IEEE Std 1635 -ASHRAE 21-2012 Guide for Ventilation and Thermal Management of Batteries for Stationary Applications

IEEE Power and Energy Society Stationary Batteries Committee

ASHRAE Guideline Project Committee 21 (GPC 21)

IEEE 3 Park Avenue New York, NY 10016-5997 USA 23 October 2012

Authorized licensed use limited to: Michigan State University. Downloaded on February 24,2015 at 06:24:59 UTC from IEEE Xplore. Restrictions apply.


IEEE Std 1635™-2012/ ASHRAE Guideline 21-2012

IEEE/ASHRAE Guide for the Ventilation and Thermal Management of Batteries for Stationary Applications Co-Sponsors

IEEE Power and Energy Society Stationary Batteries Committee and

ASHRAE Guideline Project Committee 21 (GPC 21)

Approved 27 June 2012

ASHRAE Standards Committee/Board of Directors Approved 30 August 2012

IEEE-SA Standards Board


Abstract: Vented lead-acid (VLA), valve-regulated lead-acid (VRLA), and nickel-cadmium (NiCd) stationary battery installations are discussed in this guide, written to serve as a bridge between the electrical designer and the heating, ventilation, and air-conditioning (HVAC) designer. Ventilation of stationary battery installations is critical to maximize battery life while minimizing the hazards associated with hydrogen production. This guide describes battery operating modes and the hazards associated with each. It provides the HVAC designer with the information to provide a cost effective ventilation solution. Keywords: ASHRAE 21, battery, battery cabinets, battery gassing, battery room, battery vaults, forced ventilation, hydrogen, IEEE 1635™, natural ventilation, stationary battery, thermal management, ventilation, ventilation system maintenance •

The Institute of Electrical and Electronics Engineers, Inc. 3 Park Avenue, New York, NY 10016-5997, USA ASHRAE 1791 Tullie Circle, NE, Atlanta, Georgia 30329-2305, USA Copyright © 2012 by The Institute of Electrical and Electronics Engineers, Inc., and ASHRAE. All rights reserved. Published 23 October 2012. Printed in the United States of America. IEEE is a registered trademark in the U.S. Patent & Trademark Office, owned by The Institute of Electrical and Electronics Engineers, Incorporated. ASHRAE is a registered trademark in the U.S. Patent & Trademark Office, owned by the American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc. ANSI is a registered trademark of the American National Standards Institute. EnerSys trademark is the property of EnerSys and its affiliates. ICC, International Code Council, IFC, and I nternational Fire Code are registered trademarks of the International Code Council. National Electrical Code, NEC, and NFPA 70 are registered trademarks of the National Fire Protection Association, Inc. National Electrical Safety Code and NESC are both registered trademarks and service marks of The Institute of Electrical and Electronics Engineers, Inc. Telcordia is a registered trademark of Telcordia Technologies, Inc. UL is a registered trademark of UL LLC © 2012. PDF: Print:

ISBN 978-0-7381-7317-7 ISBN 978-0-7381-7332-0

STD97292 STDPD97292

IEEE prohibits discrimination, harassment, and bullying. For more information, visit http://www.ieee.org/web/aboutus/whatis/policies/p9-26.html . No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written permission of the publisher.


Notice and Disclaimer of Liability Concerning the Use of IEEE Documents : IEEE Standards documents are developed within the IEEE Societies and the Standards Coordinating Committees of the IEEE Standards Association (IEEE-SA) Standards Board. IEEE develops its standards through a consensus development process, approved by the American National Standards Institute, which brings together volunteers representing varied viewpoints and interests to achieve the final product. Volunteers are not necessarily members of the Institute and serve without compensation. While IEEE administers the process and establishes rules to promote fairness in the consensus development process, IEEE does not independently evaluate, test, or verify the accuracy of any of the information or the soundness of any judgments contained in its standards.

Use of an IEEE Standard is wholly voluntary. IEEE disclaims liability for any personal injury, property or other damage, of any nature whatsoever, whether special, indirect, consequential, or compensatory, directly or indirectly resulting from the publication, use of, or reliance upon any IEEE Standard document. IEEE does not warrant or represent the accuracy or content of the material contained in its standards, and expressly disclaims any express or implied warranty, including any implied warranty of merchantability or fitness for a specific purpose, or that the use of the material contained in its standards is free from patent infringement. IEEE Standards documents are supplied "AS IS." The existence of an IEEE Standard does not imply that there are no other ways to produce, test, measure, purchase, market, or provide other goods and services related to the scope of the IEEE standard. Furthermore, the viewpoint expressed at the time a standard is approved and issued is subject to change brought about through developments in the state of the art and comments received from users of the standard. Every IEEE standard is subjected to review at least every ten years. When a document is more than ten years old and has not undergone a revision process, it is reasonable to conclude that its contents, although still of some value, do not wholly reflect the present state of the art. Users are cautioned to check to determine that they have the latest edition of any IEEE standard. In publishing and making its standards available, IEEE is not suggesting or rendering professional or other services for, or on behalf of, any person or entity. Nor is IEEE undertaking to perform any duty owed by any other person or entity to another. Any person utilizing any IEEE Standards document, should rely upon his or her own independent judgment in the exercise of reasonable care in any given circumstances or, as appropriate, seek the advice of a competent professional in determining the appropriateness of a given IEEE standard. Translations: The IEEE consensus development process involves the review of documents in English only. In the event that an IEEE standard is translated, only the English version published by IEEE should be considered the approved IEEE standard. Official Statements : A statement, written or oral, that is not processed in accordance with the IEEE-SA Standards Board

Operations Manual shall not be considered the official position of IEEE or any of its committees and shall not be considered to be, nor be relied upon as, a formal position of IEEE. At lectures, symposia, seminars, or educational courses, an individual presenting information on IEEE standards shall make it clear that his or her views should be considered the personal views of that individual rather than the formal position of IEEE. Comments on Standards : Comments for revision of IEEE Standards documents are welcome from any interested party, regardless of membership affiliation with IEEE. However, IEEE does not provide consulting information or advice pertaining to IEEE Standards documents. Suggestions for changes in documents should be in the form of a proposed change of text, together with appropriate supporting comments. Since IEEE standards represent a consensus of concerned interests, it is important to ensure that any responses to comments and questions also receive the concurrence of a balance of interests. For this reason, IEEE and the members of its societies and Standards Coordinating Committees are not able to provide an instant response to comments or questions except in those cases where the matter has previously been addressed. Any person who would like to participate in evaluating comments or revisions to an IEEE standard is welcome to join the relevant IEEE working group at http://standards.ieee.org/develop/wg/.

Comments on standards should be submitted to the following address: Secretary, IEEE-SA Standards Board 445 Hoes Lane Piscataway, NJ 08854 USA Photocopies: Authorization to photocopy portions of any individual standard for internal or personal use is granted by The Institute of Electrical and Electronics Engineers, Inc., provided that the appropriate fee is paid to Copyright Clearance Center. To arrange for payment of licensing fee, please contact Copyright Clearance Center, Customer Service, 222 Rosewood Drive, Danvers, MA 01923 USA; +1 978 750 8400. Permission to photocopy portions of any individual standard for educational classroom use can also be obtained through the Copyright Clearance Center.


Notice to users Laws and regulations Users of IEEE Standards documents should consult all applicable laws and regulations. Compliance with the provisions of any IEEE Standards document does not imply compliance to any applicable regulatory requirements. Implementers of the standard are responsible for observing or referring to the applicable regulatory requirements. IEEE does not, by the publication of its standards, intend to urge action that is not in compliance with applicable laws, and these documents may not be construed as doing so.

Copyrights This document is copyrighted by the IEEE. It is made available for a wide variety of both public and private uses. These include both use, by reference, in laws and regulations, and use in private selfregulation, standardization, and the promotion of engineering practices and methods. By making this document available for use and adoption by public authorities and private users, neither the IEEE nor ASHRAE waives any rights in copyright to this document.

Updating of IEEE documents Users of IEEE Standards documents should be aware that these documents may be superseded at any time by the issuance of new editions or may be amended from time to time through the issuance of amendments, corrigenda, or errata. An official IEEE document at any point in time consists of the current edition of the document together with any amendments, corrigenda, or errata then in effect. In order to determine whether a given document is the current edition and whether it has been amended through the issuance of amendments, corrigenda, or errata, visit the IEEE-SA Website at http://standards.ieee.org/index.html or contact the IEEE at the address listed previously. For more information about the IEEE Standards Association or the IEEE standards development process, visit IEEE-SA Website at http://standards.ieee.org/index.html.

Errata Errata, if any, for this and all other standards can be accessed at the following URL: http://standards.ieee.org/findstds/errata/index.html. Users are encouraged to check this URL for errata periodically.

Patents Attention is called to the possibility that implementation of this standard may require use of subject matter covered by patent rights. By publication of this standard, no position is taken by the IEEE with respect to the existence or validity of any patent rights in connection therewith. If a patent holder or patent applicant has filed a statement of assurance via an Accepted Letter of Assurance, then the statement is listed on the IEEE-SA Website at http://standards.ieee.org/about/sasb/patcom/patents.html. Letters of Assurance may indicate whether the Submitter is willing or unwilling to grant licenses under patent rights without compensation or under reasonable rates, with reasonable terms and conditions that are demonstrably free of any unfair discrimination to applicants desiring to obtain such licenses.

iv

Copyright © 2012 IEEE/ASHRAE. All rights reserved.


Essential Patent Claims may exist for which a Letter of Assurance has not been received. The IEEE is not responsible for identifying Essential Patent Claims for which a license may be required, for conducting inquiries into the legal validity or scope of Patents Claims, or determining whether any licensing terms or conditions provided in connection with submission of a Letter of Assurance, if any, or in any licensing agreements are reasonable or non-discriminatory. Users of this standard are expressly advised that determination of the validity of any patent rights, and the risk of infringement of such rights, is entirely their own responsibility. Further information may be obtained from the IEEE Standards Association.

v



Participants At the time this IEEE/ASHRAE guide was completed, the Ventilation Working Group had the following membership: M. S. (Steve) Clark, IEEE, Co-Chair Deep Ghosh, ASHRAE, Co-Chair John B. Riley, ASHRAE Vice Chair Curtis Ashton, IEEE Secretary

Phyllis Archer, IEEE

Ramesh Desai, IEEE

James A. McDowall, IEEE

Allen Byrne, IEEE William Cantor, IEEE Jay L. Chamberlin, IEEE Garth P. Corey, IEEE Thomas G. Croda, IEEE

Mark A. Hartfiel, ASHRAE Dennis G. Hellmer, ASHRAE Stephen W. McCluer, IEEE/ASHRAE

Dan McMenamin, IEEE Haissam Nasrat, IEEE Edward P. Rafter, IEEE Kristin Wilde, IEEE Lesley P. Varga, IEEE

At the time this guide was approved, ASHRAE Guideline Project Committee 21 (GPC 21) had the following membership: Cognizant TC: TC 9.2, Industrial Air Conditioning James R. Tauby,SPLS Liaison Deep Ghosh,Chair* John B. Riley,Vice Chair and Secretary*

George M. Adams

Mark A. Hartfiel* Dennis G. Hellmer*

Stephen W. McCluer*

*Denotes members of voting status when the document was approved for publication.

vi



The following members of the individual balloting committee voted on this guide. Balloters may have voted for approval, disapproval, or abstention. William Ackerman Samuel Aguirre Steven Alexanderson James Anderson Phyllis Archer Stan Arnot Curtis Ashton Gary Balash

Donald Dunn Gary Engmann Vladimir Fedkiw Charles Finin Robert Fletcher Carl Fredericks Deep Ghosh James Gleason

Kimberly Mosley Jerry Murphy Dennis Neitzel Arthur Neubauer Michael S. Newman Joe Nims Miklos Orosz Bansi Patel

Michael Bayer Robert Beavers Erich Binder Thomas Blair William Bloethe Richard Bolgeo Paul Boman Mark Bowman Steven Brockschink Terrence Burns William Byrd William Cantor Robert Carruth Leonard Casella Randy Casteel Jay L. Chamberlin Michael Champagne Keith Chow M. S. (Steve) Clark Waller Clements Kevin Coggins Tommy Cooper Garth P. Corey Charles Cotton John Coyle Alireza Daneshpooy Matthew Davis Carlo Donati Randall Dotson

Jalal Gohari Edwin Goodwin James Graham Randall C. Groves Thomas Gruzs Ajit Gwal Scott Hietpas Gary Hoffman David Horvath James Houston David Ittner Randy Jamison Alan Jensen James Jones Joseph L. Koepfinger John J. Kopera David Krause Jim Kulchisky William Kumpf Saumen Kundu Chung-Yiu Lam Daniel Lambert Daniel Levin Albert Livshitz Debra Longtin Federico Lopez Omar Mazzoni Stephen W. McCluer James A. McDowall Jeffrey Merryman

Vernon Peppers Percy Pool Edward P. Rafter Ryland Revelle John B. Riley Michael Roberts Bartien Sayogo Robert Schuerger Devki Sharma Charles Shieh David Smith James Smith John Spare Joseph Stevens Allan St. Peter Paul Sullivan David Tepen S. Thamilarasan James Thompson Wayne Timm Richard Tressler Lesley Varga Gerald Vaughn Stephen Vechy John Vergis William Wessman Kenneth White Luis Zambrano Theodore Zeiss

vii



When the IEEE-SA Standards Board approved this guide on 30 August 2012, it had the following membership: Richard H. Hulett, Chair John Kulick, Vice Chair Robert M. Grow, Past Chair Konstantinos Karachalios, Secretary

Satish Aggarwal Masayuki Ariyoshi Peter Balma William Bartley Ted Burse Clint Chaplin Wael Diab Jean-Philippe Faure

Alexander Gelman Paul Houzé Jim Hughes Young Kyun Kim Joseph L. Koepfinger* David J. Law Thomas Lee Hung Ling

Oleg Logvinov Ted Olsen Gary Robinson Jon Walter Rosdahl Mike Seavey Yatin Trivedi Phil Winston Yu Yuan

*Member Emeritus

Also included are the following nonvoting IEEE-SA Standards Board liaisons: Richard DeBlasio, DOE Representative Michael Janezic, NIST Representative Don Messina IEEE Standards Program Manager, Document Development Malia Zaman IEEE Standards Program Manager, Technical Program Development

ASHRAE Standards Committee 2011-2012 Carol E. Marriott, Chair Kenneth W. Cooper, Vice Chair Ross D. Montgomery, CO Eckhard A. Groll, BOD ExO

Douglass S. Abramson Karim Amrane Charles S. Barnaby Hoy R. Bohanon, Jr. Steven F. Bruning David R. Conover Steven J. Emmerich Allan B. Fraser

Krishnan Gowri Maureen Grasso Cecily M. Grzywacz Richard L. Hall Rita M. Harrold Adam W. Hinge Debra H. Kennoy Jay A. Kohler

Janice C. Peterson Douglas T. Reindl Boggarm S. Setty James R. Tauby James K. Vallort William F. Walter Michael W. Woodford Craig P. Wray

Stephanie C. Reiniche, Manager of Standards

viii



Introduction This introduction is not part of IEEE Std 1635-2012/ASHRAE Guideline 21-2012, IEEE/ASHRAE Guide for the Ventilation and Thermal Management of Batteries for Stationary Applications.

The primary purpose of this guide is to assist users involved in the design and management of new stationary battery installations. The focus is the environmental design and management of the installation to maximize battery reliability as well as the safety of personnel and equipment. This guide is a joint effort by the IEEE and ASHRAE.

ix



Contents 1. Overview .................................................................................................................................................... 1 1.1 Scope ................................................................................................................................................... 1 1.2 Purpose ................................................................................................................................................ 2 1.3 Exclusions............................................................................................................................................ 2 1.4 Document organization........................................................................................................................ 2 2. Normative references.................................................................................................................................. 3 3. Definitions, acronyms, and abbreviations .................................................................................................. 4 3.1 Definitions ........................................................................................................................................... 4 3.2 Acronyms and abbreviations ............................................................................................................... 4 4. Battery safety hazards and considerations.................................................................................................. 6 5. Fundamentals.............................................................................................................................................. 6 5.1 Battery types........................................................................................................................................ 6 5.2 Battery application............................................................................................................................... 9 5.3 Installation enclosure applications..................................................................................................... 10 6. Heating, ventilation, and air conditioning ................................................................................................ 11 6.1 General .............................................................................................................................................. 11 6.2 HVAC design for performance.......................................................................................................... 11 6.3 HVAC design for safety .................................................................................................................... 13 7. Environmental design............................................................................................................................... 15 7.1 General .............................................................................................................................................. 15 7.2 Operatingventilating, modes................................................................................................................................ 7.3 Heating, and air-conditioning system design requirements............................................ 15 26 7.4 HVAC system design for ventilation................................................................................................. 28 7.5 Integrated battery areas...................................................................................................................... 30 7.6 Controls and alarms ........................................................................................................................... 30 7.7 Battery room hazard classification .................................................................................................... 31 7.8 Enclosure design applications............................................................................................................ 31 8. Economics ................................................................................................................................................ 32 8.1 General .............................................................................................................................................. 32 8.2 Battery replacement factors............................................................................................................... 33 8.3 Relative importance of the installation .............................................................................................. 33 8.4 Reliability of the HVAC system........................................................................................................ 33 8.5 Availability of maintenance resources............................................................................................... 33 8.6 Cost and availability of battery replacement ..................................................................................... 33 8.7 HVAC System control based on battery operating mode.................................................................. 33 9. Environmental management (operation and maintenance)....................................................................... 34 9.1 Battery system operation and maintenance........................................................................................ 34 9.2 HVAC system operation and maintenance........................................................................................ 34 Annex A (informative) Hydrogen generation in lead-acid and nickel-cadmium batteries ...........................36 Annex B (informative) Heat generation in lead-acid batteries ..................................................................... 59

x



Annex C (informative) Existing U.S. codes and standards .......................................................................... 86 Annex D (informative) Explosive and toxic gas allowance considerations ................................................. 88 Annex E (informative) Thermal runaway..................................................................................................... 90 Annex F (normative) Hydrogen sulfide gas ................................................................................................. 92 Annex G (informative) Bibliography ........................................................................................................... 93

xi




IEEE/ASHRAE Guide for the Ventilation and Thermal Management of Batteries for Stationary Applications IMPORTANT NOTICE: IEEE Standards documents are not intended to ensure safety, health, or environmental protection, or ensure against interference with or from other devices or networks. Implementers of IEEE Standards documents are responsible for determining and complying with all appropriate safety, security, environmental, health, and interference protection practices and all applicable laws and regulations. This IEEE document is made available for use subject to important notices and legal disclaimers. These notices and disclaimers appear in all publications containing this document and may be found under the heading “Important Notice” or “Important Notices and Disclaimers Concerning IEEE Documents.” They can also be obtained on request from IEEE or viewed at http://standards.ieee.org/IPR/disclaimers.html.

1. Overview

1.1 Scope This guide discusses the ventilation and thermal management of stationary battery systems as applied to the following:

⎯

Vented (flooded) lead-acid (VLA)

⎯

Valve-regulated lead-acid (VRLA)

⎯

Nickel-cadmium (NiCd)

For each category, both the technology and the design of the battery are described in order to facilitate user understanding of the environmental issues associated with each type of technology. The scope of this document includes only stationary batteries under conditions of expected use. Multiple operating modes are identified. The ventilation practices described in this guide represent the “best practice” based on the information available at the time this document was developed. The user should evaluate these practices against their operating experience, operating conditions, number and size of battery systems, manufacturer’s recommendations, resources, and needs in developing an environment that maximizes safety and is

1



IEEE Std 1635-2012/ASHRAE Guideline 21-2012 IEEE/ASHRAE Guide for the Ventilation and Thermal Management of Batteries for Stationary Applications

conducive to optimum operation of the equipment. These recommendations were developed without consideration of economics, availability of equipment and personnel, or relative importance of the application. Design of a ventilation system for a specific battery installation requires consideration of all issues, not just the technical issues considered in this document.

1.2 Purpose The purpose of this document is to provide heating, ventilation, and air conditioning (HVAC) and battery system designers and users with information and recommendations concerning the ventilation and thermal management of stationary battery installations.

1.3 Exclusions Specifically not included in this document are the following:

⎯

Ventilation for spilled electrolyte, as a spill is considered an accident condition. Ventilation for fumes associated with spills due to handling of batteries is addressed in IEEE Std 1578™ [B25]. 1

⎯

Recharging stations for motive power or automotive batteries. These stations use different charging regimes from stationary float applications, although the charging gases produced can be managed in accordance with the guidelines in this document

⎯

Batteries that are embedded in small equipment such as desk-top uninterruptible power supply (UPS) systems and emergency lighting systems

⎯

Battery installations in classified (hazardous) environments

⎯

Ventilation for gases or other byproducts given off by battery installations involved in fires. The latter are not considered conditions of expected use, and in such circumstances ventilation could be disabled

⎯

Ventilation of the charger, UPS, or other equipment associated with the battery system

⎯

Design for fire or smoke events.

⎯

Ventilation and management of gases, such as, arsine, stibene, hydrogen sulfide, ozone, sulfur dioxide, sulfur trioxide, or chlorine, produced during extreme over charge conditions (e.g., charger malfunction, multiple shorted cells)

⎯

General design criteria for HVAC systems and/or design alternatives and schemes.

1.4 Document organization This recommended practice is divided into nine clauses. Clause 1 provides the scope of this guide. Clause 2 lists references to other standards that are useful in applying this guide. Clause 3 provides definitions, acronyms, and abbreviations that are either not found in other standards, or have been modified for use with this guide. Clause 4 discusses battery safety hazard considerations. Clause 5 provides information on battery fundamentals. Clause 6 provides information on heating, ventilation, and air conditioning. Clause 7 providesdesign. information environmental design. Clause 8 discusses the economics of battery ventilation system Clause on 9 provides information on environmental management.

1

The numbers in brackets correspond to those of the bibliography in Annex G.

2




This guide has seven annexes. Annex A provides the gassing calculations. Annex B provides heat generation calculations. Annex C provides information on codes and standards. Annex D provides information on explosive and toxic gas allowance considerations. Annex E provides information on thermal runaway. Annex F provides information on hydrogen sulfide gas. Annex G provides bibliographic references.

2. Normative references The following referenced documents are indispensable for the application of this document (i.e., they must be understood and used, so each referenced document is cited in text and its relationship to this document is explained). For dated references, only the edition cited applies. For undated references, the latest edition of the referenced document (including any amendments or corrigenda) applies. ASHRAE Standard 52.2-2007™, Method of Testing General Ventilation Air-Cleaning Devices for Removal Efficiency by Particle Size. 2

ASHRAE Terminology of Heating, Ventilation, Air Conditioning, & Refrigeration, 1991. IEEE Std 450™, IEEE Recommended Practice for Maintenance, Testing, and Replacement of Vented Lead-Acid Batteries for Stationary Applications. 3, 4 IEEE Std 937™, IEEE Recommended Practice for Installation and Maintenance of Lead-Acid Batteries for Photovoltaic (PV) Systems. IEEE Std 1106™, IEEE Recommended Practice for Installation, Maintenance, Testing, and Replacement of Vented Nickel-Cadmium Batteries for Stationary Applications. IEEE Std 1188™, IEEE Recommended Practice for Maintenance, Testing, and Replacement of ValveRegulated Lead-Acid (VRLA) Batteries for Stationary Applications. IEEE Std 1561™, IEEE Guide for Optimizing the Performance and Life of Lead-Acid Batteries in Remote Hybrid Power Systems. NFPA 1, Fire Code.5 NFPA 90A, Standard for the Installation of Air-Conditioning and Ventilation Systems.

2 3 4 5

Publications are available from ASHRAE (http://www.ashrae.org/). IEEE publications are available from The Institute of Electrical and Electronics Engineers (h ttp://standards.ieee.org/). The IEEE standards or products referred to in this clause are trademarks of The Institute of Electrical and Electronics Engineers, Inc. NFPA publications are available from the National Fire Protection Association (http://www.nfpa.org/).

3




3. Definitions, acronyms, and abbreviations For the purposes of this document, the following terms and definitions apply. The IEEE Standards Dictionary Online and ASHRAE Terminology of Heating, Ventilation, Air Conditioning, & Refrigeration should be consulted for terms not defined in this clause. 6, 7

3.1 Definitions confined space: A space that (1) is large enough and so configured that an employee can bodily enter and perform assigned work; (2) has limited or restricted means for entry or exit (for example, tanks, vessels, silos, storage bins, hoppers, vaults, and pits); and (3) is not designed for continuous employee occupancy. NOTE—See Code of Federal Regulations (29CFR1926) [B16]. 8

controlled environment vault (CEV): A remote and partially or fully buried structure to house electronic and telecommunications equipment that includes backup power. hydrogen pocket: A volume of air in which hydrogen accumulates to a concentration approaching or exceeding the lower flammability limit (LFL). hydrogen trap: A spatial feature that prevents diffusion of hydrogen, thereby leading to the creation of a hydrogen pocket. string: A grouping of interconnected cells that has the same nominal voltage as the dc system. thermal management: The use of various temperature monitoring devices, heating and cooling methods to control temperature within a defined space. thermal runaway: A condition that is caused by a battery charging current or other process that produces more internal heat than the battery can dissipate.

3.2 Acronyms and abbreviations ac

alternating current

AGM ®

ANSI

absorbed glass mat American National Standards Institute

9

ASHRAE

American Society of Heating, Refrigerating and Air-Conditioning Engineers Inc.

CEV

controlled environment vault

cfm

cubic feet per minute

6

IEEE Standards Dictionary Online subscription is available at http://www.ieee.org/portal/innovate/products/standard/standards_dictionary.html. 7

Information on references can be found in Clause 2.

8

Notes in text, tables, and figures of a standard are given for information only a nd do not contain requirements needed to implement this standard. 9

ANSI is a registered trademark of the American National Standards Institute.

4




CFR

Code of Federal Regulations

dc

direct current

FC

Fire Code

HVAC

heating, ventilation, and air conditioning

ICBO

International Conference of Building Officials

®

ICC

IEEE IFC® LEL

® 10

International Code Council

Institute of Electrical and Electronics Engineers Inc. International Fire Code

® [B27] 11

lower explosive limit

LFL

lower flammability limit

LOI

limiting oxygen index

NEC®

National Electrical Code

®

[B29] 12

NFPA

National Fire Protection Association

NiCd

nickel-cadmium battery

OSHA

Occupational Safety and Health Administration

ppb

parts per billion

ppm

parts per million

PV

photovoltaic

s.g.

specific gravity

SCBA

self-contained breathing apparatus

UEL

upper explosive limit

UFL

upper flammability limit

®

UL

Underwriters Laboratories

13

UPS

uninterruptible power supply

VLA

vented lead-acid battery

VRLA

valve-regulated lead-acid battery

10 11 12 13

ICC is a registered t rademark of the International Code Council. IFC and International Fire Code are registered trademarks of the International Code Council. National Electrical Code and NEC are registered trademarks of the National Fire Protection Association, Inc. UL is a registered trademark of UL LLC © 2012.

5




4. Battery safety hazards and considerations There are two primary safety considerations associated with the ventilation and thermal management of stationary battery installations. The first is personnel safety and the second is the reliability of the battery and the equipment located in the same or adjacent spaces. This document focuses primarily on personnel safety hazards that can be mitigated through ventilation and thermal management, although where equipment reliability is affected, it will be noted. Safety hazards associated with gas evolution from stationary battery installations include the following:

⎯

Explosive gases

⎯

Toxic gases

⎯

Corrosive gases

⎯

Acid vapor and or mist

The causes, risks, and mitigation of these safety hazards are discussed in detail in Clause 6 and Clause 7 following a discussion of battery types and typical applications for each type.

5. Fundamentals

5.1 Battery types Rechargeable batteries store and release electrical energy through reversible chemical reactions. As the batteries release or discharge their stored electrical energy, the principal inefficiency involves electrical resistance to the current flow, which is released in the form of thermal energy. The process of storing electrical energy is also accompanied by inefficiencies, released in the form of thermal energy and evolution of gases. The of release thermal energy ischemical due to electrical or exothermic reactions. The evolution gasesof is due to incidental reactions.resistance Some of these incidental chemical chemical reactions result in the release of oxygen and hydrogen gas molecules.

5.1.1 Lead-acid batteries Lead-acid is the most common battery type used in stationary applications. In the charged condition the active material in the positive plate is lead dioxide and in the negative it is sponge lead. The sulfuric acid electrolyte solution is an active component in the cell charge/discharge reaction. The plate structure is typically grids or spines of pure lead or lead alloys with various hardening and grain refining agents added such as calcium, tin, antimony, selenium, etc. The alloy composition affects the gassing characteristics of the battery (see 7.2). Ideal process: Discharge/recharge [see Equation (1)]

Pb + PbO2 + 2 H 2 SO4 ⇔ 2 PbSO4 + 2 H 2O

(1)

where

Pb is metallic lead PbO2 is lead dioxide

6




H2SO4 is sulfuric acid PbSO4 is lead sulfate H2O is water Ideal recharging reactions: Positive plate [see Equation (2)]

PbSO4 + 2 H 2O ⇒ PbO2 + SO4 −2 + 4 H + + 2e −

(2)

where

SO4–2 is the sulfate ion H+ is the hydrogen ion e– is an electron Negative plate [see Equation (3)]

PbSO4 + 2e − ⇒ Pb + SO4 −2

(3)

When a cell is subjected to an overcharge it will electrolyze water in the electrolyte. The electrolysis will produce oxygen gas at the positive plate as noted in Equation (4). As long as the oxygen gas production does not exceed the achievable oxygen gas diffusion rate of the cell, the principal reaction at the negative plate will be recombination as noted in Equation (5). The recombination reaction is exothermic, producing heat at the negative plate. To the extent the overcharge current leads to oxygen production exceeding the achievable oxygen diffusion rate of the cell, the negative plate reaction will proportionately shift to hydrogen gas production as noted in Equation (6), leading to the release of both hydrogen and oxygen gas from the cell. Overcharge reactions: Positive plate 2 H 2O ⇒ O2 + 4 H + + 4e −

(4)

where

O2 is oxygen gas Negative plate reactions: Recombination 4 H + + 4e − + O2 ⇒ 2 H 2O

(5)

Gas production: 4 H + + 4e − ⇒ 2 H 2

(6)

where

7




H2 is hydrogen gas 5.1.1.1 Vented lead-acid (VLA) batteries The VLA battery is simply a lead-acid battery with a free flowing liquid electrolyte that allows any gases generated during charging to be vented directly to the atmosphere. That volume below the cover and above the electrolyte level can be up to 67% hydrogen and 33% oxygen by volume, which is an explosive mixture. Typically these gases are vented via a flame arresting (safety) vent to prevent ignition from a spark or flame outside the cell from entering the cell container and igniting the contained gases. For VLA batteries being overcharged the principal overcharge reaction involves gas production as in Equation (6). Recombination as in Equation (5) does occur, but at such a low rate that it is can be disregarded. The reason is that oxygen gas diffuses at a very slow rate through liquid electrolyte. During a recharge cycle the rate of gas production is dependent on a number of factors involving characteristics of both the batteries and the charger. In general, peak gas production occurs at the end of the cycle as the batteries near their full state of charge. At this point, most of the charge energy results in electrolysis of water. VLA batteries use a variety of lead alloys, which impact the generation rate and types of gases produced during an overcharge condition. The alloys discussed in the tables in Clause 7 are lead-calcium, leadantimony, and lead-selenium. Lead-calcium batteries have low float currents. Lead-antimony and leadselenium have higher float currents, which increase over the life of the battery. Pure lead cells have the lowest float current. All lead-acid batteries are capable of producing hydrogen sulfide gas (toxic and corrosive) and acid mist under extreme overcharge conditions, but such failure modes are beyond the scope of this document. See Annex D for additional information.

