-critical functions, it is impossible to divide or isolate the rmware, operating system, application software, and associated hardware functions. If any of the functions of a control system are considered critical, all of the software is considered critical and should be validated. 11.6.2.2 User Use r Congura Congurable ble Software The functions of the user congurable software may be dened as critical or non-critical . The critical functions or modules require enhanced documentation, including validation. In some cases, it may be impossible to divide or isolate software adequately. adequately. In such cases, if some of the functions are critical, it may be necessary to validate all the software. The type of process control required is often the determining factor in the type of software required, and software requirements often dene the type of system selected. Major considerations include: •

number of Input/Output (I/O) points

•

mathematical or statistical functions required

•

reporting features required (particularly if the control system is to be further integrated into higher systems)

•

whether or not advanced control techniques are required (e.g., neural nets; fuzzy logic controllers; adaptive gain; dead-time compensation)

11.6.3 Control Hardware and Operation Interface Critical software requires enhanced documentation and should be designed and tested in accordance with current applicable cGMP. The water system, eld instruments, and control requirements affect control hardware selection. Site standards or a large installed base of a particular system may drive the selection of hardware.



12 Commissioning and Qualification



ISPE Baseline ® Guide:

Page 205 Commissioning and Qualication

Water and Steam Systems

12 Commissioning and Qualification 12.1

Introduction Note: in February, 2007, ISPE published a Good Practice Guide (GPG) entitled “Commissioning and Qualication Note: in of Pharmaceutical Water and Steam Systems” (Reference 14, Appendix 1). Chapter 12 is intended primarily to reference that GPG, and is not intended as a replacement for the GPG. For more complete treatment of information on this topic, the reader should refer to that GPG. Included in the GPG is discussion on risk assessment and other activities normally associated with creating and maintaining quality within pharmaceutical water and steam distribution systems. Commissioning and qualication are important components of the validation process by which a system is put into service and demonstrated to consistently produce water or steam of a specied quality, quality, under variable conditions, while operated under dened procedures. Recently, Recently, the international regulatory community, community, through ICH, has issued guidance based on the desired state for manufacturing practices. In addition, a related ASTM standard has been published (E2500) (Reference 8, Appendix 1) that describes a set of principles, concepts, terminology, terminology, and a process for delivering facilities and items of regulated manufacturing capacity. Neither of these documents is specic to water systems and both were created after the issuance of the GPG. It is also not the purpose of this chapter to bridge or contrast the differences in the various approaches for documenting or verifying the qualication of a water or steam system to consistently and reliably deliver the required quality.. The reader is simply referred to the ISPE Good Practice Guide on Commissioning and Qualication of quality Pharmaceutical Water and Steam Systems (Reference 14, Appendix 1) which is consistent with the guidance provided by the FDA’s Guide to Inspections of High Purity Water Systems (Reference 3, Appendix 1) or the appropriate ICH or ASTM documents referenced above for details. It is important to note that there is no single correct way to accomplish this assurance of consistent and reliable water quality as long as all the essential elements for achieving and maintaining that nal quality are qualied and the conguration of the entire system, including its starting water quality, are accurately documented for future reference.

12.2

Sampling for Water Systems Regardless of the approach that one uses to document the consistency and reliability of a particular water system, there are specic nished water quality attributes that must be veried as being consistently met: •

conductivity

•

TOC

•

microbial counts

•

bacterial endotoxin (only for water for injection(s) and highly puried water systems)

The conrmation of these attributes may be accomplished by grab sample testing from the distribution system’s points of use or by on-line instruments (where compendial requirements allow and the technology is available). Where on-line testing is utilized for conformance or “release” testing, an element of the qualication process is to verify that the on-line values are indeed representative of use point quality. A difculty in this regard is that grab samples may reect somewhat poorer quality than their on-line counterparts, especially for conductivity conductivity and TOC attributes. Nevertheless, Nevertheless, with careful attention to controlling external inuences, the impact can be minimized and accommodated.



ISPE Baseline® Guide:


In Chapter 9, Table 9.3, there are additional chemical attributes that must be met for EP waters if the water system is intended to comply with these compendial requirements. There are no commonly used on-line analyses for these attribute analyses, so grab sampling and laboratory testing are the usual option for assuring conformance with these attribute specications specications..

12.3

Sampling for Steam Systems For a pure steam distribution system that is well designed (e.g., steam trap locations, line sloping), maintained (e.g., frequent steam trap inspection for proper drainage), and operated (e.g., POU ushing times, sampling condenser hook-ups), there should be no microbial growth concerns because of the high system temperature. Typical Typical parameters requiring quantitative monitoring in pure steam systems include: •

superheat

•

non-condensable gases

•

dryness

•

bacterial endotoxins

•

conductivity

•

TOC

Bacterial endotoxins, conductivity conductivity,, and TOC are tested on condensate samples from this steam system in a fashion similar to testing WFI. Bacterial endotoxins testing is not required if the steam is not used in parenteral applications. The monitoring of superheat, non-condensable gases, and dryness parameters is performed on the live steam samples rather than its condensate, typically using the methods and specications of the European Standard EN 285 (Reference 15, Appendix 1). Usually these attributes are only required where the steam is used for SIP of productcontact equipment, related parts autoclaving, and for “porous” product autoclave loads. Less guidance generally exists for pure steam system validations, though performance testing, strategies, and sequencing are similar to high purity water systems. The time duration of the PQ, which is protracted in water systems because of the time associated with slow biolm development, is signicantly reduced in steam systems because of the extreme antimicrobial hostility of steam systems eliminating the potential for biolm development. For further information, see the ISPE Good Practice Guide: Commissioning and Qualication of Pharmaceutical Water and Steam Systems (Reference 14, Appendix 1).

12.4

Acceptance Criteria

12.4.1

Chemical Attributes

For products to be marketed in the US, the requirements for the associated waters and steam are found in the United States Pharmacopeia (USP 34) Puried Water Water,, Water for Injection, and Pure Steam monographs, which further reference General Test Test Chapters <643> for Total Total Organic Carbon (TOC), <645> for Water Conductivity and <85> for Bacterial Endotoxins Tests Tests (Reference 4, Appendix 1). These USP chapters provide detailed testing instructions, as well as chemical specications which are identical for both water grades and pure steam. These specications specications are summarized in Table Table 9.3 in Chapter 9. These specications are often used as the minimum acceptance criteria in a validation program; however, if a rm chooses to also use alert and action levels set at lower levels for better process control in preventing specication excursions, the action level may be used as the acceptance criteria instead of the specication.


ISPE Baseline ® Guide:



For systems used in manufacturing products to be marketed in Europe, the chemical requirements are referenced in European Pharmacopoeia (EP 7.0) monographs (Reference 5, Appendix 1). Water conductivity, conductivity, TOC, and the bacterial endotoxins test methods and specications for the EP’s Water for Injections are similar to those in the USP (Reference 4, Appendix 1). The overall EP puried water specications for water conductivity are looser than those of the USP and oxidizable substances testing is allowed as an alternative to TOC. There are also additional wet chemistry tests for these EP waters. Again, these specications are summarized in Table Table 9.3 in Chapter 9 and may be used as the minimum acceptance criteria in a validation program, but testing to USP specications specications plus the EP wet chemistry tests would satisfy both compendia for these attributes. As above, if tighter alert and action levels are used for better process control, then the action levels may be used as the acceptance criteria. For systems used in manufacturing products to be marketed in Japan, the chemical requirements are referenced in analogous Japanese Pharmacopoeia (JP 16) monographs (Reference 6, Appendix 1). These requirements are nearly identical to USP’s for TOC and water conductivity (with the exception that no Stage 3 conductivity testing is specied). There are no additional wet chemical tests. As with the other pharmacopeia’s requirements, these specications specications are summarized in Table Table 9.3 of Chapter 9 and may be used as the minimum acceptance criteria in a validation program. As with the other other pharmacopeia, if tighter alert and action action levels are used used for better process process control, then the action levels may be used as the acceptance criteria. How compendial requirements are interpreted may vary from company to company company.. The most recent version of the pharmacopeia(s) for the area(s) where products will be sold should be the basis for the selection of acceptance criteria and that rationale documented. The information provided in Table 9.3 may be used as a guide in the development of this rationale, but those specications will inevitably become outdated. 12.4.2

Microbial Attributes

Microbial testing and values that could be used as acceptance criteria are not well dened in the USP USP,, but are discussed in the General Information Chapter <1231> “Water for Pharmaceutical Purposes.” Purposes.” This chapter provides a suggested method, but also notes that this method may not be optimal and that other methods could be evaluated to nd one more suitable for enumerating the microorganisms that may be unique to a water system. Similarly, Similarly, this non-mandatory chapter also provides a suggested maximum action level that is the highest that should be considered for process control, and further suggests that lower microbial levels derived from trend analysis be employed for optimizing process control. It is typical to use such action levels as the acceptance criteria in a validation program. For new water systems, usually insufcient data is available for meaningful trending, so the suggested maximum action level in USPs <1231> (Reference 4, Appendix 1) typically is used as the acceptance criteria. If this qualication exercise is performed after more extensive prior microbial testing, sufcient data for trending may be available, in which case trend-derived action levels may provide more meaningful acceptance criteria. The EP is more prescriptive regarding microbial enumeration test methods which appear within the water monographs along with an “appropriate action level,” identical to USPs non-mandatory “maximum action level” for the respective waters. By their appearance in the EP monographs, the test method becomes a referee test and the “appropriate action level” becomes a specication limit rather than a process control value as implied by its name. Therefore, the EP microbial enumeration test method should be used during the system validation program and thereafter if marketing products in Europe, unless a superior method can be demonstrated for the user’s water system. Similarly, Similarly, the EP action levels may be used as the acceptance criteria, unless tighter acceptance criteria are more suitable. The JP is similar to USP with its placement of suggested microbial test methods and “appropriate and generally applicable” action levels in its non-mandatory informational chapter <21> “Quality Control of Water for Pharmaceutical Use.” However, the JP suggests that different test conditions be used for quality control testing versus for process control testing. The action levels mentioned in this informational chapter are the same as mentioned in USP as well as EP.



12.5

ISPE Baseline® Guide:


Change Control and Maintaining the Qualied State of the System Changes in the water or steam system should be coordinated through an appropriate change control program. A riskbased approach should be used when assessing the impact of the change and measures must be put in place that minimize the potential risk and demonstrate that the system remains in a qualied state. Where changes may alter or compromise the state of control, remedial actions should be implemented to avoid a loss of control. For systems that have gone through multiple individual changes over a period of time, it is necessary to evaluate that the accumulation of changes has not affected the qualication of the entire system. This evaluation could result in an abbreviated or full system requalication, requalication, or simply a formal historic data review, depending on the collective number and signicance of those changes. Frequent quality attribute monitoring using meaningful in-process control levels combined with an effective change control program assure that the investment in the validation of a water or steam system is protected and the continuation of consistent and reliable control is achieved.


13 Microbiological Considerations for Pharmaceutical Water Systems




Page 209 Microbiological Considerations for Pharmaceutical Water Systems

13 Microbiological Considerations for Pharmaceutical Water Systems 13.1

Introduction Issues related to controlling microbial proliferation in a water system require consideration throughout: •

conception

•

design

•

construction

•

qualication

•

operation

•

monitoring

Although the highly puried waters, such as those prepared for pharmaceutical uses, have very low levels of inorganic and organic contaminants, these waters are sufciently rich in nutrients to support the growth of several types of microorganisms. This microbial growth in water purication, storage, and distribution systems should be controlled to reasonable levels or where required, prevented. Otherwise, the water purication unit operations may fail to perform as intended and the nished water may be unsuitable or unsafe for use in pharmaceutical, biopharmaceutical, and medical device applications, or with patients and consumers.

13.2

The Microbial Growth Process in High Purity Water Systems It is useful to understand how the microorganisms are able to thrive in nutritionally austere environments, to understand both how to control or prevent this microbial growth and the consequences of not preventing their growth. This chapter focuses on the development and spread of microbial growth and how it can change over time.

13.2.1

Low Level Nutrient Behavior

To understand the growth of microorganisms, it is necessary to understand how the thermodynamic properties of nutrients affect their molecular “behavior.” When organic nutrient molecules are adsorbed to water system surfaces, hydrophobic properties of both the molecules and the surfaces are satised (at least partially). This allows the water system surfaces and dissolved organic nutrient molecules to achieve a more favorable lower energy state. Hydrophilic interactions occur to facilitate adsorption of other organic molecules; therefore, molecules that could be considered as nutrients for aquatic bacteria tend to be in higher concentrations on surfaces than they are in solution. Aquatic bacteria that can use these low organic nutrient levels (oligotrophs), can exploit the nutritional advantages of these surfaces by growing on such surfaces in preference to growing in suspension as planktonic ora. When microorganisms grow on such surfaces, the resulting biomass is called “ biolm.”



13.2.2


Planktonic Cell Characteristics

The development of biolm can start with a single cell of a microorganism, usually a motile bacterium having specic structural, genetic, genetic, and biochemical characteristics characteristics referred to as a “pseudomonad” “pseudomonad” (named after the type genus Pseudomonas Pseudomonas with with which most of these bacteria share many similarities). The cell of the microorganism planktonically oats along with the water ow, and may use its own agella for locomotion. It typically is not able to grow using planktonic nutrients because of their low levels, but maintains its cellular integrity and locomotion, and may perform one or two replications, by using its limited internal nutrient stores. The cell wall surface of this pseudomonad is covered in a lipopolysaccharide layer (also called endotoxin which is most well known for its pyrogenic or fever-causing properties) that participates in giving the cell surface a slight stickiness and hydrophobicity.

13.2.3 Biolm Initiation and Growth If this cell touches a surface, it has an initial tendency to loosely and reversibly adhere to that surface. If the layer of water adjacent to the attachment surface is moving, it can exert a shear force on the cell that is stronger than the cell’s initially weak adhering force and almost immediately detach the cell back into the owing water. The longer the cell remains loosely adhered to the surface (in terms of seconds to minutes), however, the more it “senses” the presence of those surface-concentrated nutrients and perhaps the solid surface itself, triggering a complex series of gene expressions and suppressions that prepare it for a non-motile life on the surface (where nutrients are in relative abundance). The genetic expressions cause the production of enzymes and other proteins suited to the metabolism of the particular nutrients on the surface and for anchoring the cell more rmly to the surface, making it very difcult for the shear forces of the owing water to dislodge the cell. Within hours to days of the initial adherence, depending on the species of microorganism and the surface nutrient levels and types, the nutrients are metabolized to supply the energy and building blocks for cell replication, so proliferation of additional strongly anchored cells begins. During this same period, these “naked” cells also begin to exude a sticky; mostly polysaccharide exopolymer exopolymer.. During the following days, this slimy Extracellular Polymeric Substance (EPS, also known as glycocalyx) continues to be produced as the cells continue to proliferate within the protective matrix. The stickiness of this EPS begins to trap debris particles, other planktonic cells of microorganisms, and ocs of biolm sloughed from other locations, which also begin to thrive in the localized nutrient-rich community. community. Wastes from cells already present plus wastes from the new arrivals and trapped debris are added to the growing biolm community. Biolm Biolm formation on a new surface also can begin by the deposition, possibly in an eddy or other low ow location, of a sloughed fragment or oc of intact biolm from an already established upstream biolm location; once deposited the growth process is identical.

13.2.4 Biolm Regulation and Behavior As the biolm grows larger, complexly orchestrated control, called quorum sensing, occurs based on the production and reception of signaling compounds by the various biolm cells. The resulting reaction collectively affects affects the biolm’s shape, growth rate, and death of specic cells, and may promote the development and release of planktonic “pioneer” cells. The focus is on adapting and surviving in the local environment and dispersing to other locations.

13.2.5 Biolm Microbial Selectivity The diversity of the species that participate in a biolm at a given location in a water system is affected by: •

planktonic species present in the water at that location

•

ability to adapt to using the nutrients available locally




The primary source of the planktonic organisms is the feed water. Planktonic organisms also enter the system during its construction and from biolms on replacement bed matrices, lters, and chemical additives, including regenerants. For example, the biolm member species in an activated carbon bed grow on a rich source of many different organic molecules adsorbed by the activated carbon granules. Biolm member species may have “favorite foods” among the nutrient choices. A large diversity diversity of available nutrients encourages the potential for a large diversity of biolm participants. Where ocs of biolm or pioneer cells from the biolm move downstream, to environments with more limited nutrient concentrations and diversity, diversity, many of those cells may not nd a surface or existing biolm with suitable nutrient choices or levels to accommodate their nutrient specializations specializations.. These cells eventually starve to death. Cells that have more highly adaptable nutrient utilization diversity are more likely to nd a surface or existing biolm where they can grow. This selective survival process occurs at each successive purication step until cells survive the nutritional variability and arrive at the nished water distribution system. These microorganisms may be the most nutritionally diverse of the biolm-forming microorganisms and are able to survive and grow on the types and levels of chemical impurities that exist in the nished water or may have been added from system materials, the atmosphere around the tank vent or local ora. Water systems can be unique in types and levels of impurities, favoring the emergence of a few naturally selected and resilient microorganisms. The microbial diversity can be inuenced signicantly by: •

source water, which often is the cause of nal impurities (including microorganisms)

•

variable unit operation selection and arrangement

•

quality of design and operation

Factors such as the frequency of sanitization may inuence the survival of microorganisms. The initial microbial population in a newly commissione commissioned d water system is probably very low, representing only recent arrivals. It may be distinct both in population density and diversity from an older water system.

