Quantitative Risk Assesment - 06_0160_appendix_d

Appendix D

Qu a n ti ta ti v e R i s k A s s e s sment p r e p ared by M o d u Sp e c Au str a lia Pty Li mi ted

CHAPTER13 Land Use

CALTEX REFINERIES (NSW) PTY LTD

Kurnell Refinery

Tank 632 Quantitative Risk Assessment

6th Floor, 34 Queen Street, Melbourne VIC 3000, Tel: +61 3 9614 1285 Fax: +61 3 9614 3917 Email: [email protected] Website: www.moduspec.com Enquiries regarding this report please contact the author For other ModuSpec services contact Lachlan Dreher

Ref: AUS0352.8, Release 01 7 July 2006

Page 1 of 26

ModuSpec

Caltex Refineries (NSW) Pty Ltd Tank 632 Quantitative Risk Assessment

Abstract

Project Title

Tank 632 Quantitative Risk Assessment

Client Name

Caltex Refineries (NSW) Pty Ltd

Job No.

AUS0352.8

Project Manager

Lachlan Dreher

Project Analyst (s)

Lachlan Dreher, Marian Magbiray, Patrick Walker

Report Author (s)

Marian Magbiray, Patrick Walker

ABSTRACT ModuSpec Australia Pty Ltd was engaged to undertake a quantitative risk assessment to analyse the risks associated with the installation of the proposed new crude oil tank (Tank 632). This report details the results of the individual components of the risk assessment, including the hazard identification, frequency assessment and consequence assessment. The individual risk was evaluated in terms of risk of fatality and risk of injury. These results were compared with the applicable criteria to determine the acceptability of the risks associated with the proposed installation. In determining the acceptability of the risks, the impact on the adjacent industrial facilities to the west of the refinery boundary was assessed, with particular emphasis on the Serenity Cove Development. Key Words: PETROL, QRA, BUNDFIRE, STOR Release No. Draft A

Date of Issue 19 April 2006

Draft B

15 June 2006

Draft C

22 June 2006

Draft D

29 June 2006

Release 01

7 July 2006

Reviewed by L. Dreher S. Masterton L. Dreher S. Masterton L. Dreher S. Masterton L. Dreher S. Masterton L. Dreher S. Masterton

Approved by

Reason for Update

L. Dreher

Client Review

L. Dreher

Client Review

L. Dreher

Client Review

L. Dreher

Client Review

L. Dreher

Release To Client

This report, prepared by ModuSpec, is confidential. It has been prepared on behalf of the client mentioned on the cover page (“the client”) and is issued pursuant to an agreement between ModuSpec and the client. It has been produced according to the scope of work and is only suitable for use in connection therewith. All measures and decisions based on this analysis and these findings are the sole responsibility of the client. ModuSpec does not accept: any liability for the identification, indication or elimination of dangers and non-compliances (in the broadest sense of the word), nor for any damage caused by any of these; any obligation to report all facts or circumstances established during the visit. This obligation comes completely under the authority and responsibility of the client any liability for the client’s obligations resulting from (legal) rules and/or statutes; any liability or responsibility whatsoever in respect of or reliance upon this report by any third party. The execution of improvements recommended by ModuSpec does not indemnify the client against any legal or contractual obligations and offers no safeguard against the elimination of dangers or damages resulting from the client’s products, services, company assets, et cetera. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise without prior permission, in writing, of ModuSpec, except for restricted use within the client’s organisation.


Page 2 of 26


ModuSpec Table of Contents

TABLE OF CONTENTS 1. EXECUTIVE SUMMARY ............................................................................. 5 1.1. Fatality Risk ........................................................................................... 5 1.2. Injury Risk ............................................................................................. 5 2.

ACRONYMS & GLOSSARY ......................................................................... 8

3. INTRODUCTION..................................................................................... 10 3.1. Project Scope ....................................................................................... 10 3.2. Locations ............................................................................................. 10 4.

STUDY METHODOLOGY .......................................................................... 12

5. RISK CRITERIA...................................................................................... 14 5.1. Individual Fatality Risk Criteria................................................................ 14 5.2. Individual Injury Risk Criteria.................................................................. 14 6. FACILITY AND OPERATION DESCRIPTION ............................................. 16 6.1. Facility Description ................................................................................ 16 6.2. Process Description ............................................................................... 16 6.3. Meteorological Conditions ....................................................................... 16 7. HAZARD IDENTIFICATION ..................................................................... 17 7.1. Hazardous Materials .............................................................................. 17 7.2. Hazardous Scenarios ............................................................................. 17 8. FAILURE FREQUENCY AND EVENT TREE ANALYSIS................................. 18 8.1. Failure Frequency.................................................................................. 18 8.2. Equipment Failure Scenarios ................................................................... 18 8.3. Event Tree Analysis ............................................................................... 18 8.4. Full Surface Tank Fire ............................................................................ 19 8.5. Bund Fire ............................................................................................. 20 9.

CONSEQUENCE MODELLING ................................................................... 22

10. RISK RESULTS ....................................................................................... 23 10.1. Individual Risk................................................................................... 23 10.2. Comparison with Risk Criterion ............................................................ 23 10.2.1. Fatality Risk................................................................................... 23 10.2.2. Injury Risk .................................................................................... 24 10.3. Major Risk Contributors ...................................................................... 24 10.3.1. Fatality Risk................................................................................... 25 10.3.2. Injury Risk .................................................................................... 25 10.4. Maximum Consequence Impact............................................................ 25 11.

REFERENCES.......................................................................................... 26


Page 3 of 26


ModuSpec Table of Contents

APPENDICES: APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX

A – PROJECT ASSUMPTIONS...........................................................A1 B – HAZARDOUS SCENARIOS AND PROCESS CONDITIONS ..............B1 C – FAILURE FREQUENCY DATA ...................................................... C1 D – HAZARDOUS SCENARIOS AND FAILURE CONTRIBUTORS ..........D1 E – EVENT TREE ANALYSIS ............................................................. E1 F – CONSEQUENCE LEVEL IMPACT CRITERIA .................................. F1 G – CONSEQUENCE RESULTS ..........................................................G1


Page 4 of 26


1.

ModuSpec Executive Summary

EXECUTIVE SUMMARY

Caltex Refineries (NSW) Pty Ltd has proposed the installation of an additional crude oil storage tank (Tank 632), for the Kurnell Refinery. The proposed location for the tank is in the south crude storage area, immediately to the west of Tank 633. Several industrial facilities, most notably the Serenity Cove Industrial Facility, are situated to the west of the refinery boundary, neighbouring the southern crude tank farm and the proposed location of Tank 632. Hence, a quantitative risk assessment (QRA) was undertaken to assess the risk impacts associated with the new installation, and to establish whether these risks comply with the applicable criteria. The full range of potential hazardous scenarios and consequence events associated with the installation and operation of the proposed tank was considered in the analysis. The individual risk was assessed in terms of risk of fatality and risk of injury to determine the impact the proposed installation would have on the surrounding area. 1.1.

Fatality Risk

The 5 x 10-6 per year risk criterion applied in the assessment was based on the guidelines for risk acceptance levels to neighbouring commercial developments, as published by the NSW Department of Urban Resources and Planning (DUAP) [1]. The 5 x 10-6 per year risk contour is presented in Figure 1.1. The 5 x 10-6 per year individual risk contour level is confined within the refinery boundary. This risk level represents the limit of acceptability for risk impact on the neighbouring commercial area of the Serenity Cove Development and therefore with this risk level contained within Caltex’s site, the risk criterion is satisfied. A review of the consequence events that contribute to the western region of the 5 x 10-6 per year individual risk contour indicated that a bund fire associated with the new installation constitutes a major contribution to the risk. The risk assessment was based on whole crude oil service. The modelling of whole crude oil represents the worst-case scenario. 1.2.

Injury Risk

Continued exposure to heat flux of 4.7 kW/m2 is considered sufficient to cause injury [1]. Hence, injury risk was assessed based on exposure to this level of heat flux or greater. The acceptance criteria for risk of injury posed by industrial facilities onto neighbouring residential areas is 50 chances per million per year (i.e. 50 x 10-6 per year) [1]. The 50 x 10-6 per year injury risk contour is depicted in Figure 1.2. The areas bordering the Caltex site near the proposed location of Tank 632 are not residential areas and hence application of this criterion to the neighbouring land-uses represents a high degree of conservatism. Although no injury risk criteria has formally been specified, it is reasonable to consider that less stringent criteria would be applied to commercial and industrial land use, as bounds the Caltex site near the proposed location of Tank 632. With the 50 x 10-6 per year individual injury risk contour lying inside the site boundary, the risk exposure at the site boundary with the Serenity Cove Development is therefore less than this value. This is considered acceptable.


Page 5 of 26

ModuSpec


Executive Summary

N

632

5 u 10-6

622

633 623

100 m

Figure 1.1: Individual Risk of Fatality Contour – 5 u 10-6 per year


Page 6 of 26

ModuSpec


Executive Summary

N

632

50 u 10-6

622

633 623

100 m

Figure 1.2: Individual Risk of Injury Contour – 50 u 10-6 per year


Page 7 of 26


2.