5.1.1.2 Valve-regulated lead-acid (VRLA) batteries VRLA batteries are designed to take advantage of the naturally occurring recombination cycle where internally generated oxygen gas molecules are recombined with hydrogen ions to form water molecules. Key construction elements of these batteries are pressure relief valves and special internal construction to permit oxygen gas migration to the negative plate. To minimize electrolysis of the electrolyte and the resulting water loss, VRLA batteries are designed to have low float current. Current VRLA battery technologies are based on lead-calcium, lead-tin, or purelead grid alloys. The electrolyte is immobilized in either a glass mat [absorbed glass mat (AGM)] or a silica gel (gelled cell). These construction methods achieve rates of oxygen gas migration sufficient to achieve recombination at a level to substantially reduce or eliminate the need for watering for the life of the battery. The recombination cycle comes into play during charging, especially at the end of a recharge cycle and during periods of elevated charging for equalization or freshening. In the recombination cycle oxygen gas generated at the positive plates migrates to the negative plates for recombination with hydrogen ions to form water molecules. The principal limiting factor to recombination is the sustainable rate of migration of oxygen gas molecules to the negative plate. The pressure relief valve allows excess internal pressure from generated gases to be relieved while preventing atmospheric air from entering the battery. When the valves do function to relieve internal pressure, the gases that escape directly equate to a loss of water from the battery that generally cannot be replaced. The recombination reaction is exothermic releasing heat at the negative plate. The greater the overcharge current, the greater the oxygen generated at the positive plate. This will result in greater heating at the negative plate and that heat must be dissipated. If thermal equilibrium cannot be reached, thermal runaway may occur (see Annex E). Regardless of whether or not thermal runaway occurs with increasing overcharge

8




current, at some point the oxygen generation rate will exceed the diffusion rate to the negative plate and oxygen and hydrogen gas will be expelled as with a vented cell. The vented gas will remove some of the excess heat, but this is at the expense of water loss from the cell. During normal charging of VRLA batteries a small portion of the charge current results in corrosion of their lead positive plate grid structure. This process is represented by Equation (7). Positive plate grid corrosion reaction:

Pb + 2 H 2O ⇒ PbO2 + 4 H + + 4e −

(7)

The oxygen consumed this reaction not available for recombination at thewith negative plate,(6). so This the only possible reaction at the by negative plate isishydrogen gas evolution, in accordance Equation low rate but continuous release of hydrogen results in a gradual increase in pressure in the head space of the cell; when the pressure reaches a predefined level the valve opens, some hydrogen is released, and then the valve reseals. The rate of hydrogen release from this process is significantly lower than from vented cells. See 7.2 for gassing rate information.

5.1.2 Nickel-cadmium (NiCd) batteries Vented NiCd batteries behave in a similar way to VLA batteries on overcharge, releasing hydrogen and oxygen gases in a 2-to-1 ratio. See 7.2 for gassing rate information. There are also partially recombinant designs, which may or may not be fitted with pressure valves. Consult the manufacturer for specific information regarding gassing rates under the expected operating conditions. In the absence of such information these products may be conservatively treated in the same way as vented designs. Fully sealed NiCd cells, such as those used in portable appliances, are not suitable for use in stationary float applications.

5.2 Battery application

5.2.1 General Batteries can be categorized either by the type of service or by the type of installation enclosure. Battery applications by service application and by enclosure application are discussed as follows.

5.2.2 Service applications

5.2.2.1 Standby or float service The following applications typically use lead-acid (both vented and VRLA) and NiCd batteries. Standby or float application is the typical use of stationary batteries. Float service is a battery connected to a charger that is continuously maintained at or near full charge so that it has a design energy storage potential for a specific purpose. Common application types are as follows:

⎯

Telecommunication (low rate, long duration)

⎯

UPS (high rate, short duration)

9




⎯

Switchgear (general purpose, medium duration)

⎯

Starting batteries (high rate, short duration)

5.2.2.2 Cycling service These applications are characterized by frequent discharge/charge cycles. The increase in discharge/recharge cycles typically increases the amount of gas and heat generated by the installation. However, this is not always true, particularly in photovoltaic (PV) and other applications designed to operate at a partial state of charge (PSOC).

5.3 Installation enclosure applications

5.3.1 General Personnel safety and proximity to critical equipment are important considerations when deciding whether or not a dedicated battery room is necessary. Batteries should be enclosed to prevent unauthorized access because of safety considerations. The necessary enclosure may consist of a battery cabinet, a cage, enclosing partitions or walls, or a vault. When enclosure is not feasible, personnel should be appropriately notified of any battery-related hazards present. As the nature of the facility in which the batteries are installed becomes more critical, it is more common to see the batteries more isolated from other equipment.

5.3.2 Dedicated battery rooms A dedicated battery room is a specially constructed and equipped area typically used primarily for installation of the battery and other components of the dc system. Batteries are typically installed on open racks. The room is physically separated from other areas, avoiding localized heat sources, with controlled access and doors and partitions determined to meet the required fire resistance rating for the application. For purposes of this guide, a battery room may include the charging or UPS equipment to which the batteries are connected.

5.3.3 Indoor cabinets Indoor cabinets are enclosures housing batteries and/or monitoring equipment. The purpose of the cabinet is to house the batteries in a compact space while limiting unauthorized access to live electrical components. The batteries in indoor cabinets are almost exclusively of the VRLA type. For small systems the batteries and charging system or UPS might be located in a common cabinet. While acid mist or vapor is not normally an issue during normal charge modes, if the batteries are vented type, it is preferred that the cabinet ventilation system provide for isolation of the battery exhaust from the electronics.

5.3.4 Outdoor cabinets The most common (but not the only) applications for outdoor battery cabinets are for telecommunications electronic equipment, signal cabinets, and utility substations. Most of these cabinets are equipped with VRLA or NiCd batteries for backup. Some outdoor cabinets have some sort of air conditioning, but many others do not. Those that do not have some form of air conditioning often have heat exchangers to minimize the interior cabinet temperature rise to no more than 10 °C above outdoor ambient temperatures. Usually, the electronics in the cabinet are in a sealed chamber, so the batteries are usually kept in a separate

10




chamber with some sort of ventilation. Outdoor ambient temperature extremes, radiant heating, and cabinet positioning can all affect the performance and health of the batteries.

5.3.5 Controlled environment vault (CEV) The most common application for CEVs is to house telecommunication equipment. CEVs are remote, partially or fully buried structures to house equipment including backup power. CEVs are locked and unoccupied except when maintenance or installation is taking place so they must be remotely monitored. The batteries are normally installed in relay racks and co-located with other electrical and electronic equipment. As the name implies, ventilation and temperature control is always present in order to keep the electronic equipment operating within its recommended range. Temperature tolerance of the batteries may be lower than for the electronics (see 7.8.4).

5.3.6 Integrated battery and equipment areas

⎯

Telecommunication and UPS system battery plants often share the same space as the equipment they support. Many battery systems in these environments use VRLA batteries. Telecommunications equipment often mounts batteries on trays in open relay racks below the rectifiers. UPS equipment is more often enclosed in cabinets. The batteries can be integrated into the UPS equipment, or they can be in cabinets.

⎯

Batteries used in larger utility substations are most often VLA. These systems are generally located in the substation control house along with control equipment. In smaller substations where the control house consists of a single cabinet, the batteries are typically VRLA and enclosed in the substation cabinet.

6. Heating, ventilation, and air conditioning

6.1 General A well-designed HVAC system for environments in which stationary batteries are installed should simultaneously achieve the following objectives within the budgetary constraints of the project (see Clause 8):

⎯

Optimize performance of the battery system within the budgetary constraints of the project (see Clause 8)

⎯

Provide maximum safety for personnel and infrastructure

6.2 HVAC design for performance

6.2.1 Temperature

6.2.1.1 Battery life The service life of a stationary battery is typically based on the electrolyte temperature of 25 °C (77 °F) in North America and 20 °C (68 °F) in other parts of the world. NiCd batteries are less affected than lead-acid batteries by temperature extremes. Continuous operation at elevated temperature will reduce the life of a 11




lead-acid battery by approximately 50% for every increase of 8.3 °C (15 °F) in electrolyte temperature above 25 °C (77 °F). For NiCd batteries life is reduced by 50% for every 30 °C (54 °F) above 25 °C (77 °F). Operation at reduced temperatures results in some gain in battery life, but the gain is significantly less than the loss of life above the design temperature.

6.2.1.2 Battery capacity The ability to provide power on demand decreases as electrolyte temperature is reduced. The extent of this decrease depends on battery chemistry, cell design, and discharge rate. The minimum design temperature is established during battery sizing and requires coordination between the battery system and HVAC design engineers.

6.2.1.3 Ambient temperature Changes in electrolyte temperature lag behind changes in ambient air temperature. Parameters that can affect electrolyte temperature as a function of ambient temperature include the following:

⎯

Battery mass

⎯

Spacing/airflow between batteries

⎯

Duration of temperature deviation

⎯

Charging current

The optimal operating range for a battery (VLA, VRLA, or NiCd) is an electrolyte temperature between 20 °C and 25 °C. Operation in this temperature range provides the best balance between capacity and battery life. When a battery is float charged, the heating due to the charge current and, in the case of VRLA, the recombination reaction results in the internal battery temperature being approximately 1 °C (2 °F) higher than the ambient temperature. Because of this heating effect the optimal ambient temperature range is 19 °C to 24 °C (66 °F to 75 °F). The optimal temperature range for a specific application depends on the balance between battery life and capacity, the capital and operating costs of the HVAC system, and the requirements of other equipment located in the same area. Achieving this balance requires dialogue between the battery system specifier and HVAC system designer. Gassing increases with increasing temperature. In lead-antimony batteries, gassing also increases with age. See Clause 7 for methods to calculate the gassing rates. Maintaining the battery installation within the optimal temperature range reduces the possibility of thermal runaway in VRLA batteries.

6.2.1.4 Temperature gradients Installation design should allow for radiant and convective heat dissipation around the battery. Heat sources and airflow should be managed so as to limit electrolyte differential temperature within a series battery string to a maximum of 3 °C for lead-acid and 5 °C for NiCd batteries.

6.2.2 Air humidity Batteries can function across a very wide range of humidity. High humidity is generally not a problem unless it condenses and causes oxidation of exposed metal parts, such as intercell connections. Low

12




humidity conditions are a long term issue for VRLA batteries due to increased water loss that results in reduced battery life. Low humidity presents no problem for VLA batteries, but it does increase the potential for electrostatic discharge (ESD). Personnel working in close proximity to a battery vent, regardless of humidity level, should take precautions to avoid generating an ESD as it can ignite the hydrogen and oxygen in the head space of a battery, thereby creating the possibility of an explosion even when the battery is fitted with flame arrestors. Relative humidity should be maintained above 20%, ideally between 35% and 65%. No special humidity control features are normally necessary.

6.2.3 Air contaminants

6.2.3.1 Dust Water soluble salts found in dust particulate buildup, especially when it has metallic (conductive) elements and/or is combined with humidity levels above 40%, can lead to accelerated corrosion levels, tracking, and electrical shorting or arcing problems on batteries. It is recommended that medium efficiency filters be used to filter outside air supplied to an indoor battery installation. Filters with ASHRAE 52.2 Minimum Efficiency Reporting Values (MERV) between 7 and 11 should be used to filter the outside air source (ASHRAE Standard 52.2-2007).

6.2.3.2 Other contaminants A comprehensive HVAC design should take into account not only any vapors or gases produced by the battery, but other vapors or gases present in the facility that could potentially damage the batteries or associated equipment. Chemical process industry facilities are an example of facilities likely to contain other vapors or gases of concern. Electrical contacts and connections are particularly susceptible to damage from corrosive chemicals. In more extreme environmental conditions, battery racks may also be vulnerable to corrosive attack.

6.3 HVAC design for safety

6.3.1 Flammable/explosive gases The potentially explosive gas emitted by all lead-acid and NiCd stationary batteries is hydrogen. The hydrogen gas is released in a two-thirds/one-third mixture with oxygen gas. The evolution of this gas mixture depends highly upon the charging current flowing through the battery. The relationship of current to operating mode, temperature, and other factors are covered in more depth in Clause 7. In the following discussion the terms flammable and flammable air-vapor mixture are defined as follows: flammable: Subject to easy ignition and rapid flaming combustion. flammable-air-vapor mixture: Mixture of a flammable vapor in air between the lower flammability level (LFL) and upper flammability limit (UFL).

Flammable and explosive limits apply generally to vapors and are defined as the concentration range in which a flammable substance can produce a fire or explosion when an ignition source (such as a spark or open flame) is present. The concentration is generally expressed as percent fuel by volume (Interactive Learning Paradigms [B26]).

13




Above the UFL the mixture of substance and air is too rich in fuel (deficient in oxygen) to burn (Interactive Learning Paradigms [B26]). The upper explosive limit (UEL) is the mixture of substance and air that is too rich in fuel to explode. Below the LFL the mixture of substance and air lacks sufficient fuel (substance) to burn. The lower explosive limit (LEL) is the mixture of substance and air that is too lean in fuel to explode.

WARNING

Any concentration between the LFL and UFL limits can either ignite or explode. While hydrogen and air mixtures above the UFL and UEL cannot burn or explode respectively, the introduction of air would cause the mixture to become diluted and the concentration could drop into the hazardous range (Interactive Learning Paradigms [B26]).

The LEL of hydrogen is approximately 17% and the LFL is approximately 4% (NFPA 497 [B32]). The UEL is about 56% and the UFL is approximately 75% (NFPA 497 [B32]). While technically incorrect, it is common for the flammability limits to be referred to as the explosive limits. In stationary battery installations, the UFL for hydrogen is not exceeded because the gas released from the batteries is not pure. It is a mixture that is one third oxygen gas, two thirds hydrogen gas, and accompanied by small amounts of other elements.

6.3.2 Hydrogen traps Hydrogen molecules are the smallest and lightest of all naturally occurring molecules. Accordingly, hydrogen is very easily dispersed with a minimum amount of air movement. Hydrogen also mixes with air by diffusion. Usually, unrestricted natural air movement in the vicinity of the battery is sufficient to disperse released hydrogen. If there is insufficient airflow to disperse hydrogen vented from a battery, concentrated hydrogen pockets may form near the battery vents on the lower tiers of a battery rack. Ensuring good air movement through the battery racks will prevent hydrogen concentration.

6.3.3 Confined spaces Ventilation of confined spaces is of particular importance because of the increased possibility of hydrogen traps forming flammable or explosive hydrogen concentrations. One example of such a space, where batteries are installed, is a CEV in a remote telecommunications site (see 7.8.4).

6.3.4 Thermal runaway Thermal runaway can occur in any aqueous battery design, but is most likely to occur in a VRLA battery due to its limited water content. Designs deficient in providing adequate ventilation, temperature control, or charger settings contribute to the phenomena. Operational issues such as ventilation or temperature control malfunctions, individual cell failure(s), excessive charging capacity, or charger malfunctions may aggravate the situation. During thermal runaway, hydrogen gas evolution greatly increases. The event may continue without abatement, causing other gases and particulates to become airborne. For VRLA batteries, thermal runaway is a significant safety concern, but it can be detected and controlled with appropriate safeguards in the electrical system design. See Annex E for more information on thermal runaway.

14




Thermal runaway in NiCd batteries is normally associated with separator breakdown resulting in recombination heating (see Annex E for additional information).

6.3.5 Toxic gases Toxic gases such as hydrogen sulfide, arsine and stibine (see Annex D) are normally only present during thermal runaway and are not included in the scope of this document. See 6.3.4, 7.2.10.1, and Annex E for further discussion of thermal runaway.

6.3.6 Corrosive gases and acid vapor During operational modes such as boost or equalize charge, small quantities of corrosive gases such as hydrogen sulfide and acid vapor can be generated. The quantities generated are normally too small to be detected. The amount of corrosive gas or acid vapor generated is so small (even in these modes) that this is primarily a long term concern for equipment reliability. Examples of equipment subject to potential long term equipment reliability issues when exposed to corrosive gases include cell terminations, inter-tier and inter-rack lug-to-cable connections, and electronic equipment (e.g., chargers, UPS) installed in close proximity to the battery. Since release of corrosive gases and acid vapor during boost or equalize charge is rare and the quantities are small when it does occur, they generally pose little personnel risk. Normal building ventilation is usually sufficient to disperse the corrosive gases and acid vapor. No special ventilation fan or duct construction material will generally be required. Low exhaust pickup locations areas are also not generally needed. It should be noted that the corrosive effect of gases is magnified by humidity. Corrosive gas and acid vapor generation during thermal runaway can be significant. See Annex E for a discussion on thermal runaway and Annex D for a discussion on hydrogen sulfide gas.

7. Environmental design

7.1 General The following subclauses describe the common battery operating modes and the relevant battery design parameters for each mode. The HVAC system should be designed to address the site specific features. Safety, performance, and the reliability of the HVAC system can be achieved by keeping the system design simple, but with sufficient redundancy.

7.2 Operating modes

7.2.1 General The most accurate values for current, hydrogen release, and heat generation should be obtained from the manufacturer. However, if such values cannot be obtained, the equations in the following Table 1 through Table 5 will give reasonable upper bound estimations for a given operating mode. By necessity, the equations in the tables are based on assumptions to be applicable to the broadest ranges of battery types (meaning that some individual batteries from some manufacturers may have somewhat different gassing characteristics). Annex A and Annex B describe the assumptions used in the development of the equations

15




along with the applicable references. In addition, the tables do not account for excessive abnormalities in individual batteries or in operating conditions. All of this means that while the calculations of the tables will give reasonable upper bound estimates of hydrogen gassing and heat generation for the great majority of applications, there may be individual cases where the hydrogen release or heat generation is higher. The typical mode(s) of operation will help determine which table to use for gassing and heat generation calculations. Gassing calculations under float conditions should generally not be used as the basis for ventilation design. Due to failure modes, human error, and code requirements, it is probably safest to use gas generation calculations for boost/equalize charge in Table 2 to determine normal ventilation requirements (it will be noted that even assuming these conditions, gas generation is relatively low). Additional ventilation may be needed for initial charging (particularly for vented cells) based on the calculations in Table 5 for that mode. When determining HVAC design to maintain space temperature, the heat generation calculations from Table 1 are normally the most appropriate to use for float applications. For cycling applications, the heat generation calculations from Table 3 for discharge and Table 4 for bulk recharge are normally most appropriate.

7.2.2 Assumptions for the tables

7.2.2.1 Units The equations produce results in SI units, including that for the generation rate of the hydrogen gas. For inch-pound units appropriate conversion factors should be applied [e.g., m 3/s × 2119 ≈ cubic feet per minute (cfm)].

7.2.2.2 Standard temperature and atmospheric pressure Table 1 through Table 7 are based on an operating temperature of 25 °C (77 °F) at sea level. Hydrogen gassing and heat release are directly related to current flow into the battery. Increases in temperature will increase current, and therefore increase gassing and heat release. Lower temperatures will have the opposite effect. If the battery will be operated in an elevated temperature environment, the battery manufacturer should be contacted to determine the exact effect on gassing and heat release. Gassing and the heat release rate double for approximately every 8 °C to 10 °C (15 °F to 18 °F) above 25 °C (77 °F) for lead-acid batteries due to a rise in current. For elevated temperatures up to 33 °C (92 °F), or operation at elevations up to 300 m (approximately 1000 ft), a battery on float, without an adjustment for temperature, will react similarly to a battery on equalize, and Table 2 may be used. For higher temperatures the battery manufacturer should be contacted. For elevations above 300 m, the normal atmospheric pressure at the battery location should be used to adjust the hydrogen evolution rate.

7.2.2.3 Upper bounds Conservative assumptions are made in the tables to simplify both the gassing and heating calculations. For example, a review of multiple manufacturers literature was performed and the highest recommended voltages for the majority of the manufacturer’s for a given operating mode are used, and the highest specific gravities for a specific type are used when it will give a higher hydrogen or heat release rate. The conservative estimates are such that the gassing and heat generation would not be likely to exceed the results given by the tables in over 95% of actual cases; however, for the most accurate calculations, refer to Annex A and Annex B.

16




7.2.2.4 Battery ratings Note that although some lead-acid batteries will have ratings for both the 8 h (C8) rate and the 15 min (R15) rate, most batteries are designed for either a long duration, or a short-duration discharge. For long and medium duration discharge batteries, use the C8 rate. For batteries designed for rapid discharge (such as in a UPS), use the R15 rate. It will make a difference with some calculations, especially those for internal resistance and intercell connector resistance, used in the charge and discharge heat generation rate equations. European lead-acid batteries may be rated at the 10 h ( C10) rate to 1.8 V/cell at 20 °C or at the 15 min rate to 1.65 V/cell at 20 °C. These ratings are essentially equivalent to the C8 rating or the P15 rating at 25 °C, respectively. For other nonstandard ratings, request the standard rating from the battery manufacturer. NiCd batteries are generally rated at the 5 h (C5) rate to 1.0 V/cell at 20 °C.

7.2.2.5 Current Due to subtle differences in cell chemistry from manufacturer to manufacturer and from one model type to another, the most accurate current in amperes at a given voltage can be obtained from the manufacturer. However, if such a value cannot be obtained, the equations in the following tables will give a reasonable upper bound estimation of the current for a given operating mode (see Annex A and Annex B for how these equations were determined). Note that current is not included in the hydrogen gassing calculations for VRLA batteries. This is due to the fact that not all the current generates hydrogen gas, and due to recombination, not all of the gas escapes the cell (in fact, most of it does not). If the battery manufacturer gives the actual gas release numbers for the different operating modes, then these numbers are generally more accurate than the values produced by the equations.

7.2.2.6 Oxygen evolution While vented batteries evolve both hydrogen and oxygen into the surrounding atmosphere, only the hydrogen gassing equations are given. The oxygen produced does not contribute any more to the flammability of hydrogen because there is already oxygen in the atmosphere into which the hydrogen is being released.

7.2.2.7 Counting the cells in a module The number of cells can be misleading from a visual perspective. Smaller batteries typically come in multicell units. A lead-acid battery with a nominal rating of 12 V actually has 6 cells and should be counted as 6. A nominal 12 V NiCd battery typically has 9 to 10 cells.

7.2.2.8 Estimates of resistance For the internal resistance of the cells and intercell connector resistances for heat generation calculations, the best data comes from the manufacturer. The equations in the tables are to be used if data cannot be obtained from the manufacturer or from direct readings. The resistance equations in the tables represent generic numbers obtained by averaging the data from multiple manufacturers.

17




7.2.2.9 Differing battery models in the same site Typically only one type of battery will exist in a specific location. So, only the calculations for that specific battery type need to be done. In other words, if the battery is a vented lead-calcium battery, perform the calculations only for that type. In facilities using multiple battery types located in areas with varying environmental conditions, each installation should be individually evaluated based on the battery type and the expected environmental conditions. In plants with mixed cell sizes or types (in parallel strings), the calculations should be done for each type/size, and the results added together.

7.2.3 Standby/float operation The float operating mode occurs when a constant voltage is applied to the battery terminals sufficient to maintain an approximately constant state of full charge. The standby float operating mode can be detected by observing the float/equalize mode of the chargers. Most batteries in stationary applications (with the exception of cycling batteries, such as PV/solar applications; or constant-current charging schemes) spend more than 99% of the time on float. For applications where normal cycling of batteries is expected, the heat generation calculations of Table 1 should be used in air-conditioning system sizing. The calculations in Annex B should be used for the following applications: a)

Where frequent cycling occurs

b)

Where frequent boost/equalize or finishing charges occur

c)

Where the dc system does not give the HVAC system a continuous indication of the dc system operating mode

Even though heat generation during discharge/recharge is higher, heat dissipation is much less than the generation due to the large thermal mass of batteries. Most discharges and recharges are short (except in cycling applications); so although the rate of heat generation on discharge or bulk recharge may be a significant number, the heat dissipated to the room is usually quite low (see 7.2.9 for further information on calculation of heat generation for cycling applications).

7.2.4 Accelerated charging mode In the accelerated charging mode, during recharge the voltage of the chargers is raised to an equalize (sometimes known as boost) level for a time period ranging from a few minutes to a few days, in order to more quickly replace the lost energy from the battery. This is not done for all types of applications. It is most common in cycling applications. It is least common in non-cycling applications using VRLA batteries. Accelerated recharge can be recognized by the charger being in the equalize mode, or by a bus voltage higher than the normal float voltage. See Table 2.

18




Table 1 —Standby/float operating mode Battery type

H 2 − rate = 1.27 × 10

Lead-calcium

Design parameter Toxic and corrosive gases Vented lead-acid

Flammable gases −7

Heat generation

× I × nc

where I = C8 × 9.76 × 10−5

a

or I = P15 × 0.025

H 2 − rate = 1.27 × 10 −7 × I × nc

Pure lead

where I = C8 × 6.62 × 10−5 H 2 − rate = 1.27 × 10 −7 × I × nc

Lead-selenium

qW = 0.264 × I × nc

None

where I = C8 × 3.27 × 10−4 or I = P15 × 0.0835 H 2 − rate = 1.27 × 10 −7 × I × nc

Lead-antimony

where I = C8 × 8.97 × 10 −4 or I = P15 × 0.230

Lead-tin or lead-calcium AGM Low-antimony AGM

Lead-calcium gel

qW = 0.34 × I × nc where I = C8 × 0.00121

H 2 − rate = nc × C8 × 2.69 × 10 −12

H 2− rate = nc × P15 × 6.73 × 10−10

or I = P15 × 0.326

−11

H 2 − rate = n c × C8 × 1.39 × 10

qW = 0.34 × I × nc

None

where I = C8 × 0.00626

H 2 − rate = nc × P15 × 3.48 × 10−9

or I = P15 × 1.69 qW = 0.34 × I × nc

H 2 − rate = nc × C8 × 6.72 × 10−12

where I = C8 × 6.1 × 10 −4

Nickel-cadmium

Vented NiCd

H 2 − rate = 1.27 × 10 −7 × I × nc −

None

qW = 0.122 × I × nc

where I = C5 × 5 × 10 4 where

I C8 P15 C5 H2–rate qW nc

is the current through each string in amperes (a) is the 8 h ampere-hour rating of a lead-acid cell to 1.75 V at 25 °C is the 15 min kW/cell rating of a lead-acid cell to 1.67 V at 25 °C is the 5 h ampere-hour rating of a NiCd cell to 1.0 V at 20 °C is the hydrogen gas release rate in m3/s at standard sea level atmospheric pressure and 25 °C is the total heat produced in watts is the number of cells in the plant

a

This formula is the best estimate from available data of gassing current when floating at an approximate average upper bound of 0.19 V/cell over open circuit. This covers the majority of the installed base applications, but there may be a few where the float voltage is even higher. In such cases, the formula from A.2.3.2 should be used.

19




Table 2 —Accelerated/boost/equalize charging mode

Battery type

Lead-calcium and pure lead

Design parameter Toxic and Flammable gases corrosive gases Vented lead-acid

Heat generation

H 2 − rate = 1.27 × 10 −7 × I × nc where I = C8 × 4.88 × 10−4 a or I = P15 × 0.125

H 2 − rate = 1.27 × 10 −7 × I × nc Lead-selenium

=

where I = C8 × 0.00175 or I = P15 × 0.448

None

qWh

× × 0.369 I

× nc

t

H 2 − rate = 1.27 × 10 −7 × I × nc Lead-antimony

where I = C8 × 0.00492 or I = P15 × 1.26 Valve-regulated lead acid

Lead-tin or lead-calcium AGM Low antimony AGM

Lead-calcium gel

qWh = 0.449 × I × nc × t

H2−rate = nc × C8 × 1.04 × 10−11

where I = C8 × 0.00252

H 2−rate = nc × P15 × 2.60 × 10−9

or I = P15 × 0.659

H 2− rate = nc × C8 × 5.36 × 10−11

None

H 2−rate = nc × P15 ×1.34 × 10−8

qWh = 0.449 × I × nc × t where I = C8 × 0.0130 or I = P15 × 3.41 qWh = 0.449 × I × nc × t

H 2 − rate = nc × C8 × 2.60 × 10−11

where I = C8 × 0.00126

Nickel-cadmium

I = C × 0.00305 Vented NiCd

5

None

H 2 − rate = 1.27 × 10 −7 × I × nc

qWh = 6.77 × 10 −4 × C5 × nc × t

where

qWh t

is the heat generated, in watthours (Wh) is the time, in hours (h)

a

This formula represents the best estimate from available data of gassing when the boost/equalize voltage is at an approximate upper bound of 0.33 V/cell over open circuit voltage. This covers 95% of the installed base, but there are some applications where the boost/equalize voltage may be even higher. In those cases, t he formula from A.2.3.2 should be used to calculate the higher gassing current

7.2.5 Discharge Discharge occurs whenever the batteries are called upon to deliver power to the loads, typically when ac power or other primary sources of power fail. Depending on the duration of the outage, and the design of the backup system, discharge can last from seconds to days. This mode is typically distinguished by a “battery on discharge” or “low-voltage” alarm for float applications (this is not true of cycling applications). Lead-acid batteries on discharge have a slightly endothermic chemical reaction, so the only heating is due to current through internal and intercell resistance. In a long duration discharge, the currents are relatively low, and therefore the endothermic nature of the chemical reaction outweighs the Joule heating from the I2R losses. Thus, the formulas in Table 3 are not provided for C8 rates, but only for high-rate batteries. Note that the internal resistance is not calculated for NiCd batteries, because the heating effects provided by that resistance are minimal, and are taken into account by the general equation provided.