13.2.6 Microbial Diversity Diversity as a Function Function of Seasons and Water System Ageing As the water system ages, the initial microbial populations have opportunity to develop into larger and probably more complex communities, as different survivors make their way into the nished water, perhaps from eventual unit operation maintenance procedures, a contaminated bed matrix, or chemical addition, or as a function of changing seasonal source water chemistry or microbial populations. During this system ageing process, biolms will have had opportunity to nd the most hospitable locations to grow and survive any microbial control efforts applied. Seasonal changes, coupled with minimal temperature control, can affect the water temperature in some systems, including effects from outside storage tanks or frictional heating from circulation pumps. Gradual temperature shifts, as little as a few degrees, also can cause a shift in the microbial density, density, growth rate, and species composition of associated biolms. Therefore, no two water system’ system’s s biolms could be exactly alike. They can change over time as a function of: •

system ageing

•

seasons

•

unit operation efciency changes

•

intentional changes in system maintenance practices

•

slight, unrecognized maintenance practice changes that may be related to unapproved operator “short cuts” or personnel changes




Therefore, a high purity water system that has been in operation for a period of time may have different and possibly more or less diverse ora than were observed just after commissioning. The test methods used to recover microorganisms in the water system during its original qualication may be less optimal for microbial recovery in the mature water system, possibly resulting either in lower than actual counts or taking longer than other media for countable colonies to develop. Optimizing a microbial enumeration method to suit a more mature system ora (e.g., by using different media or incubation temperatures or durations) may impact the validity of historic data, including those generated during system qualication, as well as potential regulatory consequences. Such impacts usually are not considered sufcient reason to stay with an inferior method that fails to reveal the accurate nature of the ora in a water system or delays timely response to system control issues because of inordinately long incubation; as long as there is reasonably adequate documented evidence that an alternative method is superior. To To demonstrate the relationship between the data from a new, well-justied approach to the data from a previous approach, a “bridge period” of concurrent testing with both approaches should be executed (longer is better). The microbial recovery difference observed and the microbial count differences observed, between the current data (from both methods) and the data generated during initial system qualication, should be used to determine the need to repeat a full or abbreviated requalication to establish a new baseline for current system performance. Water systems with continuously sanitizing distribution systems are unlikely to demonstrate microbial changes as a function of system ageing. Microbial changes in the purication train may be “obscured” by the sanitizing conditions in the distribution system. Ageing Ageing changes may not be evident in the distribution system microbial counts, but microbial changes within the purication train could cause chemical purity changes in the nished water. This may provide justication for root cause investigation investigation or an abbreviated abbreviated requalication requalication focused on the chemical attributes attributes in the distribution system, and microbiological and chemical evaluations in the purication train.

13.3

Detrimental Effects of Biolm

The presence of biolm in a high purity water system has consequences that are rarely benecial. These detrimental effects are the very reason why so much activity and cost are expended in attempts to control it and minimize the damage it can cause. Biolm in a high purity water system can impact both the surfaces it colonizes and the water that passes over those surfaces.

13.3.1 Potential Impact of Surface Surface Alteration If the biolm is growing on a surface that has a function within the water purication process, the “activity” of that surface can be altered. Possibly the only benecial biolm colonization phenomenon occurs with coarse ltration. Sticky biolm grows on the surface of multimedia bed grains or on the elements of coarse mesh lter cartridges, functionally increasing their adherence properties, as well as the size of the grains or mesh size. This reduces the size of the holes in the ltering sieve and increases particulate ltration efciency by trapping even smaller particles than those for which the lters were designed. Conversely, Conversely, this causes lower dirt loading and faster blockage of the ltration step, along with biolm sloughing and pioneer cell release. If the colonized surface is the upstream side of an RO membrane, water will not be able to freely permeate that portion of the membrane, effectively blocking or fouling it, leading to reduced permeate ow, which, if compensated for by increasing pressure, leads to poorer product water quality. If the colonized surface is an ion exchanging resin bead, water and its ionic impurities cannot as easily reach the active sites on and within the resin bead, effectively reducing ion exchange efciency. This surface occlusion also may signicantly slow the penetration of any regenerant chemicals and their post-regeneration free-rinsing; further compromising the purifying ion exchange functionality of the surface.




A common biolm colonization phenomenon occurs within granular activated carbon beds. These beds remove organic compounds and chlorine, which has modest antimicrobial activity against the naked pioneer cells (but not EPS-coated biolm). The highly porous surface of the carbon granules provides an ideal environment for biolm because of the: •

enormous surface area on and within each granule

•

organic nutrients concentrated on those surfaces

•

absence of any antimicrobial chlorine

If allowed to continue without control from frequent backwashing or hot water sanitization, the organic removal capacity of the carbon bed can be lost rapidly, and the bed quickly can become the source of high level downstream contamination from released biolm ocs and pioneer cells.

13.3.2 Potential Impact in Water Used When biolms develop in distribution systems, they typically do not harm the functionality or chemical inertness of the surfaces of the piping and valves on which they are growing. Harm is caused by gradual sloughing off into the water stream and exiting the water system along with the water at POUs into the applications required by the water. It is in these applications where these biolm-derived organisms or their by-products can be detrimental. If the water is used for dry products, such as solid oral dosage forms, the low water activity of the formulation, or possibly heat or solvents used in their processing, usually is lethal to the biolm-derived microorganisms, or as a minimum unfavorable for continued growth. These dosage forms are ingested, so even if the microorganisms survive, the hostility of the digestive tract usually will kill them. If the product is a liquid formulation, that is not intended to be sterile and is contaminated with sloughed ocs of biolm containing cells imbedded in protective EPS, these contaminants could evade product or preservative hostility that may kill “naked” cells. They could survive for an extended period of time or even proliferate to high numbers in products that otherwise may have been deemed as suitably preserved by the “naked” cell challenges of compendial antimicrobial effectiveness effectiveness tests;therefore, biolm should be appropriately controlled in pharmaceutical water systems used for liquid non-sterile formulations. The presence of specic biolm-derived organisms also may be potentially harmful to patients using the dosage form, especially if the dosage form is: •

directly administered to or contacts sensitive tissues, such as those of the respiratory tract or abraded/inamed skin or mucous membranes

•

used where normal immunity is compromised by an underlying disease or medical treatment, such as chemotherapy

Most aquatic biolm microorganisms are specialized for growth in their austere habitat and are unable to survive in a host or become pathogenic. A few highly adaptable species are considered opportunistic pathogens because they can cause infections in certain patient exposure situations when present in relatively high numbers. An example is Pseudomonas aeruginosa, aeruginosa, known for causing pneumonias as well as burn infections; an emerging opportunistic pathogen is Burkholderia cepacia which cepacia which also has been associated with lung and dermal infections, particularly in immuno-compromised immuno-compromis ed individuals. These microorganisms can form and live in water system biolms and also are capable of growing or as a minimum, surviving in otherwise hostile product formulations. Their presence in a water system should be a concern if the water is used for susceptible formulations or patients.




The products and uses most susceptible to the effects of water system biolms are parenteral or injected products. These are sterile products, but the water used in these applications also must be free from bacterial components, such as the lipopolysaccharides (pyrogens) from cell walls of gram-negative biolm bacteria. Pyrogens are pharmacologically active active at extremely low concentrations and should be at extremely low levels or absent in the water used for susceptible formulations. This pyrogenic cell wall component is released by biolm bacteria into the slimy EPS surrounding the cells as well as into the water outside the biolm. Killing the cells in the biolm producing this lipopolysaccharide lipopolysacchari de does not destroy it and may cause the release of more of lipopolysaccharid lipopolysaccharide, e, as the killed cells lyse or no longer retain cellular integrity. The best way to effectively control the presence of such harmful compounds in the water is to prevent the development of biolm in the distribution system, and assure that any pyrogens present in the pretreatment part of the system are prevented from entering the nished water. When such precautions are taken, the resulting endotoxin levels are likely to be suitable for water for injection.

13.4

Microbial Control Strategies Effective microbial control strategies typically involve more than one approach, involving microbial-controlli Effective microbial-controlling ng features and events in multiple locations and at different times. These controlling features and events are intended to work together to achieve control of microbial proliferation within a water system. These strategies can be applied locally to individual unit operations or more broadly, e.g., to the entire storage and distribution system. Their combined success is related to their individual effectiveness. Individual elements discussed below can be incorporated into an overall microbial control strategy strategy..

13.4.1 Design and Operational Parameters Understanding how biolm responds to commonly used design and control approaches helps predict how well they should work and which approaches justify design and operational costs, based on the level of control that should be achieved. General control strategies usually include: •

high ow rate and backwashing

•

surface smoothness and composition

•

use of hostile regenerant or passivant chemicals and sanitants

•

use of heat

•

use of UV

•

use of ltration

Individually, these approaches work with varying degrees of thoroughness. When several compatible approaches are Individually, combined; however, the effects tend to last longer. High purity water biolm-forming microorganisms are generally equivalent in their susceptibility to these microbialcontrolling conditions. In a protective, slimy biolm form, compared to their naked planktonic cell form, the differences in susceptibility between these microorganisms is substantial, by factors of as much as 10 4-fold. A microbial control program, therefore, should be aimed at the vulnerable microbial forms and preventing biolm formation. Measures that control the proliferation of biolm also may help control the presence of endotoxin in water, as these microorganisms can be a source of endotoxin. Endotoxin in the water also may have originated from bacteria upstream of the distribution system (possibly from biolms growing within the purication train as well as the incoming source water). Complete control of endotoxin should include microbial control within the system and direct removal of endotoxin from the incoming water by purication unit operations capable of separating these macromolecules from water. Unit operations suitable for this purpose include:




•

distillation

•

ultraltration

•

reverse osmosis

•

anion exchange resins and positively charged membranes (to a limited extent)

Given sufcient concentration and contact time, ozone also can destroy endotoxin. 13.4.1.1 Effects of Flow Rate on Biolm Control As a biolm grows larger, it tends to grow as tall as the ow rate shear forces allow. If left undisturbed in stagnant or slowly moving water (0 – about 0.1 m/s velocity), the biolm can form column or mushroom-like shapes that maximize its surface contact with water-borne nutrients; however, tall biolms are fragile and particularly susceptible to the shear forces of fast owing water. If a biolm that developed in slowly moving or stagnant water experiences any sudden increase in ow rate or directional change, such as from backwashing or sampling, large portions of the outer biolm structures may shear off along with any captured particulate debris, and be carried with the water ow. This is the expected and desired consequence when backwashing a granular purication bed such as activated carbon, softeners, deionizers, or other similar devices. When performed effectively with the entire granular bed being lifted and uidized by the rapidly reverse owing water, such backwashing should partially remove the build-up of biolm and trapped debris. If the biolm is given sufcient undisturbed time to accumulate in the spaces between granules, particularly downstream of chlorine removal or where chlorine levels are minimal, it effectively may ‘cement’ the granules together making them highly resistant to being fully uidized during backwashing. This agglomeration also impedes the penetration of sanitizing hot water or regenerant chemicals to bed granules. In addition, when biolm forms in POU valves and their connecting side legs (possibly as well as in downstream hoses), a preliminary ush of the outlet to drain may be required to rst remove the bulk of the loose biolm. This helps to avoid the presence of sloughed biolm in water used from the outlet. A 10 to 20 µm thick zone of laminar owing water, called the ‘boundary layer’, exists next to the solid surface, even in highly turbulent water. The less turbulent the bulk water circulation, the thicker this boundary layer. Even in the most turbulently owing water, causing the thinnest possible boundary layer, a several-cell thick tenacious biolm can develop, though with a much thinner and simplied structure compared to stagnant water biolms. If such a biolm continues to grow and extend above the laminar zone into the turbulently owing water or the turbulence momentarily penetrates into the boundary layer, the upper biolm layers may be sheared off into the water ow, probably as small multi-celled, sticky ocs ocs of intact biolm. Flow-sheared ocs of biolm along with any intentionally released pioneer cells, may nd their way to a downstream boundary layer, surface imperfection, or crevice to establish a biolm or leave the water system along with the water from a POU. Biolm growth rate is dependent on nutrient levels. In nutritionally equivalent systems with nearly stagnant ow and fast, highly turbulent ow, the biolm growth rate tends to be slightly faster in the higher ow scenario because of an increase in the passage of nutrients across the biolm and more opportunities for nutrient molecule capture by biolm microorganisms. The shape and associated tenacity of the biolm develop to accommodate the continuous shear forces of the water. Sampling creates ows in a stagnant system that can shear off sizeable amounts of erect, loosely adhering biolm; minimal biolm shearing occurs from tenacious, surface-hugging biolm when sampling turbulently owing systems. Plate counts of those samples may give the impression of large differences in levels of biolm development, where this is not the case. Highly turbulent ow is considered advantageous in reducing the shedding of biolm by inducing it to grow less erectly and attach more tenaciously to the surface. Turbulent ow also is necessary for thorough distribution of sanitizing chemicals and hot water into short side legs and valves connected to the main piping. Turbulent ow alone has very little impact on the rate of biolm initiation or growth.




13.4.1.2 13.4.1. 2 Effects of Surface Smoothness and Composition on Biolm Control Controlled studies have shown that using expensive ultra-smooth surfaces virtually free from surface scratches and imperfections, such as PVDF or electropolished stainless steel, offer only a moderate advantage in delaying the initial cell adsorption that begins biolm development. The hydrophobic surface of many high technology plastics also shows a similar moderate delay in initial colonization compared to hydrophilic stainless steel surfaces. These delays are only possibly a few hours under highly turbulent ow conditions. Once biolm growth has begun, the biolm demonstrates an equivalent tenacity for a surface, irrespective of the material or nish. Once the surface is covered in biolm, the surface of the biolm becomes the preferred site for further microbial attachment, and that surface is identical whether the biolm initially formed on a smooth or rough, hydrophilic or hydrophobic surface. The presence of surface imperfections, however, can have a signicant impact on the ability of the surface to be cleaned or sanitized chemically. Crevices caused by microscopic cracks or pits in the surface, or macroscopic gaps at gasket/seal edges of sanitary connections, at bad welds, in hoses or the pinched septa of sanitary valves offer protective areas from poorly penetrating chemical cleaners and sanitants, allowing survival of at least a portion of the cells within biolms growing in these areas. These macroscopic imperfections probably provide the greatest survival advantage to biolm from sanitization efforts. The relatively high cost of ultra-smooth and hydrophobic surfaces should be evaluated against the modest delay in biolm formation and marginal improvements in cleanability, cleanability, particularly in heated or continuously sanitized systems which tend to never form biolms anyway. Efforts to minimize macroscopic imperfections and crevices are very worthwhile and should be considered. 13.4.1.3 Effect of Water Water Purity on Biolm Control As the purity of water increases, the rate of development of the biolm decreases (and vice versa). Low levels of nutrient are still present (concentrated at surfaces), even with the purest possible water, so microbial growth will still occur although at an extremely slow rate. Very high purity water that has resistivity near 18 Megohm-cm and TOC levels in the low parts per billion (ppb) (single gures) tends to allow only very slow biolm development and growth. This purity level is better than the lesser purity generally needed for most pharmaceutical manufacturing or required by the pharmacopeia. It is not easy or economical to attain and maintain, but it has a proven track record for microbial control in the microelectronics industry and may be a consideration for use under a few pharmaceutical applications where such purity is otherwise needed, such as a few laboratory applications. 13.4.1.4 Effect of Water Water Temperatu Temperature re on Biolm Control Temperatures between about 0°C (freezing) and 40°C tend to encompass the operating temperatures of water systems that are considered to be “ambient.” The ambient air temperature of a manufacturing or utility area tends to be around 20°C to 25°C. The incoming water temperature may be lower, but frictional heating from the circulating pumps and water circulation can add 1°C to 2°C to the water temperature per pass. Rapidly recirculating water systems, whose circulating temperatures temperatures are not appropriately moderated by cooling heat exchangers, could reach the range of 30°C to 35°C, an ideal growth temperature for many water system bacteria. Warmer Warmer temperatures tend to encourage faster microbial growth; cooler temperatures tend to slow microbial growth. Most people consider distillation processes to be immune from microbial problems because of high operating temperatures intolerable to bacteria; however, wet, cooler locations within the still, such as the cooling condenser or in stills that do not operate continuously, continuously, may not be at microbial-inhibitory temperatures and appropriate microbial control precautions may be required. Chilling of water distribution systems may be used for process reasons. Pipes and tanks usually are insulated to avoid greater energy expenditure and to avoid condensation. Chilling has the benet of keeping the water temperature low enough (typically 2°C to 8°C) to dramatically slow the development of biolm. However, biolm eventually will develop, even in cold water. A continuously cold system will select for colonization by psychrophilic (cold-loving) microorganisms. microorganisms. These bacteria will preferentially grow in the warmer places in the cold water system.




Their growth rate will be slow, but faster than the typical mesophilic (moderate temperature-loving) microorganisms at the same location. Psychrophiles typically are susceptible to heat kill at lower water temperatures than mesophiles. Continuously cold systems require constant attention to maintain low temperatures throughout, particularly with circulating pumps, uninsulated POUs, or uninsulated portions of the system continuously adding heat. In addition, if enumeration of the psychrophilic microbial population uses a cultivative approach, it may require cooler incubation conditions more suited to their optimal growth temperatures. This usually will require extended incubation times (perhaps doubled) because of their slower growth rates compared to those of mesophilic microorganisms. A cold water system may not be optimal for microbial control from the perspective of capital and operational cost (as well as timelines for monitoring results). A cold water system does offer intermittent heat sanitization advantages that may be within the heat tolerance range of several plastic materials which are unable to withstand elevated water temperatures required for sanitizing mesophiles.

13.4.2 Effective Sanitization Concepts Sanitization of individual purication unit operations, including storage and distribution systems, aims to improve the functionality of that unit operation over the long term, as well as the microbial quality of the water passing through it. Sanitization normally is achieved by exposing microorganisms in the water and the biolm growing on associated surfaces to a physical condition or chemical that kills them. In practice, sanitization can be frustratingly inadequate if not performed properly if the system components prevent sanitant penetration or if the materials of construction limit sanitization options to less effective approaches. Inadequate sanitization that leads to frequent or excessive system contamination can be the cause of signicant cost in terms of labor, lost production time, and if not performed when required, potential negative product, safety, and regulatory consequences. A few basic, rational concepts should be considered when sanitizing a water system to be assured of success and improve the microbial quality of the water system in the longer term. Misinformation or ignorance of efcacious parameters often leads to ineffective sanitization. 13.4.2.1 Frequency of Sanitization Continuous Sanitization Continuously sanitizing conditions that prevent initiation of biolm development is considered the ideal sanitization frequency.. The cells targeted for these continuously sanitizing conditions are usually naked “newcomers” with no frequency protective EPS coating. This allows sanitization conditions conditions to be of minimal “potency” because of the ease of target cell destruction. For continuous sanitization to be practical; however, the sanitant needs to be rapidly removable and leave no residue to be ushed out of the system. This limits the acceptable sanitizer choices to heat and ozone (at time of publication) because not only are these sanitizers quite lethal at the right doses, but they are the two for which system features can be designed to continuously tolerate their lethal or chemically reactive properties as well as rapidly “neutralize” them in situ without situ without residue. Continuous sanitization is considered the most effective approach to microbial control and frequently is employed for WFI systems, where any level of distribution system biolm development is generally intolerable. Continuous sanitization is used for PW systems less frequently because of cost and logistical considerations and the awareness that some minimal level of biolm growth may not be problematic. Intermittent Sanitization Systems may use intermittent sanitization for microbial control, particularly where a low microbial level in the water may not be problematic. A sanitization sanitization frequency that prevents signicant biolm development during the intervening non-hostile periods between sanitizations is considered fundamental to the success of this approach. Once biolm has developed its protective EPS coating, it becomes much harder to kill with chemical agents because of poor penetration to the cells imbedded in this EPS. When biolm has had opportunity to develop in crevices which also impede sanitant penetration, this resistance is more pronounced. If signicant biolm has been allowed to develop between sanitizations, particularly within protective crevices, the ability of even the most aggressive chemical sanitants to give lasting microbial control may be permanently compromised.