ModuSpec Acronyms & Glossary

ACRONYMS & GLOSSARY

ACRONYMS ADG ALARP Caltex CDU DUAP HUM IR ModuSpec MV NSW PHA P&ID QRA

Australian Dangerous Goods As low as reasonably practicable Caltex Refineries (NSW) Pty Ltd Crude Distillation Unit Department of Urban Affairs and Planning Hold up mass Individual risk ModuSpec Australia Pty Ltd Motorised valve New South Wales Preliminary hazard analysis Piping and instrumentation diagram Quantitative risk assessment

GLOSSARY Acceptance Criteria

Bund

Consequence

Frequency

Hazard

Hazardous Scenario

Individual Risk

Individual Risk of Fatality

Individual Risk of Injury


Defines the level of risk to which an individual is exposed, as either tolerable (negligible risk), intolerable or within the ALARP region. An embankment or wall which may form part or all of the perimeter of a compound around a storage tank, intended to contain any release of liquid. This is the severity associated with an event in terms of toxic doses, fire or explosion etc., i.e. the potential effects of a hazardous event. This is the number of occurrences of an event expressed per unit time. It is usually expressed as the likelihood of an event occurring within one year. A physical situation with the potential for human injury, damage to property, damage to the environment or some combination of these. The identified isolatable sections and/or those which have been broken down into scenarios for specific items of equipment. The frequency at which an individual may be expected to sustain a given level of harm from the realisation of specified hazards. Individual risk with “harm” measured in terms of fatality. It is calculated at a particular point for a stationary, unprotected person for 24 hours per day, 365 days per year. Normally measured in chances of fatality per million years. Similar to individual risk of fatality, however with “harm” measured in terms of injury.

Page 8 of 26


ModuSpec Acronyms & Glossary

GLOSSARY Individual Risk Contours

Isolatable Section Isolation Point Probability

Quantitative Risk Assessment Risk

Risk Reduction


As IR (Individual Risk) is calculated at a point, calculating the IR at many points allows the plotting of IR contours, these being lines that indicate constant levels of risk. Most commonly used are the 1 chance per million-year contour and the 10 chances per million-year contour. A system of pipes or vessels containing the hazardous materials that are bounded by specific isolation points. A point in the process, which can be used to isolate one part of the process from the rest of the system. The expression for the likelihood of an occurrence of an event or an event sequence or the likelihood of the success or failure of an event on test or demand. By definition, probability must be expressed as a number between 0 and 1. A risk assessment undertaken by combining quantitative evaluations of event frequency and consequence. The combination of frequency and consequences, the chance of an event happening that can cause specific consequences. The process of risk assessment coupled to a systematic consideration of potential control measures and a judgement on whether they are reasonably practicable to implement.

Page 9 of 26


3.

INTRODUCTION

3.1.

Project Scope

ModuSpec Introduction

Caltex Refineries (NSW) Pty Ltd has proposed the installation of an additional crude oil storage tank, to be designated as Tank 632, for the Kurnell Refinery. A Preliminary Hazard Analysis (PHA) was initially conducted to provide a semi-quantitative assessment of the risks associated with the proposed new installation and the acceptability of these risks [2]. The PHA was unable to conclusively demonstrate that the risk impact onto the adjacent industrial facility complied with the adopted risk acceptance criterion. The PHA was conducted as a semi-quantitative analysis, based on a series of simplifying assumptions. In order to draw more definitive conclusions about the acceptability of the offsite risk exposure, more detailed quantitative analysis was conducted, i.e. a quantitative risk assessment (QRA). The QRA involved the assessment of the likelihood and consequence for scenarios associated with the process in a quantitative manner, based on data specific to the operation. 3.2.

Locations

The proposed storage tank is to be located in the southern crude storage area, adjacent to the refinery's western boundary. The location of the tank and the bunded area within which it is located was taken from information provided by Caltex [3], which indicated tank size, tank location and the configuration of the bunded area. These specifications have been reproduced in Figure 3.1.


Page 10 of 26


ModuSpec Introduction

Figure 3.1: Proposed Location of Tank 632.


Page 11 of 26


4.

ModuSpec Study Methodology

STUDY METHODOLOGY

The study methodology followed the standard risk assessment steps outlined below. Figure 4.1 presents a flow chart of the risk assessment steps followed. Hazard identification Hazard identification was carried out by a review of the proposed operations and materials handled, in order to identify the equipment and pipelines containing potentially hazardous materials and to define representative hazardous scenarios. Frequency assessment The frequency assessment stage of the analysis involved defining the potential release sources and subsequently determining the likelihood (frequency) of the various releases. The failure frequencies were determined using failure item counts for each of the failure items identified and publicly available historical failure rate data. Details of the failure rate values used are provided in Appendix C. Ignition probability data was used to estimate the probability of a release subsequently being ignited. Consequence assessment The potential consequences from the hazardous scenarios were determined and the impact zones modelled using appropriate software tools. Where possible, the effects of existing mitigation measures at the facility were also taken into account in the consequence assessment. The primary consequence type was a pool fire following a fuel spill. Details of these steps are described in the appropriate sections of the report. A number of assumptions were made during the analysis. Details of the assumptions are presented in Appendix A. Risk assessment: The frequency and consequence assessments were combined to calculate individual risk for both fatality and injury. The risk results have been presented as contours on a site plan. The risk results were then assessed against the selected risk criteria to determine risk acceptability.


Page 12 of 26

ModuSpec


Study Methodology

HAZARD IDENTIFICATION Identification of Hazardous Substance

Detailed process information including plans, process flow diagrams and emergency detection and shutdown systems

Identification of Failure Modes Definition of Failure Case

Select Appropriate Risk Criteria FREQUENCY ASSESSMENT Event Tree Analysis

Failure Rate Data Component Data Specific System Data

End Event Identification

Identify Major Risk Contributors and propose risk reduction measures to achieve acceptable risk levels

End Event Frequency Determination

Chemical Data - Flammability

CONSEQUENCE ASSESSMENT

- Specific Properties

- Fire

Meteorological Data

- Flammable Vapour Dispersion

Ignition Probabilities Explosion Probabilities Detection Strategies Isolation Strategies

Consequence Modelling

Equipment layout and release control and protection systems Determination of Impact Zones

Emergency Response Capabilities

RISK ASSESSMENT Determine Acceptability of Risk

Individual Risk, calculations

Not Acceptable

Comparison with Risk Criteria

Acceptable

Figure 4.1: Risk Assessment Study Methodology


Page 13 of 26


5.

ModuSpec Risk Criteria

RISK CRITERIA

A comparison of the risk against an appropriate target or criterion is required in order to assess the acceptability of that risk. The risk criterion applied for this assessment was obtained from the criteria published by the NSW Department of Urban Affairs and Planning (DUAP) [1]. 5.1.

Individual Fatality Risk Criteria

The NSW criteria are based on a principle that if the risk from a potentially hazardous installation is less than most risks being experienced by the community (e.g. voluntary risks, transportation risks), then that risk may be tolerated. This principle is consistent with the basis of risk criteria adopted by most authorities internationally. The individual risk criteria are as follows: x Hospitals, schools, child-care facilities and old age housing development should not be exposed to individual fatality risk levels in excess of half in one million per year (0.5 x 10-6 per year) x Residential developments and places of continuous occupancy, such as hotels and tourist resorts, should not be exposed to individual fatality risk levels in excess of one in a million per year (1 x 10-6 per year) x Commercial developments, including offices, retail centres, warehouses with showrooms, restaurants and entertainment centres, should not be exposed to individual fatality risk levels in excess of five in a million per year (5 x 10-6 per year) x Sporting complexes and active open space areas should not be exposed to individual fatality risk levels in excess of ten in a million per year (10 x 10-6 per year). These criteria apply to new industry and surrounding land use proposals. In theory, the criteria should apply to existing facilities, however this may not be possible in practice. For existing facilities, an overall planning approach is necessary. In terms of criteria, the following principles should apply [1]: x The 1 x 10-6 per year individual fatality risk level is an appropriate criterion within which no intensification of residential development should take place x Safety updates/reviews and risk reduction at facilities where resultant levels are in excess of the 10 x 10-6 per year individual fatality risk level should be implemented to ensure that operational and organisational safety measures are in place to reduce the likelihood of major hazardous events to low levels. A target level is to be established on an area basis x Intensification of hazardous activities in an existing complex accommodating a number of industries of a hazardous nature should only be allowed if the resultant 1 x 10-6 per year individual fatality risk level is not exceeded by the proposed facility and subject to cumulative risk threshold considerations x Mitigating the impact on existing residential areas from existing hazardous activities (in addition to safety review/updates) should include specific area-based emergency plans. Emergency planning should be developed on the basis of consequences for credible scenarios with emphasis on areas within the 1 x 10-6 per year risk contour. 5.2.

Individual Injury Risk Criteria

The NSW individual injury risk criterion for exposure to heat radiation is as follows [1]: x Incident heat flux radiation at residential areas should not exceed 4.7 kW/m2 at frequencies of more than 50 chances in a million per year.


Page 14 of 26


ModuSpec Risk Criteria

Exposure to heat flux of greater than 4.7 kW/m2 is considered high enough to trigger the possibility of injury for persons who are unable to be evacuated or seek shelter. This amount of heat radiation would cause injury after an exposure period of 30 seconds. This criterion is applicable to residential areas. Injury risk criteria for neighbouring commercial developments or industrial facilities have now been published. The land uses along the site boundary in the area of interest in this study are commercial and industrial. Similar to the relationship between individual fatality risk criteria for residential, commercial and industrial land uses, higher acceptability criteria for injury risk would be expected for commercial and industrial land uses, as compared to that for residential areas. On this basis, for neighbouring commercial land uses, injury risk less than 250 chances in a million per year would be deemed acceptable.