20




Table 3 —Discharge mode

Battery type

Flammable gases

Design parameter Toxic and corrosive gases

Heat generation

qWh = ( I × t × nc ) × ((0.727 × ( s.g .)) + 0.0718 −

Lead-acid where I None or miniscule

None

=

Vc −eod ) 2

IL Ns

qWh = 0.285 × I × t × nc IL

Vented NiCd where I

= Ns

where

Ns IL s.g. Vc–eod

is the number of parallel strings feeding the load is the load current in amperes (A) is the nominal specific gravity (s.g.) of a fu lly charged cell is the average expected end-of-discharge voltage per cell at the battery plant

7.2.6 Bulk recharge Bulk recharge occurs when a primary source of power returns after a discharge, and high efficiency recharging occurs until the gassing stage is reached. Depending on the duration of the primary source outage, the size of the batteries being recharged, and the charging current available, bulk recharge can last from seconds to days. For float applications, this mode is typically distinguished by the time between the return of a primary power source, and the retirement of the battery on discharge or low-voltage alarm (this statement is not true for cycling applications). During most of the recharge, the system voltage will slowly rise from its initial value until the charger set point is reached. Once the critical minimum voltage for maintaining full charge is reached (see Annex A), the bulk recharge period ends and gassing may begin. At this point, the system is either in float/standby mode (see Table 1) or accelerated/finishing charge mode (see Table 2).

Table 4 —Bulk recharge mode

Battery type

Flammable gases

Design parameter Toxic and corrosive gases

qWh = Cr − s × nc × 0.14

Vented lead-acid None or miniscule

VRLA

Heat generation

None

qWh = C r − s × nc × 0.138 qWh = Cr − s × nc × − 0.00882

Vented NiCd where

Cr–s

is the ampere-hours (Ah) removed from the string during discharge that must be replaced (see B.2.3.1.4 for further information on how to determine this factor)

7.2.7 Equalize operating mode The equalize mode may be used to correct variations of charge among individual cells and/or to ensure that the entire battery is at a full state of charge. It is rarely entered into automatically (although some battery installations may be set up to do this). Usually it is a reactive solution to a low voltage or voltage imbalance problem detected by maintenance and/or monitoring activities. 21




Generally, high-rate VRLA batteries are not boost/equalize charged. However, some UPS and PV systems are setup to do this automatically, and if the function is not turned off, it will happen. The maintenance (equalize) mode can be recognized by the charger being visibly switched to the equalize mode or by a higher bus voltage than float. Refer to Table 2 for the calculations for this mode. System limits (due to maximum equipment operating voltages) may dictate equalize voltages at the low end or below manufacturer recommended values. If calculation simplicity is desired and/or the dc system does not give the HVAC system indications of the present operating mode, the gassing calculations from Table 2 can be used as a default (in fact, some jurisdictions require that this mode be used for the gassing calculations). Not only does this calculate gassing at equalize voltage levels, but takes into account the possibility of high charger voltage (malfunction or human error), or some partially or fully shorted cells in the string. Admittedly, the gassing will be higher than that computed on float, but will still be small enough that an increase in ventilation airflow should not be needed.

7.2.8 Initial charge mode An initial charge (also known as a commissioning charge) is only applied when the battery is first placed into service. It is normally performed at equalize or higher voltage potentials (Table 5 assumes the higher voltage in order to provide the worst-case gassing and heat generation rates), but normally runs for at least double the equalize time. Due to the extended time, temperatures tend to rise higher than during an equalize charge, so that total gassing at the end of the charging period is higher. If cells have been in storage for more than 6 months (depending on the temperature), the manufacturer typically recommends a freshening charge, which is similar to an initial charge. An initial charge for VLA batteries typically runs between 100 h to 250 h (it can be more than that— consult the battery engineer or installer) and is manually initiated. Initial charging of VRLA batteries should only be done if recommended by the manufacturer and should be limited to the time specified by the manufacturer (it is almost never done for high rate VRLA batteries, so no equations are included for VRLA batteries with those ratings in Table 5). NiCd commissioning charging is not as rare as a VRLA initial charge, and also does not last more than 48 h (note that NiCd manufacturers typically recommend an initial charge at constant current and Table 5 reflects this practice). The initial charge is a once-in-a-lifetime event for any given battery type, and so the ventilation system should not be designed to compensate for this mode of operation. Instead, the ventilation system should be designed for the other modes, and temporary ventilation may be needed during the initial charge mode. The requirement for additional ventilation should be evaluated as the existing ventilation may have the additional capacity, and temporary ventilation may not be necessary.

22




Table 5 —Initial/freshening charge mode

Battery type


Lead-calcium or pure lead

H 2 − rate = 1.27 × 10 −7 × I × nc where I = C8 × 0.00157

Lead-selenium


Heat generation

or I = P × 0.401 15

None

qWh = 0.539 × I × nc × t

or I = P15 × 0.689 Lead-antimony

H 2 − rate = 1.27 × 10 −7 × I × nc where I = C8 × 0.00893 or I = P15 × 2.29 Valve-regulated lead acid

Lead-tin or lead-calcium AGM Low antimony AGM Lead-calcium gel

qWh = 0.449 × I × nc × t

H2− rate = nc × C8 × 1.04 × 10−11

where I = C8 × 0.00252

H 2− rate = nc × P15 × 2.60 × 10−9

or I = P15 × 0.659

H 2 − rate = nc × C8 × 2.68 × 10−11

None

H 2−rate = nc × P15 × 6.71×10−9

qWh = 0.449 × I × nc × t where I = C8 × 0.00650 or I = P15 × 1.70

qWh = 0.449 × I × nc × t

H 2 − rate = nc × C8 × 2.60 × 10 −11

where I = C × 0.00126 8

Nickel-cadmium

Vented NiCd

H 2 − rate = 2.54 × 10

−8

× C5 × nc

None

qW = 0.0444 × C5 × nc

7.2.9 Cycling operating mode The cycling mode occurs when batteries are repeatedly discharged and recharged. During recharge in cyclical applications, an accelerated recharge voltage level is usually applied during the early stages of the recharge cycle, but then decreased towards a float level to minimize gassing. In many cycling applications (e.g., PV), the battery may rarely get fully recharged. Batteries in cycling applications are usually in a discharge or recharge mode, and if the primary source is there for extended periods of time, they may be in a finish charge or float charge mode. The charging system should provide indications of the current mode of operation. The cycling mode is determined by the design of the installation. Refer to Table 2, Table 3, and Table 4 for the calculations for the various modes during cycling operation.

7.2.10 Failure modes (abnormal operation) The normal HVAC system does not need to accommodate the heat or gases generated during abnormal operation of the batteries. NFPA 1 (Fire Code), NFPA 90A, the International Fire Code [B27], and other

23




regulatory documents provide guidance for battery fires and extreme failure modes such as thermal runaway.

7.2.10.1 Thermal runaway Thermal runaway is described in 6.3.4. As noted there, thermal runaway should be a rare event. It is more commonly associated with VRLA cells, older lead-antimony vented cells, and high-temperature environments, but can occur in almost any cell. During thermal runaway, higher currents are drawn through the battery in float or equalize mode, thereby increasing gassing. Thermal runaway can be detected by higher than normal float currents, and/or by battery temperatures significantly higher than ambient. See Table 6.

Table 6 —Thermal runaway

Battery type

Pure lead or lead-calcium or lead-selenium Lead-antimony

Flammable gases

Design parameter Toxic and corrosive gases Vented lead-acid

Heat generation

These cells rarely go into thermal runaway without a manufacturing defect or multiple cell failures (the design criteria of Table 2 adequately cover this slight possibility).


Hydrogen sulfide arsine stibine

or I = P15 × 5.54

qW = 0.499 × I × nc

Valve-regulated lead acid

Lead-tin or lead-calcium

H 2 − rate = nc × C8 × 1.33 × 10−10

AGM

H 2 − rate = nc × P15 × 3.33 × 10

Low antimony AGM Lead-calcium gel

Vented NiCd

qW = 0.455 × I × nc where I = C8 × 0.00498

−8

Hydrogen sulfide

H 2− rate = nc × C8 × 6.82 × 10−10

H 2 − rate = nc × P15 × 1.72 × 10−7

or I = P15 × 1.29 qW = 0.455 × I × nc where I = C8 × 0.0257 or I = P15 × 6.67

qW = 0.455 × I × nc

H 2 − rate = nc × C8 × 3.32 × 10 −10

where I = C8 × 0.0025

Nickel-cadmium It is not normally possible to drive these batteries into thermal runaway unless maintenance (water additions) is neglected and the battery is severely overcharged.

7.2.10.2 Shorted cells With fewer than 10% shorted (or partially shorted) cells in the string, gassing response is similar to equalize mode. With a higher percentage of shorted cells per string, the battery can be driven into thermal runaway.

7.2.10.3 Cell reversal Due to over discharge or incorrect installation where the polarities on a cell are connected backwards in a series string, individual cells can be driven into cell reversal (where the positive becomes the negative and vice versa). This will cause extreme heating and gassing. While it is impractical to design for this

24




consideration, Table 7 is provided to show the amounts of gas and heat that may be produced as an upper bound if all cells are in reversal.

Table 7 —Cell reversal

Battery type

Lead-calcium, pure lead, and lead-selenium Lead-antimony


H 2 − rate = 1.27 × 10 −7 × I × nc where I

I ct − spare

Hydrogensulfide

I ct − spare

=

Heat generation

Ns

Hydrogensulfide arsine stibine

⎛ n ⎞ =⎜ Ic ⎟ − I L ⎜ ⎟ ⎝ c =1 ⎠

∑

qW = 0 .847 × I × nc

Valve-regulated lead acid

H 2 − rate = 1.27 × 10 −7 × I × nc where I

=

I ct − spare Hydrogensulfide

Ns

VRLA

⎛ ⎞ = ⎜∑ Ic ⎟ − IL ⎝ c =1 ⎠

qW = 0 .847 × I × nc

n

I ct − spare

Nickel-cadmium

H 2 − rate = 1.27 × 10 −7 × I × nc

I where I Vented NiCd

I ct − spare

=

ct − spare Ns

None

qW = 0.468 × I × nc

⎛ n ⎞ =⎜ Ic ⎟ − I L ⎜ ⎟ ⎝ c =1 ⎠

∑

where

Ict–spare Ic IL

is the maximum spare charging capacity available to charge the batteries is the capacity (in amperes) of an individual charger is the average load, or the load over a given time period

7.2.10.4 Charger runaway A battery charger/rectifier can “fail” into a high-voltage mode that is potentially harmful to the batteries and any paralleled equipment. Most chargers/rectifiers are equipped with alarms to alert an operator either locally and/or remotely of this condition. In addition, many charger/rectifiers and/or dc plant controllers are equipped with high-voltage shutdown (HVSD) circuitry to minimize the damage. The charger manufacturer should be consulted for default settings for both alarms and HVSD, and these settings should be made with the recommendations of the battery manufacturer and any connected equipment in mind. Using Annex A or Annex B, or other data from the battery manufacturer, gassing and heating calculations can be made based on the expected worst-case high-voltage “failure” of the charger/rectifier(s). Based on typical HVSD settings, Table 5 could be used as a “worst-case” estimate of gassing and heat generation during charger runaway for most vented batteries, and Table 6 could be used for the same purposes for VRLA batteries and vented lead-antimony batteries.

25




7.3 Heating, ventilating, and air-conditioning system design requirements

7.3.1 General Heating, ventilating, and air-conditioning (HVAC) systems for stationary battery installations should include features to address the design considerations discussed in Clause 5 and Clause 6. For safety, systems should be designed to maintain the hydrogen gas concentration at less than the design limit concentration in all areas within the ventilated space. The HVAC system designer should select margins appropriate for the installation. Refer to Annex C for a discussion of codes, standards, and NFPA guidance related to design margins. For optimum battery performance, a properly designed HVAC system should maintain an acceptable thermal environment and be credited with the management of potentially hazardous gases, which are sometimes released from the batteries. The following are considerations to prevent or mitigate thermal consequences described in 6.2.1:

⎯

Spacing between units: Maintain a minimum spacing between battery units/cells of at least 1 cm (0.2 in) or in accordance with manufacturer’s recommendations.

⎯

Mechanical remedies: Do not direct heating or cooling airflow directly on the battery to minimize temperature gradients.

⎯

Seismic blockage: Check for free flow of air. In a seismic installation, supports or seismic padding installed between the battery unit/cells are designed to interfere as little as possible with airflow, but can sometimes restrict the flow of air needed to remove heat if they are not properly installed.

⎯

Temperature gradients: Control the airflow pattern in the room to limit temperature gradient in the battery electrolyte temperatures.

⎯

External temperature sources: Do not place batteries above or adjacent to heat generating sources. When sharing enclosures with electronic equipment, place batteries below the electronics. Shield the battery compartment from radiated heat sources (such as the sun, high temperature piping, etc.).

7.3.2 Systems for heating and cooling

7.3.2.1 Battery installation heat sources As described in Annex B, batteries generate heat both during charge and discharge. Providing adequate cooling is important for maintaining acceptable battery electrolyte temperature. Cooling requirements for the different battery installations are as follows:

⎯

Integrated battery and electrical equipment room. If the batteries share the same space with the battery charger, the heat released from the batteries will be small in comparison to that of the battery charger. During normal operation, peak heat release from the batteries occurs during the discharge mode, but during this mode the charger is de-energized and thus producing no heat. For these spaces, the heat generation by the batteries can usually be ignored if the battery charger heat

⎯

Dedicated battery room or outdoor battery enclosures. When the batteries are isolated in a separate room or enclosure providing adequate cooling is important. Heat degradation occurs over a long period of time (typically months or years based on the actual operating temperature). Although normal operation maximum heat generation occurs during a discharge event, it is

generation is considered.

26




typically short in duration and mechanical cooling is usually not available during the event. Elevated temperatures that may occur during a discharge are not an issue. Cooling should be restored when recharging is commenced.

7.3.2.2 Active heating and cooling systems Battery rooms and enclosures may be provided with active cooling and heating systems. Active systems include a broad range of air-conditioning system equipment types including units with cooling and heating coils. The cooling coil can be the direct expansion (refrigerant) type or one served by a chilled water system. The heating coil can be served by sources such as a hot water or steam system or a low-power density (preferably about, 9 W/in2) electric heating coil.

7.3.2.3 Passive cooling Passive cooling systems are recommended for enclosures located outdoors and in remote locations where routine maintenance of mechanically driven systems is deemed impractical. Three primary methods for passive cooling are (1) natural air circulation, (2) radiant heat transfer, and (3) wind. Natural cooling systems employ thermal buoyancy to induce air movement through the enclosure. Radiant cooling systems utilize closed loops with natural circulation of their coolant or direct conduction of heat to outside the enclosure. Wind systems rely on the prevalence of naturally occurring winds to drive outside air through the enclosure.

7.3.2.4 Radiant cooling Thermal radiation is the transfer of heat energy in the electromagnetic spectrum through space. Radiant heat transfer occurs when there is a temperature differential between two surfaces. As a result, both surface temperatures will attempt to equalize. Radiant energy travels through space without heating the space itself. It only turns into heat when it contacts a cooler surface. Radiant heat transfer requires direct line of sight between the surfaces. For radiant battery cooling systems, the warmer radiates energypanels, to the cooler radiant panel. As long as there is a temperature difference betweenbattery the battery and heat the radiant heat transfer will occur. The design of a radiant cooling system must include means of rejecting heat from the radiant panels and consider the potential for the formation of condensation on the radiant panels. Radiant cooling method is an inefficient method of cooling used only to cool batteries and not the environment. The conductive media can be in the form of a fluid or conductive material. A liquid or air-cooled panel adjacent to the battery system components can be used to transfer heat away from the batteries. The removal of the heat from the space could be by natural circulation of the cooling medium to a heat sink. Also, mechanical cooling could be employed by use of a chilled water system. The design should consider the potential for condensation on the cooling panels in the vicinity of the batteries or the supporting electrical systems. The design features of the battery enclosure may be configured to achieve the thermal design objective. This would be accomplished with thermal insulation and the arrangement of the enclosure surfaces exposed to the environment. Pre-engineered passively cooled/heated enclosures are sometimes used for this type of application. Some installations may provide features that include radiant heat transfer characteristics relevant to the overall HVAC system design. Walls that provide large thermal mass for example may permit a smaller installed cooling capacity for transient thermal events.

27




7.4 HVAC system design for ventilation

7.4.1 General The primary focus of ventilation as it relates to batteries is upon the dilution of the hydrogen gas that the batteries release as a normal part of their operation. Hydrogen gas will disperse quickly throughout the room, but to promote mixing and to speed the dispersion of the hydrogen gas being evolved from the batteries, ventilation or airflow is necessary. The recommended maximum average concentration in the room should be less than two percent (<2%) by volume. The ventilation system can be natural or forced, and the selection of the design type will depend on the application.

7.4.2 Natural ventilation Airflow in a natural ventilation system is based on (1) thermally induced or buoyant flow based on temperature gradients within the enclosure or (2) wind induced ventilation. Battery ventilation systems that rely solely on natural wind energy are not recommended due to the inconsistent nature of wind to be present. In thermally induced ventilation systems it is essential that a heat source be present in the room. When batteries share a common enclosure with the equipment they serve, it is the heat release from the equipment that provides the driving force for the air movement. The natural ventilation system design requires a low air intake opening and a high exhaust opening to enable cold air to enter at the bottom of an enclosure and hot air to exit near the top of the enclosure. Heat generation within the enclosure is credited with generating the buoyant forces to induce airflow within the enclosure. The induced airflow assists in the removal of hydrogen that might be generated if any battery is located in the enclosure. To keep the enclosure free of pests, insect screens are usually installed at the openings. In dedicated battery cabinets, the heat generated from the battery system is not, by itself, sufficient to drive airflow within the enclosure. Such cabinets rely upon relatively large intake and exhaust air openings. Provided these openings remain open, any evolved hydrogen gas will readily be carried away by the induced airflow. Even without substantial induced airflow, the evolved hydrogen will quickly diffuse out the openings. If the openings are equipped with some form of damper, then limits or controls should be installed to assure adequate opportunity for hydrogen to diffuse to the outside. It should be noted that battery installations with a natural ventilation system may result in elevated temperatures in the room that may cause the battery electrolyte temperature to also increase. Even though the elevated temperature in the room may enhance the performance or the batteries, the service life of the batteries is compromised. The high and low exhaust openings are sized as shown in Equation (8):

Q = q / ((ti – to ) × 0.24 × air density (lb/ft 3) × 60)

(8)

where

Q is the induced airflow (cfm) q is the heat generated in the enclosure (BTU/h) ti is the inlet temperature (°F) to is the outlet temperature (°F) The air opening size may be obtained from Equation (9):

28




A = Q /(60 × Cd × 2 × g × hNPL × (Ti − To) / Ti

(9)

where

A is the area in square feet Q is the airflow rate (cfm) g is the acceleration of gravity (ft/sec2) hNPL is assumed to be 0.5 times the vertical distance between the inlet and outlet openings in feet Ti is the absolute inlet temp (°R) To is the absolute outlet temp (°R) Cd = discharge coefficient varying between 0.42 for one opening and 0.65 for multiple inlet openings During periods of hot weather, these systems loose the ability to maintain battery electrolyte temperature at their ideal temperature. The associated loss of battery life is therefore being sacrificed in exchange for the reduction in cooling system maintenance. Typically installations of this type of ventilation scheme involve common enclosures for both the batteries and the equipment they serve. Due to the unreliability of the wind to provide dilution of gases, it is not recommended as the sole source for battery area ventilation.

7.4.3 Active or forced ventilation system Forced ventilation systems use fans to mechanically move air into or out of the battery room or enclosure. The forced ventilation system usually employs a supply or an exhaust fan or a combination of the two along distribution ductwork as needed. The supply fan can be used to bring conditioned or tempered air into the battery room or enclosure. The exhaust fan removes air from the room or cabinet to the outside to maintain an acceptable hydrogen concentration level. A battery room or enclosure is often held at a slightly negative pressure relative to surrounding spaces so as to minimize exfiltration of air from the battery area to the adjoining areas. The slight negative pressure in the room will draw in air from the surroundings, and this along with any supply air to the room will provide dilution of contaminants within the room. The supply air can be conditioned, tempered, and treated as needed to control room temperatures and satisfy requirements stated in Clause 6 and in Table 2 through Table 7. System capabilities would typically be designed for the worst-case tabulated generation rates for the expected battery types and operating modes. To meet the project cost objectives (see Clause 8), controls may be used to operate the ventilation system in operating modes to match the battery operational modes. Exhausting air from the battery enclosure or room to outdoors is the preferred method for removal of battery generated gases. If the makeup air is potentially contaminated or corrosive, an acceptable method for ventilating such spaces is to recirculate the air but mixed with some filtered outside air for dilution. In a recirculation type of design, the facility owner would be required to regularly test to ensure adequate outside air supply is available for dilution. In a battery enclosure or room with an active ventilation system, the air will be well mixed, in which case it would not be necessary to provide exhausts at highest points. However, exhausting air from a high point in the room will better ensure the hottest air is removed, and may enhance removal of hydrogen gas in the absence of any air movement in the space.

29




7.5 Integrated battery areas

7.5.1 Integrated battery and charger/rectifier/inverter room The ventilation rate for these spaces is principally driven by the need to dilute any hydrogen gas evolved from the batteries. The peak rate of hydrogen evolution occurs during the initial charge mode. If special procedures or addition of temporary ventilation have been established for dealing with peak hydrogen evolution during initial charge mode, then the rate of hydrogen evolution during accelerated charging and maintenance (equalize) operating modes could be considered for use as a design basis. For systems with direct outside air supply the supply of outside air should be sufficient to dilute the hydrogen gas being evolved by the batteries. The HVAC system could potentially be interlocked with the charger to adjust the flow of outside air based on the mode of operation of the battery. For systems with a centralized ventilation system the supply of air should be continuous and mixed with sufficient outside air necessary for maintaining hydrogen concentration within acceptable limits.

7.5.2 Integrated battery and equipment areas Facilities sometimes rely on air conditioning for comfort cooling to also provide the equipment cooling requirement. More often, these applications require supplemental air handling to circulate air and to maintain temperature, humidity, and dust filtration within the range required for electronic equipment. See Annex C for a discussion of requirements that may relate to specific applications.

7.6 Controls and alarms

7.6.1 General Battery ventilation systems should operate continuously to enable dilution of generated gases. The alarm features would be dependent on (1) whether the system has multiple operating modes and (2) whether system operators would normally be present at the facility. Controls for operating modes may be manual or automatic to be consistent with the operation of the battery electrical systems as long as the design accounts for the thermal performance of the batteries and safety. Use of two speed or variable speed exhaust fans can often provide the additional flow of air when the battery operation is other than the float mode. If the charging equipment is fitted with controls to detect and prevent conditions that lead to excessive gas generation, supplemental ventilation for battery systems may not be necessary. Such controls can include the following:

⎯

Charge current limiting

⎯

Temperature compensated charging (lead-acid batteries)

⎯

High-voltage shutdowns

Alarm functions and switching to backup ventilation components would depend on operator availability, component redundancy, battery charging cycle duration and frequency, and ventilation dilution rate margins. Normally alarms are not required; however, when the batteries are utilized to provide backup power to data centers or other mission critical facilities, ventilation fan failure (loss of airflow) may be alarmed locally (audible or visual) based on the following typical scenario:

30




⎯

Maintenance personnel in the area of the batteries on a daily basis or several times a week.

⎯

Batteries infrequently in the maximum gas generation mode.

⎯

Multiple fans serving the area, either a supply and an exhaust fan combination or two 50% capacity exhaust fans.

⎯

Relatively large room volume compared to the gas generation capacity of the battery.

Automatic change over to redundant components, remote failure alarms, and automatic system shutdown would not normally be required.

7.6.2 Sensors Hydrogen and other gas sensors are not required to maximize operational safety of battery installations designed with natural or forced ventilation systems meeting the design basis dilution and reliability criteria (see 7.4). Hydrogen sensors may be used as a supplemental monitor but are not a substitute for dilution ventilation. If a decision is made to use gas sensors, the user is encouraged to consult with experts in their selection and placement and to enforce disciplined maintenance practices. Sensors require frequent maintenance and calibration, typically on less than a one year interval and periodic replacement. If regular maintenance and replacement, in accordance with the manufacturer’s recommendations, cannot be assured, hydrogen sensors should not be used. A risk analysis is recommended to provide technical data that would warrant the installation of sensors.

7.6.3 Reliability/redundancy Adequate redundancy would be provided by a system with two fans, either a supply and an exhaust fan or two half-capacity exhaust fans. Also a multispeed exhaust fan would provide a level of redundancy. For remote locations, the HVAC systems should be inspected on a regular basis and alarmed for system failure or trouble. Direct drive components should be used when practical to minimize the potential for belt failure.

7.7 Battery room hazard classification Buildings or portions thereof containing battery systems should not be classified as hazardous locations as long as they meet the minimum requirements for natural or forced ventilation and other safety features normally recommended by installation standards or guidelines for battery systems.

7.8 Enclosure design applications

7.8.1 General Cabinet, vault, or closet installations using VRLA batteries should be provided with a forced air ventilation system or with sufficient vent openings to allow natural ventilation such that the rise in temperature within the cabinet is less than or equal to 2 °C above ambient. This is due to the following facts:

31




⎯

VRLA batteries generate significantly more heat than vented batteries.

⎯

VRLA batteries are much more sensitive to thermal damage than vented batteries.

⎯

Elevated temperatures can be a contributing factor to a thermal runaway event in a VRLA battery.

7.8.2 Indoor cabinets Cabinets should be designed to prevent the accumulation of hydrogen gas pockets. Where access is not controlled, cabinets should have lockable doors. A cabinet ventilation system can be a natural or a forced air system. However, a forced ventilation system may be necessary for both temperature control and gas dissipation. See Annex C for additional discussion.

7.8.3 Outdoor cabinets Some outdoor cabinets are provided with air conditioning. Unconditioned cabinets should be louvered to allow natural ventilation sufficient to remove heat and prevent accumulation of hydrogen gas pockets while at the same time preventing intrusion by insects, vermin, rain, snow, and dirt. Outdoor ambient temperature extremes, radiant heating, filter replacement intervals, and positioning of the cabinet all affect the performance and service life of the batteries.

7.8.4 Vaults Batteries are sometimes located in equipment vaults. The batteries are normally installed on relay racks and collocated with other electrical and electronic equipment. Since vaults may be partially or completely below ground level, air exchanges are more difficult and ventilation requirements are more important. Natural convection is frequently not possible. The thermal and ventilation environment in this type housing may be maintained by a thermostatically controlled HVAC system.

8. Economics

8.1 General There are many factors to consider for determining the most cost-effective battery installation. The cost analysis should consider the initial capital cost, maintenance and operating cost for not only the batteries, but also the ventilation and thermal management systems. The lowest initial cost may meet the project financial objectives but it may not be the best economic decision. It may be necessary to perform a life-cycle cost study based on the safety, performance and projected service life of the battery installation, and the type of HVAC systems to meet some or all of those objectives. The battery installation with a simple ventilation system may provide the lowest initial cost but not the optimum service life of the batteries that could be expected with an enhanced HVAC system. Comparing the additional cost of the enhanced HVAC system that will result in increased service life of the batteries to the replacement cost of the batteries for the life of the facility may provide the necessary justification for an increase in the initial cost for that battery installation. Alternatively, a similar life-cycle cost approach may be used in the consideration of a more temperatureresistant battery technology with a simpler ventilation system. The following summarizes the factors to consider in the evaluation.

32




8.2 Battery replacement factors In particular, lead-acid battery life is significantly impacted by operating temperature. See 6.2.1.1 for a discussion of the effect of temperature on battery life and capacity. Labor costs for replacing a large stationary battery can be significant and should be considered. Another consideration is the lead time for obtaining a replacement battery. Lead times for obtaining replacement batteries can be weeks or even months for specialized applications.

8.3 Relative importance of the installation By definition, all battery installations are important. However, consideration should be given to the risk involved in failure of the battery system. As an example, a relatively inexpensive battery system that can have significant impact on economic or human risk may be able to justify significant expenditure on an HVAC system to achieve a high degree of reliability of the battery system.

8.4 Reliability of the HVAC system As the complexity of the HVAC system increases or the operating environment worsens, the maintenance activities required to keep the HVAC system functional may exceed the value added of the system. In particular, direct expansion cooling systems in coastal areas are often subjected to the effect of corrosion due to the presence of salt in the air, and consequential rapid failure of outdoor condensing units.

8.5 Availability of maintenance resources All HVAC systems require periodic maintenance. Consideration should be given to the availability of the personnel to perform routine maintenance activities.

8.6 Cost and availability of battery replacement HVAC system installation and maintenance costs should be weighed against the battery replacement cost over the life of the installation.

8.7 HVAC System control based on battery operating mode Using the various operating modes presented in Clause 7, an HVAC system could be designed to maximize facility safety as well as operation and control strategies to meet other objective such as reduced capital and operating costs and energy conservation. This would be accomplished by integrating the HVAC system controls with the battery system controls to adjust the HVAC system to meet the modes of operation of the battery system. For example, the airflow can be increased by the use of two speed or variable speed exhaust fans to meet the demand for additional airflow for changeover of battery operation mode from float mode (normal operation) to the equalize charging mode.

33




9. Environmental management (operation and maintenance)

9.1 Battery system operation and maintenance Batteries and battery systems should be maintained and operated in accordance with the following IEEE guides and standards:

⎯

IEEE Std 450 for VLA batteries

⎯

IEEE Std 1106 for NiCd batteries

⎯

IEEE Std 1188 for VRLA batteries

⎯

IEEE Std 937 for lead batteries in PV applications

⎯

IEEE Std 1561 for hybrid PV applications

9.2 HVAC system operation and maintenance HVAC systems should be operated and maintained to assure that the design objectives of dilution ventilation and thermal management are provided throughout the life of the facility. The following are minimal recommendations; other actions may be required consistent with the complexity and reliability objectives of the system.

9.2.1 Operation Battery room/area HVAC systems should be operated in accordance with documented procedures. The facility operator should be adequately trained and be aware of the implication of taking the battery HVAC system out of service for maintenance activities. The operator training should be updated on a regular interval and the training should address the following as a minimum:

⎯

How to optimize operating temperature ranges for the batteries installed in the space.

⎯

When and how to monitor battery room/area temperature.

⎯

Appropriate actions to take in response to alarms, in particular high and low temperatures and ventilation system failure alarms.

⎯

How to address heating and ventilating consequences of abnormally frequent starting and stopping of equipment.

⎯

How to coordinate HVAC maintenance and battery charging activities (e.g., avoid equalize charging while the HVAC system is inoperable).

⎯

How to manage charging functions to prevent hydrogen evolution during periods of ventilation system maintenance outages.

⎯

How to current limit charger(s) during ventilation system maintenance to prevent hydrogen evolution. If the ventilation system is off or is operating at a reduced capacity during maintenance, hydrogen evolution must be minimized but not at the risk to the critical connected load.

9.2.2 Maintenance The following are recommended maintenance activities and intervals. Operating experience and system reliability objectives may dictate more frequent maintenance.