13.4.2.2 Kill and Remove the Biolm For lasting microbial control, biolm should be killed and the dead biomass removed. This biomass is a rich source of nutrient for any pioneer cells that may appear and attach to it after the sanitization process is complete. In the presence of such rich nutrients, biolm regrowth is more rapid. Complete biolm degradation and r emoval also assures that complete biolm kill has been achieved. If the biolm is only partially killed and partially removed, then biolm regrowth by the remaining live cells, nourished by the dead biomass, may rapidly rebound after a brief period of low microbial counts. 13.4.2.3 13.4.2. 3 Use an Effective Sanitizing Agent The sanitizing agent should be deadly to the microbial cells that thrive inside biolms. Some sanitizers may work by penetrating and killing the biolm, but leaving it in place (such as heat and chlorine dioxide), while other sanitizers work by chemically degrading the biolm starting at the outside and killing cells as they are exposed (e.g., most oxidizing chemicals or caustic). The oxidation potential is a factor in the efcacy of an oxidizing sanitizer; this is related to its ability to degrade strong covalent bonds in the complex organic molecules associated with the biolm, including the EPS and cellular components, such as endotoxin. If heat is used as the sanitant, it usually has no difculty in penetrating through to the base of the biolm and then into crevices where biolm may be growing. Heat needs to be sufciently hot to be lethal to biolm microorganisms after heat losses to the system components. In addition, heat has no direct ability to remove biolm or to degrade the endotoxin present in the biolm slime and on the killed cells. 13.4.2.4 13.4.2. 4 Use an Effective Sanitization Sanitization Procedure Appropriate sanitizing agents should be used at adequate concentrations and exposure times to be effective. What constitutes “adequate” depends on the resistance of the biolm to the attack and the properties of the sanitizing agent. There are no universally effective treatment parameters. If an oxidizing chemical is used, its efcacy is related to its reactivity (for oxidizers, expressed as oxidation potential) as well as its concentration, contact time, and possibly its temperature. Chemical sanitization failure often is the result of inadequate sanitizer contact time and insufcient sanitizer concentration, leading to incomplete destruction of the biolm and a rapid post-sanitization re-growth. Another signicant signicant cause of sanitization sanitization failure is the growth of biolm in crevices crevices or cracks created by sub-optimal sub-optimal construction techniques and materials. Within such protected locations, a chemical sanitant is unlikely to penetrate sufciently to kill all the biolm developing within, so after the sanitizing conditions are removed, the biolm resumes growth, possibly accelerated by a rich supply of nutrient from killed, but still present biolm material. Heat is considered an extremely effective approach to killing biolm because of its ability to penetrate even thick biolms easily, easily, and the susceptibility of biolm microorganisms to relatively modest levels of heat. The concerns with effective hot water sanitization focus on temperature and contact time, as affected by the balance between heat distribution and heat loss. The timing of the treatment should be based on the temperature at the coolest point in the system. The temperature of the return water may not be a good indicator of this temperature. This is particularly important for systems that are intermittently hot water sanitized and may not be well insulated against heat loss, making them very slow to equilibrate to a sanitizing temperature. Plastic systems are also a concern for this approach, because plastics do not readily conduct heat to moist surfaces at the distal parts of POU valves. In these situations, momentarily ushing the hot water through these valves will allow the penetration of the sanitizing heat to these moist surfaces. A follow-up chemical treatment after an infrequent hot water sanitization may be necessary to remove sizeable dead biolm deposits which might otherwise fuel rapid biolm regrowth. Wet locations that could support microbial growth should experience a heat or chemical sanitizing treatment for sufcient time to kill the microorganisms. Sanitant contact with all surfaces needing sanitization is important to the success of the process for both chemical and heat sanitization, including:




•

the domes of holding tanks (using sprayballs)

•

valves (especially sampling and at points of use)

•

at-line or in-line unit operations like heat exchangers

•

ow metering devices

•

hoses

•

lter housings (and installed lters, if compatible)

•

parallel/back up pumps and associated piping

•

piping/tubing (direct and sidestream) to and including instruments

•

other appropriate components or hardware

Following sanitization, the sanitizing agent should be removed to render the water usable again (except with hot water when compatible with products). With continuous sanitizing agents, this process should be achieved in situ with no system ushing required. For intermittently used chemical sanitants, the sanitant, along with any associated released debris, usually needs to be ushed out of the system. In such sanitizing applications, sloped and drainable systems have the advantage of allowing more rapid and efcient sanitant removal with minimal ushing needed to completely remove any residuals. If steam is used for sanitization, sloping and use of condensate bleeding valves are considered essential. Where ushing of sanitant residues is needed, microbiologically and chemically high quality water should be used to avoid contaminating and re-inoculating the system immediately after sanitization. This water can originate from a volume of water reserved just prior to sanitization or it can be the water freshly produced by the system normally after sanitization. System System ushing and rinsing capability should be designed into the water system from the start. 13.4.2.5 Minimize Recolonization Once a system has been sanitized and biolm has been reduced signicantly signicantly,, an ongoing process to minimize recolonization may be useful to prolong the inter-sanitization period. These approaches generally work by removing or killing all or most new pioneer cells and oating biolm ocs coming from the water purication units before they can enter and recolonize the storage and distribution system surfaces, allowing for protracted periods between resanitizations. This typically typically is done by controlling the biolm growth and shedding in the step prior to the distribution system or killing the planktonic organisms before or as they enter the storage and distribution system. A number of approaches have been commonly used, such as: 1.

an in-line UV sanitizer with or without a downstream microbial retentive lter

2.

continuous ozonation of the distribution storage tank, so that both incoming and recirculated water receive germicidal ozone doses

3.

a nal purication unit operation that is incompatible with microbial growth, such as a still, a hot water RO unit, or ultralter

An in-line microbially retentive lter should be used in conjunction with other upstream control measures like UV sanitizers. In the absence of immediately upstream control measures, the lter is likely to have only a short retentive period. The live microorganisms and ocs form a biolm on and within the lter membrane and rapidly may penetrate the lter to become a source of downstream contamination, after a relatively short period (possibly only a few days), rather than acting as a permanent absolute barrier.



13.5


Sanitizer Choices Generally, sanitizers can be grouped as physical or chemical. Physical sanitization options, such as UV sanitizers, Generally, may exert only a local effect; others may exert only a temporary effect, such as microbial retentive ltration. Heat is considered a more conventional physical sanitant, either as hot water or as steam, in pharmaceutical water systems. There are a number of chemical sanitizer choices, several of which are oxidizing agents with various oxidizing capabilities and other attributes that affect their lethal, penetrative, and biolm-destruct biolm-destructive ive properties and usefulness as water system sanitizers. In addition, a number of non-oxidative sanitizers may be used with varying degrees of success and often in special sanitization applications.

13.5.1 Physical “Sanitizers” 13.5.1.1 13.5.1. 1 Ultraviolet Light Sanitizers Light is maximally absorbed by the pyrimidine bases in DNA across the ultraviolet wavelengths of 240 nm to 280 nm. The absorbed energy makes these bases reactive, causing them to covalently bind to neighboring pyrimidine bases in the DNA, which prevents microbial replication and protein synthesis and ultimately kills the cells. The UV light emitted by low pressure mercury vapor uorescent light bulbs has an intense emission at 254 nm and is the type of bulb used in UV Sanitizers. At At the appropriate ow rate, water can ow through a chamber containing such bulbs and be in contact with the UV light long enough to kill 99% or more of the microorganisms in the water. As much as 1% may not be killed and no bacteria are killed outside of the exposure chamber. Those that survive the journey through the exposure chamber may do so because of being attached to an opaque particle (such as rouge) or imbedded in the center of a biolm oc, where they are shielded or shadowed from the deadly light rays. Many users incorrectly assume that in-line UV sanitizers sanitize the entire water system constituting continuous sanitization. Such in-line UV sanitizer lights can kill only what they shine on. Surviving microorganisms can form downstream biolms. In-line sanitizers slow the downstream development of biolm from planktonic cells that pass through the chamber suspended in the water ow. This is most valuable where the downstream biolm has already been eliminated by a highly effective approach, helping to reduce the frequency of this aggressive sanitization. The remaining 1% of microorganisms that survive the UV sanitizer light can be captured or signicantly delayed from passing further downstream, by placing micro-retentive lters downstream of a UV sanitizer, further extending the period between required between sanitizations. Another type of UV sanitizer uses medium pressure UV bulbs. These bulbs emit strongly across a broader spectrum including both 254 nm and 185 nm. These shorter and more energetic wavelengths can create extremely reactive, short-lived hydroxyl free radicals from the water molecules. The hydroxyl free radicals rapidly react, within a splitsecond, with organic molecules with which they make contact, including those on bacterial cells and some of the dissolved organic molecules that constitute the TOC in the water. When the free radicals react with the organic molecules, they can oxidize C-C covalent bonds creating organic carboxylic acids and increasing the conductivity of the water. If a polishing deionizer is positioned just downstream of an 185 nm UV sanitizer or elsewhere within the tank and distribution system, these ionized organic molecules can be removed from the water, therefore reducing its TOC level. As the water continues to recirculate past the 185 nm UV sanitizer and polishing deionizer, the TOC levels continue to drop. Care should be taken when using ultra low TOC values to control biolm, as deionizer beds and electro-deionizers can provide an ideal environment for biolm to ourish. For this and other reasons, including ow limitations, a separate polishing loop circulating around a tank, which incorporates microbial control features such as UV sanitizers, may be used. For further information, see the Chapter 8 of this Guide. A separate separate low pressure 254 nm UV sanitizer normally is not required when using a medium pressure UV unit because they also strongly emit the “bactericidal” wavelength of 254 nm. If required in different location for a different purpose, such as for ozone destruction or downstream of a distribution system polishing deionizer, a separate low pressure 254 nm UV sanitizer may be justied.




13.5.1.2 Filtration Application of ltration ranges from the coarse lters used to remove multi-micron sized particles from incoming water or protect downstream unit operations from debris-shedding granular beds, to the ultra-ne RO lters whose permeability is so ne that essentially only water molecules can pass through. In water systems, there are usually several grades of lters between these two extremes intended for various retention purposes. Coarse Filtration Coarse lters are not intended to retain bacteria and are not used for microbial control. Where left in service for excessive periods; however, however, coarse lter can add to the microbial content of the passing water. The large surface area of these lters usually becomes heavily colonized by biolm (even in the presence of incoming chlorinated water), which then sheds microbial contaminants into its efuent water, particularly during the surge of “start and stop” operation. They may present a more serious microbial contribution where they are used between unit operations within the purication train, to protect downstream unit operations from escaped bed granules and fragments. The backpressure that develops over time in coarse lters may be more because of biolm occlusion than debris loading. With such heavily colonized locations able to shed large amounts of biolm bacteria, their maintenance or replacement frequency, frequency, based on backpressure, may need to be assessed on the microbial sensitivity of the downstream unit operation, rather than on the manufacturer’s replacement recommendations. Micro-Retentive Filtration Micro-retentive lters with ratings of 0.22 µm or 0.2 µm “absolute” traditionally have been used to lter sterilize processed liquids prior to aseptic packaging. There is a long-held misconception that “absolute” rated lters are absolute in their removal capability capability,, that the lter’s rating is its maximum pore size, and that the lter functions solely as a sieve to remove bacteria. None of these beliefs are true for most membrane lters. Such lters’ pore sizes is actually a range of sizes with the largest possibly being up to two to three times larger than its rating, meaning that a small percentage of particles larger than the rating could get through. Bacterial capture by the lter is based on a mixture of sieving effects and a cell adsorption phenomenon occurring within the lter matrix; organisms tend to stick to the surface of matrix “bers” as they ow through the lter along an intrinsically convoluted path created by the matrix. In fact, in spite of a lter’s validated absolute retention of the challenge organism Brevundimonas diminuta, diminuta, microbial penetration of these supposedly absolute 0.2 μm-rated lters does indeed occur, and it is especially prevalent with water system microorganisms. In water systems, it occurs well before any hint of backpressure has developed. In fact, by the time there is any even barely detectable additional back pressure, the lter may have been passing large numbers of bacteria for quite some time. The occurrence of this penetration phenomenon is unarguable, but its mechanism is still debated. One theory is that particularly small cells (smaller than B. diminuta) diminuta) and/or cells with low surface hydrophobicity (less hydrophobic than B. diminuta), diminuta), like those found in water systems, eventually happen upon one of these larger pores, avoid hydrophobic adsorption to the lter matrix, and manage to simply “go through” the lter. Another theory involves “grow through” in which the bacteria colonize the lters internal matrix surfaces and through the cellular elongation that precedes binary ssion, some of the newly formed daughter cells are incrementally pushed along these surfaces and through small gaps, eventually all the way to the downstream side. Yet Yet another theory involves a “blow through” phenomenon in which intermittent ow through a lter may either cause some lter matrix exing and momentary pore size alteration or the sudden shear forces cause some adsorbed cells to momentarily detach and incrementally work their way through the lter matrix, perhaps aided by their own motility. Microbial penetration of 0.2 µm-rated lters by water system organisms typically occur when micro-retentive lters are kept in service too long. Among other variables, the length of time for which these lters should be kept in service depends upon the size, surface properties, and level of viable bacteria impinging on the lter, as well as the composition and pore size distribution of the lter, ow rate effects, and a number of other variables.




There is no generally applicable time before lter penetration occurs. Where allowed by regulatory expectations, an appropriate usage period can be determined empirically and validated for a given ltration application. Note: the FDA disallows these lters to be used for unvalidated use periods or in situations where the lters are Note: the used to mask poor water quality because of poor system design and maintenance. European regulatory authorities generally disallow their use in water systems. 0.1 µm-rated lters, which purportedly are impenetrable by small bacteria found in water systems, may be used. The integrity challenge tests for these lter ratings are less standardized and difcult to perform, so the r etentive properties are hard to verify and not all lters with this rating are equally effective at retaining specic particularly less hydrophobic aquatic bacteria. The ner lter matrix for these lters may create greater ow resistance than 0.2 µm-rated lters, and a greater number of 0.1 µm-rated lter cartridges being need to achieve the same ow rate as 0.2 µm-rated lters. Depending on the application, the potential sacrice in ow rate and cost may be warranted, but these lters also may plug with tiny particulates or biolm more quickly than 0.2 µm-rated lters, adding to the cost. The use of in-line 0.2 µm-rated lters immediately downstream of UV sanitizers is considered to be a more effective and sustainable use of these lters. The UV sanitizers (or upstream ozone) kill the majority of cells before they impinge on the lters. The lters capture the killed cells and the remaining small viable population, preventing them, for a time, from getting through the lter and downstream to recolonize new distribution system surfaces. This tandem unit operation helps to prolong the usable life of the lter, possibly from only a few days, with an unprotected lter, lter, up to several months, depending on the water’s microbial content and the efciency of the UV sanitizer. sanitizer. The combination of UV sanitizer and 0.2 µm-rated ltration should be used without interruption or even being momentarily bypassed, and the downstream system should have been previously sanitized with almost all downstream biolm killed. If the water is allowed to bypass these units during maintenance, e.g., to replace bulbs or lter elements, or during bypass sanitization, or subsequent rinse out, biolm colonizers may be allowed downstream into the system to begin forming biolms. In such situations, system sanitization sanitization should be performed again soon afterward to re-establish a system-wide condition of near freedom from biolm, which the UV/lter combination is intended to prolong.

Ultraltration

These lters have minute pores and are able to screen out large organic molecules to a rated molecular weight cut-off. The use of ultralters for endotoxin removal from water is reasonably common for non-distillation systems, particularly downstream of RO or deionization systems. For this application, JP Information Chapter <21> (Reference 6, Appendix 1) cites a molecular weight cut-off of 6,000 Daltons; however, because endotoxin invariably exists in pure water in an agglomerated, multiple-molecule multiple-molecule state, higher molecular weight cut-offs of 10,000 or 20,000 Daltons also have been used effectively. Ultralters often are used in a tangential ow pattern which may not assure microbial impenetrability impenetrability.. There is no widely accepted pre-use integrity test (at time of publication) so the ability to determine membrane and assembly integrity generally is not possible. Theoretically these lters should not be able to pass bacteria, but their bacterial retentiveness cannot be assured. Ultralters typically are sanitized with chemicals which are a low enough molecular weight to penetrate both sides of the lter medium. They offer a barrier to microbial penetration and to penetration by organic macromolecules (and silica colloids) with molecular weights above the lter rating. Issues with their maintenance include the: •

frequency of sanitization

•

degree of biolm development

•

efcacy of the sanitizer at killing and removing biolm




Some ultralter materials and designs may be sanitized with hot water and others may be operated continuously with hot water. When the temperature is sufciently high to prevent biolm growth and development, concerns regarding biofouling and non-absolute microbial retention of ultralters are moot. Reverse Osmosis RO units can offer a barrier to microbial and endotoxin penetration, but with the same unveriable membrane and assembly integrity issues as ultralters. They typically are sanitized with chemicals, but with the added impediments that only a few sanitizers are compatible with RO membranes and the reduced ability to sanitize the permeate side of the RO membrane, which may lead to downstream biolm development and potential generation of endotoxin by that biolm. Some sanitants are small enough molecules or are in equilibration as dissolved gases that can penetrate RO membranes to sanitize permeate surfaces. This includes once commonly used traditional sanitants, such as formaldehyde, but were subsequently found to be carcinogenic, warranting extreme care in handling and are used only infrequently. infrequently. RO units may be designed to be either hot water sanitizable or continuously operated with hot water, which can contact all upstream and downstream surfaces. Concerns with ROs include: •

frequency and efcacy of chemical cleaning and sanitization

•

possibility of permeate-side biolm development

•

non-absolute microbial retention and unveriable membrane integrity

These concerns are greatly reduced if the RO can be operated continuously or intermittently with hot water. 13.5.1.3 Heat Sanitization with Hot Water In high purity water, biolm growth typically is either minimal or absent above approximately 45°C. Temperatures above approximately 50°C usually are hostile and slowly lethal. Higher temperatures such as 65°C and 80°C often are used for hot water sanitization. EU GMP (Annex 1, #59) (Reference 40, Appendix 1) suggests using greater than 70°C. The higher the temperature, the quicker will be the microbial death. Aquatic thermophiles that may tolerate or require these relatively modest heat levels (from a sterilization perspective) do not exist in high purity water systems because of the absence of their essential nutrients. Designing sanitization sanitization cycles or monitoring techniques to kill thermophiles or detect their presence is considered unnecessary and a waste of resources. Water systems operating at or above 65°C generally are considered to be self sanitizing; 80°C is an often used target temperature. When used continuously, continuously, biolm does not have an opportunity to form in locations where those temperatures are present, so the water can be maintained, fundamentally, fundamentally, in a sterile state. It is common for WFI systems to be operated in a continuously self-sanitizing condition, because because the system has little toleration for microbial/biolm growth, one of the possible sources of endotoxin which is tightly regulated in this grade of water. PW systems also may be operated under continuously sanitizing temperatures, either to assure trouble-free microbial control or because of process needs. Intermittent hot water sanitization is more common in PW systems. Hot water sanitization has the advantage of penetrating, by heat conduction, into crevices where biolm could be growing. Conduction-mediated heat penetration may reduce the concerns with microbial control in crevices associated associated with seals, gaskets, and surface imperfections. The temperatures generally used are in the range of 65°C to > 80°C. Sanitization efcacy issues can be accommodated with the timing for the treatment beginning when the target temperature is reached at a determined coolest point (which may not be the return to the tank). Treatment times of 0.5 hours to 4 hours are common for > 80°C sanitizations, with longer times required for more complex designs, to assure heat penetration to all moist surfaces. Care should be exercised with 65°C sanitizations since heat losses can cause a temperature drop of as much as 10°C to 15°C across a metal POU valve to downstream moist surfaces, yielding minimally sanitizing conditions on those distal surfaces. Longer exposures may be required if relying solely on heat conduction




for sanitizing these surfaces. In such situations, ushing the hot water for a few seconds through the valve effectively can overcome the otherwise minimally sanitizing condition. This approach is considered particularly useful for plastic systems, because of poor heat conduction through valves. Occasionally, plastic systems (and stainless steel systems) will be considered for hot water sanitization when Occasionally, chemical sanitization has failed to achieve microbial control. In such situations, the plastic systems may not have been designed to be heated and may not be fully compatible with typically used sanitization temperatures. The plastic systems theoretically theoretically may be tolerant of the stresses of marginally sanitizing temperatures of 55°C to 60°C. Care should be exercised before attempting this since there may be inordinate heat losses, due to long, uninsulated runs of piping, as well as thermal expansion, piping support (permanent sag), and physical strength issues that could cause a catastrophic failure in pipe or tank integrity, integrity, or an unanticipated incompatible component in a unit operation. Sanitizing under sub-optimal conditions may cause several additional concerns, including: 1.