Page 15 of 26

ModuSpec


Facility and Operation Description

6.

FACILITY AND OPERATION DESCRIPTION

6.1.

Facility Description

The location proposed for Tank 632 is the southern crude tank area, directly to the west of the existing Tank 633. There are several light industrial facilities neighbouring the southern crude tank area, including the Serenity Cove Development adjacent to the proposed location. The H.C.E Extractions Facility is located to the north of Serenity Cove. 6.2.

Process Description

The product to be stored in Tank 632 would be whole crude oil. Tankers transporting crude oil are unloaded at the Kurnell wharf and the oil is transferred via pipeline to the storage tanks in the southern crude tank area. The crude oil in the storage tanks is transferred to the refinery Crude Distillation Units (CDU) for further processing. The inventory of crude oil stored in Tank 632 will cycle up and down in line with the transfer of the cargo from the ships and subsequent transfer for processing. Tank 632 would tiein to the existing crude receiving and process plant suction lines. The proposed design and operation of Tank 632 has been modelled on the existing Tank 633. Therefore, the design and operating parameters associated with Tank 633 have been used in the QRA. Tank 633 is a floating roof tank in whole crude oil service. 6.3.

Meteorological Conditions

The local meteorological data was taken from the 2001 and 2002 CALMET weather data files, supplied by Caltex [4]. From this data, the probability of the various atmospheric stability conditions at the site, in addition to the average temperature and wind speed at these stability classes were determined. These values are presented in Table 6.1. The overall average temperature was 18.1qC and the average wind speed was 3.4 m/s. The value for the average humidity used in the analysis (57%) was obtained from the nearby weather station at Sydney Airport [5]. Table 6.1: Meteorological Conditions Analysed

Stability Wind speed (m/s) Temperature (qC) Probability

B 2.4

C 3.7

D 7.2

D 3.5

E 4.0

F 1.9

21.7

20.0

18.0

18.1

16.9

15.7

0.140

0.182

0.126

0.126

0.134

0.292


Page 16 of 26


7.

ModuSpec Hazard Identification

HAZARD IDENTIFICATION

The hazard identification phase involved the review of relevant process information, the site layout, and the proximity of the neighbouring industrial facilities. The following information was used: x The proposed location of Tank 632 and the layout of its bunded area x The proposed piping and instrumentation diagram (P&ID) for the facility x Process conditions, such as flow rates, temperatures, pressures x Frequency and duration of storage tank filling and emptying operations x Typical flow rates of transfer pumps x Local meteorological data. 7.1.

Hazardous Materials

The QRA was based on the storage of whole crude oil in Tank 632. Whole crude oil is a complex mixture containing many different hydrocarbon compounds. The components of the mixture may range from light hydrocarbons having low boiling points, such as ethane, propane and butane, to large heavy components, having very high boiling points. Whole crude oil is highly flammable due to the presence of the light components. It is classified as a Class 3 dangerous good (i.e. flammable liquid) under the Australian Dangerous Goods Code. Whole crude oil therefore presents a potential fire hazard. 7.2.

Hazardous Scenarios

The hazardous scenarios identified for the operation under consideration are associated with the release of crude oil from the tank or associated piping and equipment, or the transfer piping and associated connections (valves, flanges etc.). Depending on the amount of inventory released, such a scenario would result in the formation of a pool of crude oil, with the potential to extend to the full surface area of the bund. Ignition of the spill would subsequently result in a pool fire. In addition to the potential for a fire as a result of a spill, there is also the potential for a tank fire scenario. A full tank surface fire may occur as a result of: x The sinking of the floating roof tank and subsequent product ignition x The escalation of a rim seal fire x Lightning strike. The dispersion of flammable vapour from spills of crude oil to ground or from a tank (e.g. crude oil exposed to atmosphere in the case of a sunken tank roof) was considered during the analysis. The distance to which flammable vapours would extend was short for all cases. Flash fires were therefore not analysed further in the QRA.


Page 17 of 26


ModuSpec Failure Frequency and Event Tree Analysis

8.

FAILURE FREQUENCY AND EVENT TREE ANALYSIS

8.1.

Failure Frequency

The potential for the release of product is attributed to the potential for the failure of any item of equipment within the process. The hazard identification phase involved the identification of specific isolation points within the process under consideration. Any items of equipment and fittings located between these isolation points were therefore assessed as items that could potentially fail and cause a release. The frequency assessment step involved the calculation of the likelihood (ie. frequency) of releases from each of these sources, based on the failure frequency of the individual items within the isolatable section. The Caltex Kurnell Crude System P&ID, inclusive of the proposed tank, was used to identify these isolation points and associated items of equipment and fittings. The failure frequencies were estimated using generic failure frequency data obtained from industry databases. The failure rate data for different types of failure items are summarized in Appendix C. These values are dependent on the equipment/pipe sizes. The leak frequency applied for a mixer seal is detailed in Appendix A. 8.2.

Equipment Failure Scenarios

The overall failure frequency represents the rate at which an item of equipment or pipe will fail, but provides no indication of the magnitude of the failure. Hence, a distribution of hole sizes was assigned to represent the full range of potential failure scenarios. A representative selection of four hole sizes was modelled for each scenario. The guidelines for selecting the sizes were: x Select sizes that fall into the following categories: Small hole up to 10 mm Medium hole 10 mm to 75 mm Large hole 75 mm to 100 mm Rupture Size of pipeline or 1000 mm for vessel. x Examine the spread of hole sizes in conjunction with the failure rates of the contributing failure items. If necessary, adjust the selected hole sizes to ensure that the spread of hole sizes and the associated failure rate is an appropriate representation for the particular line/vessel. A summary of the failure frequency for each of the hazardous sections, according to the representative hole sizes is presented in Appendix D. 8.3.

Event Tree Analysis

Event trees are used to determine the potential consequence events that may result for a given release scenario. Event tree analysis provides a systematic means of determining which factors will influence the release, in addition to the probability associated with each of those factors. The possible outcomes of a release scenario are dependent on the following factors: x Characteristics of the release (e.g. hole size, product released etc.) x Release conditions (e.g. temperature, pressure) x Release detection systems x Intervention systems (i.e. to isolate or contain a release) Ref: AUS0352.8, Release 01 7 July 2006

Page 18 of 26


x x


Ignition sources Consequence mitigation measures.

The event tree analysis includes the following factors: x The probability of detection of a release (by personnel, the standard process control system or via automatic release detection systems) x Following detection of a release, the probability that the release can be successfully isolated x The probability of ignition – both immediate ignition and delayed ignition. The data used in the development of the event trees is presented in Appendix E. An event tree diagram depicting the frequency and probability values associated with the scenario involving a loss of containment of product from Tank 632 as a result of the rupture of the tank is presented in Figure 8.1.

Figure 8.1: Event Tree Diagram for the Rupture of Tank 632 8.4.

Full Surface Tank Fire

For the scenario of a full surface tank fire, insufficient data is available on the causation mechanisms to enable the fire frequency to be estimated directly based on the detailed tank design. Therefore, for the purposes of the QRA, generic frequency data must be used. In selecting an appropriate frequency to apply for a full surface tank fire in the QRA, an extensive review of numerous published sources was undertaken [6 ,7, 8, 9, 10]. Of these, the two sources considered most applicable was the LASTFIRE Project [6] and Technica’s “SingaporeStudy” [10].


Page 19 of 26



The Singapore Study provided data taken from three studies covering storage tank operations in the Netherlands, USA and Scotland, as well as from oil and petrochemical companies operating terminals in Singapore from 1945. The full surface tank fire frequency derived from the USA/Europe and Singapore operations were 2.0 x 10-4 per year and 9.3 x 10-4 per year respectively. The LASTFIRE Project involved the largest study to date undertaken to determine the fire frequency for large floating roof storage tanks. It involved data obtained from 16 companies, operating 2,420 tanks at 164 sites throughout 36 countries over a survey period from 1981 to 1996. The study derived a full surface tank fire frequency of 1.2 x 10-4 per year. The frequency considered most applicable for the Kurnell refinery is 1.2 x 10-4 per year, derived from the LASTFIRE project. This value has been selected because it has been derived from the widest sample set of events and tank locations. Statistically, this can be expected to provide a more appropriate representation of the true event frequency. In addition, both data sources reviewed suggested that there is a correlation between the frequency of storage tank fires and the number of thunderstorm days experienced in the area. When compared with Singapore, the number of thunderstorm days experienced in the Kurnell area is relatively low. This suggests that the expected frequency for a full surface tank fire in Singapore should be higher than at Kurnell. The selected frequency is consistent in this respect, in that it is lower than the value determined solely for operations in the Singapore area. 8.5.