34




⎯

⎯

Monthly

⎯

Inspect air filters and clean as necessary

⎯

Test the annunciation and automatic switching functions

⎯

Check refrigerant pressures of air-conditioning refrigeration system as needed

⎯

Regularly test to ensure adequate fresh air supply is available for dilution

Annually

⎯

Inspect and clean inlet and exhaust air openings

⎯

Perform lubrication and inspect belts and pulleys

⎯

Calibrate instruments and controls

35




Annex A

(informative) Hydrogen generation in lead-acid and nickel-cadmium batteries A.1 Purpose This annex exists to show how the gassing equations in the tables of 7.2 were derived, in addition to providing sample calculations. This annex also illustrates that the calculated approach results in significantly lower ventilation requirements than typical rule-of-thumb or default solutions.

A.2 Gassing equations for lead-acid batteries

A.2.1 General Not all the current flowing in a battery electrolyzes water. Battery manufacturer documentation and experiments show this. With a fully or partially discharged battery, almost all of the current entering the battery returns charge (energy) into the battery. With an only slightly discharged battery, some of the current goes into energy return, and some drives side reactions. Once the battery achieves full charge, there is still some slight self-discharge due to local action at the plate surfaces; so even on “float” a slight amount of current returns energy to the battery. However, on float, the majority of the current in the battery is driving side reactions, principally the electrolysis of water. For a constant voltage charging system, the charge voltage on each cell must be a certain amount over the open-circuit voltage to overcome the effects of self-discharge. Current provided beyond that level (typically about 2.12 V/cell for a 1.215 s.g. lead-acid battery), due to a constant-current charger or by raising the voltage, predominantly electrolyzes water. Actualeither float voltages are slightly higher than this minimum because cell life is optimized when the positive plate voltage (polarization) is at a higher value. This higher voltage also ensures that each cell in a string has at least the minimum voltage given that there are slight differences between cells. This means, that in a float world application (constant voltage plants), water is always being electrolyzed. The following equations for computing actual gas generation from a lead-acid battery are culled from IEEE documents, ANSI documents, Telcordia® documents, and battery manufacturers’ documentation. 14 It is always best to get the current at a given voltage level from the manufacturer, given the slight differences in chemistries, manufacturing techniques, etc.; however, when that is not possible, the current equation of this and the following subclauses give a reasonable upper bound approximation of the current flowing through the battery. Hydrogen evolution (in m3 per second at standard atmospheric pressure at sea level at 25 °C) is shown in Equation (A.1) (see EnerSys [B20]):

H 2 − rate @ sea lvl = 1.27 × 10−7 × I g × nc

14

(A.1)

Telcordia is a registered trademark of Telcordia Technologies, Inc.

36




where

Ig nc

is the gassing current (not necessarily the same as the charge current 15) per string is the is the total number of cells in the battery plant

Typically, ventilation fan sizes in the U.S. are rated in cubic feet per minute (cfm). Translating Equation (A.1) to cfm yields Equation (A.2):

H 2 − rate @ sea lvl ( cfm) = 0.000269 × I g × nc

(A.2)

The volume occupied by the hydrogen gas released will change with pressure (which varies slightly with altitude) and temperature. Standard thermodynamic equations apply (see Portland State Aerospace Society [B33]). For purposes of this document, those changes do not need to be considered, as the HVAC engineer will take those into account. Changes in gas produced due to temperature (due to the rise in current) are taken into account in this document (see A.2.5).

A.2.2 Gassing of vented cells w ith constant-current charging For constant-current charging (fairly uncommon in standby applications), the gassing calculations become very easy for vented cells where all the cells are of the same size. Simply take the constant-current output of the charger, divide it by the number of parallel strings (if there is more than 1 string), then plug that result into Equation (A.1) or Equation (A.2).

A.2.3 Equations for lead-calcium and pure lead vented batteries

A.2.3.1 General The gassing current is a function of the voltage applied to the cell that can be expressed in terms of the average cell voltage in relation to the specific gravity of the cell, which is as follows:

Δsg Δsg Δ sg =

is the differential between the average cell voltage and the fully charged specific gravity of the cells (ΔVcell–sg might be a more correct variable to describe this, but the variable has been shortened for ease of application) is computed as shown in Equation (A.3)

Vp nC NS

− s.g. = Vc − avg − s.g .

(A.3)

where

Vp is charging bus voltage imposed on the string(s) NS is the number of parallel strings Vc–avg is the average voltage imposed across each cell

15

During bulk recharge, all of the charging current is going into material conversion (recharging the battery) and a few other reactions, but none of it is going into gassing. As the cell becomes more fully charged, the gassing current becomes a greater and greater percentage of the t otal charge current. On float charge, or boost/equalize charging, or constant-current charge mode once the ce ll is fully charged, almost all of the charge current is gassing current, with a small amount overcoming the effects of self-discharge.

37




Due to cell type and manufacturing procedures, the gassing current may actually be less than that predicted by Equation (A.4). Equation (A.4), and the others that follow, are dependent on the difference between float voltage and specific gravity and represent an upper bound. Assuming ambient conditions of 25 °C, and a constant voltage charger operating in float mode, the gassing current for pure lead or lead-calcium vented batteries is (as long as Δsg > 0.922 if not, gassing current is 0) (see EnerSys [B20]). 16

I g 25 Pb / Ca = C8 × 10−9 × e

(11.1 × Δ sg )

(A.4)

where

Ig25Pb/Ca is the vented lead-calcium electrolyzing (gas-producing) current (in amperes) at 25 °C C8 is the 8 h ampere-hour rating of the battery to 1.75 V/cell Or as shown in Equation (A.5):

I g 25 Pb / Ca = P15 × 2.56 × 10 −7 × e

(11.1 × Δ sg )

(A.5)

where

P15

is the kW/cell rating (not the watt/cell rating) of the battery at the 15 min rate to 1.67 V/cell

A.2.3.2 Current equation for vented lead-calcium cells Equation (A.4) and Equation (A.5) can be further simplified for inclusion in the tables. The upper bound of the current on float would be dependent on the highest expected difference between the average cell float voltage and average fully charged cell specific gravity. The highest typical expected difference would be 1.035 (based on an average utility cell float voltage of 2.25 V for a 1.215 s.g. vented lead-calcium cell). Plugging this differential into Equation (A.4) and Equation (A.5) yields the results shown in Equation (A.6) and Equation (A.7) (which are the equations found in Table 1 for this cell type):

I g 25Ca = C8 × 10−9 × e

(11.1 × Δ sg )

I g 25Ca = P15 × 2.56 × 10−7 × e

= C8 × 10−9 × e(11.1 × 1.035) = C8 × 9.76 × 10−5

(11.1 × Δ sg )

= P15 × 2.56 × 10−7 × e(11.1 × 1.035) = P15 × 0.0250

(A.6)

(A.7)

Similarly the upper bound of the current on boost, equalize, or finish charge would be dependent on the highest expected difference between the average cell boost/equalize/finish voltage and the average fully charged cell specific gravity. The highest expected difference would be 1.180 (for example, an average equalize voltage of 2.43 V/cell on a 1.250 s.g. vented lead-calcium or pure lead cell). Plugging this differential into Equation (A.4) and Equation (A.5) yields the results shown in Equation (A.8) and Equation (A.9) (which are the equations found in Table 2 for this cell type and for pure lead vented cells):

I g 25Ca / Pb = C8 × 10−9 × e

(11.1 × Δ sg )

= C8 × 10−9 × e(11.1 × 1.18) = C8 × 4.88 × 10−4

(A.8)

16

Equation (A.3) and the equations developed from it do not appear directly in the reference document, but are derived from the gassing data in EnerSys [B20] and UL 94 [B38].

38




I g 25 Ca / Pb = P15 × 2.56 × 10 − 7 × e

(11 .1 × Δ sg )

= P15 × 2.56 × 10 − 7 × e (11 .1 × 1.18 ) = P15 × 0 .125

(A.9)

For some manufacturers, there is no difference between a boost, equalize, or finishing charge and an “initial” charge. However, some manufacturers allow and/or require initial charges (typically for 100 h to 250 h) at higher voltages than even a boost, equalize, or finishing charge in order to finish formation of the plates (which was mostly done in the factory). The current and gassing in this charge would be dependent on the highest expected difference between the average cell initial charge voltage and the average fully charged cell specific gravity. The highest expected typical difference would be 1.285 (for example, an average initial charge voltage of 2.50 V/cell on a 1.215 s.g. vented lead-calcium or pure lead cell). Plugging this differential into Equation (A.4) and Equation (A.5) yields the results shown in Equation (A.10) and Equation (A.11) (which are the equations found in Table 5 for these cell types):

I g 25Ca / Pb − initial = C8 × 10−9 × e

(11.1 × Δ sg )

I g 25Ca / Pb − initial = P15 × 2.56 × 10 −7 × e

= C8 × 10−9 × e (11.1 × 1.285) = C8 × 0.00157

(11 .1 × Δ sg )

= P15 × 2.56 × 10 −7 × e (11.1 × 1.285) = P15 × 0.401

(A.10) (A.11)

A.2.3.3 Current equation for vented pure lead cells As with the upper-bound equation simplification performed for lead-calcium vented cells, the same can be done for pure lead cells. The upper bound of the current on float would be dependent on the highest expected difference between the average cell float voltage and average fully charged cell specific gravity. The highest expected difference would be 1.000 (for example, an average cell float voltage of 2.25 V/cell on a 1.250 s.g. vented pure lead Planté cell). Plugging this differential into Equation (A.4) and Equation (A.5) yields the results shown in Equation (A.12) (which is the equation found in Table 1 for this cell type):

I g 25 Pb = C8 × 10−9 × e

(11.1 × Δ sg )

= C8 × 10−9 × e(11.1 × 1.000) = C8 × 6.62 × 10−5

(A.12)

Note that no equation is developed for UPS high-rate applications because pure lead-vented batteries are seldom (if ever) used in that type of 15 min or less high-rate application, and if they are, these batteries will also have a C8 rating that can use Equation (A.11).

A.2.4 Equations for lead-antimony and lead-selenium vented batteries

A.2.4.1 General Assuming ambient conditions of 25 °C and a constant voltage charger, the gassing current for relatively new lead-antimony cells is (as long as Δsg > 0.927; if not, gassing current is 0) as shown in Equation (A.13) or Equation (A.14) (see EnerSys [B20]):

I g 25Sb − new = C8 × 3 × 10 −9 × e

(11.6 × Δ sg )

(A.13)

I g 25Sb − new = P15 × 0.000000768 × e (11.6 × Δ sg )

(A.14)

where

Ig25Sb–new

is the electrolyzing current (in amperes) at 25 °C for new vented lead-antimony cells

39




Lead-selenium batteries have a much lower antimony content (typically less than 3% of the plate grid metal, while most standard antimony designs are between 4% and 8%). The lower antimony content yields a lower float current. When the lead-selenium design is new, the float current can be approximated by Equation (A.15) or Equation (A.16) (see VARTA [B39]):

I g 25Se − new = C8 × 2 × 10 −9 × e

(11.2 × Δ sg )

I g 25Se − new = P15 × 0.000000512 × e

(A.15)

(11.2 × Δ sg )

(A.16)

where

Ig25Se–new

is the electrolyzing current (in amperes) at 25 °C for new vented lead-selenium cells

For lead-antimony batteries near the end of life, see Equation (A.17) or Equation (A.18) (EnerSys [B20]):

I g 25Sb − old = C8 × 2 × 10−8 × e

(11.1 × Δ sg )

I g 25Sb − old = P15 × 0.00000512 × e

(A.17)

(11.1 × Δ sg )

(A.18)

where

Ig25Sb–old

is the electrolyzing current (in amperes) at 25 °C for older vented lead-antimony cells

For end of life lead-selenium batteries, see Equation (A.19) or Equation (A.20) (see VARTA [B39]):

I g 25Se −old = C8 × 3.2 × 10 −9 × e

(11.2 × Δ sg )

I g 25Se −old = P15 × 0.000000819 × e

(A.19)

(11.2 × Δ sg )

(A.20)

For ventilation system design, the appropriate end-of-life Equation (A.17), Equation (A.18), Equation (A.19), or Equation (A.20) should be used.

A.2.4.2 Current equations for vented lead-antimony cells The equations in the preceding subclause can be further simplified for inclusion in the tables. The upper bound of the current on float would be dependent on the highest expected difference between the average cell float voltage and average fully charged cell specific gravity. The highest expected difference would be 0.965 (for example, an average cell float voltage of 2.18 V/cell on a 1.215 s.g. vented lead-antimony cell). Plugging this differential into the end-of-life Equation (A.17) and Equation (A.18) yields the results show in Equation (A.21) and Equation (A.22) (which are the equations found in Table 1 for this cell type):

I g 25Sb − old − float = C8 × 2 × 10

−8

(11.1 × Δ )

×e

I g 25Sb − old − float = P15 × 5.12 × 10−6 × e

sg

= C8 × 2 × 10

(11.1 × Δ sg )

−8

×e

(11.1 × 0.965)

= C8 × 0.000 897

= P15 × 5.12 × 10−6 × e(11.1 × 0.965) = P15 × 0.230

(A.21) (A.22)

40




Similarly, the upper bound of the current on boost, equalize, or finishing charge would be dependent on the highest expected difference between the average cell float voltage and average fully charged cell specific gravity. The highest expected difference would be 1.118 (for example, an average cell equalize voltage of 2.33 V/cell on a 1.215 s.g. vented lead-antimony cell). Plugging this differential into the end-of-life Equation (A.15) yields the results show in Equation (A.23) and Equation (A.24) (which are the equations found in Table 2 for this cell type):

I g 25Sb − old − boost = C8 × 2 × 10−8 × e

(11.1 × Δ sg )

I g 25Sb − old − boost = P15 × 5.12 × 10−6 × e

= C8 × 2 × 10−8 × e(11.1 × 1.118) = C8 × 0.00492

(11.1 × Δ sg )

= P15 × 5.12 × 10−6 × e(11.1 × 1.118) = P15 × 1.26

(A.23)

(A.24)

The upper bound of the current an initial charge would be dependent on the highest expected difference between the average cell float voltage and average fully charged cell specific gravity. The highest expected difference would be 1.285 (for example, an average cell initial charge voltage of 2.50 V/cell on a 1.215 s.g. vented lead-selenium cell). Plugging this differential into the beginning-of-life Equation (A.13) yields the results show in Equation (A.25) and Equation (A.26) (which are the equations found in Table 5 for this cell type):

I g 25Sb − new− initial = C8 × 3 × 10−9 × e

(11.6 × Δ sg )

I g 25Sb − new− initial = P15 × 7.68 × 10 −7 × e

= C8 × 3 × 10−9 × e(11.6 × 1.285) = C8 × 0.00893

(11.6 × Δ sg )

(A.25)

= P15 × 7.68 × 10−7 × e(11.6 × 1.285) = P15 × 2.29 (A.26)

Towards end of life, when there could possibly be internal shorts due to sediment buildup or separator degradation, vented lead-antimony batteries can go into thermal runaway. This is generally accepted to occur before approximately 10% of the cells are shorted. A cell rarely experiences a bolted short in this near end-of-life condition, and even 1.0 V/cell is extremely low for a partially shorted cell. Assuming 10% of the cells are at this 1.0 V, and assuming that the batteries are in equalize mode (average cell boost voltage of 2.33 for a 1.215 s.g. cell), the increased voltage across the non-shorted cells can be computed using Equation (A.27):

Vcell − sc 1.0 ))) × Vcell −boost = (1 + (0.1 × (1 − ))) × 2.33 = 2.46 Vcell −boost 2.33

VSb − cell −TR = (1 + (10 % × (1 −

(A.27)

where

Vc–Sb–TR Vc–sc Vc–boost

is the average peak voltage across a vented lead-antimony cell in thermal runaway is the average partial short circuit voltage of a lead-acid cell is the average boost/equalize voltage applied to a lead-acid cell

Running the calculation with the values given above yields:

VSb −cell −TR = (1 + (10 % × (1 −

Vcell − sc Vcell −boost

))) × Vcell −boost = (1 + (0.1 × (1 −

1. 0 2.33

))) × 2.33 = 2.46 V

Therefore, the average cell voltage in thermal runaway condition becomes 2.46 V/cell. This means that the highest expected difference between average cell float voltage and specific gravity would be 1.25. This can be applied to Equation (A.15) for end of life antimony cells to yield the results shown in Equation (A.28) and Equation (A.29) (which are the equations found in Table 6 for this cell type):

I g 25Sb −old −TR = C8 × 2 × 10 −8 × e

(11.1 × Δ sg )

= C8 × 2 × 10 −8 × e (11.1 × 1.25) = C8 × 0.0217

(A.28)

41




I g 25 Sb −old −TR = P15 × 5.12 × 10 −6 × e

(11.1 × Δ sg )

= P15 × 5.12 × 10 −6 × e (11.1 × 1.25) = P15 × 5.54

(A.29)

A.2.4.3 Current equations for vented lead-selenium cells As with the upper-bound equation simplification performed for vented lead-antimony cells, the same can be done for lead-selenium (low antimony) cells. The upper bound of the current on float would be dependent on the highest expected difference between the average cell float voltage and average fully charged cell specific gravity. The highest expected difference would be 1.01 (for example, an average cell float voltage of 2.25 V/cell on a 1.240 s.g. vented leadselenium cell). Plugging this differential into the end-of-life Equation (A.16) yields the results shown in Equation (A.30) and Equation (A.31) (which are the equations found in Table 1 for this cell type):

I g 25Se−old − float = C8 × 3.99 × 10 −9 × e

(11.2 × Δ sg )

I g 25Se −old − float = P15 × 1.02 × 10 −6 × e

= C8 × 3.99 × 10 −9 × e(11.2 × 1.01) = C8 × 0.000 327

(11.2 × Δ sg )

(A.30)

= P15 × 1.02 × 10 −6 × e (11.2 × 1.01) = P15 × 0.0835 (A.31)

The upper bound of the current on boost, equalize, or finishing charge would be dependent on the highest expected difference between the average cell float voltage and average fully charged cell specific gravity. The highest expected difference would be 1.160 (for example, an average cell equalize voltage of 2.40 V/cell on a 1.240 s.g. vented lead-selenium cell). Plugging this differential into the end-of-life Equation (A.16) yields the results shown in Equation (A.32) and Equation (A.33) (which are the equations found in Table 2 for this cell type):

I g 25Se − old −boost = C8 × 3.99 × 10−9 × e

(11.2 × Δ sg )

= C8 × 3.99 × 10−9 × e(11.2 × 1.16) = C8 × 0.00175

(A.32)

I g 25Se − old −boost = P15 × 1.02 × 10−6 × e (11.2 × Δ sg ) = P15 × 1.02 × 10−6 × e (11.2 × 1.16) = P15 × 0.448

(A.33)

The upper bound of the current an initial charge would be dependent on the highest expected difference between the average cell float voltage and average fully charged cell specific gravity. The highest expected difference would be 1.26 (for example, an average cell initial charge voltage of 2.50 V/cell on a 1.240 s.g. vented lead-selenium cell). Plugging this differential into the beginning-of-life Equation (A.14) yields the results shown in Equation (A.34) and Equation (A.35) (which are the equations found in Table 5 for this cell type):

I g 25Se − new−initial = C8 × 2 × 10 −9 × e

(11.2 × Δ sg )

I g 25Se − new−initial = P15 × 5.12 × 10−7 × e

= C8 × 2 × 10−9 × e(11.2 × 1.26) = C8 × 0.00269

(11.2 × Δ sg )

= P15 × 5.12 × 10−7 × e(11.2 × 1.26) = P15 × 0.689

(A.34)

(A.35)

A.2.5 Temperature effects on the current Temperature also affects the charge current (and therefore the gassing current) by the Arrhenius Equation (A.31). This equation basically says that there will be a doubling of current in lead-acid batteries for about every 8 °C above 25 °C, and a halving for about every 8° below (see Proceedings of Intelec [B34]) [see Equation (A.36) and Equation (A.37)]:

42




I Pb = I Pb − 25

⎛ ⎜ 8555 × ⎛⎜ 0.00335 − 1 ⎜ ⎜ TK ⎝ × e⎝

⎞ ⎞⎟ ⎟ ⎟⎟ ⎠⎠

(A.36)

where

IPb is the full-charge current (in amperes) flowing into a lead-acid battery IPb–25 is the full-charge current (in amperes) flowing into a 25 °C lead-acid battery TK is the absolute temperature in Kelvin TK = TC + 273

(A.37)

where

TC

is the temperature in degrees Celsius

The effect (increase of the current due to temperature) is similar for NiCd batteries (see A.6).

A.2.6 Effects of shorted cells When the batteries are equalized, or otherwise charged at levels higher than float, the equations do not change. Obviously, the Δsg will increase due to the increased voltage level. If it is necessary to account for the presence of shorted or partially shorted cells in the string(s) for a worst-case scenario calculation, the Δsg can be adjusted accordingly, as shown in Equation (A.38) and Equation (A.39):

Δ sg =

(V p + VDsc )

nC NS

− s.g .

(A.38)

where

VDsc

Vp NS

is the total voltage drop across the shorted cells (i.e., the difference between their open-circuit voltages under normal conditions, and their open-circuit shorted voltage). Note that most shorted cells are not bolted dead shorts (even in a shorted cell, typical open-circuit voltage of that cell exceeds at least 1 V). is the plant voltage at the common buss is the number of paralleled strings n

V Dsc =

∑ (VOC − V sc )

(A.39)

#

#= x

where

VOC V#sc

is the open-circuit voltage of cells of that particular specific gravity is the open-circuit voltage of a particular shorted cell

Equation (A.34) represents the summation of all the lost voltage due to partially shorted cells. Open-circuit voltage can be determined from Equation (A.35), as shown in Equation (A.40):

VOC = s.g. + 0.845

(A.40)

43




A.2.7 Maximum gassing current Even under the very worst-case scenario where the charger voltage runs away, gassing current ( Ig) is limited to the most that can be put out by the charger(s). The same is true for a constant-current charger, or for reversed cells, as shown in Equation (A.41) and Equation (A.42):

I g = I S max

(A.41)

where

ISmax I S max =

is the maximum string current

I ct

(A.42)

Ns

where

Ict

is the maximum charge current available to the batteries from the charger (or from all paralleled chargers if there is more than one)—in some cases, this is limited only by the charger output maximum setting, while in other cases there are charge current limiting devices (either in the charger, the controller, or in series with the battery string)

In some cases, the charger(s) are in parallel with the batteries and with the loads on a common dc buss. In these cases, ICCmax is limited also by the load, as shown in Equation (A.43):

I ct − spare = I ct − I L

(A.43)

where

IL

is the average load, or the load over a given timeframe

A.2.8 Equations for VRLA batteries

A.2.8.1 General All of the preceding equations also could apply to VRLA batteries, but only if it were assumed that the VRLAs were operating as vented cells. In actuality, VRLA batteries normally operate in a recombinant mode where typically 95% to 99% (see C&D [B12]) of the gas generated by electrolysis is not released from the battery. To get a better idea of actual gas release for constant voltage charging applications, manufacturer experiments offer the best data. For VRLA batteries using catalysts in their valves, the amount of hydrogen gas given off may decrease due to the catalyst’s effect on plate polarization. However, with some products, it is nearly impossible to tell if there is a catalyst in it or not, as the manufacturer sells it both ways (with and without a catalyst) and one cannot tell the difference visually. Also, the possibility exists that towards the end of life, the catalyst may become poisoned. In that case, gassing would become the same as any other VRLA at the end of life. For this reason, the effect of catalysts should generally be ignored in gassing calculations. Note that particularly for smaller VRLA cells (and often for high-rate vented cells) the number of cells is not necessarily the same as the number of battery units. Particularly for VRLAs, many products come in multi-cell units of nominal 12 V, 6 V, or 4 V. The number of cells in those multi-cell units are 6, 3, and 2; respectively. Therefore, the number of cells (not units) must be taken into account in all calculations.

44




A.2.8.2 AGM cells without antimony

A.2.8.2.1 General Comparing data from various manufacturers yields results for gas release that can be expressed in the following relationships as an upper bound (some alloys may be much less, but the gassing rate is so small that it is not going to make a difference for ventilation design purposes, so it is not necessary to have separate equations) for pure lead, lead-calcium, or lead-calcium-tin alloys (as long as Δsg > 0.922; if not, gassing current is 0) with AGM separators [see Equation (A.44)]:

= H 2−VRLACa25

× nC

×

−18

×

(15 × Δ sg )

×

C8 1.11 10

e

=

× nC

× P15

×

−16

(15 × Δ sg )

×

2.78 10

e

(A.44)

Correcting for temperature [modifying the Arrhenius Equation (A.31) to account for the increased current due to temperature, but ignoring the expansion of the gas due to temperature] yields Equation (A.45):

H 2−VRLA = H 2 −VRLACa25

⎛ ⎜ 8555 × ⎛⎜ 0.00335 − 1 ⎜ ⎜ TK ⎝ × e⎝

⎞ ⎞⎟ ⎟ ⎟⎟ ⎠⎠

(A.45)

A.2.8.2.2 Upper-bound gassing equations for the tables Equation (A.38) can be further simplified for inclusion in the tables. The upper bound of the hydrogen release on float would be dependent on the highest expected difference between the average cell float voltage and average fully charged cell specific gravity. The highest expected difference would be 0.980 (for example, an average cell float voltage of 2.28 V/cell on a 1.300 s.g. VRLA cell). Plugging this differential into Equation (A.40) yields the results shown in Equation (A.46) and Equation (A.47) (which are the equations found in Table 1 for this cell type): H 2−VRLACa25 = nc × C8 × 1.11 × 10−18 × e H 2 −VRLACa25 = nc × P15 × 2.78 × 10 −16 × e

(15 × Δ sg )

(15 × Δ sg )

= nc × C8 × 1.11 × 10−18 × e(15 × 0.980) = nc × C8 × 2.69 × 10−12

= nc × P15 × 2.78 × 10−16 × e(15 × 0.980) = nc × P15 × 6.73 × 10−10

(A.46) (A.47)

VRLA batteries in cycling applications may be subjected to an accelerated recharge control regime. They are rarely (if ever) subjected to boost/equalize charging, but if/when they are, it is at an average voltage no higher than that used for accelerated recharge. That peak charging voltage is bounded at the top by a maximum Δsg of approximately 1.070 (e.g., 2.35 V/cell for 1.280 s.g. cells). Plugging this upper bound into Equation (A.40) yields the following peak gassing release equations [Equation (A.48) and Equation (A.49)] on accelerated recharge for lead-calcium or lead-tin AGM batteries (which are the equations found in Table 2 for this cell type): H 2 −VRLA−Ca −boost = nc × C8 × 1.11 × 10−18 × e

(15 × Δ sg )

H 2 −VRLA−Ca −boost = nc × P15 × 2.78 × 10−16 × e

= nc × C8 × 1.11 × 10−18 × e(15 × 1.07) = nc × C8 × 1.04 × 10−11

(15 × Δ sg )

= nc × P15 × 2.78 × 10−16 × e (15 × 1.07) = nc × P15 × 2.60 × 10−9

(A.48) (A.49)

Formation of VRLA batteries is almost always completed in the factory, so while a freshening charge might be necessary (even that is rare) before installation if the interval between manufacture and installation is relatively long, an initial charge at a voltage or rate higher than the freshening charge level is not necessary. Therefore, for lead-tin or lead-calcium AGM cells, the equations in Table 5 for gas generation on an initial charge are the same as the equations found in Table 2 for boost/equalize charging.