It usually requires the treatment time to be quite long to accommodate a slow heat up, prolonging the risk period to system integrity.

2.

The heat losses in the system system may require water hotter than the treatment treatment temperature to to be present in the initial stretch of piping downstream of the heat exchanger so heat compatibility in this location is particularly important

3.

Plastic tanks may need to be sanitized in a minimally lled condition to minimize hydrostatic stress on the tank walls when hot.

13.5.1.4 13.5.1. 4 Heat Sanitization with Steam Steam can be used to sanitize water systems. Traditionally Traditionally,, steam under pressure has been used in sterilization methods, and generally is perceived as more effective than using hot water for water system sanitization. Water Water systems can be designed to be steam sanitized and the temperatures this approach can achieve are far in excess of those necessary to kill biolm organisms. Sterilizing conditions of steam may be required to kill extremely heat resistant exogenous organisms that may enter a water system through a compromised vent lter or rupture disc, or prior to initial system start-up. Such organisms are not aquatic, typically do not produce endotoxin, do not have opportunity to become established or proliferate, and are transients that do not survive very long in water systems. A hot hot water sanitization and system ushing is usually all that is required to eliminate these microorganisms, but system sanitization sanitization with steam is an alternative approach. If steam is used for in situ sanitization situ sanitization within a water system, USP pure steam should be used. If this purity of steam is not available, precautions should be taken to assure component compatibility and post-sanitization post-sanitization ushing or chemical neutralization to assure that all steam additive residues and particulates have been removed. During design and construction, consideration should be given to: •

the choice of materials and design features to accommodate the extreme temperature differentials between ambient routine use and sanitization

•

piping insulation

•

complete system drainability prior to sanitization

•

sanitary air venting during sanitization

•

pipe sloping

•

condensate drainage and removal

•

sanitary steam injection valve design



•


the quantity of steam that may be needed for large water systems

Insulation is important for safety and to avoid heat loss and poor heat distribution in long piping runs, large tanks, and granular beds, and to reduce the amount of steam needed. If such features are improperly installed or fail (e.g., condensate accumulation near a clogged condensate drain), the accumulating water may not be sufciently hot for even “hot water” sanitization in that section of a system. The use of steam for sanitization may increase the maintenance requirements of a system, including issues related to faster rouge development. If steam is used to sanitize unit operations, such as Granular Activated Carbon (GAC) beds, without the proper steam injection design, the heating of the carbon in the bed may not be uniform and the sanitization will be ineffective because of the tendency for steam to channel through the bed rather than ow evenly through it. GAC beds are more thoroughly sanitized by rst backwashing to separate any biolm-agglomerated granules, followed by a ow of hot water to kill that biolm; adequate temperature and contact time are essential.

13.5.2 Chemical Oxidizing Sanitizers The effectivenes effectiveness s of oxidizing sanitizers depends on the combination of their oxidation potential, concentration, stability,, and contact time. Generally, stability Generally, the higher the oxidation potential, the more reactive it is against strong covalent bonds in organics and the more rapid it is against the weaker bonds. It penetrates poorly into thick, well-developed biolm because of its reactivity at the outside layers. The penetrability can be improved by using high concentrations and long contact times. This apparent disadvantage is balanced by the potential to completely degrade and remove the biolm, which is desirable for prolonged microbial control in water systems. Biolms growing into crevices may be almost impossible to kill with chemical sanitizers, because of the limited exposed surface area. Understanding the properties of a chemical sanitant helps in using it to greatest effectiveness. effectiveness. Table Table 13.1 gives the absolute oxidation potential of various sanitants and the active components of several of these sanitants. The table also depicts a relative oxidative effectiveness, effectiveness, as compared to chlorine gas ( commonly used as a sanitant in potable water). Table 13.1 Oxidative Effectiveness of Vario Various us Sanitants and their Reactive Components Oxidation Potential(Note a)

Relative Oxidative Activity(Note b)

Fluorine (F2)

3.06

2.25

Hydroxyl Free Radical (•OH)

2.80

2.05

Ozone (O3)

2.07

1.52

Peroxyacetic Acid (Note c) (CH2COOOH)

1.81

1.33

Hydrogen Peroxide (H2O2)

1.77

1.30

Perhydroxyl Free Radical (•OOH)

1.70

1.25

Hypochlorous Acid (HOCl)

1.49

1.10

Chlorine (Cl2)

1.36

1.00(Note b)

Bromine (Br 2)

1.09

0.80

Chlorine Dioxide (ClO2)

1.57 (~1.0)(Note d)

1.15 (~0.7)(Note d)

Hypochlorite Ion (OCl-)

0.94

0.69

Iodine (I2)

0.54

0.40

Agent/Chemical

Notes: a. Oxidation potential expressed in Electron Volts Volts.. b. Oxidative Activity expressed relative to Chlorine as 1.00. c. Commonly known as Peracetic Acid. d. Possesses multiple oxidation oxidation states with with the lower value value most functional functional in water with with biolm penetration penetration as a gas. gas.




13.5.2.1 Ozone Ozone is triatomic oxygen (O 3). Its instability and explosive reactivity at concentrations above 10% prevent it being supplied as a compressed gas. It is frequently and economically generated at the POU by passing diatomic (O 2) oxygen gas through an electrical coronal discharge. The resulting ozone gas is then sparged into water to dissolve it. The oxygen gas should be as pure and as dry as possible, because the presence of even small amounts of nitrogen gas can lead to the formation of nitrogen oxides within the coronal discharge that, when dissolved, become nitrates and degrade the water conductivity. Electrolytic technologies also are used, require no feed gas, and create ozone from the electrolytic splitting of water molecules, with the ozone dissolved directly in the water as it is formed. Ozone created through electrolytic technology usually is more pure that that created by coronal discharge and may be able to achieve higher dissolved concentrations, though at a higher component cost than corona discharge. Ozone is a sparingly water soluble gas and an aggressive oxidizer in its own right, but when it reacts with water, it generates more aggressively oxidative hydroxyl free radicals by the reaction: O3 + H2O → 2 •OH + O2 Both the ozone and the hydroxyl free radicals react with organic molecules with sufcient energy to break most types of covalent bonds, and if sufcient dissolved oxygen is present, to insert an oxygen atom where a bond is broken. The organic degradation tends to form carboxylic acids which affect the water conductivity, conductivity, and may need to be addressed with system ushing or water use, or with deionizer polishing and appropriate precautions. See Section 13.5.1.1 of this Guide. Ozone used as a water system sanitizer usually is in concentrations ranging from 0.02 ppm to 2 ppm, with concentrations of 0.08 to 0.2 ppm commonly used. Because of ozone’s instability and tendency to react with the TOC in the water, it may require injection into a long-looped or multi-looped system at more than one point to assure the target concentration is reached throughout. When used continuously continuously,, ozone usually is used at or below 0.1 ppm levels because it generally attacks very susceptible pioneer cells or small biolm ocs entering the distribution system from a nal unit operation in the purication portion of the system. When ozone is used as an intermittent sanitant, increased concentrations usually are required because the ozone may be attacking EPS-coated biolm which it penetrates poorly. Biolm is killed by ozone via attack from the outside. It penetrates the biolm only as the outer layers of EPS, endotoxin, and bacterial cells are degraded. If the biolm has had ample time to develop and is thick, ozone has little chance to completely kill it without prohibitively lengthy, repeated applications. Only thin biolm can be completely killed by ozone, so using ozone as a periodic sanitant should be timed such that any EPS-coated biolm formation is in its very early stages. The frequency depends on factors unique to each water system, but daily to weekly treatments have been found to be effective at minimizing or preventing recolonization and interrupting biolm regrowth. Continuous ozonation of the storage tank into which the loop return and make-up water from the purication system is fed may be used to prevent recolonization. This normally is achieved either by injecting the ozone through a venturi into the recirculated water just prior to its return to the storage tank or by sparging ozone into the tank, either technique creates an ozone-lled hostile environment in the tank water. The ozone can be destroyed by 254nm UV before the rst POU. Periodic loop sanitization with ozone can be achieved by stopping its destruction by de-energizing the ozone destruct UV.. Consideration should be given to assuring that sanitizing levels of ozone are present throughout the loop to UV compensate for its natural degradation and reactivity with any biolm or organics present in the loop. The minimal ozone concentrations that are effective as a continuous sanitant in the tank may be ineffective as an intermittent sanitant in the loop; therefore, the ability to considerably increase the ozone concentration during loop sanitization may be signicant to this approach.




When ozone is used as a sanitant, its destruction is as important as its formation. It should not be present in water used from the system. If ozone is present in this water, it both violates the pharmacopeial water monograph requirement of containing no added substance, and is so reactive that it could degrade ingredients or surfaces the water contacts. In addition, it may create a health hazard to the operators in the environment, as the ozone outgases into the room air. Ozone is degraded easily by germicidal 254nm UV sanitizers; however, it is usual to size these units at 2.5 to 3 times the intensity used for germicidal activity to assure complete ozone destruction. The irradiation activates the ozone to react with water to form transient, highly reactive hydroxyl free radicals and diatomic oxygen by the equation: O3 + H2O + UV 254 254 nm → 2 •OH + O2 The ozone destroying process adds a nal surge in lethality; viable organisms in this water stream have are unlikely to survive an encounter with the free radicals or the UV. Although Although the free radicals exist only for a fraction of a second in less pure waters, care should be exercised with very high purity water to avoid de-ozonated water contacting sensitive applications in the rst few seconds after ozone inactivation. The hydroxyl free radicals also react with any dissolved CO2 or bicarbonate in the water so less pure de-ozonated waters do not have this concern. Additional details on design and use of ozone is discussed in the ISPE Good Practice Guide: Ozone Sanitization of Pharmaceutical Water Systems (Reference 41, Appendix 1).

Advantages: Ozone is an extremely potent and reactive oxidizer oxidizer,, which is quickly lethal to naked bacteria cells not imbedded in slimy ocs, and given sufcient time, it can degrade biolm directly along with its components, such as EPS, endotoxin, and the bacteria cells. Ozone can be used to help reduce TOC and endotoxin levels. Ozone normally is used for continuous sanitization of complete or parts of water distribution systems (for eliminating biolm re-colonizing pioneer cells) at low concentrations. It is removed easily by germicidal UV irradiation, eliminating the requirement for removal by system ushing. As an intermittent sanitant, ozone is effective against naked planktonic bacterial cells and on early (thin) biolm formations that have produced minimal EPS slime.

Disadvantages: Despite its reactivity reactivity,, several factors compromise the ability of ozone to penetrate and kill existing (thick) biolm formations effectively, including: •

unstable

•

short-lived and poorly soluble

•

able to attain only very low ppm to sub-ppm concentrations in water

Ozone tends to react supercially on thick biolms and does not penetrate well because of its extreme reactivity with all organic molecules and its low concentrations. Ozone has minimal efcacy against thicker, thicker, older biolms embedded in substantial EPS, even after many hours of contact time. The key to the efcacious use of ozone as an intermittent sanitant is to use it frequently, before thick biolm has an opportunity to form. Ozone may need to be added at several locations throughout long distribution systems for lower chemical purity waters, as it reacts and degrades so quickly. quickly. The extreme reactivity of ozone also is directed toward system materials; although it is well tolerated by stainless steel, PVDF and PTFE plastics, EPDM, Viton, or PTFE-coated elastomers, it is particularly aggressive toward many other plastics and elastomers. The incompatibility of ozone with system components is a signicant limiting factor associated with its use. Ozone is toxic to humans at low atmospheric concentrations and its use, particularly around ozone generators and when ushing use points with




ozonated water to sanitize outlets, should be controlled carefully and monitored for undue indoor airborne releases. UV ozone destruct units, airborne detectors, controls, and alarms should be used to ensure worker safety. 13.5.2.2 13.5.2. 2 Hydrogen Peroxide Hydrogen Peroxide (H2O2) is used as an intermittent sanitant. The energetic and unstable peroxy (-O-O-) bond in this molecule is its source of oxidative reactivity. reactivity. It also can form the highly reactive hydroxyl free radical in water upon UV 254 254 nm irradiation: H2O2 + UV 254 254 nm → 2 •OH Hydrogen peroxide is widely available as a 30% solution, but technical grades are stabilized with ppb levels of certain metals which may be unsuitable for use in high purity water systems. Reagent grades contain the less objectionable 1ppm of zinc or tin and 0.5ppm of iron as stabilizers. Hydrogen peroxide has been used effectively for water distribution system sanitization sanitization at concentrations ranging from 3% to 10% with contact times of several hours (the longer the better). It is compatible with most surfaces, including thin lm composite polyamide RO membranes. Hydrogen peroxide decomposes over time to water and oxygen gas, which would not be considered added substances as they are already present in water. 2 H2O2 → 2 H2O + O2 After hydrogen peroxide sanitization, the sanitant solution should be ushed out of the system because of the metal ion stabilizers in the concentrate and the organic molecules and debris potentially released by biolm degradation. Hydrogen peroxide is not particularly temperature stable so its use is limited to ambient temperatures, which should not exceed 25°C for maximum stability. It is pH stable; this attribute has been used by combining hydrogen peroxide at a less than 1% concentration with a 1% caustic solution (pH 14) to make an extremely effective effective remover of organic matter for compatible surfaces. Hydrogen peroxide also can be combined with ozone use to make a particularly aggressive sanitant based on the prolic formation of hydroxyl free radicals. The compatibility of hydrogen peroxide with other similar acting oxidizers and their free radical-potentiating effects also has been exploited with other proprietary sanitants and sterilants. When hydrogen peroxide is used, depending on the amount of biolm present, there may be a substantial evolution of oxygen gas facilitated by cellular catalase; relief of developing pressure should be considered. Although the lethal activity is almost instantaneous for naked bacterial cells, the process is much slower for developed biolm that is degraded from the outside, so sufcient exposure time should be allowed for the sanitization and biolm r emoval to be as complete as possible. The higher the concentration of the sanitant, the more rapid the action; treatment durations of many hours are common. The amount of time required for a given system will depend on the level of biolm development allowed since the previous sanitization; an efcacious treatment time can be determined only by experience.

Advantages: Hydrogen peroxide is an aggressive sanitant that is available without difculty as a concentrate. It can be diluted for use at high concentrations to maximize efcacy, efcacy, though 3% to 5% is typical. At the correct concentrations, it may be compatible with sensitive surfaces such as TFC polyamide RO membranes. It possesses reasonably good stability, stability, so adding additional chemical to maintain an effective concentration concentration over long treatments usually is unnecessary if the initial concentration is sufciently high. The pH stability of hydrogen peroxide allows it to be combined with caustic to form an exceptionally good sanitizer and biolm remover. Its interactivity interactivity with other oxidizing sanitizers creates an abundant release of extremely reactive and lethal hydroxyl free radicals making the mixture more reactive than either sanitizer alone.




Disadvantages: Hydrogen peroxide is unstable at elevated temperatures, limiting its use to temperatures no warmer than approximately 25°C. After After a sanitization cycle, the sanitant should be ushed from the system. Quick test kits and test strips are available for conrmation of post-sanitization rinsing efcacy, efcacy, but are usually quantitative only down to 1ppm, which may not be sufciently low.

13.5.2.3 Peracetic Pe racetic Acid Peracetic acid also is known as peroxyacetic acid (CH 3COOOH), this may be a less commonly used intermittent water system sanitant. (It is often used as a vaporized or aerosolized sterilant for enclosed environmental surfaces like sterility test isolation chambers.) An energetic peroxy bond is the source of its oxidative reactivity. It is available as a 40% solution, but should be stabilized with heavy metals. It may work in a similar way to hydrogen peroxide with similar if not better effectiveness. It naturally decomposes to acetic acid and oxygen: 2 CH3COOOH → 2 CH3COOH + O2 Peracetic acid has been used at concentrations ranging from 50 ppm to 4%. It is considered safe for use with polyamide RO membranes at a 1% concentration. It is not susceptible to the temperature instability of hydrogen peroxide, but is less stable than hydrogen peroxide at high pH, being susceptible to alkali dissociation. Peracetic acid is suitable for mixing with other oxidants for potentiated oxidizing activity, activity, and because of this, its use as a solo sanitant in water systems is largely being replaced by proprietary oxidant mixtures. Sanitization treatments usually are several hours, the duration of which should be determined by experience and is a function of the concentration used and the amount of biolm that has developed.