Bund Fire

A bund fire is generated by the ignition of a major release of flammable liquid from a pipe or storage tank into a bunded area. The QRA assessed the frequency of a bund fire based on the release of product from a failure of the tanks or associated fittings in conjunction with the likelihood of ignition. The intervention measures implemented are also considered in the derivation of the consequence frequency values. The likelihood of ignition is dependent on the release rate of the product. The failure frequencies were determined by identifying the various items associated with the tank that may fail. The failure rate for these items, in combination with the failure rate data for the tank itself were combined to determine the overall failure frequency. The LASTFIRE Project estimated the frequency for a large bund fire resulting from a major spill to be 6 x 10-5 per tank per year [6]. The bund fire frequency used in the QRA was determined from a specific analysis of the proposed tank and associated equipment. Failure events that would lead to a major spill into the tank bund, would include large and catastrophic equipment failures. A bund fire would result if such a spill was subsequently ignited. Of the two large bund fire events described in the LASTFIRE Project, one had a release rate1 of ~4.5 m3/min and the other had a pool fire surface area of 232 m2. The failure cases analysed as part of the QRA with an equivalent hole size greater than 100 mm would result in spills of this magnitude. The total frequency of bund fires caused by

1

This event is described as resulting in an 8,000 m2 bund fire, however the magnitude of this bund fire was due to the flooding of the bund with water, which was done as part of the efforts to control the fire. If the bund had not been flooded, the fire area would have been limited to well below this value.


Page 20 of 26


these failures is 6.1 x 10-5 per year. determined by the LASTFIRE Project.


This is comparable to the bund fire value

The bund fire frequency calculated as part of the QRA were considered a more reasonable representation of the scenarios at the proposed facility, as they were derived from a specific analysis of the proposed design. Specific design information was used in the analysis, including equipment parts counts and the proposed failure detection and mitigation measures. This gives an assessment that is more specific to the system under consideration, rather than a more generic frequency value that may be based on widely differing systems.


Page 21 of 26


9.

ModuSpec Consequence Modelling

CONSEQUENCE MODELLING

The consequence scenarios associated with the installation of the proposed Tank 632 were modelled to determine their potential impact on the surrounding area. The modelling took into account the chemical properties of the product released and the meteorological conditions, where applicable. The consequences modelled were based on a release of whole crude oil from the following: x Tank 632 or associated fittings x Transfer piping or associated fittings. The Bernoulli equation was used to determine the liquid release rate for the scenarios considered. The release rate provides a measure of the magnitude of the spill. For releases that are bunded, the size of the liquid pool would be contained, thereby limiting the magnitude of any subsequent pool fire. The effect of bunding has been accounted for in the consequence analysis. The consequence types that could result from the scenarios under consideration included: x Pool fires x Full surface tank fires. Pool fires result from the ignition of a flammable liquid spill. The heat radiation emitted by pool fires was modelled using the Mudan & Croce model [11]. The results of the analysis were used to determine the impact on personnel (fatality rate). The probability of fatality from exposure to heat radiation from a pool fire was determined using a probit equation. The heat flux exposure was calculated at 2 m above ground level, to represent an upper limit of heat flux exposure to a person present near the flame. The heat radiation levels modelled, the resulting fatality probabilities, along with a detailed description of the criteria, are presented in Appendix F. Full surface tank fires were also modelled using the Mudan & Croce model. The likelihood of fatality from heat radiation was determined in a similar manner to pool fires. Tables presenting the impact distances at the specified heat criteria for all the individual events analysed are presented in Appendix G. Although large storage tank fires and bund fires are very hazardous events, if managed correctly, the likelihood of fatality is low. The most likely mechanism for fatality is when a person is involved in the initial flash that ignites the fire. The likelihood of fatalities for members of the public or fire fighters is extremely low and such fatalities are rare. A notable exception to this was an incident in which a number of people were killed when too close to a tank fire when a boilover occurred. This phenomenon is now well understood and any such risks can be well managed by appropriate emergency response procedures.


Page 22 of 26


ModuSpec Risk Results

10. RISK RESULTS 10.1. Individual Risk The overall risk was determined by combining the frequency and consequence data for the individual scenarios examined. The results of the analysis are presented as individual risk contours. Contours were generated for individual risk of fatality and individual risk of injury. These are illustrated in Figure 1.1 and Figure 1.2 respectively. 10.2. Comparison with Risk Criterion The quantified risk results were compared with the applicable risk criteria for land use established by DUAP. This provides the basis for determining the acceptability of the risk. 10.2.1.Fatality Risk Focus was placed on the risk exposure on the Serenity Cove Development located to the west of the refinery border, adjacent to the proposed Tank 632. The Serenity Cove Development is an office building and hence the applicable risk criterion for individual risk of fatality level is 5 chances in a million per year (5 x 10-6 per year). The 5 x 10-6 per year individual risk contour is shown in Figure 1.1. This contour is confined within the beyond the site boundary. Therefore, the risk criterion for individual risk of fatality is satisfied. An overall individual risk contour for the Caltex Refinery has not been developed at this point and hence was not available to enable a risk contour to be developed representing the cumulative risk, including the addition of the proposed tank. In the absence of a risk contour for the existing activities, the cumulative risk exposure on neighbouring facilities from the addition of Tank 632 has been assessed qualitatively. The existing refinery activities in the area where the proposed tank is to be installed include other atmospheric storage tanks and several pipelines. The distance between these other storage tanks and their bunds and the Serenity Cove Development is large (i.e. shortest distance is well in excess of 100 m). Based on this separation distance, there would be negligible contribution to risk of fatality from events involving these tanks and their bunds because the impact distances for most of the scenarios will not extend this far. Several pipelines run near the site boundary. These are as follows: x Lube oil shipping line x Flare line. The lube oil shipping line runs along the site boundary, whereas the flare line is separated from the site boundary by an internal road. The risk exposure associated with pipeline operations is typically low compared to the risk from other items such as process vessels, storage tanks, pumps etc. High risk exposure normally only results when large numbers of pipes run together.


Page 23 of 26



Considering the lube oil shipping line, the risk exposure may considered lower again for the following reasons: x The pipeline is not in continuous use for lube oil transfers x The ignition probability is very low due to the high flash point of the oils being transferred. The risk exposure to adjacent areas from the flare line may be considered low for the following reasons: x The pressure in the flare line would normally be low x The pipeline is separated from the boundary by a short distance. The pressure in the flare line would normally be low as the there would be minimal transfer through the line unless a process upset was in progress. Therefore, if there were a failure of the line, the release rate of vapour from the leak would be low and have a very limited impact zone. The separation distance between the flare line and the Caltex boundary would reduce the likelihood of offsite impacts even further. Based on the discussions above, the existing risk exposure to the Serenity Cove Development and the HCE site from operations in the vicinity of the proposed storage tank is expected to be well within the risk acceptance criteria applicable for these land uses. Given that the existing operations on the Caltex site are not expected to impose high levels of risk to the neighbouring areas, and the risk associated with the addition of Tank 632 does not exceed the criteria, it can reasonably be assumed that the cumulative individual fatality risk from the existing and proposed operations does not exceed the acceptance criteria. For the Serenity Cove Development, the individual risk of fatality from Caltex’s operations considering the addition of the proposed storage tank would be less than 5 x 10-6 per year, and the cumulative individual risk of fatality risk imposed from Caltex on HCE site would be less than 50 x 10-6 per year (i.e. the applicable criteria for neighbouring industrial sites). 10.2.2.Injury Risk The 50 u 10-6 per year individual risk of injury contour is presented in Figure 1.2. This contour represents the criterion applicable to residential areas. The risk contour is contained within the site boundary west of Tank 632. Its containment within the site boundary therefore satisfies the injury risk criterion, regardless of the land use beyond the site boundary in this region. 10.3. Major Risk Contributors The following events contribute to the individual risk to the west of Tank 632, both in terms of risk of fatality and injury: x Significant release from the crude oil tank resulting in a large ignited spill x Full surface tank fire. The contribution to the risk associated with these events is primarily attributed to the magnitude of the resulting fires. In the case of major spills into the bunded area, the entire surface of the bund would be covered and resulting fires would have large impact distances. A full surface tank fire is essentially a large pool fire located at the top of the storage tank. Such fires generate a large amount of heat. The proposed size and location of Tank 632 mean that a full surface tank fire would result in significant heat radiation at the western boundary of the refinery. Ref: AUS0352.8, Release 01 7 July 2006

Page 24 of 26



10.3.1.Fatality Risk The major risk contributor to the western sector of the 5 x 10-6 per year fatality risk contour is a major release from Tank 632, resulting in a large bund fire. As the bund surrounding Tank 632 has a very large surface area, the heat radiation from a full bund fire will extend well beyond the western boundary of the refinery into the neighbouring area. The heat radiation impact from smaller fires will typically not extend a sufficient distance to contribute to the risk at this location. The events that contribute the most to the risk in the eastern sector of the 5 x 10-6 per year risk contour are releases from the transfer piping. These risks impact the adjacent Tanks 622, 623 and 633, however are limited to the bunded areas through which they pass. 10.3.2.Injury Risk The major risk contributor to the 50 x 10-6 per year injury risk contour is a full surface tank fire. The 4.7 kW/m2 heat flux associated with this event has the potential to extend a distance of 83.6 m downwind from the tank centre. Another significant risk contributor is a pool fires resulting from the ignition of significant releases (representative a range in hole sizes starting from 100 mm). 10.4. Maximum Consequence Impact Based on the heat radiation impact distances, the installation of the proposed crude tank at the location nominated will have the potential to generate offsite heat radiation impacts. The major consequence event with the potential to generate offsite impact is a full bund fire resulting from a major release of whole crude oil from Tank 632. This event also represents the maximum extent of heat flux to the west of the refinery boundary. Although this worst-case event will produce large impact zones, the frequency of the event is low (calculated to be in the order of 1 x 10-7 per year). The low event frequency leads to a low contribution to the overall individual risk from this event. Consequently, the impacts of a bund fire can be considered acceptable from a risk perspective.