45




All VRLA batteries are susceptible to thermal runaway. Similar to vented lead-antimony designs, it mostly occurs towards the end of life, when there could possibly be internal shorts. Thermal runaway is generally accepted to occur before approximately 10% of the cells are shorted. A cell rarely experiences a bolted short in this near end of life condition, and even 1.0 V/cell is extremely low for a partially shorted cell. Assuming 10% of the cells are at this 1.0 V, and assuming that the batteries are in boost mode (average cell boost voltage of 2.40 for a 1.300 s.g. cell), the increased voltage across the non-shorted cells can be computed as follows: VVRLA− AGM −Ca−cell −TR = (1 + (10 % × (1 −

Vcell − sc Vcell −boost

))) × Vcell −boost = (1 + (0.1 × (1 −

1.0 2.40

))) × 2.40 = 2.54

Therefore, the average cell voltage in thermal runaway condition becomes 2.54 V/cell. This means that the highest expected difference between average cell float voltage and specific gravity would be 1.24. This can be applied to Equation (A.40) for AGM VRLA lead-calcium or lead-tin cells to yield the results shown in Equation (A.50) and Equation (A.51) (which are the equations found in Table 6 for this cell type): H 2−VRLA−Ca −TR = nc × C8 × 1.11 × 10−18 × e

(15 × Δ sg )

H 2−VRLA−Ca−TR = nc × P15 × 2.78 × 10−16 × e

= nc × C8 × 1.11 × 10−18 × e(15 × 1.24) = nc × C8 × 1.33 × 10−10

(15 × Δ sg )

= nc × P15 × 2.78 × 10−16 × e(15 × 1.24) = nc × P15 × 3.33 × 10−8

(A.50) (A.51)

A.2.8.3 Low antimony VRLA designs

A.2.8.3.1 General For VRLAs using low-antimony designs, the hydrogen release rate will increase by a factor of about 2 to 6 (depending on applied charging voltage and the age of the cell). From manufacturer data (Berndt [B10] and Exide-GNB [B21]), the upper bound hydrogen release equations for new and old low antimony VRLA designs can be determined (relative to their counterparts that do not contain antimony). The current drawn by a low-antimony VRLA when new is about 2 to 3 times that drawn by a lead-tin or lead-calcium cell (Berndt [B10]) as shown incell Equation (A.52):

H 2 AGM −lowSb−new = 2.59 × H 2 AGM −Ca

(A.52)

As the low-antimony cell ages, the current draw increases, until at near the end of life (Exide-GNB [B21]), as shown in Equation (A.53):

H 2 AGM −lowSb −old = 5.17 × H 2 AGM −Ca

(A.53)

A.2.8.3.2 Upper-bound gassing equations for the tables Equation (A.47) and Equation (A.48) can be further simplified for inclusion in the tables. The upper bound of the hydrogen release on float for low antimony VRLA cells would be towards the end of life. Using Equation (A.48), Equation (A.41) and Equation (A.42) yield Equation (A.54) and Equation (A.55) (which are the equations found in Table 1 for this cell type): −12

H 2 − AGM −lowSb = 5.17 × H 2 − AGM −Ca = nc × C8 × 5.17 × 2.69 × 10

−11

= nc × C8 × 1.39 × 10

H 2− AGM −lowSb = 5.17 × H 2 − AGM −Ca = nc × P15 × 5.17 × 6.73 × 10−10 = nc × P15 × 3.48 × 10 −9

(A.54) (A.55)

46




Using the same ratio, but applying it to an accelerated/boost/equalize charge, applying Equation (A.48), Equation (A.43) and Equation (A.44) yield Equation (A.56) and Equation A.57) (which are the equations found in Table 2 for this cell type): H 2− AGM − lowSb −boost = 5.17 × H 2 − AGM − Ca −boost = nc × C8 × 5.17 × 1.04 × 10 −11 = nc × C8 × 5.36 × 10−11

(A.56)

H 2 − AGM −lowSb −boost = 5.17 × H 2− AGM −Ca −boost = nc × P15 × 5.17 × 2.60 × 10−9 = nc × P15 × 1.34 × 10−8

(A.57)

As noted, formation of VRLA batteries is almost always completed in the factory, so while a freshening charge might be necessary (even that is rare) before installation if the interval between manufacture and installation is relatively long, an initial charge at a voltage or rate higher than the freshening charge level is not necessary. However, because the battery is new (as opposed to old), the equations found in in Table 5 for a freshening/initial charge for a low antimony VRLA battery are different than those found Table 2. Applying Equation (A.47), Equation (A.43) and Equation (A.44) yield Equation (A.58) and Equation (A.59) (which are the equations found in Table 5 for this cell type): H 2 − AGM −lowSb−initial = 2.59 × H 2 − AGM −Ca −boost = nc × C8 × 2.59 × 1.04 × 10 −11 = nc × C8 × 2.68 × 10−11

(A.58)

H 2− AGM − lowSb −initial = 2.59 × H 2− AGM − Ca −boost = nc × P15 × 2.59 × 2.60 × 10 −9 = nc × P15 × 6.71 × 10 −9

(A.59)

Using the end-of-life Equation (A.48), and applying it to the thermal runaway Equation (A.45) and Equation (A.46) yields Equation (A60) and Equation (A.61) (which are the equations found in Table 6 for this cell type): H 2 − AGM − lowSb−TR = 5.17 × H 2 − AGM −Ca − boost = nc × C8 × 5.17 × 1.33 × 10 −10 = nc × C8 × 6.87 × 10−10

(A.60)

H 2− AGM − lowSb −TR = 5.17 × H 2 − AGM −Ca − boost = nc × P15 × 5.17 × 3.33 × 10 −8 = nc × P15 × 1.72 × 10 −7

(A.61)

A.2.8.4 Gel cells Although gelled VRLAs do not have as good of a recombination efficiency as AGM VRLAs [instead of a typical 99% recombination efficiency seen for AGM cells, the efficiency at the beginning of life (it gets better as time goes on) can be as low as 95%], the gassing is not proportionally higher than that of the AGMs because their charge current is lower (Bode [B11]). The charge current is lower by approximately half. Assuming half the charge current, but up to 5 times the recombination inefficiency of AGM cells yields a gel cell gas release of approximately 2.5 times for the same cell chemistries and valve pressures. This can be developed into upper-bound equations as shown in Equation (A.62) and Equation (A.63) for the different operating modes found in Table 1 and Table 2. H 2− gel− float = 2.5 × H 2− AGM − float = nc × C8 × 2.5 × 2.69 × 10−12 = nc × C8 × 6.72 × 10−12

(A.62)

H 2− gel−boost = 2.5 × H 2− AGM −boost = nc × C8 × 2.5 × 1.04 × 10−11 = nc × C8 × 2.60 × 10−11

(A.63)

As noted previously, formation of VRLA batteries is almost always completed in the factory, so while a freshening charge might be necessary (even that is rare) before installation if the interval between manufacture and installation is relatively long, an initial charge at a voltage or rate higher than the freshening not necessary. Therefore, cells, theboost/equalize equation in Table 5 for gas generation on an initialcharge chargelevel is theissame as the equation foundfor in gel Table 2 for charging. Applying similar principles yields the result shown in Equation (A.64) for thermal runaway of gel cells (the equation found in Table 6 for this battery type):

47




H 2− gel−TR = 2.5 × H 2− AGM −boost = nc × C8 × 2.5 × 1.33 × 10−10 = nc × C8 × 3.32 × 10−10

(A.64)

Gelled VRLAs are generally not used as high rate application batteries because they are space inefficient in this mode. For this reason, no equations are developed using kW/cell (P15) capacity ratings.

A.3 Sample gassing calculations for vented lead-acid batteries

A.3.1 Assumptions It may be helpful to take a real-world installation and calculate gas generation under different operating modes, and different assumptions, as follows:

⎯

–48 Vdc plant with the batteries, rectifiers, and loads connected in parallel

⎯

Float voltage of –52.8 V (average of 2.2 V/cell)

⎯

1600 A load

⎯

Three parallel strings of batteries (24 lead-calcium, 1.215 s.g. cells per string, 4000 Ah)

⎯

Sixteen 200 A rectifiers to feed the load and charge/recharge the batteries

⎯

30 m × 30 m (100 ft × 100 ft) room

⎯

Normal operating temperature of 25 °C (77 °F)

⎯

Worst-case operating temperature of 49 °C (120 °F)

⎯

Equipment operating voltage limits are –42.75 to –56 V

⎯

Initial charger size is 200 A

Note that while the plant and cell voltages (and ranges) are negative with respect to ground for these assumptions, for calculation purposes, the absolute value is used in order to obtain proper results.

A.3.2 Fire code default Most fire codes specify that if gas generation from lead-acid batteries is not calculated, the ventilation system must be designed to provide air exchange of 1 cfm per square foot of floor area. In this case, that requires ventilation sizing of 10 000 cfm.

A.3.3 Worst-case calculation Although all charging current does not electrolyze water, it theoretically could in a worst-case scenario. The worst-case scenario in the example is if all of the excess rectifier capacity above that needed for the load were being delivered to already charged cells due to extreme overvoltage, massive cell short circuits, or some other reason. This means that the three parallel strings share 1600 A (16 rectifiers × 200 A – 1600 A of load). That averages to about 533 A per string. Plugging in the appropriate numbers yields:

H 2 − rate = 1.27 × 10−7 × I g × nc = 1.27 × 10−7 × 533 × 72 = 0.00488 m3 s (10.3 cfm) In order to maintain 2% hydrogen concentration, the ventilation rate needs to be a minimum of 517 cfm (10.3 × 100% ÷ 2% = 10.3 × 50).

48




A.3.4 Normal gassing Normal gassing rates on float are extremely low:

I g 25Ca = C8 × 10 −9 × e

(11.1 × Δ sg )

= 4000 × 10 −9 × e (11.1 × ( 2.2 − 1.215)) = 0.224 A

H 2 − rate = 1.27 × 10−7 × I g × nc = 1.27 × 10−7 × 0.224 × 72 = 2.05 × 10−6 m3 s (0.00434 cfm) In order to maintain 2% hydrogen concentration, the ventilation rate needs to be a minimum of 0.217 cfm.

A.3.5 Gassing during initial charging Initial charging may be done at many voltages, but for these particular cells, the manufacturer typically suggests initial charging voltages in the range of 2.37 V/cell to 2.50 V/cell. Taking the worst gassing case would be an initial charging voltage averaging 2.50 V/cell:

I g 25Ca = C8 × 10−9 × e

(11.1 × Δ sg )

= 4000 × 10−9 × e(11.1 × ( 2.50 − 1.215)) = 6.26 A

H 2 − rate = 1.27 × 10−7 × I g × nc = 1.27 × 10−7 × 6.26 × 72 = 5.72 × 10−4 m3 s (0.121 cfm) In order to maintain 2% hydrogen concentration, the ventilation rate needs to be a minimum of 6.06 cfm. Some would argue that the capability of the chargers is the current that should be used in the calculation. In that case, the gassing current per string would be 66.67 A (200 A ÷ 3 strings). Redoing the calculations:

H 2 − rate = 1.27 × 10−7 × I g × nc = 1.27 × 10−7 × 66.7 × 72 = 0.000610 m3 s (1.29 cfm) Using this number, to maintain 2% hydrogen concentration requires a ventilation rate of 64.6 cfm. In some cases, users bring in extra fans during initial charging for safety.

A.3.6 Fire code worst-case calculation NFPA 1 (Fire Code) and the IFC (International Fire Code [B27]) suggest calculating gassing rates for ventilation purposes at equalize (boost) voltages. The use of an equalizing voltage, on a per cell average, also simulates a float condition with some partially shorted cells in the string. Although equalizing can be done at many different voltages, as noted in 7.1, it is often limited by the maximum voltage that the equipment can withstand (when the batteries are operating in parallel with the load and the rectifiers/chargers). In this case, the maximum average cell equalize voltage is 2.33 V/cell (56.0 V ÷ 24 cells per string). Rerunning the calculations:

I g 25Ca = C8 × 10−9 × e

(11.1 × Δ sg )

= 4000 × 10−9 × e(11.1 × ( 2.33 − 1.215)) = 0.984 A

Assuming worst-case temperature (49 °C for this example), this gassing current becomes:

I g = I g 25

⎛ ⎜ 8555 × ⎛⎜ 0.00335 − 1 ⎜ ⎜ TK ⎝ × e⎝

⎞ ⎞⎟ ⎟ ⎟⎟ ⎠⎠

⎛ ⎞⎞ ⎜ 8555 × ⎛⎜ 0.00335 − 1 ⎟⎟ ⎟ ⎜ 322 ⎠ ⎟⎠ ⎝ ⎝ = 0.984 × e

= 7.97 A

49




H 2−rate = 1.27 × 10−7 × I g × nc = 1.27 × 10−7 × 7.97 × 72 = 7.28 × 10−5 m3 s (0.154 cfm) In order to maintain a 2% hydrogen concentration, the ventilation rate needs to be a minimum of 7.72 cfm. (Note that the preceding calculation does not take into account the expansion of the gas due to the higher temperature.)

A.4 Sample gassing calculations for lead-calcium-tin VRLA batteries

A.4.1 Assumptions Similar to the preceding subclauses on VLA batteries, it may be helpful to take a real-world installation of VRLA batteries and calculate gas generation under different operating modes, and different assumptions, as follows:

⎯


⎯


⎯

1600 A load

⎯

Ten strings (24 cells per string) of 1400 Ah 1.300 s.g. lead-calcium AGM VRLA batteries

⎯


⎯

10 m × 10 m (33 ft × 33 ft) room

⎯


⎯


⎯

Equipment operating voltage limits are –42.75 to –56 V

⎯



A.4.2 Fire code default Most fire codes specify that if gas generation from lead-acid batteries is not calculated, the ventilation system must be designed to provide air exchange of 1 cfm per square foot of floor area. In this case, that requires ventilation sizing of 1100 cfm.

A.4.3 Worst-case calculation Although all charging current does not electrolyze water, it theoretically could in a worst-case scenario (such as extreme thermal runaway). The worst-case scenario in the example is if all of the excess rectifier capacity above that needed for the load were being delivered to already charged cells due to extreme overvoltage, massive cell short-circuits, or some other reason. This would drive the cells completely out of recombinant mode, with their valves wide open. This means that the three parallel strings share 1600 A (16 rectifiers × 200 A – 1600 A of load). That averages to about 533 A per string. (Note that under most thermal runaway conditions, not all of the potentially available charge current is drawn from the chargers, nor are the valves always wide open, however, it is instructive to run the worst-case calculation.)

50




H 2− rate = 1.27 × 10−7 × I g × nc = 1.27 × 10−7 × 533 × 72 = .00488 m3 s (10.3 cfm) Note that this worst case is the same as the worst case for the vented cells, even though the batteries have slightly different capacities. In order to maintain 2% hydrogen concentration, the ventilation rate needs to be a minimum of 515 cfm (10.3 × 100% ÷ 2% = 10.3 × 50).

A.4.4 Normal gassing Normal gassing rates on float are extremely low: H 2 −VRLACa 25 = nc × C8 × 1.11 × 10 −18 × e

(15 × Δ sg )

= 72 × 1400 × 1.11 × 10 −18 × e (15 × (2.27 − 1.300)) = 2.36 × 10 −7 m 3 s (0.0005 cfm)

Multiplying by 50 to maintain 2% hydrogen concentration, the ventilation rate needs to be a minimum of 0.025 cfm. This is extremely small, and what would be expected of a recombinant battery.

A.4.5 Gassing during initial charging Initial charging is not normally suggested for VRLAs. However, if done, it could be done at voltages as high as 2.40 V/cell. H 2−VRLACa25 = nc × C8 × 1.11× 10−18 × e

(15 × Δ sg )

= 72 × 1400 × 1.11 × 10−18 × e(15 × ( 2.50 − 1.300)) = 7.35 × 10−6 m3 s (0.0156 cfm)

In order to maintain 2% hydrogen concentration, the ventilation rate needs to be a minimum of 0.579 cfm. Some would argue that the capability of the chargers is the current that should be used in the calculation. In that case, the gassing current per string would be 20 A (200 A ÷ 10 strings). Redoing the calculations (assuming failed, wide open valves):

H 2−rate = 1.27 × 10−7 × I g × nc = 1.27 × 10−7 × 66.7 × 72 = .000610 m3 s (1.29 cfm) Using this number, to maintain 2% hydrogen concentration requires a ventilation rate of 64.6 cfm. In some cases, users bring in extra fans during initial charging for safety.

A.4.6 Fire code worst-case calculation NFPA 1 and the IFC suggest calculating gassing rates for ventilation purposes at boost voltages. The use of boost voltage, on a per cell average, also simulates a float condition with some partially shorted cells in the string. Although boosting can be done at many different voltages, as noted in 7.1, it is often limited by the maximum voltage that the equipment can withstand. In this case, the maximum average cell equalize voltage is 2.33 V/cell (56 V ÷ 24 cells per string). Rerunning the calculations: H 2 −VRLACa25 = nc × C8 × 1.11 × 10−18 × e

(15 × Δ sg )

= 72 × 1400 × 1.11 × 10−18 × e(15 × (2.33 − 1.300)) = 6.03 × 10−7 m3 s (0.00128 cfm)

Assuming worst-case temperature (49 °C for this example), this hydrogen release becomes (neglecting the minor effects of the expansion of gas at the higher temperature):

51




H 2 −VRLA = H 2 −VRLA25

⎛ ⎜ 8555 × ⎛⎜ 0.00335 − 1 ⎜ ⎜ TK ⎝ × e⎝

⎞ ⎞⎟ ⎟ ⎟⎟ ⎠⎠

⎛ ⎞⎞ ⎜ 8555 × ⎛⎜ .000335 − 1 ⎟⎟ ⎟ ⎜ 322 ⎠ ⎟⎠ ⎝

= 5.7 × 10− 7 × e⎝

= 4.88 × 10− 6 m3 s (0.0103 cfm)

In order to maintain a 2% hydrogen concentration, the ventilation rate needs to be a minimum of 1.74 cfm.

A.5 Sample gassing calculations for vented lead-antimony batteries

A.5.1 Assumptions For the calculations in the subsequent subclauses, the following assumptions apply:

⎯


⎯


⎯

6000 A load

⎯

Ten parallel strings (24 cells per string) of 3900 Ah 1.215 s.g. lead-antimony batteries

⎯

Ten 800 A rectifiers to feed the load and charge/recharge the batteries

⎯

30 m × 30 m (100 ft × 100 ft) room

⎯


⎯


⎯

Equipment operating voltage limits are –42.75 to –56.0 V

⎯



A.5.2 Fire code default Most fire codes specify that if gas generation from lead-acid batteries is not calculated, the ventilation system must be designed to provide air exchange of 1 cfm per square foot of floor area. In this case, that requires ventilation sizing of 10 000 cfm.

A.5.3 Worst-case calculation Although all charging current does not electrolyze water, it theoretically could in a worst-case scenario. The worst-case scenario in the example is if all of the excess rectifier capacity above that needed for the load were being delivered to already charged cells due to extreme over voltage, massive cell short-circuits, or some other reason. This means that the 10 parallel strings share 2000 A (10 rectifiers × 800 A – 6000 A of load). That averages to about 200 A per string.

H 2− rate = 1.27 × 10−7 × I g × nc = 1.27 × 10−7 × 200 × 240 = 0.00610 m3 s (12.9 cfm) In order to maintain 2% hydrogen concentration, the ventilation rate needs to be a minimum of 646 cfm (12.9 × 100% ÷ 2% = 12.9 × 50).

52




A.5.4 Normal gassing Normal gassing rates on float are low (although not nearly as low as their lead-calcium counterparts); especially when the lead-antimony batteries are new:

I g 25 Sb − new = C8 × 3 × 10 −9 × e

(11.6 × Δ sg )

= 3900 × 3 × 10 −9 × e (11.6 × ( 2.17 − 1.215)) = 0.757 A

H 2−rate = 1.27 × 10−7 × I g × nc = 1.27 × 10−7 × 0.757 × 240 = 2.31 × 10−5 m3 s (0.0489 cfm) In order to maintain 2% hydrogen concentration, the ventilation rate needs to be a minimum of 2.45 cfm. Towards the end of life, as the electrolyte becomes poisoned by the antimony, the current (and therefore gassing) rise:

I g 25Sb −old = C8 × 2 × 10 −8 × e

(11.1 × Δ sg )

= 3900 × 2 × 10 −8 × e (11.1 × ( 2.17 −1.215)) = 3.13 A

H 2−rate = 1.27 × 10−7 × I g × nc = 1.27 × 10−7 × 3.13 × 240 = 9.55 × 10−5 m3 s (0.202 cfm) In order to maintain 2% hydrogen concentration, ventilation rate needs to be a minimum of 10.1 cfm.

A.5.5 Gassing during initial charging Initial charging may be done at many voltages, but for these particular EnerSys™ cells, the manufacturer typically suggests initial charging voltages in the range of 2.33 V/cell to 2.50 V/cell. 17 Taking the worst gassing case would be an initial charging voltage averaging 2.50 V/cell: −9

I g 25Sb − new = C8 × 3 × 10

×e

(11.6 × Δ sg )

−9

= 3900 × 3 × 10

(11.6 × ( 2.50 − 1.215))

×e

= 34.8 A

The initial charger size is 800 A, split across ten strings (average of 80 A per string). This means that the charger is capable of putting out all of the 34.8 A per string on initial charge that these batteries will require.

H 2 − rate = 1.27 × 10 −7 × I g × nc = 1.27 × 10 −7 × 34.8 × 240 = 0.00106 m 3 s ( 2.25 cfm) In order to maintain 2% hydrogen concentration, air exchange during this initial charging period needs to be a minimum of 112 cfm. This might entail bringing in some temporary fans for the 4 day to 10 day initial charging period if the room is not normally sized for this amount of air exchange. Some would argue that the capability of the chargers is the current that should be used in the calculation. In that case, the gassing current per string would be 80 A, as noted above. Redoing the calculations:

H 2−rate = 1.27 × 10 −7 × I g × nc = 1.27 × 10−7 × 80 × 240 = 0.00244 m 3 s (5.17 cfm) Using this number, to maintain 2% hydrogen concentration requires a ventilation rate of 258 cfm.

17

EnerSys trademark is the property of EnerSys and its affiliates. This information is given for the convenience of users of this standard and does not constitute an endorsement by the IEEE of these products. Equivalent products may be used if they can be shown to lead to the same results.

53




A.5.6 Fire code worst-case calculation In the U.S., NFPA 1 and the IFC suggest to calculate gassing rates for ventilation purposes at equalize voltages. The use of an equalizing voltage, on a per cell average, also simulates a float condition with some partially shorted cells in the string. Although equalizing can be done at many different voltages, as noted in 7.2, it is often limited by the maximum voltage that the equipment can withstand. In this case, the maximum average cell equalize voltage is 2.33 V/cell (56 V ÷ 24 cells per string). Rerunning the calculations:

I g 25 Sb −old = C8 × 2 × 10 −8 × e

(11.1 × Δ sg )

= 3900 × 2 × 10 −8 × e (11.1 × ( 2.33−1.215)) = 19.2 A

Assuming worst-case temperature (49 °C for this example), this gassing current becomes:

I g = I g 25

⎛ ⎜ 8555 × ⎛⎜ 0.00335 − 1 ⎜ ⎜ TK ⎝ × e⎝

⎞ ⎞⎟ ⎟ ⎟⎟ ⎠⎠

⎛ 1 ⎞ ⎞⎟ ⎜ 8555 × ⎛⎜ 0.00335 − ⎟ ⎜ 322 ⎟⎠ ⎟⎠ ⎝

= 19.2 × e⎝

= 155 A

H 2−rate = 1.27 × 10 −7 × I g × nc = 1.27 × 10 −7 × 155 × 240 = 0.00473 m3 s (10.0 cfm) In order to maintain a 2% hydrogen concentration the ventilation rate needs to be a minimum of 502 cfm.

A.6 Battery gassing calculations for NiCd batteries

A.6.1 General Gases are only generated during overcharging of batteries. When the cell is charged to about 80%, there is a sudden increase in voltage due to the start of hydrogen evolution at the negative plates. The last 20% of the charge is recovered during the gassing phase. The charging current is tapered during this phase as the battery string voltage increases. Part of this charging current electrolyzes the water in the electrolyte and the rest of it recharges the negative electrode. Most of the energy lost during the charging process is spent on the electrolysis of water evolving hydrogen and oxygen gases. The amount of gases evolved will depend on the charging voltage, battery design, and battery temperature. An increase of charging voltage by 0.046 V/cell, will double the float current or the stabilized charging current. The stabilized charging current is also affected by temperature. The stabilized charging current then can be calculated approximately by Equation (A.65) and Equation (A.66): V − V0

I Π = I Π 0 × 2 0.046

(A.65)

where

IΠ IΠo Vo

is the stabilized charging current towards the end of charge is the stabilized charging current at the beginning of charge is the charging voltage at the beginning of charge all the way through the bulk charging phase

IT = I Π × 2

T − T0 10

(A.66)

54




where

IT T T0

is the temperature-adjusted charging current is the later temperature in Kelvin or degrees C is the initial temperature in Kelvin or degrees C

The factor of 10 in Equation (A.61) represents the doubling of float current for every 10 °C rise. This is the worst-case scenario. Depending on the cell design, this factor varies between 10° and 15° (Alcad [B2], [B3], [B4], and [B5]). However, to be conservative, and estimate worst-case gassing, it is best to assume a current doubling for every 10 °C rise. The maximum hydrogen evolution rate in cubic meters per second is calculated using the Equation (A.67):

H 2−rate = 1.27 × 10−7 × I chg × nc

(A.67)

where

Ichg

is the charging current in amperes

Converting Equation (A.62) to the more standard cubic feet per minute (cfm) at standard temperature and atmospheric pressure (STP) yields Equation (A.68):

H 2−cfm = 0.000269 × I chg × nc

(A.68)

A.6.2 Float charging Stabilized charging current for a NiCd couple at the most common float voltage (at 25 °C) of 1.43 V/cell (it could be as high as 1.45 V/cell) is approximately 0.5 mA/Ah (at the C5 Ah rating).

A.6.3 Finish/boost/equalize charging Equalizing charge when required is typically conducted at a boost voltage like 1.55 V/cell. Normally this charging is conducted at a current equal to 10% of the 5 h capacity rating of the battery throughout the charge, as shown in Equation (A.69): V − V0

1.55 − 1.43

I Π = I Π 0 × 2 0.046 = (C5 × 5 × 10−4 ) × 2

0.046

= C5 × 0.00305

(A.69)

The hydrogen evolution is calculated with Equation (A.61) based on this charging current.

A.6.4 Initial charging Initial charging of NiCd cells (when required) is conducted using a constant-current charger. Normally this charging is conducted at a current equal to 20% of the 5 h capacity rating of the battery throughout the charge. The hydrogen evolution is calculated based on this charging current, as shown in Equation (A.70):

H 2 − rate = 1.27 × 10 −7 × I × nc = 1.27 × 10 −7 × 0.2 × C5 = 2.54 × 10 −8 × C5 × nc

(A.70)

55




A.6.5 Worst-case scenarios The stabilized charging current (float current) is also affected by factors such as shorted cells within the battery strings. The battery string may operate with as many as 10% shorted cells. This scenario may allow the battery string to operate with the battery charger in current limit. The charger will feed the connected load first and the rest of the charger current will feed into the battery string. Such conditions should be considered.

A.7 Sample gassing calculations for vented NiCd batteries

A.7.1 General The state of charge of NiCd batteries can be determined by placing the charger in a high-rate charge mode. As illustrated in the following example, stabilized charging current would increase and hydrogen evolution will proportionally increase. For a 100 Ah 38-cell string, the stabilized float current is 50.0 mA at 1.43 V/cell at 25 °C. If we assume that the charging voltage is increased to 1.45 V/cell, the new stabilized charging current becomes: V − V0

I Π = I 0 × 2 0.046 = 0.050 × 2

(1.45 − 1.43) 0.046

= 0.0676 A

Therefore, hydrogen evolution at this higher float voltage is (neglecting the minor effects of the expansion of gas due to temperature):

H 2 − rate = 1.27 × 10 −7 × 0.0676 × 38 = 3.26 × 10 −7 m 3 s ( 0.000 691 cfm )

In float can calculate battery-stabilizing currentasatina different temperature equations from theoperation, precedingwe subclause. Using the same assumptions the previous example,using if wethe assume that the temperature is increased to 45 °C, the new stabilized charging current becomes: 45 − 25

I T = 0.050 × 2

10

= 0.050 × 2 2 = 0.200 A

Therefore, hydrogen evolution at this higher temperature is (neglecting the minor effects of the expansion of gas due to temperature):

H 2 − rate = 1.27 × 10 −7 × 0.200 × 38 = 9.65 × 10 −7 m 3 s (0 .00205 cfm )

A.7.2 Assumptions For the calculations in the subsequent subclauses, the following assumptions apply:

⎯


⎯

Float voltage of -54.4 V (average of –1.43 V/cell)

⎯

20 A load

⎯

Two parallel strings (38 cells per string) of 125 Ah vented NiCd cells

56




⎯

Three 20 A rectifiers to feed the load and charge/recharge the batteries

⎯

164 cm × 78 cm (5.5 ft × 2.5 ft) battery cabinet compartment

⎯


⎯


⎯


⎯



A.7.3 Fire Code default Because these batteries are in an outdoor compartment, rather than a room, the Fire Code does not apply. However, if the cabinet were indoors, the Fire Code could be interpreted to specify that if gas generation is not calculated, the cabinet ventilation system must be designed to provide air exchange of 1 cfm per square foot of compartment area. In this case, that requires ventilation sizing of approximately 14 cfm (5.5 ft × 2.5 ft × 1 cfm/ft2).

A.7.4 Worst-case calculation Although all charging current does not electrolyze water, it theoretically could in a worst-case scenario. The worst-case scenario in the example is if all of the excess rectifier capacity above that needed for the load were being delivered to already charged cells due to extreme over voltage, massive cell short-circuits, or some other reason. This means that the 2 parallel strings share 40 A (3 rectifiers × 20 A/rectifier – 20 A of load). That averages to 20 A per string.

H 2−rate = 1.27 × 10 −7 × I g × nc = 1.27 × 10 −7 × 20 × 76 = 1.93 × 10 −4 m 3 s (0.409 cfm) In order to maintain 2% hydrogen concentration, the ventilation rate needs to be a minimum of 20.5 cfm (0.409 × 100% ÷ 2% = 0.409 × 50).

A.7.5 Normal gassing Normal gassing rates on float are low:

I g − Ni −Cd = C5 × 5 × 10−4 = 125 × 5 × 10−4 = 0.0625 A H 2− rate = 1.27 × 10−7 × I g × nc = 1.27 × 10 −7 × 0.0625 × 76 = 6.03 × 10 −7 m 3 s (0.00128 cfm) In order to maintain 2% hydrogen concentration, the ventilation rate needs to be a minimum of 0.0639 cfm.

A.7.6 Gassing during initial charging As noted in A.6.4, initial charging of NiCd cells is usually done with a charger that puts out 20% of the C5 rating of the batteries. For these 125 Ah cells, if both strings are connected in parallel and charged, that is a

57




charger output of 50 A, as noted in the assumptions. Using the Equation (A.64) developed in A.6.4 yields the following gassing results during initial charging:

H 2 − rate = 2.54 × 10 −8 × C5 × nc = 2.54 × 10 −8 × 125 × 76 = 2.41 × 10 −4 m 3 /s (0.511 cfm ) In order to maintain 2% hydrogen concentration, air exchange during this initial charging period needs to be a minimum of 25.6 cfm.

A.7.7 Fire Code worst-case calculation In the U.S., 1 and theboost/equalize IFC suggest to calculate for isventilation purposes equalize voltages. As NFPA noted in A.6.3, charging forgassing a NiCdrates battery done at 1.55 V/cell.atUsing the Equation (A.64) developed in A.6.3 yields the following results: H 2 − rate = 1.27 × 10 −7 × I × nc = 1.27 × 10 −7 × 0.00305 × C5 × nc = 1.27 × 10 −7 × 0.00305 × 125 × 76 = 3.68 × 10 −6 m 3 /s (0.0078 cfm)

In order to maintain a 2% hydrogen concentration, the ventilation rate needs to be a minimum of 0.39 cfm.

58




Annex B

(informative) Heat generation in lead-acid batteries B.1 Purpose This annex exists to show how the heat release/generation equations in the tables of 7.2 were derived, in addition to providing sample calculations.

B.2 Basics of battery heat generation

B.2.1 General Battery capacity is important to heat estimations only in that size dictates float currents and heat capacities in most calculations. However, for estimating discharge and charge reaction heat, knowledge of a battery’s expected electrical capacity proves important. Battery design influences the heat generation calculations; however, in some cases, heat equations can be treated as generic across a range of designs. Finally, each added charger or float rectifier algorithm can significantly complicate the heat estimation process. In thermodynamics the standard mathematical sign convention conforms to a perspective from the system understudy. If that system loses or gives up energy to the surrounding environment, the system has lost internal energy, and the standard sign convention treats the magnitude of that energy loss as a negative number. This means that in rigorous thermodynamics, the magnitude of heat energy transferred to the surroundings has a negative mathematical sign that indicates an exothermic process (i.e., production of waste heat). thermodynamic In systems undersign study that absorb energyand from during a reaction process, the standard convention inverses, thetheir heat surroundings absorbed is given a positive sign, which indicates an endothermic process (i.e., system is absorbing energy or cooling). For battery research and engineering this can be very confusing since from a battery energy perspective positive signs associate themselves with energy delivered to an application load. This situation is similar to the conflict in mathematical assignment of positive and negative signs between battery technologists and electroplaters. To further confuse the use of mathematical signs, the Joule effect always produces waste heat and would normally always be given a negative quantitative heat value. This makes the equations a little more complicated to track since the processes of charging and discharging normally introduce differentials in voltage that can result in negative values and positive values depending on direction of current flow. For most battery heat calculations, including those presented here, it is normally more convenient to assign positive values to waste heat generated by the battery and negative values to heat absorbed by the battery. This view is consistent with Berndt’s [B10] recommendations for defining qgen, and as a consequence, the following equations and tables are modified to present waste heat as a positive number and absorbed heat, or cooling, as a negative number. The following subclauses generally produce heat generation (or cooling if the sign is negative) in watthours (Wh) or watts. In case the HVAC engineer needs to use BTUs in their calculations, watthours can be converted to BTUs (or watts to BTUs/hr) by the following simple relationship: 1 Wh = 3.41 BTU

59




B.2.2 Sources of heat

B.2.2.1 General All of the battery designs and cell chemistries included in this document exhibit the same fundamental sources of heat. Heat generation in a cell stems from two basic sources: heat of reaction and Joule effect heat.