Advantages: Peracetic acid is an aggressive sanitant that can be used at high concentrations to maximize efcacy if desired, though 1% to 4% is typical. At At the correct concentrations peracetic acid is compatible with sensitive surfaces, such as TFC polyamide RO membranes. It possesses moderate stability, stability, including at elevated temperatures, so adding additional chemical to maintain an effective concentration over long treatments usually is unnecessary if the initial concentration is sufciently high. Using peracetic acid at higher temperatures signicantly increased its efcacy. efcacy. Its interactivity with other oxidizing sanitizers creates an exceptional release of extremely reactive and lethal hydroxyl free radicals making the mixture more reactive than either sanitizer alone.

Disadvantages: Peracetic acid is unstable at high pH so combining it with certain other high pH agents such as un-neutralized sodium hypochlorite or caustic is not feasible. It is also moderately volatile and toxic to inhale, so use at elevated temperatures, which increases volatility, volatility, should be accompanied by appropriate precautions. After a sanitization cycle, the sanitant should be ushed from the system. Depending on local codes, neutralization of the acidity and oxidative reactivity may be required before discharge to the sewer. Redox, conductivity, conductivity, or TOC instruments should detect its presence when assuring post-sanitization rinse-out completeness.

13.5.2.4 13.5.2. 4 Hydrogen Peroxide and Peracetic Acid Mixtures Hydrogen peroxide and peracetic acid mixtures are obtainable frequently as proprietary combinations, which claim to be more effective than either ingredient alone, even when used at much lower concentrations. According According to material safety data sheet information, several products contain an approximately 5:1 ratio of hydrogen peroxide to peracetic acid (22%: 4.5%) that is intended to be used at a 1:100 dilution of the concentrated mixture. At this concentration, it reportedly is safe for most water contact materials often found in high purity water systems and is considered to be highly effective at killing naked aquatic bacterial cells, and killing and removing biolm. The correct contact time should be allowed for complete biolm kill and removal, which may be affected by TOC content of the water and the




depth of the biolm that is being treated. An effective contact time should be determined by experience. For minimal or very thin biolm development and high water purity situations, contact times of one to two hours are common; signicantly longer times usually are required where thicker biolms have developed.

Advantages: Commercia lly availabl Commercially available e mixtures commonly are used in water systems with testimonia testimonials ls for efcacy. Test strips are available for assessing the correct treatment concentrations and post sanitization rinsing efcacy (for the latter, sensitivity may not be sufcient). The treatment concentration is effective and not harmful to sensitive surfaces, such as TFC polyamide RO membranes.

Disadvantages: The diluted mixture has a limited shelf life. Hydrogen peroxide and peracetic acid combinations are unstable after dilution to concentrations appropriate for use, at elevated temperatures, and high pHs. At the conclusion of treatment, the chemical and any associated organic debris should be ushed from the system. The acidic nature of the chemical may require pH neutralization before discharge to the sewer, depending on local codes.

13.5.2.5 13.5.2. 5 Chlorine Gas The use of chlorine for disinfecting water has a long and accepted history history,, primarily for potable or drinking water. When chlorine gas is dissolved in water, it reacts to form hypochlorous acid which is the sanitant molecule: Cl2 + H2O → HOCl + H + + ClThe use of chloramines has largely replaced the use of free chlorine because of the more r eactive chlorine’s tendency to break down organics in the water and form carcinogenic trihalomethanes (THMs) and haloacetic acids. Chlorine gas is rarely used in the pharmaceutical industry to sanitize PW systems because of issues of safety and practicality. practicality. It may be used occasionally in a pharmaceutical organization’s private source water purication processes or to supplement low chlorine levels in incoming municipal drinking water. The larger concern with chlorine in water is the water’s further purication and the requirement to remove it to avoid damaging sensitive unit operations. Highly reactive chlorine (as well as chloramines) can seriously damage anion exchange resins, EDI units, and TFC polyamide RO membranes. It also can cause a pitting corrosion of stainless steel, which is particularly serious at high temperatures, such as in distillation units. There are several processes that may be used to remove chlorine (and chloramines), including activated carbon beds, reducing agents, and strong UV irradiation.

Advantages: The use of chlorine gas as a water sanitizer is very economical in terms of materials cost. There is a long history of its use in treating large quantities of water with low ppm doses and it is a familiar compound. When used at EPAregulated levels for purifying private sourced water to become potable or to supplement municipal water chlorine levels, it does not need to be removed prior to direct human consumption, and its removal prior to further purication typically is already designed into water systems.

Disadvantages: Chlorine is extremely hazardous to humans in concentrated gas form and requires special handling equipment to prevent leaks, as well as safety alarms for airborne discharges. It is impractical to use for creating the high dissolved chlorine levels required for water distribution system sanitization. The low chlorine levels associated with the potable feed water to the system can kill pioneer cells given sufcient contact time, but otherwise does very little to control biolm growth.




13.5.2.6 Sodium Hypochlorite Sodium hypochlorite is used frequently and is familiar in the form of household disinfectants and laundry bleach. Typically Typi cally,, it is obtained as concentrated solutions of sodium hypochlorite (from the 6% concentration in household bleach to industrial solutions at as high as 21%). Such solutions are usually buffered to a pH of approximately pH 12 to enhance its stability during shelf life. The most active moiety is hypochlorous acid (HOCl) with the hypochlorite ion having less than 2/3 of its oxidizing activity. activity. It has a pK a of 7.4, so at pH 7.4 it exists in equal amounts as undissociated hypochlorous acid and dissociated hypochlorite anion. HOCl → H+ + OClThis equilibrium makes the activity of this sanitant highly pH dependent with neutral to acidic pHs yielding more oxidative reactivity than alkaline pHs. Acidic pHs reduce its stability. stability. When diluted for use as a sanitant, its efcacy can be enhanced by adjusting the pH of the nal solution to below neutrality for maximal activity in its non-ionized form, as hypochlorous acid. However, in this “acid” form, it is in equilibrium with chlorine gas which tends to outgas from the solutions, so care should be exercised when neutralizing these solutions in their concentrated forms. It is moderately heat stable, so increasing the system temperature during sanitization, where compatible with the materials of construction, can signicantly increase its activity activity (the general rule is 2-fold for every 10°C increase). Simply increasing the treatment concentration may provide an equivalent effect to that of heating or pH neutralization, as there is no limitation on the maximal concentration that can be used during sanitization, except perhaps by cost (this compound is widely available and relatively inexpensive). Sanitant use concentrations generally range from 50 ppm to 5000 ppm (0.5%) as a function of the required contact time, whether or not pH-neutralized, the TOC content of the water, and the depth and location of the biolm that is being treated. Sufcient contact time should be allowed for it to degrade and remove developed biolm. At the completion of treatment, the chemical should be ushed from the system along with any organic debris it may have released, and if the hypochlorite is used in an alkaline form, the amount of ushing required may be extensive. Simple chlorine test kits may be used and redox instrumentation also is available to determine when system ushing is complete. Water conductivity conductivity can be used to reinforce an indication of rinsing completion. If large amounts of hypochlorite have been used, it may require inactivation with reducing agents, such as sodium sulte or bisulte before discharge to a sewer. This sanitant is incompatible with prolonged contact, even at low concentrations and particularly at high temperatures with stainless steel, as it causes chloride corrosion. In addition, it is not compatible with TFC polyamide RO or ultralter membranes; however, however, it is compatible at a 10 ppm concentration with cellulosic RO membranes.

Advantages: Sodium hypochlorite is generally available, relatively inexpensive, reasonably safe to use, and familiar familiar.. Although it is not the most potent oxidizing agent that may be used, the lesser oxidative power can be partially compensated for by using higher concentrations, making it a commonly used and effective sanitant. sanitant. It is heat stable, so its efcacy can be further increased by treatments at elevated temperatures, system materials permitting, as well as by neutralizing the pH of the diluted treatment solution. It is inactivated easily by reducing agents for easy disposal and tests for chlorine residues after rinsing are generally available.

Disadvantages: Sodium hypochlorite should not be used in stainless steel systems or in contact with chlorine-sensitiv chlorine-sensitive e or oxidizersensitive equipment. Post-sanitization Post-sanitization removal is by system ushing, but the volume required could be extensive if the hypochlorite is used in its alkaline form.




13.5.2.7 13.5.2. 7 Chlorine Dioxide Chlorine dioxide gas is freely soluble in water. It has been used widely in other industries and applications for effectively effectivel y killing biolm accumulations and for correcting odor and taste problems in drinking water. It should be generated at the point of use, as it is explosive in a concentrated state. Normal generation approaches include passing chlorine gas through a bed of sodium chlorite, bubbling the resulting chlorine dioxide gas into the water, or by mixing solutions of chlorine gas (or pH-neutralized hypochlorite) with solutions of sodium chlorite and then adding them to the high purity water distribution system to be sanitized. Once the dissolved chlorine dioxide gas is present, it is effective over a broad pH range of 1 to 10. Chlorine dioxide possesses multiple oxidation states, but its relatively low oxidation potential gives it selective organic reactivity,, which combined with its existence as a gas rather than a dissolved ion, allow it to deeply penetrate biolm reactivity without reacting with the EPS. There it subsequently enters the embedded cells to oxidize sulfhydryl-containing proteins, as well as complex amines and other macromolecules found in living cells, quickly killing the cells. Its low effective oxidation potential, though penetrative and lethal to biolm cells, does not allow it to degrade and remove biolm. Its penetrative properties start start at a 1.5 ppm level, but can be used at levels as high as 150 ppm. Treatment times vary depending on the concentration used and the depth of biolm needing to be killed, which can be established only from experience. Such treatments may need to be performed by contractors specializing in this work, because of the specialized equipment which may be needed for treatments, as well as human exposure limitations to the moderately toxic gas. Disposal of the sanitant solution directly to the sewer without treatment may be possible, depending on the quantity and local codes. Advantages: Chlorine dioxide’s ability to penetrate and kill even thick biolm is its most useful attribute. Test kits to assess residues for rinsing efcacy verication are available.

Disadvantages: There is not a long history of previous experience for this sanitant in the pharmaceutical industry industry.. If biolm removal is needed to avoid post-sanitization accelerated biolm regrowth, other oxidative or caustic treatments should be used. Post-sanitization Post-sanitiz ation removal is by system ushing.

13.5.2.8 13.5.2. 8 Bromine and Iodine Bromine and iodine are halogens that have been used occasionally for water sanitization. The reactions of bromine and iodine in water are analogous to the behavior of chlorine in their formation of hypobromous or hypoiodous acids; however, these compounds are far less reactive than hypochlorous acid. Although there are applications for these compounds in sanitizing drinking water, particularly in emergency situations, their use in sanitizing pharmaceutical water systems is relatively rare because of their relative lack of efcacy compared to other sanitizers, as well as their high cost.

13.5.3 Other Chemical Sanitizers 13.5.3.1 13.5.3. 1 Extreme pH Regenerants Frequently regenerated ion exchange resin beds tend to be relatively free from high microbial counts, but infrequently regenerated beds experience exactly the opposite phenomenon. The extreme pH chemicals used for regeneration (typically HCl and NaOH) create hostile conditions on and in the resin beads. If cells that would form slimy biolms have only recently adsorbed to those resin surfaces and not started producing EPS slime, they are very susceptible to those hostile pH regeneration conditions and are likely killed. The key to such regenerants being effective sanitizers sanitizers is their frequency of use. Once a layer of slime has been produced, (e.g., on the resin surfaces




of infrequently regenerated deionization beds), short extreme pH exposures have little detrimental effect on biolm which, after regeneration, continues to grow further out of control, possibly becoming a serious source of water system contamination. Biolm problems may be worse on anion resins, even if these resins are regenerated with caustic, because anion resins tend to adsorb more organic matter from the water (i.e., more food), leading to faster and more luxuriant biolm growth than on cation resins. Thicker biolms lead to poorer chemical penetration, possibly affecting regeneration effectiveness. Extreme regenerant pH occurs on the active surfaces within the electrical cells of Continuous Electro-Deionization (CEDI) units. The difference is that those extreme pHs are continuously present on the surfaces of the charged resin beads and membranes, as a result of the electrolytic effects on the water molecules that keep the resins and membranes continuously regenerated.

Advantages: Very little microbial colonization or resulting biolm formation is likely to occur on deionizer surfaces, because of frequent or continuously hostile pHs. This advantage applies only when continuous or frequent regenerations occur and are unlikely for exchange ( off-site regenerated) resins.

Disadvantages: Frequent regenerations can be costly for conventional deionization units, create frequent non-use periods, and produce more waste chemicals. Where service deionization is used, the resin bottles may be replaced every day or two, which may seem to promote this microbial control phenomenon. The deionization resin bottles may have been regenerated many weeks after exhaustion and then stored wet and warm for yet many more weeks awaiting use, negating the entire microbial control effect from regeneration. It also could create a signicant microbiological problem as fully developed biolm in a “fresh” resin bottle sheds into the nished water. Continuously Continuously hostile extreme pH conditions do not exist within an EDI unit outside of the electrical eld in the discharge paths of the concentrated waste water or in the downstream product water discharge path, so biolm can ourish, if allowed to take hold, fouling membranes and other wet surfaces before the water exits the EDI unit. This is particularly signicant, as an EDI unit often is the last unit operation in a purication train; any immediately downstream biolm within an EDI unit may have a direct impact on the quality of water in a distribution system.

13.5.3.2 Caustic Caustic, usually in the form of sodium hydroxide, has been used successfully to remove biolm slime in water systems. It frequently is used at a concentration of 3% to 5%, often in a heated state. It is compatible with TFC polyamide RO membranes and is useful in restoring biofouled membranes, but will dissolve cellulosic RO membranes. It works by hydrolyzing the EPS material of biolm, killing the exposed biolm cells by its extreme alkalinity.. Its effectiveness can be substantially enhanced by combining it with pH-compatible oxidative sanitizers such alkalinity as hydrogen peroxide. It can be used only on materials compatible with high pH. Treatment times vary depending on concentration, temperature, the presence of “activity enhancing” ingredients, and the thickness and accessibility of the biolm to be killed and removed. Before disposing of the sanitant to the sewer, it should be pH neutralized.

Advantages: Caustic is a frequently handled material in pharmaceutical facilities; familiarity is an asset to its use. Caustic is relatively cheap and moderately effective at removing biolm slime. It is compatible with the majority of pharmaceutical water system materials. Caustic can be heated or combined with pH-compatible oxidizers to enhance is activity. Rinsing efcacy can be veried by a simple pH test or conductivity.




Disadvantages: The high pH of caustic is not compatible with some water system materials. Accommodat Accommodations ions should be made for the exothermic dissolution/dilution dissolution/dilution process of caustic which also may not be compatible with a number of system materials. Post-sanitization Post-sanitization removal is by system ushing, and the volume r equired may be extensive. Postsanitization neutralization neutralization is usually needed before discharge to a sewer. 13.5.3.3 Formaldehyde Formaldehyde is a simple 1-carbon aldehyde that acts as an alkylating agent (adds methyl groups to susceptible organic molecule locations that typically kill the cells). It volatizes easily as a gas. (It was frequently used as a gaseous environmental sterilant in aseptic suites and in water systems, but its identication as a carcinogen has reduced its usefulness to those situations where there are few equivalently effective alternatives, alternatives, and only with appropriate safety precautions.) One application of formaldehyde is in the sanitization of RO membranes. Although solutions of formaldehyde do little to remove existing biolm, they can penetrate as a gas and kill the cells within. Formaldehyde gas also can penetrate the RO membrane and kill biolm cells growing on the reject as well as permeate side of the membrane, which is difcult to reach by most RO sanitants. Formaldehyde is used rarely for sanitizing other locations in water systems owing to its toxicity and cost.

Advantages: Formaldehyde is useful for sanitizing both the reject and permeate sides of RO membranes, as it penetrates from the upstream side as a gas.

Disadvantages: Formaldehyde is carcinogenic to humans, costly in large amounts, and not particularly penetrative to thick biolms. Post-sanitization Post-sanitiz ation removal is by system ushing and depending on the quantity and concentration, may need to be treated before discharge to a sewer.

13.5.3.4 Glutaraldehyde Glutaraldehyde use increased with the concerns over formaldehyde’s carcinogenicity carcinogenicity,, and with its wider use as a hard surface sterilant. Glutaraldehyde is a 5-carbon dialdehyde with a different mechanism of action than formaldehyde, though it continues to be classied as a sterilant. It is a larger molecule and does not volatize as a gas under normal use conditions. Its effectiveness effectiveness in sanitizing the entire RO membrane is considerably less than formaldehyde, because it cannot penetrate the membrane as a gas to sanitize the permeate side of the membranes. Although Although glutaraldehyde’s toxicity toxicity and potential for carcinogenicity in humans is less than formaldehyde, its ability to penetrate and kill biolms is also less, so it is used only occasionally for RO membrane sterilization. Its removal is by rinsing, usually by unit operation ushing and rinsing efcacy may be demonstrated by TOC or conductivity analyses. 13.5.3.5 13.5.3. 5 Cationic Detergents Cationic detergents are used as the active ingredients in a number of disinfectants intended for hard surface decontamination, and their mode of action is to disrupt the cellular membranes of bacterial and fungal cells. These agents have little effect on the cells in biolm because of poor penetration through the EPS in which the cells are imbedded. Cationic detergents are effective on naked cells, which may be planktonic pioneer cells, as well as on recently surface-attached surface-attached cells prior to development of the EPS. If used for sanitization, cationic detergents should be ushed out of the system. Neutralizing prior to discharge to sewer usually is not required, depending on local codes. Rinsing efcacy may be demonstrated by TOC or conductivity analyses.



13.6


Assessing Microbial Control Success Processes that may affect product quality should be justied by verication or other indications of their effectivenes effectiveness; s; this applies to microbial control activities for high purity water systems. Assessments Assessments are a regulatory expectation and considered good business practice. How these assessments should be performed and how the resulting data should be used to enhance microbial control and assure the water ’s suitability for use are discussed. The intention is to clarify the purpose and limitations of these assessments.