Page 25 of 26


ModuSpec References

11. REFERENCES 1

Department of Urban Affairs and Planning – Planning NSW, “Risk Criteria for Land Use Safety Planning”, Hazardous Industry Planning Advisory Paper No. 4, March 2002.

2

ModuSpec Australia Pty Ltd, “Caltex Refineries (NSW) Pty Ltd Preliminary Hazard Analysis, Storage Tank 632”, Reference: AUS 0352.1, August 31 2005.

3

Email from Kevin Houlihan (Caltex RPIP Project Engineer (Kurnell) Shedden Uhde Pty Ltd), Lachlan Dreher to (General Manager, ModuSpec Australia Pty Ltd) “Caltex Crude Tank QRA”, 4th May 2006.

4

Email from Ramez Aziz (Senior Risk Engineer, Caltex Refineries (NSW) Pty Ltd), to Marian Magbiray (Risk Engineer, ModuSpec Australia Pty Ltd) “MET Data for Kurnell” (AUSPLUME files for Caltex Kurnell site), 2nd June 2005.

5

Bureau of Meteorology, “Climate Averages for Australian http://www.bom.gov.au/climate/averages/tables/cw_066037.shtml, June 14 2005.

6

LASTFIRE PROJECT, Large Atmospheric Storage Tank Fire Project – LASTFIRE Technical Working Group, June 1997.

7

Offshore Hydrocarbon Release Statistics, Offshore Technology Report – OTO 97 950, UK Health and Safety Executive, December 1997

8

Guidelines for Process Equipment Reliability Data, Center for Chemical Process Safety of the American Institute of Chemical Engineers, 1989, Vessels – Atmospheric-Metallic, page 203, Lower Value.

9

Quantitative Risk Assessment Datasheet Directory, E&P Forum Report No 11.8/250, October 1996, Chapter 10 “Storage Tank Incidents”, Table 3.1 page 6, Atmospheric storage tank-mild steel.

10

Atmospheric Storage Tank Study for Oil and Petrochemical Industries Technical and Safety Committee Singapore, by Technica Ltd, London, April 1990

11

Mudan, K.S. & Croce P.A., "Fire Hazard Calculations for Large Open Hydrocarbon Fires", The SFPE Handbook of Fire Protection Engineering, 1st Edition, 1988.


Sites”;

Page 26 of 26


ModuSpec Appendix A: Project Assumptions

APPENDIX A: PROJECT ASSUMPTIONS TABLE OF CONTENTS 1.

INTRODUCTION........................................................................................ 2

2.

ASSUMPTIONS ......................................................................................... 3 Assumption 1: Modelling Boundaries .................................................................... 3 Assumption 2: Scenario 1 – Crude Receiving ......................................................... 3 Assumption 3: Scenario 2 – Transfer of Crude Oil from Tank 632 ............................. 3 Assumption 4: Scenario 3 – Static Tank 632 ......................................................... 4 Assumption 5: Pipelines ..................................................................................... 4 Assumption 6: Frequency Assessment – Tank and Piping Utilisation.......................... 4 Assumption 7: Frequency Assessment – Full Surface Tank Fires ............................... 4 Assumption 8: Frequency Assessment – Tank Mixers............................................. 5 Assumption 9: Meteorological Data ...................................................................... 6

3.

REFERENCES ............................................................................................ 7


Page A.1 of 7


1.


INTRODUCTION

This appendix documents the assumptions made during the risk analysis. The assumptions have been based on information provided by Caltex. The justification for the assumptions has been included where applicable.


Page A.2 of 7

ModuSpec


2.

Appendix A: Project Assumptions

ASSUMPTIONS

Assumption 1: Modelling Boundaries The analysis is bounded by the proposed Tank 632 and the associated crude receiving and discharge piping to the eastern bund wall of Tank 633. ModuSpec Analyst: Marian Magbiray

Date: 21/09/05

Assumption 2: Scenario 1 – Crude Receiving Isolation section: Crude Receiving Line Pipe 9-P610-KA1-650 at the boundary of the eastern bund wall of Tank 633 to the motorised valve (MV) at the inlet of Tank 632. Product: Whole Crude Oil Temperature: 30qC Pressure: 88 kPa(g) Flowrate: 4000 m3/hr [1]. Utilisation: See Assumption 6. Justification Due to the significant distance from the ship pump, the pressure in the pipeline was assumed to be the same as the hydrostatic head in the tank. ModuSpec Analyst: Marian Magbiray

Date: 21/09/05

Assumption 3: Scenario 2 – Transfer of Crude Oil from Tank 632 Crude Suction Line 9-P611-KA1-450 Isolation section: MV at Tank 632 outlet along the suction piping 9-P611-KA1-450 to the eastern bund wall of Tank 633. Product: Whole Crude Oil Temperature: 30qC Pressure: 88 kPa(g) Flowrate: 400 m3/hr [1] Utilisation: See Assumptions 6 Isolation section: MV at Tank 632 outlet along the suction piping 9-P962-B4-250 to the eastern bund wall of Tank 633. Product: Whole Crude Oil Temperature: 30qC Pressure: 88 kPa(g) Flowrate: 400 m3/hr [1] Utilisation: See Assumptions 6 ModuSpec Analyst: Marian Magbiray


Date: 21/09/05

Page A.3 of 7

ModuSpec



Assumption 4: Scenario 3 – Static Tank 632 Tank 632 was modelled at 85% level [2] Product: Whole Crude Oil Temperature: 30qC Pressure: atmospheric Utilisation: See Assumptions 6 ModuSpec Analyst: Patrick Walker

Date: 09/06/06

Assumption 5: Pipelines x

The pipe lengths were estimated from the site plan provided [3]. The largest pipe diameter in each isolatable section was used to represent the pipe sizing during the modelling.

ModuSpec Analyst: Marian Magbiray

Date: 21/09/05

Assumption 6: Frequency Assessment – Tank and Piping Utilisation The tank utilisation was assumed to be 100%. Based on the frequency and duration of the tank filling and emptying operations, the utilisation for the crude receiving pipeline was estimated to be 10%. The transfer piping for the delivery of crude to the CDU was assumed to be 90%. ModuSpec Analyst: Marian Magbiray

Date: 21/09/05

Assumption 7: Frequency Assessment – Full Surface Tank Fires Based on a review of several sources that have published the frequency values for a full surface tank fire, the frequency considered most applicable for the Kurnell refinery is 8.9 x 10-5 per year, derived from the LASTFIRE project [4]. This value has been selected because it has been derived from the widest sample set of events and tank locations. Statistically, this can be expected to provide a more appropriate representation of the true event frequency.

ModuSpec Analyst: Marian Magbiray


Date: 21/09/05

Page A.4 of 7

ModuSpec



Assumption 8: Frequency Assessment – Tank Mixers Assumption Failure of a tank mixer was considered in terms of the failure of the mixer seal and a leak from that seal. It was assumed that a mixer seal is comparable to a pump seal and that external leaks are considered to be the only reasonable foreseeable failure mode. Referring to OREDA-97 [5], the mean failure rate for pumps and the pump seal failure percentage of the overall failure rate is as follows:

Failure Rate (per 10-6 hrs)

Seal Failure Mode Percentage

Total Failure Mode Percentage

External Leakage

3.64

9.02%

22.92%

Significant External Leakage

0.63

0.26%

0.26%

External Leakage

9.61

9.02%

22.92%


0.63

0.26%

0.26%

Critical Failure Mode

Degraded Failure Mode

Applying the seal failure and total failure mode percentages presented above to the mean failure rate gives the following failure rates that have been applied for seal failures for each failure mode: Failure Mode (Seal)

Critical Failure

Degraded Failure

External Leakage

1.19 x 10-2 per year





This corresponds to a total seal mean failure frequency of 5.61 x 10-2 per year. assumed that a mixer seal failure equates to a maximum hole size of 12 mm.


It is

Page A.5 of 7

ModuSpec



Assumption 9: Meteorological Data The local meteorological data, extracted from the 2001 and 2002 CALMET files were supplied by Caltex Kurnell. [6]. The average temperature is 18.1qC and the average humidity is 57%. was taken from the nearby weather station at Sydney Airport [7].