B.2.2.2 Heat of reaction This is the reversible heat generated or absorbed as a function of the cell reactions. For lead-acid batteries the heat of reaction is endothermic for discharge and exothermic for charging. For NiCd batteries the heat of reaction is exothermic for discharge and endothermic for charging. This heat relates quantitatively to the energy associated with the conversion of active materials during charge or discharge; consequently, this heat is directly proportional to ampere-hours and is not dependent on the battery design. The fundamental equation for the heat of reaction is shown in Equation (B.1):

qrev = I × t × TK × ΔS

(B.1)

where

qrev I t TK ΔS

is the heat of reaction, in Joules is the average current in the cell (for charging, it is the portion of the current going to material conversion) is the charge or discharge time, normally in hours is the temperature at the start of discharge, in Kelvin is the entropy of reaction in Joules/mole-K

Although excellent for scientific analysis, this version of the equation is not particularly convenient for battery engineers and field installation personnel. Appropriate factor conversions permit the transformation of the heat of reaction equation and the associated entropy values into forms more easily used by battery technologists, field service, and installation personnel. Using the entropy data provided in Table B.1, Equation (B.2) allows determination of this reversible reaction heat:

qrev −Wh = E Ah × nc / s × ΔST

(B.2)

where

qrev–Wh is the heat of reaction, in watthours EAh is the ampere-hours of charge or discharge nc/s is the number of cells per string ΔST is the entropy of reaction in Wh/Ah at 25 °C In the NiCd system, the dependency of the entropy term on the potassium hydroxide electrolyte’s ionic strength is small, but for lead-acid batteries there is a significant shift as a function of the sulfuric acid electrolyte’s specific gravity. This dependency is due to the role of the acid as a core reactant in the discharge and charging reactions. Lower sulfuric acid specific gravities yield a larger heat of reaction in the sulfuric acid system. Higher sulfuric acid specific gravities generate necessary heat to compensate for the endothermic discharge process via heat liberation due to heat of mixing of the water produced in the discharge process. This means that VRLA cells, with their generally higher specific gravities, are more efficient during the discharge process than either VLA or vented NiCd cells. For reference comparative entropy data is given in Table B.1, using lead-acid entropy data from Bode [B11] and nickel-cadmium entropy data from Berndt [B10]:

60




Table B.1—Comparative heat of reaction data

Lead-acid

1.215 ± 0.020 1.250 ± 0.020 1.285 ± 0.020 1.330 ± 0.020

± 48.8 ± 43.3 ± 37.8 ± 32.0

–0.000253 –0.000224 –0.000196 –0.000166

0.000253 0.000224 0.000196 0.000166

± 0.0754 ± 0.0669 ± 0.0584 ± 0.0494

Reaction energy heat 3.6% 3.2% 2.7% 2.6%

Nickelcadmium

all

± 87.2

0.000452

–0.000452

± 0.1347

10.2%

Battery type

Specific gravity

ΔS (J/mol-K)

Δ S disch. (Wh/Ah-K)

Δ S charge (Wh/Ah-K)

Δ ST

Efficiency

96.5% 96.9% 97.3% 97.8% 90.7%

Table B.1 does not provide a comprehensive of entropy values as a function ofvary lead-acid sulfuric acid electrolyte’s specific gravity. The changes inlist entropy for sulfuric acid solutions considerably as a function of electrolyte molality, but these large variations predominately occur in the specific gravity range from about 1.010 up to about 1.200. At higher specific gravities the shift in entropy values are smaller, and even close to linear between 1.210 and 1.330 (almost all lead-acid batteries commonly in use fall into this specific gravity range). Given the comparatively small shifts in entropy in the practical window of acid concentrations used in all modern lead-acid batteries, selection of a specific gravity within 20 or 30 points (i.e., 0.02 to 0.03 s.g. points) will provide an adequate approximation of the heat of reaction for sizing of cooling systems and predicting relative temperature changes. Also, as noted previously, in Table B.1, the mathematical signs for the entropy values associated with discharging and charging have been reversed when compared to the rules of standardized thermodynamics to allow waste heat to appear as a positive number and absorbed heat (i.e., cooling) as a negative number. For every 100 Ah discharged or charged, the heat of reaction produces a thermal response in lead-acid chemistry of between ±5 Wh and about ±7 Wh (i.e., 17 BTU to 25 BTU). Similarly, every 100 Ah of discharge or charge in NiCd chemistry produces a thermal response of about ±13 Wh (i.e., about 46 BTU). While these values are small, for discharges lasting more than about 8 h the heat of reaction will be the main heat effect in either battery chemistry, and for NiCd batteries, it can still dominate even at discharge times shorter than 30 min.

B.2.2.3 Joule effect heat

B.2.2.3.1 General For most electrical devices waste heat is generated in association with the electrical resistance of the electrical components. Batteries contain additional factors that must be included and that are usually substantially larger in effect than the measured internal resistance. As electrons flow through a battery, its voltage deviates from the open-circuit equilibrium value. The current-dependent voltage deviation yields waste heat, often referred to as Joule heat or Joule effect heat. Joule effect heat is the product of the difference in voltage between the operating cell voltage and the open-circuit voltage multiplied by the current. In addition, a special case exists for rechargeable aqueous battery systems. In this special case, as a battery approaches full charge, sections or areas of the active electrodes go into an operational condition called overcharge. Once the entire battery reaches full charge, then, assuming the charging current is still present, the entire battery goes into overcharge. The overcharge processes start with electrolysis of water (i.e., water decomposition). Depending on battery design the gases produced can either leave a cell via the cell’s venting system or can be retained through processes of oxygen recombination, and to a much lesser extent hydrogen recombination. The oxygen recombination process in nearly all VRLA cells becomes initiated at a point in the cell charging processes that precedes any significant hydrogen gassing. This behavior has the effect of strongly suppressing the generation of additional hydrogen in the system during overcharge. In some limited cases for both vented and VRLA batteries, an added catalyst cap allows for hydrogen to 61




catalytically react with the oxygen gas to reform water and retain it in the cell. In these cases of electrolysis and gas recombination, the equilibrium reference voltage is markedly different from a cell’s open-circuit voltage. The following subclauses address each reaction process associated with the Joule effect.

B.2.2.3.2 Joule effect discharge heat During discharge, all of the Joule effect heat is a by-product of current. The voltage deviation from the open-circuit voltage during discharge current flow is called discharge polarization and is composed of three components, as shown in Equation (B.3): η

= η a + ηc × ηo

(B.3)

where η ηa

ηc

ηo

is the discharge polarization is the activation polarization (charge transfer limiting), which is a voltage drop associated with the initiation of the reactions and the energy losses due to electron movement from a reacting atom to an ion in solution or into an adjacent electrical conductor is the concentration polarization (mass transfer limiting), which is a voltage drop associated with feeding the reaction with active materials, and when the reaction occurs faster than the reactive materials can supply it, there is a voltage drop at the reacting surface due to the absence of adequate reactants to fully support current demand is ohmic polarization, which is the series sum of voltage drops associated with component electrical resistances in a cell or battery string, including internal cell resistances and external intercell connection resistances

Since voltage drops can be converted into equivalent resistance values the polarization values can be converted into an equivalent resistance for computational purposes and the equation would have the simple expression shown in Equation (B.4):

R = Ra + Rc + Ro

(B.4)

where

R Ra Rc Ro

is the total resistance is the resistance due to activation polarization is resistance due to concentration polarization is the ohmic resistance of the conductive paths

Unfortunately, this approach oversimplifies the analysis, because the effective reactant concentration resistances and parts of the ohmic internal resistances are actually variable resistors, or the equivalent of variable resistors, and not fixed resistors; therefore, it is often much simpler to employ voltage differences to calculate heat generation. Further, for Joule effect discharge heat it is not necessary to isolate the individual contributions to heat generation from each other, but rather, to use their summed impact as a single parameter. Thus, to compensate for the fixed and variable resistance components it becomes convenient to use a single average discharge voltage over the period of the discharge time. In battery systems where the true average discharge voltage is not known, an approximation that tends to slightly overstate Joule effect heat generation at low rates of discharges (e.g., longer than 2 h) and usually understates Joule effect heat from high-rate discharges (e.g., 1 h or less) can be substituted. This substitution consists of calculating an average voltage from the difference between a battery’s open-circuit voltage and its disconnect termination voltage. If discharge voltage behavior were linear, this would match the true average voltage, but since battery voltage behavior on discharge is not linear, this approach can only provide an approximation of the heat generation. Due to the smaller slope in the discharge curves for

62




NiCd batteries this method usually overstates its heat generation more than for lead-acid systems. For a conservative estimate of heat generation use the battery’s low-voltage load termination, or disconnect voltage in place of the average discharge voltage. This approach can be used to simplify determination of Joule effect heat during discharge, and further, partially accommodates some of the changes in component resistances associated with battery aging (e.g., positive grid corrosion in lead-acid and electrolyte carbonation uptake in nickel-cadmium). Thus the basic, idealized calculation of Joule heat generation from the discharge process is shown in Equation (B.5) and Equation (B.6):

q J − r − disch = I × t × (Vstring − oc − Vs − avg − disch )

(B.5)

where

qJ–r–disch is the Joule effect reaction discharge heat Vstring–oc is the open-circuit voltage of the string Vs–avg–disch is the average voltage of the string during discharge

Vs −avg − disch ≈

Vstring −oc + Vstring −eod

(B.6)

2

where

Vstring–eod

is the voltage of the string at the end of discharge

B.2.2.3.3 Joule effect charging heat Although charging processes have more factors to account for than discharging processes, the basic causes of charging voltage shifts are the same as those for discharge polarization. Therefore, the discussion of voltage changes in the preceding subclause directly applies. So, in an ideal process the charging heat equation, for constant-current charging, would be essentially identical to Equation (B.6), except for the fact that instead of average discharge voltage one would substitute average charging voltage. No equation for this is shown here, because the ideal condition is never met, even when a battery company supplies the necessary average charge voltage for a given constant current. When considering heat generation, the deviation in behavior from the discharging circumstance is due to the assignment of electron flow in the string. If a cell accepted charge at 100% current efficiency, all Joule effect heat would come from the charging reaction. This condition does not exist, except at the beginning of charge following a complete discharge. The key factor becomes having a method to make reasonable or appropriate proportional assignment of current between Joule effect charging heat for conversion of the active materials and for the companion overcharge heat. Charging systems impose a further complication since they can be constant voltage control, constantcurrent control, mixtures of both, and may impose multiple time bases, including pulse profiles. In the following equations, even if charge algorithms or other criteria dictate a mixture of charging parameters, usually either constant voltage or constant current will still dominate depending on the system. Consequently, two basic equations usually suffice to estimate waste heat from charging. For convenience the term CC is used to represent constant-current charge control and the term CV is used to represent constant voltage charge control. Also IAM is used to represent the share of current associated directly with conversion of the active materials. A method to estimate or determine quantitatively the proportion of total current that can be assigned to material conversion as opposed to electrolysis and/or gas recombination will be presented following these fundamental equations [see Equation (B.7) and Equation (B.8)]:

63




q J − r − chg − CC = I AM × t × (Vs − avg − chg − Vstring − oc )

(B.7)

where

qJ–r–chg–CC is the Joule effect charge reaction heat when a constant-current charger is used Vs–avg–chg is the average charging voltage applied to the string q J − r − chg − CV = I AM × t × (Vstring − chg − Vstring − oc )

(B.8)

where

qJ–r–chg–CV is the Joule effect charge reaction heat when a constant voltage charger is used Vstring–chg is the charge voltage setting of the constant voltage charger Average string voltages, under constant-current conditions, are often not precisely known, and this creates imprecision in heat generation estimates. In addition, under constant-current conditions, and most significantly, the portion of current to assign to active material conversion is variable, changing as the charge input progresses. While constant voltage charging makes the charge voltage a straightforward assigned value, the variability in current now is compounded by having two mixed behaviors where the current is varying, and at the same time, the proportion of current going into material conversion is also changing. These conditions are further compounded by the need to create separate assigned values for each battery chemistry and product design, including such factors in the lead-acid chemistry as positive and negative grid alloys. To sort and assign current distribution during charging to various chemical reactions requires extensive investigative efforts. In place of investigation and direct measurements on a battery system, examination of battery literature, including battery manufacturer’s publications, can usually provide sufficient information to estimate ampere-hour (i.e., coulombic) charge efficiency. Knowledge of the relative charge efficiencies then allows the creation of proportionality factors for splitting and assigning current distributions between active material conversion and overcharge processes. Table B.2 provides approximations of proportional current distributions that can be used to yield estimates of Joule effect heat generation during charging.

Table B.2—Comparative charge factors

Battery type

VRLA Vented lead-acid Vented NiCd

Typical charge factor range

1.02 to 1.1

Charge factor for use in heat calculations ( Kc) 1.1 1.2

Charge current factor assigned to active material conversion ( KAM) 0.91 0.83

Charge current factor assigned to overcharge processes ( KOC) 0.09 0.17

1.1 to 1.4 1.25

0.80

0.20

A battery manufacturer should be able to supply a more precise charge factor for each product type and may be able to provide additional corrections to the charge factor for specific levels of depth-of-discharge, battery aging, and operating temperature. When capacity is removed and a battery is recharged, the empirical results from many battery tests typically show that an overcharge of 5% to 50% is required to restore 95% to 100% of the batteries’ rated or specified capacity. The charge factor multipliers in Table B.2 allow a computational estimate of the split in current during recharge between active material conversion and the overcharge processes. Now that a method of assigning current has been defined, the ability to make that assignment useful depends on knowing the total ampere-hours of charge put into a battery string. If an ampere-hour meter or coulometer is part of the charge control system, then it is simply a matter of using the reported amperehours in place of multiplying current by charge time, or integrating current over charge time in the case of

64




constant voltage recharge. In some, if not many, instances these tools are not integral to the charging system and multiplying current by time is expected as an external task. For heat estimates in battery system configurations it is usually not necessary to have this data, but rather, to have the appropriate rated capacity data from the battery manufacturer for the temperature and load expected in the design of the power backup system. This greatly simplifies the determination of expected heat generation since it allows replacement of current and time with a capacity value and appropriate charge factor multipliers. Given a defined rated capacity for a battery string’s application load condition, then by combining this information with the multipliers available in Table B.2, allows modification of Equation (B.7) and Equation (B.8) as shown in Equation (B.9) and Equation (B.10) for calculating Joule effect charging heat.

q J − r − chg − CC = (C8 / 5 × K c × K AM ) × (Vs − avg − chg − Vstring − oc )

(B.9)

where

qJ–r–chg–CC is the Joule effect charge reaction heat when a constant-current charger is used C8/5 is the ampere-hour rating of the batteries at the 8 h rate to 1.75 V/cell @ 25 °C for leadacid batteries, or at the 5 h rate to 1.0 V/cell @ 20 °C for NiCd batteries Kc is the charge factor for use in heat generation calculations from the battery manufacturer or Table B.2 KAM is the charge factor assigned to active material conversion, from the battery manufacturer or Table B.2 q J − r − chg −CV = (C8 / 5 × K c × K AM ) × (Vstring − chg − Vstring − oc )

(B.10)

A careful analysis of these equations would reveal an apparent redundancy with Kc and KAM. These terms are retained, without mathematical cancellations, because these same equation structures will be used for calculating the overcharge portion of the recharge heat, and in that case, the terms will not cancel. Therefore, for the sake of maintaining symmetry and consistency in the computations, the possible mathematical term cancellations are not recommended. For constant voltage charging, the inputs are straightforward and require no special considerations. In cases where constant-current charging is employed, and the average string charging voltage is unknown, a conservative approach that will modestly overestimate the heat generation can be used by substituting the maximum charging voltage allowed for the battery string in place of the average string voltage. Battery manufacturers with proprietary data on average string voltage can provide better estimates for computing heat generation in Equation (B.10), but absent their input, thermal management considerations will often dictate use of the maximum allowable string charging voltage in the constant-current Joule effect charging heat estimates.

B.2.2.3.4 Joule effect overcharge heat

B.2.2.3.4.1 General Distinct from Joule effect heat associated with the discharge process, the charging process has a second component of heat generation driven by the overcharge reactions. As previously noted, the charging process does not operate with 100% current efficiency. As a battery charges, a portion of the current is shunted to the overcharge reactions in parts of the electrodes that had previously reached full charge. Typically, a lead-acid or NiCd will operate near 100% charge acceptance until the battery reaches about 60% state-of-charge (SOC). At that time, a portion of the current begins water electrolysis. For some leadacid designs, such as VRLA cells, the charge efficiency remains near 100% until about 80% SOC. Regardless of the overcharge’s initiation SOC, once the overcharge reactions start, a second source of heat generation is initiated.

65




B.2.2.3.4.2 Joule effect electrolysis heat In vented cell designs water electrolysis starts towards the end of bulk recharge and continues during float service. The oxygen and hydrogen generated eventually leave the system (except for cases where a catalyst cap is present). The magnitude of the heat generation depends on the effective equilibrium voltage for the water decomposition reaction (Berndt [B10]). The ideal reaction voltage where heat is neither gained nor lost defines the thermal neutral voltage for a given reaction. For the water decomposition reaction the thermal neutral equilibrium voltage (also known as the “fictive voltage,” see Berndt [B10]), combines water’s thermodynamic decomposition voltage of 1.229 V plus compensation for the decomposition reaction’s endothermic heat of reaction. Therefore, the thermal neutral or “fictive” decomposition voltage for water electrolysis in a lead-acid cell may be defined by increasing the equilibrium decomposition voltage by about 252 mV (or about 20%). The adjusted equilibrium voltage then becomes:

Ve − Pb = 1.229 + 0.252 = 1.481 V/cell where

Ve–Pb is the thermal neutral equilibrium voltage for water decomposition in a lead-acid cell For a NiCd couple, this same thermal neutral equilibrium voltage is 1.44 V/cell. This thermal neutral voltage serves as the water electrolysis equilibrium voltage. This equilibrium voltage is the baseline reference for heat generation associated with electrolysis heat. Imposed charging voltages above this value result in the production of waste heat, and imposed voltages below this value require heat extraction from the surroundings to drive the reaction. Thus, electrolysis heat generation can be calculated using Equation (B.11):

q J − e = I g × (Vc − Ve ) × t × nc / s

(B.11)

where

qJ-e Ig Vc

is the Joule effect heat generated from electrolysis is the electrolysis current is the average cell voltage

The current term in this expression is not the string current, but rather, the fraction, or proportion of the total current that splits off to support the electrolysis reactions. This is an aspect that is often missed when doing heat generation calculations for the electrolysis process in batteries. Based on the charge factors listed in Table B.2, one can compose electrolysis heat equations as companions to Equation (B.9) and Equation (B.10). The resultant equations for heat generation due to electrolysis during recharge for vented cell designs are shown in Equation (B.12) and Equation (B.13):

q J − e − chg −CC = (C8 / 5 × K c × K OC ) × (Vs − avg − chg − ( nc / s × Ve ))

(B.12)

where J–ee–chg–CC q KOC

is the effect constant-current charging heat generation electrolysis is the Joule charge factor assigned to overcharge reactions, fromdue thetobattery manufacturer or Table B.2

q J − e − chg − CV = (C8 / 5 × K c × K OC ) × (Vstring − chg − ( nc / s × Ve ))

(B.13)

66




where

qJ–e–chg–CV

is the Joule effect constant voltage charging heat generation due to electrolysis

As will be discussed in the next subclause and regardless of battery chemistry, the extent of electrolysisbased heat generation becomes miniscule and negligible in valve-regulated batteries as long as the main reactions after the initial electrolysis are gas recombination reactions inside each cell.

B.2.2.3.4.3 Joule effect recombination heat Similar to conditions in vented cell designs, water electrolysis in VRLA batteries also starts toward the end of bulk charging and continues during float service. The oxygen evolution reaction starts before hydrogen evolution in a VRLA cell design. Consequently, oxygen reaches the surfaces of the negative plates where it is recombined with either hydrogen or hydrogen ions to regenerate the water previously decomposed. This action at the negative has the effect of suppressing hydrogen evolution and significantly lowers water loss from the battery. In VRLA cells, the oxygen recombination cycle represents a significant share of the current during the later stages of bulk charging or during float service. With no net reaction, the equilibrium voltage is zero volts and all the energy supplied for oxygen recombination becomes heat. This view (Berndt [B10]) is somewhat controversial, but represents the safest position to avoid underestimating heat generation. The current used for calculating the heat from the recombination process should be based on the current flowing through the cell toward the end of charging. The heat from the oxygen recombination process may be represented as shown in Equation (B.14):

q J − O = I r × Vstring × t

(B.14)

where J–O q is the Joule effect heat generation due to oxygen recombination Ir is the portion of the charging current involved in the recombination reaction Vstring is the voltage across the string

Again, like electrolysis heat, the current is only that fraction or proportion that contributes to recombination and is not the total current in a cell or battery string. For VRLA batteries the overcharge processes are predominately oxygen recombination heat, and for estimating total heat from recombination, Equation (B.15) and Equation (B.16) are suggested.

q J −O−chg −CC = C8 / 5 × K c × K OC × Vs −avg−chg

(B.15)

where

qJ–O–chg–CC

is the Joule effect constant-current charging heat generation due to oxygen recombination

q J −O − chg −CV = C8 / 5 × K c × K OC × Vstring − chg

(B.16)

where

qJ–O–chg–CV

is the Joule effect constant voltage charging heat generation due to oxygen recombination

67




B.2.2.3.4.4 Total Joule effect charging heat The total heat generation in charging is the sum of the Joule effect heat generation associated with the electrochemical conversion of the active materials and the overcharge processes. Due to variations in battery chemistries and cell designs there are a minimum of six equations to consider for determining the heat contributions associated with the Joule effect when applied to discharging and charging reactions. Also, the need to distinguish between various lead alloys is eliminated using these equations since their differences will be reflected in operating voltage differences and larger current flows for alloys with lower oxygen and hydrogen overvoltage values. For purposes of cell recharge the ampere-hour input required to convert materials and provide overcharge, including gassing induced electrolyte stirring for VLA batteries, is similar for the various alloys; however, those equivalent ampere-hour values occur at different operating voltages. In VRLA cells the alloy and electrolyte immobilization methods make only small differences in heat generation calculations and tend to be negligible given the generally greater variability in the facility’s ability to remove heat via radiation, natural convection, forced-air convection, and conduction. Rather than list all the equations here, the next subclauses list all combinations of heat related factors in sets of equations for various battery operating conditions.

B.2.2.4 Total heat generation Equation (B.17) and Equation (B.18) outline the basic thermal components for estimating a cell’s total heat generation:

qW = qrev + q J

(B.17)

where

qW qJ

is the total heat generated, in watts is the total Joule effect heat

As shown inwhich the preceding subclauses, components, yields Equation (B.18):Joule effect heat can be further broken down into its various

qW = qrev + q J − r + q J − e + q J − O

(B.18)

where

qJ–r qJ–e qJ–O

is the total Joule effect heat of the reaction is the total Joule effect heat from electrolysis is the total Joule effect heat from oxygen recombination

Based on these generalized equations the heat generation of any aqueous battery system can be estimated if the appropriate data on each battery chemistry is known or available.

B.2.2.5 Dependency of heat on current

B.2.2.5.1 General Except for the heat of reaction, all other heat generation equations depend on current flow and the Joule effect. The current flowing through a cell must be split among the three current-dependent heat equations.

68




B.2.2.5.2 VRLA current

B.2.2.5.2.1 General Because the hydrogen release of a VRLA is not as directly related to current as it is for a vented cell, equations for currents for VRLA batteries were not developed in Annex A. VRLA battery current varies by the immobilization mode (AGM or gel) and the plate chemistry. While the current is not required for the simplified heat generation calculations found later in this subclause when the discharge time is similar to the rated capacity, it is necessary when charge or discharge times or float times greatly differ from those of the rated cell capacity.

B.2.2.5.2.2 Typical AGM cell current For typical lead-tin or lead-calcium AGM cells, Equation (B.19) gives the upper bound of current through a fully charged cell at 25 °C (C&D [B13]):

I 25 AGM − Ca = C8 × (( 0.0145 × Δ sg ) − 0.013) = P15 × ((3.7 × Δ sg ) − 3.3)

(B.19)

The equations for lead-tin and lead-calcium AGM cells can be further simplified for inclusion in the float mode of Table 1. The upper bound of the current on float would be dependent on the highest expected difference between the average cell float voltage and average fully charged cell specific gravity. The highest expected difference would be 0.980 (for example, an average cell float voltage of 2.28 on a 1.300 s.g. lead-calcium VRLA cell). Plugging this differential into Equation (B.19) yields the results shown in Equation (B.20) and Equation (B.21) (which are the equations found in Table 1 for this cell type):

I AGM − Ca = C8 × (( 0.0145 × Δ sg ) − 0.013) = C8 × (( 0.0145 × 0.980 ) − 0.013) = 0.00121 × C8

(B.20)

I AGM − Ca = P15 × ((3.7 × Δ sg ) − 3.3) = P15 × ((3.7 × 0.980 ) − 3.3) = 0.326 × P15

(B.21)

While VRLA batteries are rarely boost or equalize charged, it does occur; plus in cycling applications, they are often given an accelerated finishing charge. The upper bound of the current in this mode would be dependent on the highest expected difference between the average cell boost voltage and average fully charged cell specific gravity. The highest expected difference would be 1.070 (for example, an average cell maximum boost voltage of 2.35 on a 1.280 s.g. lead-calcium VRLA cell). Plugging this differential into Equation (B.19) from above yields the results shown in Equation (B.22) and Equation (B.23) (which are the equations found in Table 2 for this cell type): I AGM −Ca − boost = C8 × (( 0.0145 × Δ sg ) − 0.013) = C8 × (( 0.0145 × 1.07 ) − 0.013) = 0.00252 × C8

(B.22)

I AGM − Ca − boost = P15 × ((3.7 × Δ sg ) − 3.3) = P15 × ((3.7 × 1.07 ) − 3.3) = 0.659 × P15

(B.23)

Formation of VRLA batteries is almost always completed in the factory, so while a freshening charge might be necessary (even that is rare) before installation if the interval between manufacture and installation is relatively long, an initial charge at a voltage or rate higher than the boost/equalize charge level is not necessary. Therefore, for lead-tin or lead-calcium AGM cells, the equations in Table 5 for current on an initial charge are the same as the equations found in Table 2 for boost/equalize charging. Work done in A.2.8.2.2 showed that in thermal runaway, the highest expected difference between the average cell voltage and average fully charged cell specific gravity would be 1.24. Plugging this differential into the Equation (B.19) from the beginning of this subclause yields the results shown in Equation (B.24) and Equation (B.25) (which are the equations found in Table 6 for this cell type):

69




I AGM − Ca −TR = C8 × ((0.0145 × Δ sg ) − 0.013) = C8 × ((0.0145 × 1.24) − 0.013) = 0.00498 × C8

(B.24)

I AGM − Ca −TR = P15 × ((3.7 × Δ sg ) − 3.3) = P15 × ((3.7 × 1.24 ) − 3.3) = 1.29 × P15

(B.25)

B.2.2.5.2.3 Low-antimony AGM cell current The current drawn by a low-antimony VRLA cell when new is about 2 to 3 times that drawn by a lead-tin or lead-calcium cell (ASHRAE[B9]) [see Equation (B.26)]:

I AGM − lowSb − new = 2.59 × I AGM − Ca

(B.26)

As the low-antimony cell ages, the current draw increases, until at near the end of life (Exide-GNB [B21]) [see Equation (B.27)]:

I AGM − lowSb − old = 5.17 × I AGM − Ca

(B.27)

The equation for low antimony AGM cells can be further simplified for inclusion in the float mode of Table 1. The upper bound of the current on float would be dependent on the highest expected difference between the average cell float voltage and average fully charged cell specific gravity at the end of life. The highest expected difference between average cell float voltage and specific gravity on float would be 0.980 (for example, an average cell float voltage of 2.28 on a 1.300 s.g. VRLA cell). Plugging this differential into Equation (B.20), Equation (B.21), and Equation (B.27) for the current for older low antimony AGM cells yields the results shown in Equation (B.28) and Equation (B.29) (found in Table 1):

I AGM −lowSb − old = 5.17 × I AGM − Ca = 5.17 × 0.00121 × C8 = 0.00626 × C8

(B.28)

I AGM − lowSb − old = 5.17 × I AGM − Ca = 5.17 × 0.326 × P15 = 1.69 × P15

(B.29)

Using the same ratio, but applying it to the accelerated recharge Equation (B.22), Equation (B.23) and Equation (B.27) yield Equation (B.30) and Equation (B.31):

I AGM −lowSb −boost = 5.17 × I AGM −Ca −boost = 5.17 × 0.00252 × C8 = 0.0130 × C8

(B.30)

I AGM − lowSb − boost = 5.17 × I AGM − Ca − boost = 5.17 × 0.659 × P15 = 3.98 × P15

(B.31)

As noted, formation of VRLA batteries is almost always completed in the factory, so while a freshening charge might be necessary (even that is rare) before installation if the interval between manufacture and installation is relatively long, an initial charge at a voltage or rate higher than the freshening charge level is not necessary. However, because the battery is new (as opposed to old), the equations found in Table 5 for a freshening/initial charge for a low antimony VRLA battery are different than those found in Table 2 [see Equation (B.32) and Equation (B.33)]:

I AGM −lowSb −init = 2.59 × I AGM −Ca − boost = 2.59 × 0.00252 × C8 = 0.00650 × C8

(B.32)

I AGM −lowSb −init = 2.59 × I AGM − Ca −boost = 2.59 × 0.659 × P15 = 1.70 × P15

(B.33)

Using the ratio from the beginning of this subclause, but applying it to thermal runaway Equation (B.24) and Equation (B.25), yields Equation (B.34) and Equation (B.35):

I AGM −lowSb −TR = 5.17 × I AGM −Ca −TR = 5.17 × 0.00498 × C8 = 0.0257 × C8

(B.34)

70




I AGM −lowSb −TR = 5.17 × I AGM −Ca −TR = 5.17 × 1.29 × P15 = 6.67 × P15

(B.35)

B.2.2.5.2.4 Gel cell current Gel cells draw half the current of their equivalent AGM counterparts (C&D [B13]) [see Equation (B.36)]:

I gel =

I AGM 2

(B.36)

This leads to Equation (B.37) for float mode, found in Table 1:

I gel − float = 50 % × I AGM − float = 50 % × 0.00121 × C8 = 0.000 605 × C8

(B.37)

It also leads to Equation (B.38) for boost or initial charge mode, found in Table 2 and Table 5:

I gel − boost = 50 % × I AGM − boost = 50 % × 0 .00252 × C8 = 0.00126 × C8

(B.38)

And, it leads to Equation (B.39) for thermal runaway, found in Table 6:

I gel −TR = 50 % × I AGM −TR = 50 % × 0.00498 × C8 = 0.00249 × C8

(B.39)

Gelled VRLAs are generally not used as high rate application batteries because they are space inefficient in this mode. For this reason, no equations are developed using kW/cell (P15) capacity ratings.