13.6.1 Microbial Enumeratio Enumeration n Issues Biolm organisms have a strong preference to grow in water systems only when attached to a surface as part of a biolm. When biolm organisms are present in the owing water, they are either in the form of individual pioneer cells released by biolms to colonize other surfaces or as multi-celled ow-sheared biolm fragments or ocs. At a given point in time, a small minority minority,, probably far less than 1%, of the total number of microorganisms in a water system are planktonic, with the remainder being associated with the attached biolms; however, neither the release of pioneer cells nor the shearing of biolm ocs is continuous over time. Minor transient changes in water purity can affect the release rate of pioneer cells, and minor ow changes or even shock waves from sudden valve closures or vibrations from processing equipment can cause momentary releases of biolm ocs. The act of collecting a sample or using the water at a given valve can cause ow pattern changes in piping or valves that may cause biolm ocs to be released locally. To quantify the amount of biolm developing in a system, the surface growth of biolm should be assessed directly directly,, by examining typical surfaces from the water system. Sampling devices designed to perform this examination are available. Operational issues associated with retrieving samples from these devices, as well as concerns about biolm development in these devices being unrepresentative of biolm development throughout the system, however, have reduced their popularity in pharmaceutical applications. There continues to be a strong preference and tradition (dating back to pre-biolm awareness days) for collecting water samples as an indication of the microbial status of a water system, despite microbial variability problems in collecting planktonic samples. The inability to cultivate all viable organisms present in a tested sample is also problematic. Specialized nutritional requirements or other factors tend to prevent a single cultivative approach from detecting all the viable cells that may be present. In addition, traditional cultivative approaches can take between 2 and 7+ days to have countable colonies, depending on the conditions used and the ora present. This is an extremely long time to wait for test results that could identify an urgent microbial problem and a requirement to re-sanitize a water system “immediately.” “immediately.” It also may represent an extremely long delay before product made with this water may be released. There are a number of off-line rapid microbiological tests (and a few on-line versions as of publication) that can shorten this waiting time, but most of these rapid tests are destructive tests. The tests prevent subculturing and identication of the enumerated isolates, because they have been destroyed in the process of enumeration. Where organism identity is important, a cultivative approach or a non-destructive rapid test is usually necessary necessary.. Various rapid microbiological methods are available (at time of publication) and a number of them may have uses in water operations for shortening the time to receive microbial enumeration data.

13.6.2 Use of Microbial Enumeration Data for Quality Control Control Microbial enumeration data can have two distinct uses that can determine the testing approach used, as well as where and how the samples are collected. If the use of the data is for Quality Control (QC) purposes (i.e., to assess the suitability for use of the water exiting the distribution system at POUs), the water should be collected for testing in an equivalent manner to how it exits the system for use, so that the same microbial populations may be collected. This usually involves using the same hose, the same outlet sanitizing procedures, if any, any, and performing equivalent preliminary line ushing procedures, if any, to exactly duplicate the conditions that can affect the microbial content of the water during use.




If a POU outlet is hard-piped or otherwise permanently connected via automatic valves to equipment, it cannot be sampled in this way. See Section 13.6.6 of this Guide for more information regarding this scenario. In QC applications, in addition to the microbial count, it also may be important to know the identity of organisms typically found. There may be specic objectionable organisms known to be problematic with the uses of the water and may be required to be absent from the water or present only at specic allowable low levels. The use of the data for QC purposes often is associated with cultivative enumeration approaches or non-destructive rapid microbiological approaches that allow for identication of the isolates since identication may be considered important in assessing the suitability of the water for use. In QC applications, as with the pharmaceutical QC testing of any other raw material, the test results are compared against a pre-determined tness for use specication to determine if the water is acceptable. If a deviation from a specication is found, just like with any other raw material, an OOS-type of investigation should be triggered to assess the validity of the test result (i.e., whether it was caused by an assignable laboratory or sampling error) and to assess product impact, to investigate and correct the root cause of the deviation, and prevent its recurrence. There often is concern associated with the application of pass/fail specications using slow test procedures whose data are usually available some time after the water has been used. The maintenance practices that keep the microbial growth in the water system under control should be designed to maintain the microbial content of the system at levels well below the QC specication. Delayed QC test results should not be a concern where a water system is appropriately controlled and monitored relative to established suitable process controlling alert and action levels. Chapter 2 of this Guide also discusses the concepts of specications, action levels, levels, and alert levels. One of the primary reasons why the concept of water specications has not been widely used is the misuse of the process control terminology alert and action levels. These levels may have been inappropriately established (typically too high and unrelated to process capability). Action Action levels may have become a pseudo-specication rather than a process control value to trigger actions that prevent specication excursions. This misuse has encouraged sub-optimal microbial control, resulting in biom development that could grow well out of control without a timely corrective intervention, and yielding water potentially bordering on (or beyond) tness for use. Terminology Terminology misuse also relates to poor water POU practices, including sampling, which can create unpredictable and articially high local microbial levels unrelated to actual water system control. These data anomalies are most likely caused by illdened/poor pre-use outlet ushing or poor production hose storage. Improving water use (and identical sampling) practices and use of process capability-derived process control triggers should allow acceptable use of pass/fail QC specications for water.

13.6.3 Use of Microbial Enumeration Data for Process Control Control If the use of the data is for Process Control (PC) monitoring purposes only only,, the quality of the water within the pipes is the principal concern. The most important information for process control is the total microbial content of the distribution system (reected proportionally by a planktonic microbial count or some correlate to the microbial count). Microbial identication of the isolates is not particularly important, unless the appearance of specic species has been correlated with local or upstream microbial control insufciencies or unless excluded by specication. A water sample typically needs to be removed from the system for analysis, and efforts should be made to reduce or eliminate additional contamination inuences that may be caused by the sampling process, so that the water quality of the sample is as close as possible to that of the water within the pipes. Such sampling does not have to imitate a procedure for removing water from the system for production purposes. The water can be sampled using specially designed valves to minimize valve contamination and may be positioned in specic locations that are not manufacturing use points. Such sampling can be performed through sterile connector hoses and use exhaustive outlet ushing or stringent outlet sanitization procedures. In these data uses, it is the change in the relative numbers of recovered organisms, not their absolute numbers or their identity, identity, that gives an indication of relative control. The number of organisms enumerated in the planktonic phase represents a small portion of the total number of organisms present in biolms in a system. This approach




using the assumption that the greater the amount of biolm in the system, the higher will be the planktonic counts. A cultivative cultivative enumeration enumeration approach that allows viable recovery recovery of isolates, isolates, facilitating microbial identication, identication, usually is unnecessary. unnecessary. Rapid, non-cultivative approaches or destructive analyses that preclude the possibility of microbial identication could be useful. Limited characterization of the microorganisms may have some value in determining the source of a microbial problem if such a source could have a unique ora. For example, if the isolates are Gram positive cocci (typical of skin ora, not water system ora), there may be a sampling or testing issue needing to be addressed. Likewise, if the isolates are Gram positive, sporeforming rods and molds, there may either be a tank lter or rupture disk problem or environmental contamination during sampling since these are not water system ora either. Nevertheless, full microbial identication is not signicant as a routine activity activity.. Organizations may use abnormal changes in total counts from non-cultivative enumeration approaches as a signal to perform a cultivative approach to generate required investigative data, including microbial identications. In process control applications, in both the distribution system and the purication train, the enumeration test results should be compared against alert and ation levels that have been derived from analyses of the data trends over time. For further information, see Chapter 2 of this Guide. Trend analyses should describe a system’s normal microbial count variability variability,, which infers its microbial control capability.. When an action level excursion occurs, it signals the need for immediate investigative capability investigative and possible remedial action to bring the microbial content of the water system under improved control. Investigational activities should determine if high test results are a result of a sampling or laboratory error, (which are common causes of such excursions), or reect actual water system microbial levels. If the latter is true, further testing or a special local or system-wide sanitization processes may be triggered to determine the location and cause of the high counts and correct the problem at its source. By knowing the cause, appropriate preventive measures also can be instituted. Investigation Investigation of product impact should not be required if the action level is set considerably below the QC specication and the action level excursion did not exceed that specication. Action level responses also should be triggered if excessively frequent or prevalent alert level excursions have occurred. A recurring alert level excursion usually is a better indicator of a microbial control issue than is a single high action level excursion. Process control microbial monitoring also may be performed between critical purication steps, as an indication of adequate microbial control of the upstream unit operation and as a predictor of impact on microbially sensitive downstream unit operations. In both cases, abnormal microbial counts can trigger remedial actions that can correct a developing problem before nished water quality or unit operation functionality is compromised. Sampling valves should be designed to provide accurate and reliable monitoring results for a purication train and not be inuenced by ora resulting from an improper valve installation.

13.6.4 Mixed Use of Microbial Enumeration Data Organizations conceptually may use data derived from distribution system monitoring for both QC and process control purposes; the sample collection process may be mainly from POU valves where the collection of samples should mimic the use of water from those locations, as required for QC testing. These data would include any contamination associated with the sample collection process, potentially obscuring the actual microbial status of the water within the system. The technique associated with water outlet use is signicant in reducing contamination of otherwise acceptable water as it exits the system for routine use. Therefore, Therefore, when the data are used for both QC and PC, the procedures for POU water utilization must be well specied and consistently executed to avoid extraneous contamination of samples (or used water).

13.6.5 The Importance of Correct Sampling The technique for removing water from the system affects the quality of the water that is delivered; high microbial count water samples may result frequently from insufcient outlet ushing and poorly maintained hoses. The hoses used for manufacturing applications should be used for QC sampling, regardless of whether they are found to be connected at the time of sampling.




It is considered extremely important to maintain manufacturing hoses, gaskets, and other appurtenances appropriately to minimize biolm growth or environmental contamination on their surfaces during storage. Organizations may address these concerns by using only sterilized hoses or periodically sterilized hoses, and still others may rarely, if ever, ever, use sterile hoses. Each approach has benets and risks related directly to how the hoses (and their gaskets) are maintained during and after use. A hose that is stored wet and in an orientation that prevents draining and drying will eventually grow biolm that will contaminate the water that ows through it during production use or sampling. Hence, frequent hose replacement with a sterile hose is a good idea. It is also a good practice to have a hose end tting design and care in handling both the hose end and its gasket during connection to prevent inadvertent contamination of these surfaces. Where possible, hoses should not remain attached to POU valves for extended periods of time where they remain wet near the attachment point. This increases the probability of biolm growth within them signicantly. signicantly. When not in use, hoses should be disconnected from water system outlets. Where hoses are not sterilized between each use, they should be hung in an upside down “∩” orientation to promote rapid and complete drainage and drying of the hose lumen and minimize the potential for biolm growth. The hoses should be of length that permits drainage and drying in an acceptable time frame. It may be necessary to blow ltered dry compressed air or nitrogen though hoses to facilitate drying, particularly of long hoses. If there is a potential for environmental or product dust contamination to enter the hose ends (or the POU outlet valves) between uses, it is preferable to use an air-permeable “bonnet” or dust cover over the open ends, rather than using a sealing end cap that would prevent drainage and drying and trap biolm-growing surface moisture. Hoses should be taken out of service periodically to be cleaned thoroughly and sanitized or sterilized; inspected for internal cracks or other signs of wear that would justify their replacement to help to minimize the development of biolm within the hose. Preliminary ushing should be used when operating a POU outlet, particularly when the outlet also employs a ow metering device, heat exchanger, or a hose to transfer water to a manufacturing vessel or sample container. This ush is intended to slough off any loose biolm from inside the valve, POU device or hose, and to remove external contaminants that may have been introduced between uses or during the hose or device connection process. A given volume or time for ushing should be specied. Variability Variability in the execution of this ushing procedure may occur (possibly the largest contributor to microbial count variability) if an indication of the ushing ow rate is not also provided. The shear forces of the water during this ush should be sufcient to remove loose biolm and external contaminants; if the shear forces are minimal because of a slow ow, the ush may be ineffective. The number of valve turns, valve handle angle, or opening percentage should be dened for this preliminary ush. The ow rate should be extremely vigorous; the more vigorous the ow rate, the less the time or volume needed to achieve the hose and valve contaminant ushing. The number of variables that could affect its efcacy prevents the development of a universally acceptable set of ushing parameters. The correct preliminary ushing rate and duration/volume can be determined experimentally. experimentally. The ushing rate and duration or volume should be included in sampling and water system use procedures so that they can be reproduced consistently. consistently. Outlet sanitization is used often, typically with hydrogen peroxide, alcohol, or sodium hypochlorite. An outlet sanitization approach can be effective if the sanitant is applied thoroughly to all potentially contaminated surfaces inside a valve, hose, or ttings. It should be allowed an adequate contact time to work; 30 to 60 seconds for minimal efcacy and signicantly longer if the sanitization target is biolm. This level of contact time may be impractical; therefore, using chemical sanitizers on outlets or hose ends may have minimal benecial effect and may be detrimental to test results if sampling includes collecting TOC or conductivity test samples. An appropriately executed outlet ushing may be more effective than an outlet sanitization procedure. Identical procedures should be used for both QC test sampling from a POU valve and during routine water use, including the use of the same manufacturing hoses. This also applies to testing during water system qualication. qualication. If a manufacturing hose is suspected or determined to be a source of microbial contamination, manufacturing hose maintenance and sanitization should be addressed; hoses specically for sampling should not be used.




13.6.6 Sampling Valves in Quality Control Testing An automated POU valve may be permanently connected to a piece of equipment and cannot be detached for sampling purposes without compromising the process or potentially the safety of personnel. Design and maintenance of this equipment attachment conduit or associated equipment should avoid creating a potential microbiological problem within this unsampleable conduit. Microbiological problems may be prevented within this water conduit using a number of approaches including, but not limited to, the following: •

heat-tracing

•

steam or chemical in-place sanitization approaches

•

a readily drainable design

•

the use of a suitable dry compressed gas for drying the drainable connection

•

a pre-use ush programmed into the equipment automation

•

the replacement of this water conduit with a sterilized twin between uses.

When hard-piped POU valves cannot be sampled directly for QC testing purposes, sampling valves may be used. The sampling valves may be: •

incorporated into the body of the POU valve

•

located in the loop immediately up or downstream downstream of the POU valve

•

located within the equipment connection conduit.

Since sampling valves are not considered to be POU valves, there is no requirement to be used in an identical manner as POU valves. Depending on the sampling valve design and location, the timing of its use may need to be coordinated with POU activity, activity, in order to reect the quality of the water being delivered by the POU. Sampling valves should be designed and operated to avoid becoming a source of contamination of the water owing past them or of the water sampled through them, including any attached sample delivery hoses or tubing. In some sampling valve designs, the valve is intended to be oriented downward so that it is completely drainable, like POU valves. However, However, in other designs they are intended to be oriented horizontally to facilitate lling with sanitant and capping off after use to avoid biolm growth on the valve’s internal downstream surfaces between samplings. This sanitant should be purged from the valve prior to the next sample collection. For further information, see Chapter 8 of this Guide.

13.7

Functional Microbiolog Microbiological ical Pharmacopeial Compliance Each of the global pharmacopeial references is different in almost all aspects of how to manufacture, test, and control PW and water for injection(s). This presents problems when trying to comply with the different global pharmacopeial pharmacopeial references. Differences include: •

how the waters can be prepared

•

chemical quality specication specications s and tests required to assess those specications

•

microbiological attributes required or implied and how to assess those attributes




A degree of harmoni harmonization zation has occurred with some of the chemical attribute attributes s and tests among the USP USP,, EP EP,, and JP (References 4, 5, and 6, Appendix 1) (see Table Table 9.3). There remain many differences, particularly for microbiological attributes and testing. The EP puts their microbial test methods and “specications” directly in the monographs for these waters, which confers a certain level of mandate for their use, but both the USP and JP place different suggested or recommended microbial test method options and suggested action level options in their respective informational chapters, conferring a different level of mandate. So a user that markets products globally must comply with all the “disagreeing” pharmacopeias, including choosing the most appropriate microbiological tests and control criteria for their particular water system.

13.7.1 Microbial Enumeration Test Method The pharmacopeias agree on microbiological issues only to a limited degree, such as for the enumeration method to be used. The microbial enumeration test method used may have an enormous impact on the numbers and types of microorganisms recovered. There is no universally optimal test method because each water system is different and potentially selects for different combinations of ora and levels of ora; however, all pharmacopeias have “General Notices” sections that allow the use of alternative test methods if proven equivalent or superior to a referee or suggested procedure. An alternative test method may be used, e.g., one that takes less time to yield equivalent counts or one that yields higher counts, as long as it can be shown that the test method is as good (dened as resulting in no fewer counts or species) as the reference test method. The justication for an alternative test method generally is achieved by concurrent testing of the equivalent water samples, using one or more candidate test methods, including the reference method, over a suitable period of time. The potential uniqueness of the ora in a given water system and the timeliness of the data availability may justify the use of methods other than those listed in a particular compendium.

13.7.2 Establishing Appropriate Action Levels for Process Control Action levels are intended for process control purposes, not quality control purposes, and ideally are established from normal data trends for each system that take into account the innate variability of microbial enumeration data. In addition, there may be seasonal variations in the microbial control of a given system so that “normalcy” has cyclic variation over a given year. The intent of action levels is to prevent a water system from deviating out of control sufciently to generate water that is microbiologically unsuitable for its uses. These “suitability for use” values, also known as the “specications” or “limits” used for QC testing, should be considerably higher than the action levels for a system. If an action level is exceeded, microbial-controlling remedial actions can be taken to preclude microbial levels from exceeding a specication or being sufciently high to harm products or patients. The occurrence of multiple excursions of a lower microbial control level, sometimes called an alert level may provide a better indicator of microbial process control. Multiple lower level excursions also may trigger equivalent system evaluations and control responses as those triggered by an action level excursion. Sampling procedures should be effective and executed without fault, as poor or inconsistent sampling issues often cause unnecessary excursion responses.

13.7.3 Using Action Levels from from Pharmacopeia Pharmacopeia Levels of microbial control that are routinely achievable are considerably below the action levels stated (with varying degrees of mandate) in the USP USP,, EP, EP, and JP (References 4, 5, and 6, Appendix 1). “Maximum” or “reasonable” compendial action levels are: •

100 cfu/mL for PW

•

10 cfu/100mL for WFI




Waters exceeding these levels are not acceptable to regulators as being suitable for use with pharmaceutical products; therefore, these compendial action levels may be taken as functional QC specications specications.. Functional alert and action levels should be established at lower levels to allow microbial control at levels that should never exceed these functional QC specications. The most functional alert and action levels are those derived from an organization’s data trends while the system is operating optimally. They are set at levels that are triggered when the monitoring data deviates from the normal levels that are reective of suitable microbial control. Regardless of the terminology used for these various levels or the approach used to establish these levels, process controlling values should be established at levels that are useful in controlling the microbiological content of a water system so that it is able to consistently and reliably make water that is compliant with regulatory expectations and suitable for the intended use.