Humidity data

A total of 16 wind directions and 6 stability classes were used in the analysis. values used are listed below. B

C

D

D

E

F

2.4

3.7

7.2

3.5

4.0

1.9

Average of all Stabilities 3.4

0.140

0.182

0.126

0.126

0.134

0.292

1.0

Stability Wind Speed (m/s) Probability of Atmospheric Conditions Wind Direction N NNE NE ENE E ESE SE SSE S SSW SW WSW W WNW NW NNW

Probability of wind direction 0.031 0.050 0.128 0.128 0.078 0.094 0.084 0.052 0.071 0.025 0.020 0.025 0.051 0.067 0.069 0.029


0.029 0.085 0.113 0.094 0.051 0.044 0.065 0.064 0.116 0.056 0.027 0.051 0.069 0.061 0.049 0.026

0.012 0.062 0.077 0.031 0.002 0.005 0.035 0.082 0.221 0.101 0.053 0.123 0.105 0.054 0.027 0.010

0.021 0.058 0.079 0.067 0.048 0.072 0.111 0.127 0.135 0.064 0.019 0.041 0.048 0.046 0.036 0.029

0.051 0.108 0.069 0.025 0.024 0.025 0.059 0.045 0.090 0.059 0.041 0.107 0.127 0.100 0.038 0.031

0.072 0.078 0.062 0.038 0.038 0.045 0.044 0.030 0.046 0.049 0.060 0.081 0.085 0.099 0.106 0.066

0.042 0.075 0.085 0.062 0.041 0.047 0.063 0.060 0.102 0.057 0.040 0.071 0.081 0.075 0.063 0.037

Page A.6 of 7

The



3.

REFERENCES

1

Caltex Refineries (NSW) Pty Ltd, “Additional Crude Storage Project, Process Data for Phase 2 Preliminary Hazard Analysis”.

2

Email from Tracey Hyland (Environmental Engineer, Kurnell, Caltex Refineries (NSW) Pty Ltd), Lachlan Dreher to (General Manager, ModuSpec Australia Pty Ltd) “Fw: Tank 632 QRA Report”, 9th June 2006.

3

Caltex Kurnell Refinery Neighbourhood Layout for Risk Assessment, Drawing number 127103, Revision 0.

4

LASTFIRE Technical Working Group, “LASTFIRE PROJECT, Large Atmospheric Storage Tank Fire Project” June 1997.

5

Sintef Industrial Management, “OREDA Offshore Reliability Data Handbook”, OREDA Participants, 3rd Edition, 1997.

6

Email from Ramez Aziz (Senior Risk Engineer, Caltex Refineries (NSW) Pty Ltd), to Marian Magbiray (Risk Engineer, ModuSpec Australia Pty Ltd), “MET data for Kurnell” (AUSPLUME files for Caltex Kurnell site), 2nd June 2005.

7

Bureau of Meteorology, “Climate Averages for Australian Sites”; http://www.bom.gov.au/climate/averages/tables/cw_066037.shtml, 14th June 2005.


Page A.7 of 7


ModuSpec Appendix B: Hazardous Scenarios and Process Conditions

APPENDIX B: HAZARDOUS SCENARIOS AND PROCESS CONDITIONS TABLE OF CONTENTS 1.

INTRODUCTION........................................................................................ 2

2.

HAZARDOUS SCENARIOS AND PROCESS CONDITIONS.............................. 3


Page B.1 of 3


1.

ModuSpec Appendix B: Hazardous Scenarios and Process Conditions

INTRODUCTION

This appendix provides details of the hazardous scenarios identified in the analysis and the process conditions relating to them. The definition of the data presented in the tables is outlined below: Scenario Name Product Utilisation (%) Temperature (qC) Pressure (kPa) Diameter (mm or m) Length (m) Height (m) Liquid Level (%) Total HUM Vessel Capacity (t)

The name of the isolatable section or specific equipment considered as the scenario The product representing the material in the scenario The percentage of time the scenario is in use Temperature in the process Pressure in the process Diameter of pipeline or vessel Approximate length of pipeline Height of the vessel Liquid level in vessel The hold up mass for the isolatable section The capacity of the vessel


Page B.2 of 3

ModuSpec


2.

Appendix B: Hazardous Scenarios and Process Conditions

HAZARDOUS SCENARIOS AND PROCESS CONDITIONS

Table 2.1 details the tank scenario modelled. Table 2.2 presents the process conditions for the transfer pipelines (crude receiving line and transfer lines to CDU for further processing). Table 2.1: Summary of Vessel Scenario Scenario Name

Product

Diameter (m)

Height (m)

Tank 632

Whole crude oil

77.5

20.5

Vessel Capacity (t) 5.4 x 104

Utilisation (%)

Temperature (qC)

Pressure (kPa)

Liquid Level (%)

100

30

101

85

Table 2.2: Summary of Representative Pipeline Scenarios Scenario Name

Product

Crude receiving (Pipe 9-P610-KA1-650) Crude suction to CDU (Pipe 9-P611-KA1-450) Crude suction to CDU (Pipe 9-P962-B4-250)

Whole crude oil Whole crude oil Whole crude oil


Diameter (mm) 650

Length (m) 240

Total HUM (kg) 55,748

Utilisation (%) 10

Temperature (qC) 30

Pressure (kPa) 189

450

220

24,493

45

30

189

450

220

24,493

45

30

189

Page B.3 of 3


ModuSpec Appendix C: Failure Frequency Data

APPENDIX C: FAILURE FREQUENCY DATA TABLE OF CONTENTS 1.

INTRODUCTION........................................................................................ 2

2. 2.1. 2.2. 2.3. 2.4. 2.5. 2.6. 2.7. 3.

FAILURE FREQUENCY DATA ...................................................................... 3 Process Pipes........................................................................................... 3 Valves .................................................................................................... 4 Flanges................................................................................................... 6 Small Bore Fittings ................................................................................... 7 Tank Mixers............................................................................................. 7 Storage Tanks ......................................................................................... 7 Full Surface Tank Fires .............................................................................. 7 REFERENCES ............................................................................................ 8


Page C.1 of 8


1.


INTRODUCTION

For each of the hazardous scenarios examined, items such as tanks, pipework, valves, flanges and fittings termed “failure items” associated with the scenario were identified. Failure modes of each failure item were represented as a range of hole size releases. Frequencies of hole size releases of each failure item were obtained using historical industry data. This appendix presents the failure frequency values applied in the analysis.


Page C.2 of 8

ModuSpec


Appendix C: Failure Frequency Data

2.

FAILURE FREQUENCY DATA

2.1.

Process Pipes

Table 2.1: Process Piping Failure Frequency [1] Pipe Size (mm)

Hole Size (mm)

25

5

Failure Frequency (x10-6 /year per 10 m length) 317.26

25

25

58.74

40

5

178.88

40

25

50.81

40

40

5.31

50

5

135.40

50

25

45.39

50

50

7.21

60

5

107.58

60

25

40.44

60

60

8.65

75

5

80.98

75

25

34.23

75

50

7.74

75

75

2.39

80

5

74.56

80

25

32.48

80

50

7.81

80

80

2.66

100

5

55.95

100

25

26.70

100

50

7.75

100

100

3.60

125

5

41.91

125

25

21.54

125

50

7.27

125

125

4.47

150

5

33.06

150

25

17.87

150

50

6.68

150

150

5.06

200

5

22.70


Page C.3 of 8

ModuSpec


2.2.

Appendix C: Failure Frequency Data

Pipe Size (mm)

Hole Size (mm)

200

25

Failure Frequency (x10-6 /year per 10 m length) 13.07

200

100

9.42

200

200

1.80

250

5

16.94

250

25

10.14

250

75

6.96

250

250

3.56

300

5

13.33

300

25

8.19

300

100

7.29

300

300

2.53

350

5

10.88

350

25

6.81

350

100

6.45

350

350

2.72

400

5

9.12

400

50

8.77

400

200

4.71

400

400

0.90

450

5

7.81

450

50

7.64

450

200

4.42

450

450

1.02

Valves

Table 2.2: Valve Failure Frequency Data [2] Valve Size (mm) 25

Hole Size (mm) 5

Failure Frequency (x10-6 /year) 108.16

25

25

2.84

40

5

102.70

40

25

8.29

40

40

0.01

50

5

98.88

50

25

12.06

50

50

0.05

60

5

95.26


Page C.4 of 8



Valve Size (mm) 60

Hole Size (mm) 25


60

60

0.16

75

5

90.33

75

25

20.20

75

50

0.46

75

75

0.01

80

5

88.83

80

25

21.55

80

50

0.61

80

80

0.01

100

5

83.43

100

25

26.14

100

50

1.38

100

100

0.05

125

5

77.87

125

25

30.29

125

50

2.64

125

125

0.20

150

5

73.30

150

25

33.17

150

50

4.06

150

150

0.47

200

5

66.20

200

25

36.50

200

50

6.87

200

200

1.43

250

5

60.90

250

25

37.99

250

75

11.38

250

250

0.73

300

5

28.37

300

25

19.26

300

100

7.63

300

300

0.24

350

5

26.69

350

25

19.26

350

100

9.11

350

350

0.44


Page C.5 of 8


2.3.


Valve Size (mm) 400

Hole Size (mm) 5


400

25

19.12

400

200

11.06

400

400

0.03

450

5

24.10

450

25

18.90

450

200

12.44

450

450

0.06

500

5

23.08

500

25

18.64

500

200

13.68

500

500

0.10

Flanges

Table 2.3: Flange Failure Frequency [2] Flange Size (mm) 25 25 40 40 50 50 60 60 75 75 100 100 125 125 150 150 200 200 250 250 300 300 350 350


Hole Size (mm)

Failure Frequency (x10-6 /year)

5 12 5 12 5 12 5 12 5 12 5 12 5 12 5 12 5 12 5 12 5 12 5 12

108.38 2.61 106.70 4.26 110.06 0.94 109.65 1.35 109.03 1.97 107.84 3.14 108.99 2.01 108.61 2.38 107.37 3.60 108.70 2.29 53.91 1.58 54.28 1.21

Page C.6 of 8


Flange Size (mm) 400 400 450 450 500 500 600 600 2.4.