B.2.3 Heat generation calculations for various operating modes

B.2.3.1 Charging B.2.3.1.1 General For the charging process the current distribution changes as a function of the SOC. Although the current distribution among the heat generating processes changes continuously during charging some approximation rules may be applied to simplify heat generation estimates. For up to 60% SOC for vented NiCd, 70% SOC for VLA, and 80% SOC for VRLA batteries all current flow goes to the charging reactions, and only the heat of reaction and Joule effect charging heat generation occur up to this point of charging. This is generally referred to in this document as bulk charging. Above 60% SOC for NiCd, 70% SOC for VLA, and 80% SOC for VRLA batteries, the appropriate overcharge reactions of electrolysis and recombination begin to participate and the current distribution splits between the overcharge reactions and the charging reactions. The values of KAM and KOC in Table B.2 provide an approximation of the fractional distribution of the current based on starting the charge at 0% SOC. Even when starting charge at a higher SOC, these values will give a reasonable estimate unless the battery is continuously being cycled at a high SOC (e.g., 70% to 100% SOC) or at a low SOC (e.g., 30% to 60% SOC). Although the true, chemical SOC can vary with discharge rates, the relative SOC for different rates still produces acceptable results with these approximations.

71




B.2.3.1.2 Float charging For float service the heat generation involves principally the overcharge reactions, but not exclusively. There are small, but important side reactions that must be included to appropriately estimate heat generation during periods of float service. A small component of the total current must provide charge compensation for the self-discharge reactions. A listing of known self-discharge rate ranges at 25 °C, and recommended values to use will be provided in Table B.3 following this discussion. As an example, the float current for some VRLA products divert 10% to 15% of the float current into active material charging. For vented lead-acid products that diversion is even larger. In addition to self-discharge correction, the lead-acid systems all experience positive grid corrosion. The corrosion process eventually creates more active material, and this process also diverts electrons from the total float current into the conversion of positive grid lead into lead dioxide. This process can represent as much as 6% of the float current at 25 °C. Due to the significant differences in the corrosion rates among the families of lead alloys, such as pure lead, lead-tin, lead-calcium, lead-selenium, and lead-antimony, the relative fraction of current diverted to this side reaction shows large variations. This important, but small, side reaction will be disregarded or dropped from float heat generation estimates to avoid making the heat calculations more complicated and to allow for a more system protective estimate of heat generation (i.e., calculations will predict heat generation values for lead-acid that will be greater than what will actually occur during normal float service by up to 6%) For VRLA products the fraction of heat associated with electrolysis is negligible and may be ignored. This assumption may not apply early in cell life if the cell contains excessive electrolyte after the manufacturing formation step, or early in the life of gel cells (until oxygen transport cracks develop). Under these conditions, part of the recombination current must be reassigned to electrolysis. Since electrolysis produces less heat than recombination, the heat production calculated decreases when incorporating an electrolysis component; therefore, for heat estimates that apply across a cell’s life the electrolysis component in VRLA should be omitted. For float service considerations Table B.3 gives estimates for the amount of current diverted from overcharge processes to charging processes for counter-acting self-discharge at 25 °C.

Table B.3—Self-discharge rates and adjustment factors for float service heat generation Battery type

VRLA Vented lead-acid Vented NiCd

Self-discharge range

2% to 4 %/mo 3% to 6 %/mo 6% to 12 %/mo

Self-discharge rate to use Approx. current assigned in float heat calculations to self-discharge 3 %/mo 6 × 10–5 A/Ah 5 %/mo 8 × 10–5 A/Ah 8 %/mo 1 × 10–4 A/Ah

The Ah rating from the current column in Table B.3 is calibrated on the C8 rate for lead-acid cells, and the C5 rate for NiCd cells. As temperatures shift away from 25 °C these current values will generally follow Arrhenius kinetic rules for diffusion controlled processes (as described in A.2.5 and A.6.1). With the information already covered, the equations used in Table 1 for the float operating mode can now be derived. Equation (B.18) can be used. For vented cells, the contribution of recombination is small enough to be ignored, as is the contribution of the reversible (non-Joule) heat of reaction. Because the length of time in a float charge is unknown, when substituting Equation (B.8) and Equation (B.11) into Equation (B.18), the time element is removed (so that the results are in watts, rather than watthours ), and the equations are

72




modified to account for proportional currents and individual cell voltages (rather than string voltages). This yields Equation (B.40): qW −vented − flt = q J − r + qJ − e = nc × ((( I × K AM ) × (Vc − avg − flt − Vc −oc )) + (( I × KOC ) × (Vc − avg − flt − Ve )))

(B.40)

where

Vc–avg–flt is the average float voltage per cell Vc–oc is the individual cell open-circuit voltage The maximum difference between average cell float voltage and cell open-circuit voltage for a VLA cell would be 0.16 V (e.g., 2.22 V/cell float for a 2.06 V open-circuit voltage on a 1.215 s.g. cell). The expected highest average float voltage for a stationary vented cell of any specific gravity is usually 2.25 V (on a 1.240 to 1.250 s.g. cell). Substituting these values and the appropriate ones from Table B.2 yields Equation (B.41) for heat generation, found in Table 1 for VLA cells. qW − flt − vented − Pb − acid = nc × (((I × 0.83) × (0.16)) + (( I × 0.17) × (2.25 − 1.48))) = nc × I × 0.264

(B.41)

The typical maximum difference between average cell float voltage and cell open-circuit voltage for a vented NiCd cell would be 0.15 V (e.g., 1.45 V/cell float for a 1.30 V open-circuit voltage). Substituting these values and the appropriate ones from Table B.2 yields Equation (B.42) for heat generation, found in Table 1 for NiCd cells. qW − flt − Ni −Cd = nc × ((( I × 0.80) × (0.15)) + (( I × 0.20) × (1.45 − 1.44))) = nc × I × 0.122

(B.42)

Equation (B.18) can also be used for VRLA batteries. As noted, the contribution of the reversible heat of reaction is small enough to be ignored, as is the contribution of electrolysis. Because the length of time in a float charge is unknown, when substituting Equation (B.7) and Equation (B.13) into Equation (B.18), the time element is removed (so that the results are in watts, rather than watthours ), and the equations are modified to account for individual proportional currents, and individual cell voltages (rather than string voltages). This yields Equation (B.43): qW − flt −VRLA = q J − r + q J −O = nc × ((( I × K AM ) × (Vc − avg − flt − Vc −oc )) + ( I × K c × KOC × Vc − avg − flt ))

(B.43)

The maximum difference between average cell float voltage and cell open-circuit voltage for a VRLA would be 0.135 V (e.g., 2.28 V/cell float for a 2.145 V open-circuit voltage on a 1.300 s.g. cell). The expected highest average float voltage for a VRLA cell of any specific gravity is usually 2.30 V (on a 1.325 s.g. cell). Substituting these values and the appropriate ones from Table B.2 yields Equation (B.44) for heat generation, found in Table 1 for VRLA cells. qW − flt −VRLA = nc × (((I × 0.83) × (0.135)) + ( I × 1.1 × 0.09 × 2.30)) = nc × I × 0.34

(B.44)

B.2.3.1.3 Accelerated/boost/equalize charging As with float charging, a modified version of Equation (B.18) can be used for determining heat generation on an accelerated finishing, boost, or equalize charge. Because the time of this charge is usually set on the charger or plant controller by the user, the output will be in watthours, rather than watts. Because cells subjected to this type of charging are already essentially fully charged, the reversible heat of reaction is not in the equation. For vented cells, Equation (B.38) can be reused and modified, with the substitution of the appropriate voltages as shown in Equation (B.45):

73




qWh − vented − boost = q J − r + q J − e = (nc × t ) × (((I × K AM ) × (Vc − boost − Vc − oc )) + (( I × K OC ) × (Vc − boost − Ve )))

(B.45)

where

qWh

is the total heat generated in watthours

The maximum difference between average cell boost voltage and cell open-circuit voltage for a VLA cell would be 0.27 V (e.g., 2.33 V/cell equalize charge for a 2.06 V open-circuit voltage on a 1.215 s.g. cell). Substituting these values and the appropriate ones from Table B.2 yields Equation (B.46) for heat generation, found in Table 2 for VLA cells. qWh − boost − vented − Pb − acid = t × nc × ((( I × 0.83) × (0.27 )) + (( I × 0.17 ) × ( 2.33 − 1.48))) = nc × I × 0.369 × t

(B.46)

The maximum difference between average cell finishing charge voltage and cell open-circuit voltage for a NiCd cell would be 0.25 V (e.g., 1.55 V/cell finishing charge for a 1.30 V open-circuit voltage). Substituting these values, the current Equation (A.63), and the appropriate values from Table B.2 yields Equation (B.47) for heat generation, found in Table 1 for NiCd cells. qWh−boost − Ni −Cd = t × nc × (((C 5 × 0.00305 × 0.80) × (0.25)) + ((C 5 × 0.00305 × 0.2) × (1.55 − 1.44))) = nc × C 5 × 6.77 × 10 −4 × t

(B.47) For VRLA cells, Equation (B.42) can be reused and modified, with the substitution of the appropriate voltages as shown in Equation (B.48): qWh − boost −VRLA = q J − r + q J − O = ( nc × t ) × ((( I × K AM ) × (Vc − boost − Vc − oc )) + ( I × K c × K OC × Vc − boostt ))

(B.48)

The maximum difference between average cell finishing charge voltage and cell open-circuit voltage for a VRLA cell would be 0.255 V (e.g., 2.40 V/cell finishing charge for a 2.145 V open-circuit voltage on a 1.300 s.g. cell). Substituting these values and the appropriate ones from Table B.2 yields Equation (B.49) for heat generation, found in Table 2 for VLA cells. qWh − boost −VRLA = t × nc × (((I × 0.83) × (0.255)) + ( I × 1.1 × 0.09 × 2.40)) = t × nc × I × 0.449

(B.49)

B.2.3.1.4 Bulk recharge Because the bulk phase of recharge does not involve overcharge or electrolysis, Equation (B.18) can be modified for bulk recharge as shown in Equation (B.50):

qWh = q rev + q J − r − chg

(B.50)

The reversible heat of reaction is exothermic (positive) for lead-acid batteries on charge. Substituting and modifying Equation (B.2) and Equation (B.8) into Equation (B.50) yields Equation (B.51) for a typical bulk recharge at 25 °C:

qWh − Pb − chg = ( I × t × nc × ΔST ) + ( I × K AM × t × nc × (Vc − chg − Vc − oc ))

(B.51)

where

Vc–chg is the charge voltage per cell If the current and the ampere-hour rating of the battery are known, the worst-case bulk-recharge time can be computed as shown in Equation (B.52) (assumes a recharge after a complete discharge):

74




t = Kb − r ×

Cr − s

(B.52)

I

where

Kb–r Cr–s

is the % (in decimal form) SOC of the battery before side reactions begin is the ampere-hours removed (per string) during the discharge that must be replaced

The ampere-hours removed can be calculated very simply as shown in Equation (B.53):

Cr − s = I disch − s × t disch

(B.53)

where

Idisch-s is the average discharge current per string tdisch is the time of the discharge, in hours Because the specific gravity of fully charged lead-acid batteries significantly affects the entropy, the entropy can be represented by a fairly linear equation, based on the values from Table B.1:

ΔST = ( −0.227 × ( s.g.)) + 0.351 Typically, the bulk charge voltage is set at the same level as the float voltage, but it takes it a while to get there. That said, from a worst-case heating perspective, it can be assumed that Vc–chg is equal to Vc–flt. As noted in B.2.3.1.2, the maximum difference between average cell float voltage and cell open-circuit voltage for a VLA cell would be 0.16 V (e.g., 2.22 V/cell float for a 2.06 V open-circuit voltage on a 1.215 s.g. cell). Substituting all of these values (plus the time and entropy equations) yields Equation (B.54):

qWh −vented − Pb− chg = ( K b − r × C8 × nc ) × (( −0.227 × ( s.g .)) + 0.351 + (0.16 × K AM ))

(B.54)

Subclause B.2.3.1.1 points out that Kb–r is approximately 0.70 for VLA batteries, and Table B.2 expresses KAM as 0.83. While VLA batteries can have specific gravities as high as 1.280, for large stationary applications, the specific gravities usually range from 1.215 to 1.250 (use this latter number as the worst case for heat generation). Substituting these values leads to Equation (B.55) found in Table 4:

qWh − vented − Pb − chg = C8 × nc × 0.140

(B.55)

The maximum difference between average cell float voltage and cell open-circuit voltage for a VRLA would be 0.135 V (e.g., 2.28 V/cell float for a 2.145 V open-circuit voltage on a 1.300 s.g. cell). Subclause B.2.3.1.1 points out that Kb-r is approximately 0.80 for VRLA batteries, and Table B.2 expresses KAM as 0.91. The highest specific gravity used in VRLA technology is 1.325. Substituting these values leads to Equation (B.56) found in Table 4.

qWh −VRLA − chg = C8 × nc × 0.138

(B.56)

The reversible heat of reaction is endothermic (negative) for NiCd batteries on charge. Taking this into account,voltage as wellforasa the fact NiCd that the average float andopen-circuit cell opencircuit vented celllargest woulddifference be 0.15 Vbetween (e.g., 1.45 V/cellcell float forvoltage a 1.30 V voltage), Equation (B.49) can be modified as shown in Equation (B.57):

qWh − Ni −Cd − chg = ( K b − r × C5 × nc × t ) × ( − ΔST + (0.15 × K AM ))

(B.57)

75




Subclause B.2.3.1.1 points out that Kb-r is approximately 0.60 for NiCd batteries, Table B.1 gives 0.1347 for the entropy of NiCd batteries at 25°C, and Table B.2 expresses KAM as 0.80. Substituting these values leads to the Equation (B.58), found in Table 4:

qWh − Ni −Cd − chg = (0.60 × C5 × nc ) × ( −0.1347 + ( 0.15 × 0.80)) = − 0.00882 × C5 × nc

(B.58)

B.2.3.1.5 Freshening/initial charging As with float and boost/equalize charging, a modified version of Equation (B.18) can be used for determining heat generation on an initial or refresh charge. Because the time of this charge is usually set on the charger or plant controller by the user, the output will be in watthours rather than watts. Because cells subjected to this type of charging are already essentially fully charged, the reversible heat of reaction is not in the equation. For vented cells, Equation (B.38) can be reused and modified, with the substitution of the appropriate voltages as shown in Equation (B.59): qWh− vented −init = q J − r + q J − e = ( nc × t ) × (( I × K AM ) × (Vc −boost − Vc − oc )) + (( I × K OC ) × (Vc −boost − Ve ))

(B.59)

The maximum difference between average cell initial charge voltage and cell open-circuit voltage for a VLA cell would be 0.44 V (e.g., 2.50 V/cell equalize charge for a 2.06 V open-circuit voltage on a 1.215 s.g. cell). Substituting these values and the appropriate ones from Table B.2 yields Equation (B.60) for heat generation, found in Table 2 for VLA cells. qWh −init − vented − Pb − acid = t × nc × I × 0.83 × 0.44 + (( I × 0.17 ) × ( 2.50 − 1.48)) = nc × I × 0.539 × t

(B.60)

The maximum difference between average cell initial charge voltage and cell open-circuit voltage for a NiCd cell would be 0.25 V (e.g., 1.55 V/cell initial charge for a 1.30 V open-circuit voltage). As noted in A.6.4, the boost charge is typically conducted with a constant-current charger putting out 20% of the rated C5 current. Substituting these values and the appropriate ones from Table B.2 yields Equation (B.61) for heat generation, found in Table 1 for NiCd cells. qWh− init − chg − Ni −Cd = t × nc ×

C5 5

× 0.80 × 0.25 + ((

C5 5

× 0.20) × (1.55 − 1.44)) = nc × C5 × 0.0444 × t

(B.61)

For VRLA cells, as noted in A.2.8.2.2, initial/freshening charges (when done—they are rare for VRLA cells) are at the same voltage level as boost/equalize charging. Therefore the same equations found in Table 2 are found in Table 5 for heat generation for VRLA cells.

B.2.3.1.6 Thermal runaway As with float, boost/equalize, or initial charging, a modified version of Equation (B.18) can be used for determining heat generation during thermal runaway. Because the time that the cells will be in thermal runaway is unknown, the output will be in watts. Because cells subjected to this type of charging are already essentially fully charged, the reversible heat of reaction is not in the equation. Nickel-cadmium; and pure lead, lead-selenium, or lead-calcium VLA cells will rarely, if ever, go into thermal runaway. lead-antimony cells more and susceptible towith thisthe phenomenon, particularly towardsHowever, the end of vented their life. Equation (B.38) canare be reused modified, substitution of the appropriate voltages as shown in Equation (B.62): qW − Pb − Sb −TR = q J − r + q J − e = nc × ((( I × K AM ) × (Vc −TR − Vc − oc )) + (( I × K OC ) × (Vc −TR − Ve )))

(B.62)

76




where

Vc–TR is the average voltage across a cell during thermal runaway Subclause A.2.4.2 showed that the worst voltage expected across a vented lead-antimony cell in thermal runaway would probably be 2.46 V, with a worst-case (from a heat generation perspective) open-circuit voltage of 2.06 V/cell. Substituting these values and the appropriate values from Table B.2 yields the result shown in Equation (B.63): qW − Pb − Sb −TR = nc × ((( I × 0.83) × (2.46 − 2.06)) + (( I × 0.17 ) × ( 2.46 − 1.48))) = nc × I × 0.499

(B.63)

For VRLA cells Equation (B.59) can be reused and modified. A.2.8.2.2 derived a maximum thermal runaway per cell voltage for VRLA cells of 2.54 V (with a specific gravity of 1.300; i.e., an open-circuit voltage of 2.145 V/cell). Substituting these values, as well as the appropriate values from Table B.2 yields Equation (B.64): qW −VRLA −TR = nc × ((( I × 0.91) × ( 2.54 − 2.145 )) + (( I × 0.09 ) × ( 2.54 − 1.48))) = nc × I × 0.455

(B.64)

B.2.3.1.7 Cell reversal During cell reversal, the cells that are reversed essentially appear discharged to the charger; therefore, almost all of Equation (B.18) will apply in calculating heat generation. Because the time the cells will be reversed is unknown, time is not included in the equation. For vented cells, as noted previously, the contribution of recombination is small enough to be ignored; thus Equation (B.18) is modified as shown in Equation (B.65) for cell reversal conditions: (B.65)

qW − vented − Pb − cr = q rev + q J − r − chg + q J − e

Substituting appropriately yields Equation (B.66): qW − vented − Pb − cr = ( I × nc ) × (( ΔST + ( K AM × (Vc − flt − Vc − r ))) + (( K c × K OC ) × (Vc − flt − 1.48)))

(B.66)

where

Vc–r

is the typical voltage of a reversed lead-acid cell

The voltage across a reversed lead-acid cell is approximately 1.50 V. As noted earlier, the highest expected float voltage for a VLA cell is 2.25 V. Substituting these values, the appropriate values from Table B.2, and Equation (B.54) yields the worst-case VLA battery heat generation result as shown in Equation (B.67):

qW −vented − Pb −cr = ( I × nc ) × ((−0.227 × ( s.g.)) + 1.13)

(B.67)

Inserting a typical worst-case specific gravity for large stationary vented cells of 1.250 yields Equation (B.68) used in Table 7:

q W −vented − Pb −cr

= I × n × 0.847

(B.68)

c

For vented NiCd cells, because the charge reaction is endothermic, Equation (B.18) is modified as shown in Equation (B.69) for cell reversal conditions:

qW − Ni −Cd −cr = − qrev + q J − r −chg + q J −e

(B.69)

77




Substituting appropriately yields Equation (B.70): qW − Ni − Cd − cr = ( I × nc ) × (( − ΔST + ( K AM × (Vc − flt − Vc − r − Ni − Cd ))) + (( K c × K OC ) × (Vc − flt − 1.44)))

(B.70)

where

Vc–r–Ni-Cd

is the typical voltage of a reversed NiCd cell

The voltage across a reversed lead-acid cell is approximately 0.70 V. As noted earlier, the highest expected float voltage for a vented NiCd cell is 1.45 V. Substituting these values and the appropriate values from Table B.1 and Table B.2 yields the worst-case vented NiCd battery heat generation result shown in Equation (B.71): qW − Ni − Cd − cr = ( I × nc ) × ((−0.1347 + (0.80 × (1.45 − 0.70))) + ((1.25 × 0.20) × (1.45 − 1.44))) = 0.468 × I × nc

(B.71)

Inserting a typical worst-case specific gravity for large stationary vented cells of 1.250 yields Equation (B.72) used in Table 7:

qW −vented − Pb −cr = I × nc × 0.847

(B.72)

For VRLA cells, as noted previously, the contribution of electrolysis is small enough to be ignored; thus Equation (B.18) is modified as shown in Equation (B.73) for cell reversal conditions: (B.73)

qW −VRLA − cr = qrev + q J − r − chg + q J − O

Substituting appropriately yields Equation (B.74): qW −VRLA − cr = ( I × nc ) × ( ΔST + ( K AM × (Vc − flt − Vc − r ))) + ( K c × K OC × Vc − flt )

(B.74)

As noted previously, the highest expected float voltage would be 2.30 V/cell. Substituting this value and the appropriate values from Table B.2 and Equation (B.54) yields the following worst-case equation used for VRLA heat generation [see Equation (B.75)]: qW −VRLA − cr = ( I × nc ) × (( −0.227 × ( s. g .)) + 0.351) + (0.91 × ( 2.30 − 1.50 )) + (1.1 × 0.09 × 2.30)

(B.75)

Inserting a worst-case specific gravity for stationary VRLA cells of 1.325 yields Equation (B.76) used in Table 7: qW −VRLA − cr = I × nc × 1.01

(B.76)

B.2.3.2 Discharge During discharge only, the Joule effect discharge heat must be included with the heat of reaction and no splits in current distribution occur. When calculating discharge heat use the average discharge current and average discharge voltage in the Joule effect heat equations. If those parameters are unknown, or not supplied by a battery manufacturer, then for generating an estimate of heat production employ Equation (B.5) and Equation (B.6), and for the current term calculate an average current based on discharge time and the battery’s capacity rating for that discharge time. Because discharge does not involve overcharge or electrolysis, Equation (B.18) can be modified for discharge as shown in Equation (B.77):

78




qWh = qrev + q J − r − disch

(B.77)

The reversible heat of reaction is endothermic (negative) for lead-acid batteries on discharge. Because the length of the discharge is often unknown, it is safest (from a heat generation perspective) to make the assumption that the discharge lasts until the rated end voltage is reached. Substituting and modifying Equation (B.2), Equation (B.5), and Equation (B.6) into Equation (B.77) yields Equation (B.78) for a typical discharge at 25 °C:

qWh− Pb−disch = − ( I × t × nc × ΔST ) + ( I × t × nc × (Vc −oc − (

Vc −oc + Vc −eod ))) 2

(B.78)

where

Vc–eod is the end-of-discharge voltage for the cell The specific gravity to entropy relationship described in the bulk recharge section can be substituted into Equation (B.78). It is also known that specific gravity of a fully charged lead-acid cell is related to opencircuit voltage by Equation (B79):

Vc−oc = 0.845 + ( s.g.)

(B.79)

Using 25°C as the typical temperature yields Equation (B.80): qWh− Pb− disch = ( I × t × nc ) × ( −((−0.227 × ( s.g.)) + 0.351) + ((0.845 + ( s.g .)) − (

(0.845 + ( s.g .)) + Vc −eod 2

)))

(B.80)

Further simplification of Equation (B.80) yields Equation (B.81): qWh − Pb − disch = ( I × t × nc ) × (((0.227 × ( s.g .)) − 0.351) + (

( s.g .) − Vc − eod 2

+ 0.423 ))

(B.81)

The results are very dependent on the end-of-discharge voltage and that can vary widely depending on the application (e.g., high rate or long duration) and the expected depth of discharge. Further simplification of Equation (B.80) yields Equation (B.82):

V qWh− Pb − disch = ( I × t × nc ) × ((0.727 × ( s.g.) + 0.0718 − c −eod )) 2

(B.82)

The reversible heat of reaction is exothermic (positive) for NiCd batteries on discharge. The heat generation equation becomes Equation (B.83):

qWh− Ni −Cd − disch = ( I × t × nc × ΔST ) + ( I × t × nc × (Vc −oc − (

Vc −oc + Vc −eod 2

)))

(B.83)

Using a worst-case end-of-discharge voltage of 1.0 V/cell (high rate discharge) yields Equation (B.84): qWh− Ni −Cd − disch = ( I × t × nc ) × (0.1347 + (1.30 − (

1.30 + 1.0 2

))) = 0.285 × I × t × nc

(B.84)

For all of these equations, the current on discharge is a function of the load and the number of parallel strings supporting it. For constant power loads, the current will rise slightly as the voltage falls during the discharge, but a load current at an average discharge voltage can be used. For varying load demands, each portion of the discharge curve can be evaluated separately based on the approximate time the current

79




remains at that level. In sum, the discharge current to be used in the equations can be represented by Equation (B.85):

I=

IL Ns

(B.85)

When the parallel strings are of differing capacities, the currents may need to be weighted for each string accordingly.

B.2.3.3 Heating for batteries on constant-current chargers For constant-current charging (fairly uncommon in standby applications), the heating calculations become very easy. Simply take the constant-current output of the charger, divide it by the number of equivalent parallel strings (if there is more than 1 string), then plug that result into the equations of the tables in 7.2.

B.3 Sample battery heat generation calculations for vented lead-acid batteries

B.3.1 Assumptions It may be helpful to take a real-world installation and calculate heat generation under different operating modes, and different assumptions, as follows:

⎯


⎯


⎯

1600 A load

⎯ ⎯

Three parallel strings (24 cells per string) of 4000 Ah 1.215 s.g. vented lead-calcium

⎯


⎯


⎯


⎯

Maximum engineered voltage drop between the batteries and the equipment is 1.70 V

⎯


⎯

Initial charge time is 100 h

⎯

Annual equalize charging for 48 h


B.3.2 Worst-case discharge calculation As batteries discharge, on average, they will share the current equally between similar parallel strings. The following series of equations give the heat release from the batteries during discharge at 25 °C:

I=

I L 1600 = = 533.3 A/string Ns 3

+ Veod V qWh = − q rev + q J − r − disch = ( I × t × nc ) × (− ΔST + (Vc − oc − ( c − oc ))) 2

(B.86)

80




Open-circuit voltage for a 1.215 s.g. cell is 2.06. The end-of-discharge average cell voltage is 1.852 [(42.75 + 1.7 V) ÷ 24 cells/string]. Using discharge tables for this battery, the discharge time to this end voltage at the calculated current is 5.60 h. The entropy for a 1.215 s.g. cell is obtained from Table B.1. Substituting these values into Equation (B.86) yields the following:

qWh = (533.3 × 5.60 × 72) × ( −0.0754 + (2.06 − (

2.06 + 1.852 2

))) = 6150 Wh

B.3.3 Bulk recharge calculation Equation (B.51) gives the base equation for calculating the heat generation during the bulk recharge phase. Equation (B.52) shows how time and current are related to the ampere-hours removed so that they do not have to figure into the equation. Performing the appropriate substitutions yields Equation (B.87):

qWh − Pb − chg = K b − r × C r − s × nc × ( ΔST + ( K AM × (Vc −chg − Vc − oc )))

(B.87)

The ampere-hours removed per string can be calculated from the preceding subclause:

Cr − s = I disch− s × t disch = 533.3 × 5.60 = 2990 Ah/string Substituting the additional appropriate values from Table B.1 and Table B.2, the appropriate bulk recharge SOC factor fromB.2.3.1.1, and some assumptions of B.3.1 yields the following:

qWh − Pb − chg = 0.70 × 2990 × 72 × (0.0754 + (0.83 × ( 2.20 − 2.06))) = 28 900 Wh Note that this heat value will be spread out over a minimum of 6 h.