13.8

Microbial and Endotoxin Control in Pure Steam Systems Pure steam generation systems are designed to remove any endotoxin from the source water from which the steam is generated. Their distribution systems are designed to bleed all steam condensate from its piping to avoid creating cool wet areas where endotoxin could be generated from microbial growth. Pure steam distribution systems are too hot for microbial proliferation or survival, except possibly in areas of substantial condensate accumulation, which is indicative of poor design or system component malfunction malfunction (see Chapter 7 of this Guide). In-house specications for PW (for pure steam used in non-parenteral applications) or WFI (for pure steam used in parenteral applications) often are applied to pure steam that has been freshly condensed at a POU during sampling. These specications may include limits, specications, specications, or action levels for microbial content. For a well-designed and operating pure steam system, testing freshly condensed steam for microbial attributes is not considered appropriate, since the steam, as it is collected, will be immediately lethal to aquatic organisms that may have been present in coexiting condensate. Only where there may be a substantial ush of accumulated condensate from a given POU prior to the exit of live steam may there be some chance of any recoveries in microbial testing. In such situations, however, however, rather than waiting for microbial data to signal remedial action in the steam system, the discovery of accumulations of cooled condensate in the steam lines is cause for immediate remedial action, including possible quarantine of all or parts of the system. Such a discovery is an indication of a serious and intolerable aw in system operation that may lead to microbial proliferation and associated generation of bacterial endotoxins within the steam distribution system that, unlike the microorganisms, would not be destroyed by the live steam. For well-designed and operating pure steam systems, microbial microbial testing of freshly condensed steam is essentially assessing whether or not the sampler can collect an aseptic sample, not whether or not the steam system contains any microbial contaminants. Therefore, microbial attributes for pure steam may not be an important quality indicator for pure steam testing. Microbial attributes are absent from the USP pure steam monograph. At time of publication, there is a proposed change to the discussion of pure steam in Chapter <1231> of USP35, Supplement 1 (Reference 4, Appendix 1) which states that microbial analysis of pure steam condensate is unnecessary.



Appendices




Page 243 Appendix 1

14 Appendix 1 – References 1.

FDA Pharmaceutical cGMPs for the 21st Century – A Risk-Based Approach Final Report, (September 2004), www.fda.gov.

2.

ISPE Baseline® Pharmaceutical Engineering Guide Series, Volume 5 – Commissioning and Qualication, Qualication , First Edition, March 2001, International Society for Pharmaceutical Engineering (ISPE), www.ispe.org. www.ispe.org.

3.

Guide to Inspections of High Purity Water Systems July July,, 1993, The Division of Field Investigations Investigations,, Ofce of Regional Operations, Ofce of Regulatory Affairs, U.S. Food and Drug Administration, www.fda.gov. www.fda.gov.

4.

United States Pharmacopeia–National Formulary (USP-NF), www.usp www.usp.org/USPNF .org/USPNF..

5.

European Pharmacopoeia, www.edqm www.edqm.eu/. .eu/.

6.

Japanese Pharmacopoeia (English Version), jpdb.nihs.go.jp/j jpdb.nihs.go.jp/jp15e/. p15e/.

7.

ICH Pharmaceutical Development – Q8 , International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use), www.ich.org.

8.

ASTM Standard E2500, 2007, “Standard Guide for Specication, Design, and Veri Verication cation of Pharmaceutical and Biopharmaceutical Manufacturing Systems and Equipment,” ASTM International, West Conshohocken, PA PA,, www. astm.org.

9.

FDA US Code of Federal Regulations, Title 40 – Protection of Environment Chapter I – Environmental Protection Agency Subchapter Subchapter D – Water Programs Part 141 – National Primary Drinking Water Water Regulations, www.fda.gov www.fda.gov..

10. Guidelines for Drinking-Water Quality Quality,, Fourth Edition, World Health Organization (WHO) 2011, ISBN: 978-92-4154815-1, www.who.int www.who.int.. 11. EMEA Note for Guidance on Quality of Water for Pharmaceutical Use (2002), www.em www.ema.europa.eu. a.europa.eu. 12. BPE-2009 Bioprocessing Equipment, American Society of Mechanical Engineers (ASME), www.as www.asme.org. me.org. 13. Good Manufacturing Practices (GMP) Guidelines – 2009 Edition, Version 2 (GUI-0001), Health Canada, www. hc-sc.gc.ca.

14. ISPE Good Practice Guide: Commissioning and Qualication of Pharmaceutical Water and Steam Systems Systems,, International Society for Pharmaceutical Engineering (ISPE), First Edition, February 2007, www.ispe.org. 15. European Standard EN 285:2006+A1:2008, Sterilization. Steam Sterilizers. Large Sterilizers. 16. Health Technical Technical Memorandum HTM 2010, Sterilization, http://www.spaceforhealth.nhs http://www.spaceforhealth.nhs.uk/. .uk/. 17. FDA US Code of Federal Regulations, Title 21 – Food and Drugs, Chapter I – 21 CFR Part 173 Secondary Direct Food Additives Permitted In Food For Human Consumption, www.fda.gov. www.fda.gov. 18. Public Health Service/Dairy Industry Committee, 3A Sanitary Standards, Number 609-02.

19. ISPE Good Practice Guide: Heating, Ventilati Ventilation, on, and Air Conditioning , International Society for Pharmaceutical Engineering (ISPE), First Edition, September 2009, www.isp www.ispe.org. e.org.


ISPE Baseline® Guide: Water and Steam Systems Systems

Page 244 Appendix 1

20. ASTM Standard A270, “Specication for Seamless and Welded Austenitic Stainless Steel Sanitary Tubing,” ASTM International, International, West Conshohocken, Conshohocken, PA, PA, www.astm.org. www.astm.org. 21. US Code of Federal Regulations, Title 21, Food and Drugs: 21 CFR Part 211 – 4122 Current Good Manufacturing Practice for Finished Pharmaceuticals, www.fda.gov. www.fda.gov. 22. International Organization for Standardization (ISO), www.is www.iso.org. o.org. 23. European Standard EN ISO 15607:2003 Specication and Qualication of Welding Procedures for Metallic Materials. General Rules. 24. European Standard EN 287-1:201 287-1:2011 1 Qualication Test of Welders. Fusion Welding. Steels. 25. European Standard EN 1418 Welding Personnel. Approval Testing of Welding Operators for Fusion Welding and Resistance Weld Setters for Fully Mechanized and Automatic Welding of Metallic Materials. 26. US Occupational Safety and Health Administration (OSHA), http://www http://www.osha.gov/. .osha.gov/. 27. US Code of Federal Regulations, Title 21, Food and Drugs: 21 CFR Part 211.65, www.fda.gov www.fda.gov.. 28. US Code of Federal Regulations, Title 21, Food and Drugs: 21 CFR Part 211. 211.67, 67, www.fda. www.fda.gov gov.. 29. Guideline on the Specication Limits for Residues of Metal Catalysts or Metal Reagents Doc. Ref. EMEA/CHMP/ SWP/4446/2000 (February 2008), www.ema.europa.eu. 30. Wegrelius, Lena and Birgitta Sjödén, “Passivation “Passivation Treatment of Stainless Stainless Steel,” ACOM ACOM (A Corrosion Management and Applications Engineering magazine from Outokumpu Stainless), Issue 4-2004, www. outokumpu.com. 31. ASTM Standard A380, “Practice for Cleaning, Descaling, and Passivation of Stainless Steel Parts, Equipment, and Systems,” ASTM International, West Conshohocken, PA, www.astm.org. 32. ASTM Standard A967, “Specication for Chemical Passivation Treatments for Stainless Steel Parts,” ASTM International, West Conshohocken, PA, www.astm.org. 33. Henkel, MSE, PhD, Georg and Benedikt Henkel, MSE, “WFI Systems with Supplemental Monitoring System for Quality Control in Connection with Rouging,” Essay No. 71, 2007, Henkel Pickling and Electropolishing, www. henkel-epol.com. 34. ASTM D4189-07, “Standard Test Method for Silt Density Index (SDI) of Water Water,” ,” ASTM International, West Conshohocken, PA, www.astm.org. 35. “PA “PAT T – A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance,” Guidance for Industry Industry,, September 2004, US Food and Drug Administration (FDA), www.fda www.fda.gov .gov.. 36. US Code of Federal Regulations, Title 21, Food and Drugs: 21 CFR Part 11 – Electronic Records, Electronic Signatures, www.fda.gov.

37. ISPE GAMP ® 5: A Risk-Based Approach to Compliant GxP Computerized Systems, Systems, International Society for Pharmaceutical Engineering (ISPE), Fifth Edition, February 2008, www.is www.ispe.org. pe.org.

38. ISPE GAMP ® Good Practice Guide: A Risk-Based Approach to GxP Process Control Systems , International Society for Pharmaceutical Engineering (ISPE), Second Edition, February 2011, www.ispe.org. www.ispe.org.



Page 245 Appendix 1

39. Instrument Society of America (ISA), www.is www.isa.org. a.org. 40. EU GMP Volume 4 “EU Guidelines to Good Manufacturing Practice,” Annex 1: Manufacture of Sterile Medicinal Products, www.ec www.ec.europa.eu. .europa.eu. 41. ISPE Good Practice Guide: Ozone Sanitization of Pharmaceutical Water Systems, Systems , International Society for Pharmaceutical Engineering (ISPE), under development at time of publication, www.ispe.org.




Page 247 Appendix 2

15 Appendix 2 – Glossary 15.1

Acronyms and Abbreviation Abbreviations s API

Active Pharmaceutical Ingredient (also known as Bulk Pharmaceutical Chemicals)

ASME

American Society of Mechanical Engineers

ASTM

American Society for Testing and Materials

BS

British Standard

BPC

Bulk Pharmaceutical Chemicals

CEDI

Continuous Electrodeionizat Electrodeionization ion

CFS

Chemical Free Steam

CFU

Colony Forming Units, i.e., viable bacteria

CGMP

Current Good Manufacturing Practice

CIP

Clean-In-Place (system)

CLSI

Clinical and Laboratory Standards Institute

CPVC

Chlorinated Polyvinyl Chloride

CS

Clean Steam

DCS

Distributed Control System

DNA

Deoxyribose Nucleic Acid

DI

Deionized, Deionizing, Deionization

EDR

Electrodialysis Electrodialysi s Reversal (Osmonics)

EDI

Electrodeionization Electrodeionizati on (Osmonics and Generic)

EPA EP A

Environmental Protection Agency (US)

EP EP

Electropolishing

EP

European Pharmacopoeia

EPDM

Ethylene Propylene Diemer

FRP

Fiberglass Reinforced Plastic

GRAS

Generally Recognized as Safe (FDA)



Page 248 Appendix 2

ISO

International Organization for Standardization

JIS

Japanese Industrial Standards

JP

Japanese Pharmacopoeia

LAL

Limulus Amebocyte Lysate

LVP

Large Volum Volume e Parenteral

MF

Microltration or Micro-Filter

ME

Multiple-Effect Multiple-Eff ect (still)

MF MF

Microltration

MM

Multimedia Filter

MP

Mechanical Polishing

MSDS

Material Safety Data Sheet

NIST

National Institute of Standards and Technology

NTU

Nephelometric Turbidity Unit

P&ID

Process and Instrument Diagram

PE PE

Polyethylene

PLC

Programmable Logic Controller

POU

Points of Use

PP PP

Polypropylene

PTFE PTFE

Polytetrauoroethylene

PVDF PVDF

Polyvinylidene

RO

Reverse Osmosis

SDI

Silt Density Index

SE

Single Effect

SOP

Standard Operating Procedure

SS

Stainless Steel

TDS

Total Dissolved Solid

THM THM

Trihalomethanes



Page 249 Appendix 2

TIG

Tungsten Inert Gas

TOC

Total Organic Carbon

UF

Ultraltration or Ultra-lter

USP

United States Pharmacopeia

UV

Ultraviolet (light)

VC

Vapor Compression (still)

WFI

Water for Injection

15.2

Defnitions

Absorption Assimilation of molecules or other substances into the physical structure of a liquid or solid without chemical reaction. Aerobic Bacteria Bacteria capable of growing in the presence of oxygen. Anaerobic Bacteria Bacteria capable of growing in the absence of oxygen. ASME Bioprocessing Equipment (BPE) An American National Standard that provides the requirements applicable to the design of equipment used in the bioprocessing, pharmaceutical, pharmaceutical, and personal care product industries, including aspects related to sterility and cleanability,, materials, dimensions and tolerances, surface nish, material joining, and seals. cleanability

Backwash The process of owing water in the opposite direction from normal service ow through a lter bed or ion exchange bed. The purpose of backwashing a sand lter is to clean it by washing away all the material it has collected during its service cycle. The purpose of backwashing a carbon lter is also to clean it, but primarily to eliminate ow channels that might have formed and to expose new absorption sites.

Bacteria Single-celled microorganisms measured in high purity water by several means: culturing, high power microscope, or Scanning Electron Microscope (SEM). The value is reported as Colony Forming Units (CFU), or colonies per milliliter or per liter. The bacteria in the water act as particle contamination on the surface of the product, or as a source of detrimental by-products. See Pyrogen.

Blowdown The withdrawal of water from an evaporating water system to maintain a solids balance within specied limits of concentration of those solids.



Page 250 Appendix 2

BOD Biological oxygen demand of water. This is the oxygen required by bacteria for oxidation of the soluble organic matter under controlled test conditions.

Breakthrough Passage of a substance through a bed, lter, or process designed to eliminate it. For ion exchange processes, the rst signs are leakage of ions (in mixed beds, usually silica) and the resultant increase in conductivity. conductivity. For organic removal beds, usually small, volatile compounds (THMs are common in activated carbon). Cation Exchange Resin An ion exchange resin which removes removes positively charged ions. Colony Forming Unit (CFU) A measure of the number of bacteria present in the environment or on the surfaces of an aseptic processing room; measured as part of qualication and ongoing monitoring. Also applied to the testing of puried water samples.

Compendial Ofcial; purpor purported ted to comply with USP USP,, EP EP,, or JP JP..

Conductivity A measure of ow of electrical current through water. This conductance is high with high Total Dissolved Solids (TDS) water and very low with ultrapure deionized water. Conductivity is the reciprocal of resistivity (C=1/R) and is measured in micromho/cm (µmho/cm) or microsiemens/cm (µS/cm).

Contaminant Any foreign component present in another substance. For example, anything in water that is not H 2O is a contaminant. Critical Instrument These are the instruments used to measure critical parameters. Dissolved Solids The amount of non-volatile matter dissolved in a water sample, usually expressed in parts per million (PPM) by weight. Drinking Water EPA EP A primary drinking water or comparable regulations of the European Union or Japan.

Electropolishing Controlled electro-chemical process utilizing acid electrolyte, DC current, anode and cathode to smooth the surface by removal of metal.



Page 251 Appendix 2

Endotoxins Pyrogens from certain Gram negative bacteria. Generally highly toxic Lipopolysaccharide Lipopolysaccharide-protein -protein complexes (fat, linked sugars, and protein) from cell walls. A marker for these bacteria with a reputation for persistent contamination because they tend to adhere to surfaces. See Pyrogen.

Extractable Undesirable foreign substances that are leached or dissolved by water or process streams from the materials of construction used in lters, storage vessels, distribution piping, and other product contact surfaces.

Ferrite A solid solution of one or more elements in body-centered cubic iron. Unless otherwise designated (for instance, as chromium ferrite), the solute is generally assumed to be carbon. On some equilibrium diagrams there are two ferrite regions separated by an austenite area. The lower area is alpha ferrite; the upper, delta ferrite. If there is no designation, alpha ferrite is assumed. Good Engineering Practices (GEP) Standards, specications specications,, codes, regulatory and industrial guidelines and accepted engineering and design methods to design, erect, operate, and maintain a pharmaceutical facilities taking into account not only regulatory compliance, but also safety, economics, environment protection, and operability. operability. Standards and specications are provided by recognized sources such as established engineering contractors and pharmaceutical companies. Codes are provided by local, state or federal jurisdictions, or insurance companies. Guidelines are issued by professional societies, industrial organizations, organizations, or regulatory agencies. Engineering design methods have been established in the engineering educational system. Grains Per Gallon A unit of concentration. 1 grain/gal = 17.1 mg/l. Gram Negative Bacteria A basic classicat classication ion of bacterial type, along with “Gram positive.” These organisms resist straining by the Gram technique. Sometimes considered “bad” bacteria when discussing pollution or contamination; however, this is an articial and quite broad classicati classication. on. Gram Positive Of bacteria, holding the color of the primary stain when treated with Gram’s stain.

Halogens Atoms of the chlorine family which also include uorine, bromine, and iodine.

Hardness The concentration of calcium and magnesium salts in water. Hardness is a term originally referring to the soapconsuming power of water; as such it is sometimes also taken to include iron and manganese. “Permanent hardness” is the excess of hardness over alkalinity. “Temporary “Temporary hardness” is hardness equal to or less than the alkalinity alkalinity.. These also are referred to as “non-carbonated” or “carbonate” hardness, respectively.



Page 252 Appendix 2

Heavy Metals High molecular weight metal ions, such as lead. Known for their interference with many processes, and “poisoning” of catalysts, membranes, and resins. Highly Purifed Water (EP) (EP) Water intended for use in the preparation of products where water of high biological quality is needed, except where Water for Injection is required. Highly Puried water is obtained from water that complies with the regulations on water intended for human consumption laid down by the competent authority. Current production methods include, for example, double-pass reverse osmosis coupled with other suitable techniques such as ultraltration and deionization. Highly Puried water meets the same quality standards as WFI but the production methods are considered less reliable than distillation and thus it is considered unacceptable for use as WFI. High Purity Water Water conforming to USP Monographs or equivalent. Humic Acid The classical method for fractionating the humic colloids that disperse in the sodium hydroxide extract is to acidify the suspension with sulfuric or hydrochloric acid, which causes a part of the dispersed organic matter to precipitate. The part that stays in solution is known as fulvic acid, that which precipitates out as humic acid, and that part of the organic matter which does not disperse in the alkali but remains in the soil as humin.

Hydrocarbons Organic compounds containing only carbon and hydrogen. Sometimes broadened to include compounds or mixtures of compounds with small amounts of oxygen also.

Hydrophilic Having an afnity for water. Its opposite, non-water-wettable, is hydrophobic.

Hydrophobic The extent of insolubility; not readily absorbing water; resisting or repelling water, wetting, or hydration; or being adversely affected by water. water. Hydrophobic bonding is an attraction between the hydrophobic or non-polar portions of molecules, causing them to aggregate and exclude water from between them.

Impurity Any component present in the intermediate or API that is not the desired entity entity.. It may be either process or product related.

Inorganics Chemical compounds which are not organic in nature. Inorganics that are soluble in water generally split into negative and positive ions, allowing their removal by deionization. Ion An atom or radical in solution carrying an integral electric charge, either positive (cation) or negative (anion).