Hole Size (mm)

Failure Frequency (x10-6 /year)

5 12 5 12 5 12 5 12

54.00 1.49 54.36 1.13 54.00 1.49 53.49 1.99

Small Bore Fittings

The failure frequency value used for small bore fittings is 7.19 X 10-4 [3]. This has been applied for fittings having a diameter less than 25 mm. 2.5.

Tank Mixers

Failure of a tank mixer was considered in terms of the failure of the mixer seal and a leak from that seal. The total seal mean failure frequency used was 5.61 x 10-2 per year. Refer to Appendix A for further details. 2.6.

Storage Tanks

Table 2.4: Storage Tank Failure Frequency Data [4].

2.7.

Hole Size (mm)

Failure Frequency

10 75 100 Rupture

1543.75 617.50 302.75 6.00

(x10-6 /year)

Full Surface Tank Fires

The frequency of full surface tank fires used in the analysis was 8.9 x 10-5 per year [5]. Refer to Appendix A for additional details of the derivation of this frequency.


Page C.7 of 8



3.

REFERENCES

1

“Hydrocarbon Leak and Ignition Data Base”, E&P Forum, February 1992, N658/Final Report, Section III.3, Appendix III, page 2.

2

“Classification of Hazardous Locations”, A.W. Cox, F.P. Lees and M.L. Ang, IChemE, 1993, Table 18.1, page 68.

3

Geometric mean of data obtained from source 2 and “Hydrocarbon Leak and Ignition Data Base”, E&P Forum, February 1992, N658/Final Report, page 25.

4

“Offshore Hydrocarbon Release Statistics, Offshore Technology Report – OTO 97 950”, UK Health and Safety Executive, December 1997.

5

LASTFIRE PROJECT, Large Atmospheric Storage Tank Fire Project – LASTFIRE Technical Working Group, June 1997.


Page C.8 of 8


APPENDIX D: HAZARDOUS FREQUENCY

ModuSpec Appendix D: Hazardous Scenarios and Failure Frequency

SCENARIO

AND

FAILURE

TABLE OF CONTENTS 1.

INTRODUCTION........................................................................................ 2

2. 2.1. 2.2. 2.3.

HAZARDOUS SCENARIOS AND FAILURE CONTRIBUTORS .......................... 3 Tank 632 ................................................................................................ 3 Crude Receiving ....................................................................................... 4 Crude Suction to CDU ............................................................................... 5


Page D.1 of 6


1.

ModuSpec Appendix D: Hazardous Scenarios and Failure Frequency

INTRODUCTION

This appendix contains a listing of the hazardous scenarios identified and the process failure items and overall failure frequency relating to them. The definitions for each of the headings presented are detailed below: Item Count Diameter Hole Size 1 FFreq. 1 Hole Size 2 FFreq. 2 Hole Size 3 FFreq. 3 Hole Size 4 FFreq. 4 Frequency total

Failure item or specific equipment included in the isolatable section. The number of failure items associated with the isolatable section. Diameter of failure item (mm). Representative hole size for a small release. Failure Frequency of pipeline release for hole size 1 (x 10-6 per year). Representative hole size for a medium release. Failure Frequency of liquid release for hole size 2 (x 10-6 per year). Representative hole size for a large release. Failure Frequency of liquid release for hole size 3 (x 10-6 per year). Representative hole size for a rupture scenario. This equates to the maximum size of the equipment, or 1000 mm for vessels. Failure Frequency of liquid release for hole size 4 (x 10-6 per year). Sum of frequencies at individual hole sizes.

The rupture case for storage tanks was represented by a hole size of 1000 mm. The rupture cases for the pipelines were represented by the hole size equivalent to the pipe diameter.


Page D.2 of 6

ModuSpec


2.

Appendix D: Hazardous Scenarios and Failure Contributors

HAZARDOUS SCENARIOS AND FAILURE CONTRIBUTORS

The following section lists the failure item parts count within each respective isolatable section. 2.1.

Tank 632

Table 2.1: Failure Items for Tank 632

Item

Count

Hole Diameter Size 1 (mm) (mm)

FFreq 1 (x10-6/y)

Hole Hole Hole FFreq 2 FFreq 3 FFreq 4 Size 2 Size 3 Size 4 -6 -6 (x10 /y) (x10 /y) (x10-6/y) (mm) (mm) (mm)

Associated pipework

60 m

350

25

106.14

100

38.7

350

16.32

1,000

0

Associated pipework Flange

40 m 2

500 25

25 25

27.16 221.98

100 100

27 0

350 350

16.56 0

1,000 1,000

4.4800 0

Flange Flange Flange Flange Small bore fitting

1 4 1 1 6

100 350 500 600 N/A

25 25 25 25 25

110.98 221.96 55.49 55.48 4,314.18

100 100 100 100 100

0 0 0 0 0

350 350 350 350 350

0 0 0 0 0

1,000 1,000 1,000 1,000 1,000

0 0 0 0 0

Tank Mixer Seal Valve Valve Valve Valve

5 2 1 3 1

N/A 25 100 350 500

25 25 25 25 25

280,500.00 222.00 109.57 137.85 41.72

100 100 100 100 100

0 0 1.43 27.33 0

350 350 350 350 350

0 0 0 1.32 13.68

1,000 1,000 1,000 1,000 1,000

0 0 0 0 0.1

Process Vessel -floating roof

1

77.5 m

25

1,543.75

100

920.25

350

0

1,000

6

Frequency Total


287,668.26

1,014.71

47.88

10.58

Page D.3 of 6

ModuSpec


2.2.


Crude Receiving

Table 2.2: Failure Items for the Crude Receiving Line (Pipe 9-P610-KA1-650)

Item

Count

Diameter Hole Size FFreq 1 1 (mm) (x10-6/y) (mm)

Hole Size Hole Size Hole Size FFreq 2 FFreq 3 FFreq 4 2 3 4 -6 -6 (x10 /y) (x10 /y) (x10-6/y) (mm) (mm) (mm)

Flange

1

25

10

10.84

25

0.26

100

0

650

0

Flange

4

100

10

43.14

25

1.26

100

0

650

0

4 4 240 m 11 1

500 600 650 N/A 25

10 10 10 10 10

21.60 21.40 9.53 158.19 10.82

25 25 25 25 25

0.60 0.80 0 632.74 0.28

100 100 100 100 100

0 0 9.91 0 0

650 650 650 650 650

0 0 9.50 0 0

4

100

10

33.37

25

10.46

100

0.57

650

0

Flange Flange Process Pipe Small bore fitting Valve Valve Frequency Total


308.89

646.40

10.48

9.50

Page D.4 of 6

ModuSpec


2.3.


Crude Suction to CDU

Table 2.3: Failure Items for the Crude Suction Line to CDU (Pipe 9-P611-KA1-450) Hole Size Hole Size Hole Size Hole Size FFreq 1 FFreq 2 FFreq 3 FFreq 4 1 2 3 4 -6 -6 -6 (x10 /y) (x10 /y) (x10 /y) (x10-6/y) (mm) (mm) (mm) (mm)

Count

Diameter (mm)

Flange

2

300

25

49.94

100

0

250

0

450

0

Flange

1

350

25

24.97

100

0

250

0

450

0

1 3 220 m 6 2

350 450 450 N/A 300

25 25 25 25 25

24.97 74.91 77.32 1,941.38 42.87

100 100 100 100 100

0 0 75.64 0 6.87

250 250 250 250 250

0 0 43.76 0 0

450 450 450 450 450

0 0 10.1 0 0.22

1

350

25

20.68

100

4.10

250

0

450

0.2

Item

Flange Flange Process Pipe Small bore fitting Valve Valve Frequency Total


2,257.04

86.61

43.76

10.52

Page D.5 of 6

ModuSpec



Table 2.4: Failure Items for Crude Suction Line to CDU (Pipe 9-P962-B4-250) Hole Size Hole Size Hole Size Hole Size FFreq 2 FFreq 3 FFreq 4 FFreq 1 2 3 4 1 -6 -6 -6 (x10 /y) (x10 /y) (x10-6/y) (x10 /y) (mm) (mm) (mm) (mm)

Count

Diameter (mm)

Flange

1

25

25

49.95

100

0

250

0

450

0

Flange Flange

1 2

250 250

25 25

49.95 99.89

100 100

0 0

250 250

0 0

450 450

0 0

1 1 220 m 2

300 350 450 N/A

25 25 25 25

24.97 24.97 77.32 647.13

100 100 100 100

0 0 75.64 0

250 250 250 250

0 0 43.76 0

450 450 450 450

0 0 10.10 0

Valve Valve

1 2

25 250

25 25

49.95 89.00

100 100

0 10.24

250 250

0 0.66

450 450

0 0

Valve

1

300

25

21.43

100

3.43

250

0

450

0.11

Item

Flange Flange Process Pipe Small bore fitting

Frequency Total


1,134.56

89.31

44.42

10.21

Page D.6 of 6


ModuSpec Appendix E: Event Tree Analysis

APPENDIX E: EVENT TREE ANALYSIS TABLE OF CONTENTS 1.