B.3.4 Normal heat release Normal heat release from batteries on float is extremely low because the float current is extremely low. Modifying Equation (B.40) yields Equation (B.88):

qW − vented − flt = ( nc × I ) × (( K AM × (Vc − flt − Vc − oc )) + ( K OC × (Vc − flt − Ve )))

(B.88)

The current for this same example was calculated in A.3.4 as 0.224 A per string. Substituting the other appropriate values yields:

qW = (72 × 0.224) × ((0.83 × (2.20 − 2.06)) + (0.17 × ( 2.20 − 1.48))) = 3.85 W B.3.5 Battery heat release during initial charging Heat generation due to initial charging may be computed by modifying Equation (B.59) as shown in Equation (B.89):

qWh = (t × nc × I ) × (( K AM × (Vc −init − Vc −oc )) + ( K OC × (Vc −init − Ve )))

(B.89)


qWh = (100 × 72 × 6.26) × ((0.83 × (2.50 − 2.06)) + (0.17 × (2.50 − 1.48))) = 24 300 Wh

81




B.3.6 Boost/equalize charging heat release calculation Although equalizing can be done at many different voltages, as noted in 7.1, it is often limited by the maximum voltage that the equipment can withstand (when the batteries are operating in parallel with the load and the rectifiers/chargers). In this case, the maximum average cell equalize voltage is 2.33 V/cell (56.0 V ÷ 24 cells per string). Modifying Equation (B.59) yields Equation (B.90):

qWh = (t × nc × I ) × (( K AM × (Vc −boost − Vc −oc )) + ( K OC × (Vc −boost − Ve )))

(B.90)


qWh = (48 × 72 × 0.984) × ((0.83 × ( 2.33 − 2.06)) + (0.17 × ( 2.33 − 1.48))) = 1260 Wh

B.4 Sample heat generation calculations for lead-calcium tin VRLA batteries

B.4.1 Assumptions Similar to the preceding subclause, it may be helpful to take a real-world installation and calculate heat generation under different operating modes, and different assumptions, as follows:

⎯


⎯


⎯

1600 A load

⎯

Ten strings (24 cells per string) of 1400 Ah 1.300 s.g. lead-calcium VRLA AGM batteries

⎯


⎯ ⎯


⎯


⎯

Maximum engineered voltage drop between the batteries and the equipment is 2.00 V

⎯


⎯

Initial charge time and/or boost/equalize time is 48 h


B.4.2 Worst-case discharge calculation As batteries discharge, on average, they will share the current equally between similar parallel strings. The following series of equations give the heat release from the batteries during discharge at 25 °C:

I=

I L 1600 = = 160 A/string Ns 10

V − +V qWh = − q rev + q J − r − disch = ( I × t × nc ) × (− ΔST + (Vc − oc − ( c oc 2 eod )))

(B.91)

Open-circuit voltage for a 1.300 s.g. cell is 2.145. The end-of-discharge average cell voltage is 1.865 [(42.75 + 2.00 V) ÷ 24 cells/string]. Using discharge tables for this battery, the discharge time to this end

82




voltage at the calculated current is 7.93 h. The entropy for a 1.300 s.g. cell is obtained from Table B.1. Substituting these values yields the following:

qWh = (160 × 7.93 × 240) × ( −0.0584 + ( 2.145 − (

2.145 + 1.865 2

))) = 24 900 Wh

B.4.3 Bulk recharge calculation Equation (B.51) gives the base equation for calculating the heat generation during the bulk recharge phase. Equation (B.52) shows how time and current are related to the ampere-hours removed so that they do not have to figure into the equation. Performing the appropriate substitutions yields Equation (B.92):

qWh −VRLA − chg = K b − r × C r − s × nc × ( ΔST + ( K AM × (Vc − chg − Vc − oc )))

(B.92)

The ampere-hours removed per string can be calculated from the preceding subclause:

Cr − s = I disch− s × tdisch = 160 × 7.93 = 1270 Ah/string Substituting the additional appropriate values from Table B.1 and Table B.2, the appropriate bulk recharge SOC factor fromB.2.3.1.1, and some assumptions of B.4.1 yields the following:

qWh −VRLA − chg = 0.80 × 1270 × 240 × ( 0.0584 + (0.91 × ( 2.27 − 2.145 ))) = 42 100 Wh Note that this heat value will be spread out over a minimum of 8 h.

B.4.4 Normal heat release Normal heat release from batteries on float is extremely low because the float current is extremely low. Modifying Equation (B.43) yields Equation (B.93):

qW −VRLA − flt = ( nc × I ) × (( K AM × (Vc − flt − Vc − oc )) + ( K OC × K c × Vc − flt ))

(B.93)

The current for this same example can be calculated from Equation (B.20): I 25 AGM −Ca = C8 × ((0.0145 × Δ sg ) − 0.013) = 1400 × ((0.0145 × ( 2.27 − 1.300 )) − 0.013) = 1.51 A

Substituting the other appropriate values yields:

qW = (240 × 1.51) × ((0.91 × ( 2.27 − 2.145)) + (0.09 × 1.1 × 2.27)) = 123 W B.4.5 Heat release during initial charging Initial charging is not typically suggested for VRLAs. However, if done, it could be done under manufacturer guidance for short periods at voltages as high as 2.40 V/cell [see Equation (B.94)]:

q

−

−

−

Wh VRLA init chg

= (t × n × I ) × (( K c

× (V AM

−

c flt

−V

−

c oc

)) + ( K

× K ×V OC

c

−

))

(B.94)

c flt

The current for this same example can be calculated from Equation (B.20): I AGM − Ca = C8 × (( 0.0145 × Δ sg ) − 0.013) = 1400 × (( 0.0145 × ( 2.40 − 1.300 )) − 0.013) = 4.13 A

83




Substituting the other appropriate values yields:

qWh = ( 48 × 240 × 4.13) × ((0.91 × ( 2.40 − 2.145)) + (0.09 × 1.1 × 2.40)) = 22 300 Wh B.4.6 Heat release calculations for equalize/boost charging Although equalizing VRLA cells is not often done, partially shorted cells in VRLA strings are not uncommon. Although equalizing can be done at many different voltages, as noted in Clause 7, it is often limited by the maximum voltage that the equipment can withstand. In this case, the maximum average cell equalize voltage is 2.33 V/cell (56.0 V ÷ 24 cells per string). Rerunning the calculations:

I

−

= C × ((0.0145 × Δ ) − 0.013) = 1400 × ((0.0145 × (2.33 − 1.300)) − 0.013) = 2.78 A

AGM Ca

8

sg

qWh −VRLA −boost = (t × nc × I ) × (( K AM × (Vc − flt − Vc − oc )) + ( K OC × K c × Vc − flt ))

qWh = ( 48 × 240 × 2.78) × ((0.91 × ( 2.33 − 2.145)) + (0.09 × 1.1 × 2.33)) = 12 900 Wh Assuming worst-case temperature (49°C for this example), the calculations are redone as follows: ⎛ ⎜ 8555 × ⎛⎜ .00335 − 1 ⎜ ⎜ TK ⎝

I S = I 25 × e ⎝

⎞ ⎞⎟ ⎟⎟ ⎠ ⎟⎠

⎛ ⎞ ⎜ 8555 × ⎛⎜ .00335 − 1 ⎞⎟ ⎟ ⎜ ⎜ 322 ⎠ ⎟⎠ ⎝

= 2.78 × e ⎝

= 22.5 A

qWh = (48 × 240 × 22.5) × ((0.91 × (2.33 − 2.145)) + (0.09 × 1.1 × 2.33)) = 104 000 Wh

B.5 Sample heat generation calculations for vented NiCd batteries

B.5.1 Assumptions For the calculations in the subsequent subclauses, the following assumptions apply:

⎯

15 min backup design UPS with a nominal –48 Vdc bus

⎯

Charge voltage of 58.0 V (average of 1.45 V/cell)

⎯

640 A average discharge load

⎯

Two parallel strings (40 cells per string) of 115 Ah ( C5 rate) vented NiCd cells

⎯


⎯


⎯

UPS inverter operating voltage limits are 40.0 V to 60.0 V

B.5.2 Sample heat release during discharge for a UPS NiCd battery From B.2.3.2, the Equation (B.83) for computing heat release of a NiCd battery system during discharge is:

V +V qWh− Ni −Cd −disch = ( I × t × nc × ΔST ) + ( I × t × nc × (Vc −oc − ( c −oc c −eod ))) 2 There are two strings, so the current used in the equation should be 320 A. At this current, for a 15 min (0.25 h) discharge, using the manufacturer’s tables, the end-of-discharge voltage is 1.023 V/cell. As noted 84




previously, the open-circuit voltage of a NiCd cell is 1.30 V. From Table B.1, the entropy at the stated temperature is 0.1347. Substituting the appropriate values into Equation (B.83) yields the following: qWh − Ni − Cd − disch = (320 × 0.25 × 80) × (0.1347 + (1.30 − (

1.30 + 1.023 2

))) = 1748 Wh

B.5.3 Sample heat release during bulk recharge for a UPS NiCd battery Modifying Equation (B.50), Equation (B.51), and Equation (B.57) to compute heat release of a NiCd battery system during bulk recharge yields:

qWh − Ni −Cd − chg = ( K b − r × C r − s × nc × t ) × ( − ΔS T + ((Vc − chg − Vc − oc ) × K AM )) where

Cr–s

is the ampere-hours removed during the discharge

The ampere-hours removed can be calculated as follows:

C r − s = I s × t = 320 × 0.25 = 80 Ah Substituting the appropriate values yields:

qWh − Ni −Cd − chg = (60 % × 80 × 80) × ( −0.1347 + ((1.45 − 1.30 ) × 0.80 )) = − 56 .4 Wh )

85




Annex C

(informative) Existing U.S. codes and standards There are multiple codes and standards relating to batteries, but most of them only have limited information regarding ventilation. The general guidelines call for limiting the buildup of explosive and toxic gases to certain levels (see Annex D). Some codes call for taking measures that will limit abnormal conditions (such as measures that will mitigate the possibility of thermal runaway or fire). Other IEEE practices (including this document) acknowledge additional methods of thermal management of batteries that will also mitigate the possibility of thermal runaway. The National Electrical Code (NEC) (NFPA 70®, 2011 Edition) [B29], Article 480, and the National Electrical Safety Code® (NESC®) (Accredited Standards Committee C2-2012) [B1] require ventilation that will avoid buildup of gases to an explosive level. 18 This is the least stringent guideline, since the lower explosive limit (LEL) of hydrogen, as defined in the codes, is 4%. NOTE that he actual explosive level of hydrogen is generally considered to be approximately 15% to 18% concentration, but most fire codes consider the lower flammability limit (LFL), which is approximately 4% and the LEL to be the same.) Other codes are more stringent, so generally, this will not be the governing code for ventilation. The NEC Handbook [B30] provides further useful guidance in noting that forced ventilation systems are not always needed to meet the requirement. It also notes that VRLA batteries still need ventilation (refer to Clause 7 for the gassing rates of different types of batteries under different operating conditions). Failure of continuously operated or automatically powered ventilation systems should be annunciated. UL 1778 [B37] requires that hydrogen buildup be limited to 2% volume in the space (a safety margin is built in). IEEE recommended practices also specify this level. NFPA 1 (Fire Code) and the International Fire Code (IFC) [B27] (Articles 608 and/or 609, depending on which version is referenced/enforced) require that hydrogen buildup be limited to 1%. Where these codes are adopted, this will be the general design criteria. OSHA (29CFR1910 [B15] and 29CFR1926 [B16]) has special requirements for the buildup of explosive gases in “confined spaces.” Confined spaces where batteries are used are generally limited to vaults. The Federal requirement is that personnel not enter these spaces until testing shows that explosive gases are less than 10% of the LEL (0.4% hydrogen). Clause 7 gives the appropriate equations for computing the hydrogen emitted from batteries under differing operational modes. The mechanical engineer can then design forced or natural ventilation to meet the aforementioned standard limits. IEC EN 50272-2:2001 [B22] also provides calculations using Faraday’s law, modified by exacerbating conditions, such as faulty cells, aged cells, etc. Under conditions of extreme thermal runaway, toxic gases might be produced. For lead-acid batteries, the most common of these is hydrogen sulfide. OSHA (29CFR1910 [B15]) limits the buildup of this gas to 50 ppm (this is half the level at which the gas becomes less detectible to the nose, and toxic to the human body) for a 10 min exposure and 20 ppm for an 8 h exposure. At these levels (and as low as 5 ppb), this gas has a distinctive “rotten egg” odor. If the ventilation system is designed to limit the hydrogen buildup to less than explosive levels, it will easily be able to limit hydrogen sulfide (and other toxic gases) to levels below the OSHA maximum exposure limits. The OSHA limit for hydrogen sulfide is 20 ppm. For more information about hydrogen sulfide, see Annex F.

18

National Electrical Safety Code and NESC are both registered trademarks and service marks of the Institute of Electrical and Electronics Engineers, Inc.

86




It should be noted that hand-held gas sensors not specifically designed to detect hydrogen sulfide may incorrectly register hydrogen sulfide as carbon monoxide or other gases. While ventilation is the response of choice for accumulations of any unwanted gas, personnel in the space should be aware of the distinctive rotten egg smell and evacuate the space unless they are emergency responders wearing self-contained breathing apparatus (SCBA), also called air packs. If people are sent for medical evaluation the suspected exposure to hydrogen sulfide should be reported to the medical staff based on an awareness of the odor along with any suspected exposure to carbon monoxide or other toxins as the meter may report. The meter reading may be false but the point is not to confuse the medical staff by withholding the possible hydrogen sulfide exposure when the rotten egg odor was present simply because some meter identified other gases. Carbon monoxide is colorless and odorless. The rotten egg odor is distinctive of hydrogen sulfide and should be reported as such. The IFC and NFPA 1 require temperature compensation or another method of charge current limiting for VRLA batteries. IEEE Std 1187™[B23], IEEE Std 1188™, and IEEE Std 1189™[B24] also recommend these methods. This should help mitigate the possibility of thermal runaway. This document, in conjunction with IEEE Std 1187™ [B23], IEEE Std 1188™, and IEEE Std 1189™ [B24], notes that other thermal management ventilation-related measures can be taken to mitigate the possibility of thermal runaway. These include air spacing and/or forced airflow between individual VRLA containers. Note that most fire codes (NFPA 1, the IFC, etc.) require shutdown of forced ventilation systems when a fire is detected in a battery room. Therefore, to get rid of toxic gases produced by the combustion of battery materials, other methods of ventilation are needed before a person can enter that room unprotected. Appropriate response and ventilation for fire and elements of combustion are beyond the scope of this document, but it is recommended that any personnel (preferably firefighters) entering a burning (or recently burned) battery room be wearing appropriate breathing apparatus. NFPA 75 [B31] specifies ventilation coordination between rooms in a data center environment. Refer to Department of Transportation Title 49 (49CFR622 [B18]), sections 229.43, 248.425, and 393.30, for ventilation requirements of stationary batteries during transport. Environmental Protection Agency (EPA) Title 40 (40CFR300–399 [B17]), section 68, regulates the air quality required during an electrolyte spill and/or neutralization.

87




Annex D

(informative) Explosive and toxic gas allowance considerations D.1 Permissible hydrogen concentrations As noted in Annex C, various standards and codes in the U.S. cite at least four different levels of maximum allowable hydrogen concentrations: 4%, 2%, 1%, and 0.4%. Which one is correct? From a scientific perspective, the lower flammability level (LFL) of hydrogen is approximately 4% (refer to CRC [B14]). Regardless of codes and standards, for personnel and equipment safety, hydrogen pockets should not be allowed to exceed 4% (the LFL). Good engineering practice dictates the use of a safety factor. It is acceptable practice that a 50% margin of safety is reasonable, therefore a 2% maximum concentration is recommended. As described in Annex C, some fire and building codes set the threshold at only 25% of the LFL, or a 1% concentration. A decision on whether or not to design the ventilation system to prevent buildup to less than 2% should consider several factors. If the installation is subject (or might ever be subject) to a fire code, then the code dictates the design requirements. If the installation could be considered to be a confined space, then OSHA’s 0.4% limit should be the criteria. In the absence of code requirements, economic factors can be considered. Each point of reduction in the threshold requires a significant effort and cost to achieve it. The risk should be weighed against the cost. Risk is inversely proportionate to the volume of the space to be ventilated, but mitigation costs can increase exponentially.

D.2 Permissible hydrogen sulfide concentrations and responsive actions Under all normal operating modes, batteries do not produce hydrogen sulfide. Hydrogen sulfide is produced in cases of extreme thermal runaway in lead-acid batteries. Various studies and sources have produced the following commonly-accepted human exposure thresholds (expanding on the information already found in Annex C):

⎯

5 ppb is the recognition threshold, the concentration at which 50% of humans can detect an odor resembling rotten egg

⎯

10 ppm to 20 ppm is the borderline concentration for eye irritation

⎯

50 ppm to 100 ppm leads to eye damage

⎯

At 150 ppm to 250 ppm, the olfactory nerve is paralyzed after a few inhalations, and the sense of smell disappears, often together with awareness of danger

⎯

320 ppm to 530 ppm leads to pulmonary edema with the possibility of death

⎯

530 ppm to 1000 ppm causes strong stimulation of the central nervous system and rapid breathing, leading to loss of breathing

⎯ ⎯

800 ppm is the lethal concentration for 50% of humans for 5 min of exposure (LC50) Concentrations over 1000 ppm cause immediate collapse with loss of breathing, even after inhalation of a single breath

While it is highly dependent on the severity of the thermal runaway, and the volume and confinement of the space, typical concentrations of hydrogen sulfide from thermal runaway generally do not exceed 20 ppm,

88




and rarely 50 ppm; although concentrations of 70 ppm to 200 ppm have been reported in a few extreme cases. So far, there are no known cases of death due to hydrogen sulfide poisoning from battery thermal runaway. However, there have been several hospitalizations. In case of detection of hydrogen sulfide emission by lead-acid batteries, charging should be immediately disconnected/stopped minimizing safety risks to the extent possible, and the space should be ventilated to the outside atmosphere. If there are worries about exposure while trying to disconnect charging, qualified personnel with SCBA [which may be a hazardous materials (hazmat) response team or a fire department] should perform that function and ventilate the space. For additional information on hydrogen sulfide permissible concentrations, see 29CFR1910.1000, Table Z-2 [B15].

D.3 Permissible arsine and stibine concentrations In cases of extreme thermal runaway of lead-acid batteries containing antimony, arsine and stibine toxic gases can be produced. Various studies and sources have produced the following commonly accepted human exposure thresholds for arsine:

⎯

0.05 ppm is the OSHA limit for arsine

⎯

0.5 ppm is the recognition threshold, the concentration at which 50% of humans can detect the characteristic odor resembling a slight garlic smell

⎯

0.5 ppm is the borderline extended exposure concentration for poisoning

⎯

3 ppm exposure for 1 h can lead to poisoning

⎯

10 ppm can be fatal for long exposure times

⎯

25 ppm to 30 ppm for 30 min can be fatal

⎯

Exposure to concentrations of 250 ppm or higher is rapidly fatal

For additional information on permissible arsine concentrations see 29CFR 1910.1000, Table Z-1 [B15]. Various studies and sources have produced the following commonly accepted human exposure thresholds for stibine:

⎯

0.1 ppm is the recognition threshold, the concentration at which 50% of humans can detect the characteristic odor resembling a rotten egg

⎯

0.1 ppm is the OSHA limit for stibine

⎯

5 ppm is the immediate dangerous to life and health (IDLH) limit with a maximum exposure time of 30 minutes

⎯

40 ppm to 45 ppm for 1 h can be fatal

For additional information on permissible arsine concentrations, see 29 CFR 1910.1000, Table Z-1 [B15]. Arsine and stibine are produced at extremely high temperatures in extreme thermal runaway. While these are toxicrunaway gases, they would produced in quantities than sulfide during extreme thermal event. It isbe highly unlikely that they much wouldlower be fatal inhydrogen such an event before theanhydrogen sulfide. What this means in practical terms is that if the user detects and properly mitigates the environment for hydrogen sulfide (which is the easiest of these toxic gases to detect due to its very strong odor), they will also take care of any arsine or stibine problem that might also exist.

89




Annex E

(informative) Thermal runaway When a VRLA cell is operating on float or overcharge in a fully recombinant mode, there is virtually no net chemical reaction and almost all of the overcharge energy results in heat generation. If the design of the system and its environment are such that the heat produced can be dissipated and thermal equilibrium can be reached, then there is no thermal runaway problem. However, if the recombination reaction gives rise to a rate of heat generation that exceeds the rate of heat dissipation, the battery temperature will rise and more current will be required to maintain the float voltage. The additional current results in still more recombination and heat generation, which further raises battery temperature and so on. The net effect can be accelerated dry out and/or melting of the battery. This potential problem is further aggravated by elevated ambient temperatures, preexisting dry out, or by charging system malfunctions. Conditions contributing to thermal runaway can be minimized by ventilation between and around the cells and by limiting the charger output current and voltage. One such technique is the use of temperaturecompensated chargers. The ability of temperature-compensation process is limited by the minimum voltage required to keep the battery charged. This limits the practical ability of the compensation process for elevated electrolyte temperatures to approximately 38 °C (100 °F). In a gelled electrolyte battery, the gel has intimate contact with the plates and container walls. This design allows better heat dissipation characteristics than an absorbed glass mat (AGM) electrolyte battery, but not as well as the vented (flooded) system. In vented designs, thermal runaway is most probable in lead-antimony designs (antimony content >4%). In lead-antimony batteries, the float current naturally increases over time. This natural increase is caused by antimony poisoning of the negative plate. At end of life the float current increases rapidly and makes the batteries susceptible to thermal runaway. Other vented designs (lead-calcium, pure lead, and NiCds) can still suffer thermal runaway. Vented thermal runaway is generally associated with a reduction in the cell’s internal resistance due to increasing temperature but can also be caused by a breakdown of the separator membranes allowing oxygen to reach the negative plate resulting in recombination heating. Other causes include charger malfunction, shorted cells, and lack of maintenance. Conditions that can lead to or contribute to thermal runaway include the following:

⎯

High operating ambient temperature without compensation of float voltage or other methods of controlling charge current

⎯

Improper float voltage adjustment

⎯

Individual cell failure within a battery string

⎯

Charger failure resulting in high output voltage, current, or ripple

⎯ ⎯

Oversized or excessive number of chargers (which would supply too much recharge current)

⎯

A ground fault in the battery in a grounded system

Insufficient cell/unit spacing and/or ventilation

90




In an uncorrected thermal runaway condition:

⎯

High charging current and recombination inefficiencies result in excessive gas evolution, venting, eventual dry out, and failure.

⎯

Ultimately cells vent, dry out, and fail. When charging current is at maximum levels, the battery temperature can cause cell meltdown leading to a fire or explosion.

The possibility of thermal runaway can be minimized through periodic surveillance and the use of temperature compensated charging (or other charge-current-limiting methods) and the incorporation of HVSD circuitry in the charger or rectifier.

E.1 Thermal runaway in NiCd batteries Phenomena such as thermal runaway normally do not occur in vented NiCd batteries. In fact, thermal runaway is rare in any vented battery design. This phenomenon is further reduced in NiCd batteries due to a steep rise in battery voltage when the battery gets fully charged. See Figure E.1. This rise in battery voltage will not allow the rectifier to sustain high battery current, which will be required to get the battery into thermal runaway condition. For these reasons, thermal runaway in NiCd batteries is nearly always the result of separator failure resulting in recombination heating or charger runaway in conjunction with low electrolyte levels.

70 68 66 ) 64 (V 62 e g60 a tl 58 o V56 g54 n i r 52 t S 50 48 46 44 42

0.500

There's not enough voltage on a Telecom power s ystem to sustain high charge currents in the NiCd.The large “charge voltage step” protects against high charge currents and minimizes “THERMAL RUNAWAY”.

0.450 0.400 Large "charge voltage step" required to0.350 maintain high charge currents when fully charged. It's the nature 0.300 of the chemistry.

Voltage trend, with NO voltage control

0.250 54.4V Setpoint

te a R C

0.200 Voltage

0.150 12.5 Amps for NCX125

Current Under normal operation, current decreases to float levels due to voltage setpoint

0.100 0.050 0.000

0

2

4

6

8

10

12

h) Charge Time (hrs) Figure E.1—100Ah NiCd recharge characteristics

91




Annex F

(normative) Hydrogen sulfide gas Following a lead-acid battery thermal runaway event people might complain of a bad or “rotten egg” smell, or tingling of the nose. They are most likely reacting to hydrogen sulfide (H 2S) gas. Darkening of copper is also an indication of H2S. Thermal runaway in a battery does not always expel H 2S, and reports of health effects (other than reaction to unpleasant smell) are rare. H 2S poisoning outside of industrial settings is extremely rare. H2S is formed from the breakdown of the sulfuric acid electrolyte, but the required conditions for its formation are not present in every thermal runaway event. H2S is common in nature, frequently as a result of rotting vegetation or animal manure. The human nose can detect H 2S at levels as low as 0.005 ppm to 0.02 ppm. The National Institute of Environmental Health Sciences says H 2S can be detected at about 1/400 of the threshold for harmful human effects. OSHA Personnel Exposure Limit (PEL) for hydrogen sulfide (29CFR1910 [B15]) is an 8 h time weighted average at a concentration of 2 mg/m3. While there is some evidence of risk from long-term exposure to H2S, there is no evidence of risk from short-term, moderate levels of exposure. Symptoms of exposure include eye, nose and throat irritation, and sometimes headaches. Chronic exposure can cause asthenia and bronchitis. At extreme concentrations serious illness or death can result (>800 ppm). For exposures below 250 ppm, recovery occurs quickly if exposure to H2S is brief, and there should be no long-lasting effects (eMedicine [B19]). The concentration of H2S in a battery would be greatest within a few millimeters of the battery vents and will dissipate rapidly with distance. Because the amount of H 2S given off during a VRLA thermal event is so tiny, the risk is usually insignificant. H2S can be detected well before it is harmful. However, when H 2S is detected it is prudent to provide supplementary ventilation to the room and/or to exit the area (APC [B6]).

92




Annex G

(informative) Bibliography [B1] Accredited Standards Committee C2-2012, National Electrical Safety Code® (NESC®). 19 [B2] Alcad NH-1, Short Circuit Currents of Alcad High-Rate Nickel-Cadmium Batteries , Stockholm, Sweden, 1990, p. 1. [B3] Alcad NL-1, Short Circuit Currents of Alcad Long-Duration Nickel-Cadmium Batteries, Stockholm, Sweden, 1990, p. 1. [B4] Alcad NM-1, Short Circuit Currents of Alcad General Purpose Nickel-Cadmium Batteries , Stockholm, Sweden, 1990, p. 1. [B5] Alcad NX-1, Short Circuit Currents of Alcad Engine Starting Nickel-Cadmium Batteries, Stockholm, Sweden, 1990, p. 1. [B6] APC White Paper 39, “Battery Technology for Data Centers and Network Rooms: VRLA Reliability and Safety,” Providence, RI, 2002. [B7] Archives of Environmental Health, Health effects from chronic low-level exposure to hydrogen sulfide, Legator et al., (2001 Mar-Apr). [B8] ASHRAE Handbook, Chapter 27: Fundamentals, Ventilation and Infiltration, Atlanta, GA, 2005. 20 [B9] ASHRAE TC 1.6, Terminology of HVAC and Refrigeration, Atlanta, GA, 1991. [B10] Berndt, Dietrich; Maintenance-Free Batteries: Lead-Acid, Nickel/Cadmium, Nickel/Hydride: A Handbook of Battery Technology. Taylor & Francis Group, London UK, 1997, ISBN-10: 0471939609. [B11] Bode, Hans; Lead-acid Batteries (Electrochemical Society). John Wiley & Sons Inc., Hoboken, NJ, October 1977, ISBN-10: 0471084557. [B12] C&D Form 41-6739, Valve Regulated Lead-acid Battery Gassing and Ventilation, Milwaukee, WI, 1995, p. 2. [B13] C&D Form 41-7329, Valve Regulated Lead-acid Battery Life Expectancy and Temperature , Milwaukee, WI, 1999, p. 1. [B14] CRC Press Taylor and Francis Group, CRC Handbook of Chemistry and Physics, 91 st Edition. Edited by William M. Haynes, National Institute of Standards and Technology, Boulder, Colorado, USA, ISBN 9781439820773. [B15] Code of Federal Regulations Title 29 (29CFR1910) Chapter 17, Occupational Health and Safety Agency, Subpart Z, Toxic and Hazardous Substances.21 [B16] Code of Federal Regulations Title 29 (29CFR1926) Chapter 17, Occupational Health and Safety Agency, Safety and Health Regulations for Construction. [B17] Code of Federal Regulations Title 40 (40CFR300–399) Chapter I, Environmental Protection Agency, Subchapter J, Parts 300–399. [B18] Code of Federal Regulations Title 49 (49CFR622) Subpart B, Part 622, Department of Transportation, Environmental Impact and Related Procedures.

19 20 21

The NESC is available from The Institute of Electrical and Electronics Engineers (http://standards.ieee.org/). Publications are available from ASHRAE (http://www.ashrae.org/). CFR publications are available from the U.S. Government Printing Office (http://www.gpo.gov/).

93




[B19] eMedicine, Emergency: Toxicity of Hydrogen Sulfide, Omaha, NE, 2005. [B20] EnerSys US-FL-IOM-001, Safety, Storage, Installation, Operation & Maintenance Manual: Flooded Lead-acid Batteries C, D, E, F, and G; Reading, PA, 2007, p. 45. [B21] Exide-GNB Section 26.10, Absolyte IIP Specifications, Schaumburg, IL, 1990, p. 17. [B22] IEC EN 50272-2:2001, Safety Requirements for Secondary Batteries and Battery Installations: Stationary Batteries. 22 [B23] IEEE Std 1187™-2002, IEEE Recommended Practice for Installation Design and Installation of Valve-Regulated Lead-Acid Batteries for Stationary Applications.23, 24 [B24] IEEE Std 1189™-2007, IEEE Guide for Selection of Valve-Regulated Lead-Acid (VRLA) Batteries for Stationary Applications. [B25] IEEE Std 1578™, IEEE Recommended Practice for Stationary Battery Electrolyte Spill Containment and Management. [B26] Interactive Learning Paradigms, MSDS Glossary. Blackwood, NJ, 2005. [B27] International Fire Code (IFC).25 [B28] Johnson Controls 41-2128, Dynasty Charging Manual: Sealed Rechargeable Batteries in Standby and Portable Power Applications, Milwaukee, WI, 1990, pp. 2–5. [B29] NFPA 70®, 2011 Edition, National Electrical Code ® (NEC®). 26 [B30] NFPA 70®, NEC Handbook, 2011 Edition. 27 [B31] NFPA 75 (2009), Standard for the Protection of Information Technology Equipment. [B32] NFPA 497 (2004), Recommended Practice for the Classification of Flammable Liquids, Gasses, or Vapors and of Hazardous (Classified) Locations for Electrical Installations in Chemical Process Areas. [B33] Portland State Aerospace Society Version 1.03, A Quick Derivation Relating Altitude to Air Pressure, Portland, OR, 2004, pp. 1–4. [B34] Proceedings of Intelec (IEEE PELS 94CH3469-4), Feder, D., and Carosella, G., The Never-Ending Pursuit of Float Voltage Uniformity in Stationary Reserve Battery Plants, Vancouver, BC, 1994, pp. 609–617. [B35] Telcordia GR-1200, Accelerated Life Testing for VRLA Batteries at High Temperatures, Piscataway, NJ, 1992. [B36] Transtronics, Battery Backup Application Handbook. Lawrence, KS, 2010. [B37] UL 1778 (2006), Uninterruptible Power Supplies.

28

[B38] UL 94 (2006), Tests for Flammability of Plastic Materials for Parts in Devices and Appliances. [B39] VARTA 1/e, Special Report: Stationary Lead-acid Batteries with Selenium Alloys , Hanover, Germany, 1980.

22

IEC publications are available from the International Electrotechnical Commission (http://www.iec.ch/). IEC publications are also available in the United States from the American National Standards Institute (http://www.ansi.org/). 23

IEEE publications are available from The Institute of Electrical and Electronics Engineers (h ttp://standards.ieee.org/).

24

The IEEE standards or products referred to in this clause are trademarks of The Institute of Electrical and Electronics Engineers, Inc. 25

Available at http://www.iccsafe.org/.

26

The NEC is published by the National Fire Protection Association (http://www.nfpa.org/). It is also available from the IEEE at http://www.techstreet.com/ieeegate.html. 27 28

NFPA publications are available from the National Fire Protection Association (http://www.nfpa.org/). UL standards are available from Global Engineering Documents (http://www.global.ihs.com/).

94



IEEE Std 1635 -ASHRAE 21-2012 Guide for Ventilation and Thermal Management of Batteries for Stationary Applications

Recommend Documents