Page 253 Appendix 2

Ion Exchange (IX) One of the processes used to further reduce the concentration of ions in water supplies referred to as total dissolved solids removal. The process uses anion and cation exchange resin to chemically react with and remove the remaining ions or TDS in the water. This process results results in water with virtually no TDS. Ion Exchange Regeneration The process by which ion exchange resin that can no longer effectively remove ions from the water is recharged. This recharging or regeneration is performed by adding an excess of caustic (NaOH) to the anion resin and an excess of either hydrochloric acid (HCl) or sulfuric acid (H2SO4) to the anion resin. These regenerant solutions are allowed to ow through the resin beds at specic ow rates for specic periods of time depending on the type of resin, the ionic load, and the nal purity desired. The regenerant solutions react with the ion exchange resin releasing the removed cations and anions which are then carried away to drain by the ow of the regenerant chemicals. The excess chemical is rinsed from the ion exchange resin with puried water when the bed is ready for another service cycle. Ion Exchange Resin A styrene-divinylbenz styrene-divinylbenzene ene or acrylic copolymer copolymer formed into into small, spherical, spherical, and highly porous beads about the size size of a pinhead. These inert beads are chemically treated so that they perform as if they were chemical compounds. Langelier Index A measure of the degree of saturation of calcium carbonate in water that is based on pH, alkalinity alkalinity,, and hardness. If the Langelier Index is negative, the water is corrosive (pH value below 7 or acidic). If the Langelier Index is positive, calcium carbonate can precipitate out of solution to form scale (pH value above 7 or basic). The Langelier Index will vary for cold water and for warm water. Material Safety Data Sheet (MSDS) Document produced by the manufacturer that contains the chemical and physical properties of a substance that are pertinent to safe handling and storage.

Megohm-cm A measure of ionic purity in water.

Membrane A barrier, usually thin, that permits the passage only of particles up to a certain size or of special nature.

Micron The same as a micrometer or 1000th of a millimeter millimeter.. The typical particle size of importance in deionized water is less than 0.2 µm.

Microorganism Organisms (microbes) observable only through a microscope. Larger, visible types are called organisms.



Page 254 Appendix 2

Milligrams Per Liter (mg/l) A term used to report chemical analyses. Milligrams per liter refer to the milligrams of the compound or element present in 1 liter (1000 milliliters) of water. Another Another term often used is parts per million (ppm) which is the same for substances in water. 1 mg/l = 1000 µg/l = 1 ppm. Mixed Bed Ion Exchange The use of both cation and anion exchange resin mixed together in one tank. Noncarbonate Hardness Hardness in water caused by chlorides, sulfates, and nitrates of calcium and magnesium. NPDES Permit (US) The National Pollution Discharge Elimination System permit required by and issued by EP EPA. A. NPDWR Water (US) (US) Potable water meeting EP EPA A National Primary Drinking Water Regulations. Operating Parameter Any information entered into an automated system used for automated equipment operation.

Organics Short for organic chemicals; those compounds that contain carbon to hydrogen bonds and are not carbonate related.

Orifce An opening through which a uid can pass; a restriction placed in a pipe to provide a means of measuring ow.

Osmosis The passage of water through a permeable membrane separating two solutions of different concentrations; the water passes into the more concentrated solution.

Oxidizer A chemical which readily oxidizes more reduced substances. Examples of strong oxidizers are ozone, hydrogen peroxide, chloride, persulfates, and oxygen itself.

Ozone Ozone is a very strong gaseous oxidizing agent. It is used in deionized water systems to kill bacteria and to reduce, by oxidation, the amount of TOC in the water. Ozone is O 3 and due to reaction with other things rapidly becomes oxygen (O2). Therefore, it has a short but effective oxidizing potential. It can be destructive to ion exchange using membrane lters and other plastic materials in the system.



Page 255 Appendix 2

Particles A physically measurable contaminant in deionized water. Particles can be bacteria, colloidal material or any other insoluble material. Particle counts are reported as number of particles per liter of a particular size measured in micrometers (microns).

Passivation Removal of exogenous iron or iron compounds from the surface of a stainless steel by means of a chemical dissolution, most typically by a treatment with an acid solution that will remove the surface contamination but will not signicantly affect the stainless steel itself. Unless otherwise specied, it is this denition of passivation that is taken as the meaning of a specied requirement for passivation. Passive Layer A passive oxidized lm that forms naturally on a stainless steel surface when exposed to air or similar oxidizing environment thus protecting the underlying base metal from corrosion.

Pathogens Disease-producing microbes. Percent Rejection In reverse osmosis or ultraltration, the ratio of impurities removed to total impurities in the incoming feed water. For example, RO membranes typically remove (reject) 90% of the dissolved inorganic contaminants in water.

Permeability The ability of a body to pass a uid under pressure. pH pH, the negative log of the hydrogen ion concentration, is a measure of the concentration of hydrogen ions (H+) in a water-based solution. The more hydrogen ions that are present, the lower the pH and the more acidic the solution. Photo Oxidation The mechanism by which ultraviolet light reduces Total Organic Carbon (TOC) to Carbon Dioxide. If halogenated organics are present, both CO2 and mineral acids can be formed. Polished Water High purity water after it has undergone a second treatment step. Ultrapure water usually undergoes two or more treatment steps. More economical pretreatment processes processes (e.g., reverse osmosis) are used to remove all but a very small fraction of the impurities. Highly efcient polishing processes processes (e.g., mixed-bed deionization) are used to remove the impurities that remain.



Page 256 Appendix 2

Polypropylene (PP) A crystalline polymer with high heat resistance (for piping an upper limit of 212°F (100°C), stiffness, and chemical resistance with respect to handling caustics, solvents, acids, and other organic chemicals. It is not recommended for use with oxidizing type acids, detergents, low boiling hydrocarbons, alcohols, and some chlorinated organic materials. Polypropylene is a relatively inert material and contributes little in the way of contamination to pharmaceutical water. Polyvinyl Chloride (PVC) The largest volume of the vinyl family of plastics. Overall it has excellent basic properties, may be easily processed and welded, and is exceptionally economical in cost. Because PVC is a thermally sensitive thermoplastic compounding ingredients such as heat stabilizers, lubricants, llers, plasticizers, plasticizers, impact modiers, pigments, and processing aids must be added to make it processible. PVC is prone to produce extractables during start-up in high purity water. Polyvinylidene Polyvinyliden e Fluoride (PVDF) A thermoplastic uoropolymer that has a very linear chemical structure, and is similar to PTFE with the exception of not being fully uorinated, i.e. having 3% hydrogen by weight. Its drawbacks in the area of chemical resistance include unsuitability with strong alkalis, fuming acids, polar solvents, amines, ketones, and esters. It has a high tensile strength as well as a high heat deection temperature. It is readily weldable, offers high purity qualities, and is resistant to permeation of gases. PVDF is a relatively inert material and contributes little in the way of contamination to pharmaceutical water. Potable Water Water that is suitable for drinking. Potable Water (EMEA) (EMEA) It is not covered by a pharmacopeial monograph but must comply with the regulations on water laid down by the competent authority. authority. Testing Testing should be carried out at the manufacturing site to conrm the quality of the water. Potable water may be used in chemical synthesis and in the early stages of cleaning pharmaceutical manufacturing equipment unless there are specic technical or quality requirements for higher grades of water. It is the prescribed source feed water for the production of pharmacopeial grade waters.

Precipitate An insoluble reaction product; in an aqueous chemical reaction, usually a crystalline compound that grows in size to become settleable. Product Contact Surface A surface surface that contacts contacts raw materials, materials, process materials, materials, and/or product. product. Pure Steam Steam (USP) (USP) Water that has been heated above 100°C (212°F) and vaporized in a manner that prevents source water entrainment. It is prepared from water complying with the U.S. EPA Primary Drinking Water Regulations, or with drinking water regulations of the European Union or Japan, or with WHO drinking water guidelines. It contains no added substance. The level of steam saturation or dryness, and the amount of noncondensable gases are to be determined by the Pure Steam application. Note: Note: Pure Pure Steam is intended for use where steam or its condensate comes in contact with the article of the preparation.



Page 257 Appendix 2

Purifed Water (USP) (USP) Water rendered suitable for pharmaceutical purposes by using unit operations that include deionization, distillation, ion-exchange, reverse osmosis, ltration, or other suitable purication procedures. It meets rigid specications for chemical purity, purity, the requirements of the Federal Environmental Protection Agency Agency (EPA) with respect to drinking water, and it contains no added substances. Cannot be used as raw material for parenterals. Common uses are: a rinse for equipment, vials, and ampoules, and as make up for cosmetics, bulk chemicals, and oral products. For FDA acceptance, puried water must contain less than 0.5 mg/l of TOC (Total (Total Organic Carbon), and less than 100 CFU (Colony Forming Units). Purifed Water (EMEA) (EMEA) Water for the preparation of medicinal products other than those that require the use of water which is sterile and/ or apyrogenic. Puried water which satises the test for endotoxins may be used in the manufacture of dialysis solutions. Puried Water is prepared by distillation, by ion exchange or by any other suitable method, that complies with the regulations on water intended for human consumption laid down by the competent authority.

Pyrogen Trace organics which are used as markers of bacterial growth or contamination. Produced by various bacteria and fungi. Critical pharmaceutical and biotechnological processes have restrictions on contamination by these substances, usually at levels near the limit of detection. Primarily polysaccharide (made of linked sugars) in nature. Fever producing substances when administered parenterally to man and certain animals.

Resistivity The measure of the resistance to the ow of electrical current through high purity water. This is measured in millions of ohms-cm or Megohm-cm (Mohm-cm). Resistivity is the reciprocal of Conductivity (R=1/C, 1 Mohm-cm = 1 µS/cm). This provides an easy means of continuously measuring the purity of very low TDS water or ionic concentration. Reverse Osmosis A process that reverses (by the application of pressure) the ow of water in the natural process of osmosis so that it passes from the more concentrated to the more dilute solution. This is one of the processes used to reduce the ionic TDS, TOC, and suspended materials of feed water through a semipermeable membrane leaving dissolved and suspended materials behind. These are swept away in a waste stream to drain.

Rouge Rouge in stainless steel systems utilized in the biopharmaceutical/life science industry, industry, is a general term used to describe a variety of discolorations on the product contact surfaces, caused by variations in hydration agents and the formation of metallic (primarily iron) oxides and/or hydroxides from either external sources, or from alteration of the chromium rich “passive” layer.

Salinity The presence of soluble minerals in water.

Salt Neutral compound formed of two or more ions. The salt disassociates into cations and anions when dissolved in water.



Page 258 Appendix 2

Sanitary Design A system of design that meets standard, specication, codes, regulatory and industrial guidelines, and acceptable engineering design methods to reach a degree of sanitation required by food, pharmaceutical, and cosmetics processing. Saturation Index The relating of calcium carbonate to the pH, alkalinity alkalinity,, and hardness of a water to determine its scale-forming tendency.

Scale The precipitate that forms on surfaces in contact with water as the result of a physical of chemical change.

Sedimentation Gravitational settling of solid particles in a liquid system.

Softening The removal of hardness (calcium and magnesium) from water. This is a PRETREA PRETREATMENT TMENT process which used cation exchange resin to remove the hardness elements from the water. The hardness elements are calcium and magnesium. The cation resin is regenerated with sodium chloride and during the exchange process, the calcium and magnesium are removed from the water and replaces with sodium ions (Na+). The resulting sodium salts are much more soluble than the salts of calcium and magnesium and do not precipitate which provides better feed water to the RO system. Stability Index An empirical modication of the saturation index used to predict scaling or corrosive tendencies in water systems. Stainless Steel Steel to which a signicant amount of chromium and nickel has been added to inhibit corrosion.

Sterilization Refers to the killing of microorganisms in the distribution system. This is normally done periodically by ushing a sterilizing solution, such as hydrogen peroxide or ozone, through the distribution piping system. In some systems, ozone is continuously injected at low levels for continuous sterilization. Surface Finishes This term shall apply to all interior surface nishes accessible and inaccessible, that directly or indirectly come in contact with the designated product in bioprocessing equipment and distribution system components. components. Final criteria shall be determined by Ra Ra values values rather than polishing methods. Note: Note: For For commonly utilized Ra Ra readings readings on stainless steel product contact surfaces for the biopharmaceutical industry refer to Table Table SF-3 of the ASME Bioprocessing Equipment (BPE – 2007) an International Standard. Surface Water Surface water is any water where the sources is above ground. This can be rivers, lakes, or reservoirs. Surface waters are usually higher in suspended matter and organic material and lower in dissolved minerals than well water.



Page 259 Appendix 2

Thermal Fusion The joining of two materials (usually metal or plastic) by use of heat only, without any additional material. Usually done by the use of automatic TIG welding in alloy steel tubing welding or with specially designed melting equipment for plastics. Total Dissolve Dissolved d Solids (TDS) The term used to describe inorganic ions in the water. Usually measured by measuring the electrical conductance of the water corrected to 25°C. Total Organic Carbon (TOC) Measure of organics in water by their Carbon content. This is one of the parameters used to determine the purity of Semiconductor Grade water. Feed water will have TOC measured in parts per million. UPW will have TOC measured in parts per billion (ppb). Trihalomethanes (THM) Compounds present in the feed water that are formed by the reaction of chlorine and the organic material in the water. The most common THM found in water is chloroform which is quite difcult to remove. Activated Activated carbon and degasication can serve to reduce THMs.

Turbidity A suspension suspension of ne ne particles that obscures light rays, rays, but requires many many days for sedimentation sedimentation because because of the small small particle size.

Ultrafltration Filter technology similar to reverse osmosis that is capable of ltering colloids and large molecular weight organics out of the water. The lter capability of ultraltration lters to 0.005 µm particle size. Ultraltration also will remove organic material down to about 1,000 to 10, 000 molecular weight. Ultraviolet (UV) Sterilizer Ultraviolet lamps used to kill microorganisms in water. These can be placed anywhere in the water system. The wavelength used for control is 254 nanometers (nm). Ultraviolet TOC Reduction A UV source which partially oxidized organic compounds to ionic species which can be removed. Relies on 185 nm radiation from “ozone producing” mercury lamps (along with 254 nm germicidal radiation). Generally has a longer contact time than for sterilization alone. USP Purifed Water (see Puried Water (USP))



Page 260 Appendix 2

Vacuum Degasifca Degasifcation tion The process of removing dissolved and entrained gases from the reverse osmosis product water by creating a vacuum in a tower through which the RO product water ows. The degasier may be located before the reverse osmosis system, system, but the majority of the time it will be located after. The most prevalent gas present is carbon dioxide which may be have been generated during pH adjustment of the reverse osmosis feed water. Carbon dioxide can be removed by the anion exchange resin, but that load can be reduced by using the vacuum degasier. The other gas of concern is the water is oxygen which also is removed by a vacuum degasier. degasier. Water for Injection (WFI) (WFI ) (USP) Prepared from water complying with the quality attributes of “Drinking Water.” Water.” Puried by distillation or a purication process that is equivalent or superior to distillation in the removal of chemicals and microorganisms. Conductivity in accordance with Stage 1, 2, and 3 tests and Conductivity Tables. Total Organic Carbon limit is at 0.5mg/l. Typically less than 10 CFU/100ml for microbiological acceptability. acceptability. Less than 0.25 USP EU/ml. Water for Injection (WFI) (EP) (WFI) (EP) WFI in bulk is obtained from water that complies with the regulation on water intended for human consumption laid down by the competent authority, or from puried water, by distillation in an apparatus of which the parts in contact with the water are of neutral glass, quartz or suitable metal and which is tted with an effective device to prevent the entrainment of droplets. The correct maintenance of the apparatus is essential. During production and storage, appropriate measures are taken to ensure that the total viable aerobic count is adequately controlled and monitored. WFI complies with the tests for Puried water with additional requirements for bacterial endotoxins (not more than (nmt) 0.25 IU endotoxins per ml), conductivity and Total Total Organic Carbon.


The Ultimate Non-Metallic Piping Solutions for High-Purity Water W ater Applications Markets: Life Sciences, Biopharmaceutical, Pharmaceutical, Pharmaceut ical, Laboratory • High Purity Water Loops • USP Water • RO/DI Water

High Speed Smooth Bore Fusion Welding • Eliminates Areas for Entrapment • Simple to Use • Industry’s Industry ’s Fastest Weld Weld Cycle Time

Features & Benefits • Complete Line of Valves, Fittings and Accessorie Accessoriess • PVDF and Polypropylene Materials in Size Ranges from ¾” to 3” • USP/FDA Compliant Materials • Non-metallic, Rouge Free • Eliminates the Need to Passivate • Superior Flow Characteristics • Laser Etched for Complete Lot Traceability • Web-based 3D Models and Drawings Available Please Contact Us At: +1 800.435.3992 +1 908.218.8888

biopharm.saint-gobain.com Sani-Pro®, Sani-Tech® and S.I.B® are Saint-Gobain registered trademarks.

Made In

USA

©2011 Saint-Gobain

ELETTRACQUA pure water technologies

SINCE 1966

PHARMACEUTICAL WATER WATER SYSTEMS E L B A Z I T I N A S R E T A W T O H T I N U I D E + S I S O M S O

t i n u n o i t a l l i t s i d s t c e f f e e l p i t l u m

M E T S Y S Y R E V O C E R ) E m T p A g R T 2 N 6 - E C h / N ³ O m C 4 1 ( + N S I O I S T O A M Z I S T I O N E A S S R E E V R E U T R A S R S E A P P M E E L T B H U G I O D H

r o t a r e n e g m a e t s e r u p

DESIGN AND CONSTRUCTION OF PURIFIED PURIFI ED WATER PACKAGES PACKAGES  DOUBLE PASS REVERSE OSMOSIS  R.O. + ELECTRODEIONIZATION HOT WATER SANITIZABLE  ULTRAFILTRATION 

MULTIPLE - EFFECT DISTILLATION UNITS  PURE STEAM GENERATORS  STORAGE AND

DISTRIBUTION LOOP



COMPLETE TURN KEY PROJECTS

w w w . e l e t t r a c q u a . c o m

VALIDATIONS IQ, OQ



Step up your knowledge with ISPE ISPE..

3 Ways to Learn to increase manufacturing efficiency, maintain product quality, and improve GMP compliance.

Courses offered at national and international training events.

Hone your skills at your desk, immediately, with online training courses and webinars.

Bring ISPE’s expert instructors to your facility for customized training.

Register Now! www www.ISPE.org/T .ISPE.org/Training raining • www.ISPE.org/eLearning www.ISPE.org/eLearning

ISPE_Water and steam systems_Vol4.pdf

Recommend Documents