INTRODUCTION........................................................................................ 2

2. 2.1. 2.2. 3.

EVENT TREE DATA .................................................................................... 3 Operator Nearby ...................................................................................... 3 General Assumptions ................................................................................ 4 REFERENCES ............................................................................................ 6


Page E.1 of 6


1.


INTRODUCTION

Data acquired for the event tree considered the following for both a release and fire: x Probability of detection x Probability of isolation x Time to detect a release x Isolation time. The assumptions relating to leak detection and intervention measures for the proposed Tank 632 were based on previous analysis for the gasoline storage tanks at the Kurnell Refinery. The inlet and outlet lines of Tank 632 are to be fitted with motorised valves, whereas the event tree data for the gasoline tanks was based on the absence of remote isolation. Therefore, the isolation strategies for Tank 632 were modified accordingly to incorporate the provision for remote isolation. This is reflected in the time allowed for “1st isolation”. The event tree input data was collected by Caltex, based on discussions with relevant Kurnell Refinery operations personnel [1].


Page E.2 of 6


2.


EVENT TREE DATA

Descriptions of the main headings on the event tree table are provided below. % chance of seeing or % time person near unit - This is related to detection of the release or fire. There are two options for detecting a release or fire. The first is “Process” which is the chance that the basic process control system (e.g. an alarm or other indication) will alert a process operator to an anomaly. The second is “Person”, which is the chance that a person will detect the release or fire if they are in or near the unit. The “% time” is the percentage of time that a person will be in or near the unit. Note that for the case where the probability of detection is quoted as 100%, the value used in the calculation was 99.5%, to allow for human error. % chance of achieving isolation/time taken (min) – This is the probability of achieving isolation within a specific time frame. This occurs only once the release or fire has been detected. There are two chances for isolation (1st isol and 2nd isol) and each has an associated time. The 1st isolation time is typically associated with the automated isolation of the release or fire. The 2nd isolation time is related to the manual isolation, in the absence or failure of any automatic isolation. Worse case time to detect - This is the maximum time that the leak could proceed undetected. Generally this is for a release occurring in the middle of the night, when the likelihood of detection is low. 2.1.

Operator Nearby

The proportion of time that an operator / personnel are within the vicinity of the relevant unit was estimated based on weekday and weekend operations. Table 2.1 presents the number of hours that different personnel are in the vicinity of the particular plant areas. The average value, based on weekday and weekend operation is listed in the final row. The data acquired for the gasoline tank farm quantitative risk assessment was applied to the south crude tank farm, the area where Tank 632 is to be located.


Page E.3 of 6

ModuSpec


Appendix E: Event Tree Analysis

Table 2.1: Typical Presence of Personnel in the Crude Tank Farm. Number of Hours WEEKDAYS

Crude Tank Farm

Operator Security General maintenance

3.6 0.5 1

Maintenance, vessel maint. routine Personnel drive pass Public Total

0 1.5 0 6.6

% time a person at or nearby unit

27.5

WEEKEND

Crude Tank Farm

Operator Security

3.6 0.5

General maintenance Maintenance, vessel maint. routine Personnel drive pass Public Total

0 0 1 0 5.1

% time a person at or nearby unit

21.25

Average % time a person at or nearby the unit

2.2.

25.71

General Assumptions

The following general assumptions associated with detection and isolation of a release and fire in the Kurnell refinery were applied during the analysis: x The operator has no or minimal chance of detecting a small release or fire through indications on the basic process control system x There are no gas detectors at the facility x The probability of detection (process and person) increases as hole size increases x Fire is generally more readily detected than a release for any given hole size x For any given hole size, isolation times will be greater for fires than releases, as access to isolation points may be more difficult x There are no heat or fire detectors at the facility x All releases are liquid. Caltex has supplied data for the time to detect a release or fire for each of the individual scenarios analysed [1]. The values used in the analysis for the “time to detect” were the average of the time to detect a release and the time to detect a fire. The “worst case” time for detection (i.e. longest duration for which a release may go undetected) was assumed to be the time to detect a release.


Page E.4 of 6

ModuSpec


Appendix E: Event Tree Analysis

Table 2.2: Event Tree Data (probability of detection and successful isolation for a release or fire at each section) No. 1 1 1 2 2 2 Name Tank 633Tank 633Tank 633 Pipelines Pipelines Pipelines Hole Size s m r s m r Operator % time on/near unit 26 26 26 26 26 26

% chance of seeing

Detector

0

0

0

0

0

0

Release Process

10

15

80

0

15

80

Person

20

60

90

20

60

100

Detector

0

0

0

0

0

0

Process

10

15

80

0

20

80

Person

80

90

100

80

100

100

67.5 120/15

17.5 30/5

247.5 480/15

122.5 240/5

17.5 30/5

Fire

% chance of achieving isolation / time taken (mins)

Detect Time min 727.5 Time R/F Note 1440/15 1st isol.

90

90

90

90

90

90

Time min

1

1

1

2

2

2

2nd isol.

95

95

95

95

95

95

Time min

15

15

15

15

15

15

1st isol.

70

60

50

90

80

80

Time min

2

2

2

2

2

2

2nd isol.

15

15

15

80

80

80

Time min

30

30

30

20

20

20

1440

120

30

480

240

30

Release

Fire

Worst case Time min detection


Page E.5 of 6



3.

REFERENCES

1

E-mail from Ramez Aziz (Senior Risk Engineer, Caltex Refineries NSW, Pty Ltd) to Kate Filippin (Principal Risk Engineer, ModuSpec Australia) and Marian Magbiray (Risk Engineer, ModuSpec Australia), “Event Tree Data (again),” 12th July 2005.


Page E.6 of 6


ModuSpec Appendix F: Consequence Impact Level Criteria

APPENDIX F: CONSEQUENCE LEVEL IMPACT CRITERIA TABLE OF CONTENTS 1.

INTRODUCTION........................................................................................ 2

2.

HEAT FLUX CRITERIA ............................................................................... 3

3.

REFERENCES ............................................................................................ 4


Page F.1 of 4


1.


INTRODUCTION

This appendix provides details of the criteria used to evaluate the risk of fatality for pool fires. The general effects of thermal radiation are summarised in Table 1.1 [1]. Table 1.1: Thermal Radiation Effects. Heat Flux (kW/m2) 1.2 2.1 4.7 12.6

23

35

Effect Received from the sun at noon in summer. Minimum to cause pain after 1 minute. Will cause pain in 15-20 seconds and injury after 30 seconds exposure (at least second degree burns). Significant chance of fatality for extended exposure. High chance of injury. After extended exposure, causes the temperature of wood to rise to a point where it can be readily ignited by a naked flame. Thin steel with insulation on the side away from the fire may reach a thermal stress level high enough to cause structural failure. Likely fatality for extended exposure and chance of fatality for instantaneous exposure. Spontaneous ignition of wood after long exposure. Unprotected steel will reach thermal stress temperatures which can cause failure. Pressure vessel needs to be relieved or failure will occur. Cellulosic material will pilot ignite within one minute’s exposure. Significant chance of fatality for people exposed instantaneously.


Page F.2 of 4


2.


HEAT FLUX CRITERIA

The effect of human exposure to a fire is a function of both the intensity of heat radiation and the duration of exposure. The harmful effect can be characterised by a thermal dose that is defined by the function (tI4/3), where I is the heat radiation intensity and t is the exposure duration. A probit function has been used to evaluate the likelihood of fatality for different heat flux exposures. The probit equation utilised is the Eisenberg equation [2]: Equation 2.1:

Y

§ 4 · 14.9 2.56 ln¨tI 3 ¸ © ¹

where:Y= probit value t = exposure time (seconds) I = radiation intensity from fire (kW/m2) In terms of human exposure, it is generally accepted that an exposure of 12.6 kW/m2 will result in a 50% chance of fatality. Based on Equation 2.1, the required exposure time would be 81 seconds. To account for all the possible means that adverse outcomes can occur, a range of heat flux levels need to be assessed. The values used in the analysis were based on an exposure time of 81 seconds and are presented in Table 2.1. Table 2.1: Heat Flux Levels and Corresponding Fatality Probability Heat Flux (kW/m2)

8.7

9.7

11.7

13.6

16.1

% Fatality

10

20

40

60

80

The actual heat flux received at a point located between two contours will be a value between the heat flux values corresponding to each contour. Similarly, the probability of fatality at this location will be between the fatality probabilities for two contours. In determining the risk at such a location, the higher fatality probability is applied in the calculation process. For example, all points lying between the 13.6 kW/m2 and 16.1 kW/m2 heat flux contours for a given scenario are treated as resulting in fatality to 80% of the exposed population. For the case of heat flux exceeding 16.1 kW/m2, the a probability of fatality of 100% is assumed.


Page F.3 of 4


3.


REFERENCES

1

NSW Department of Urban Affairs and Planning; “Risk Criteria for Land Use Safety Planning”, Hazardous Industry Planning Advisory Paper No. 4”, Sydney, 1990.

2

Lees, F.P., "Loss Prevention in the Process Industries, Hazard Identification Assessment and Control", Butterworth & Heinemann, 1996, 2nd Edition, Volume 1, p9/64.


Page F.4 of 4


ModuSpec Appendix G: Consequence Results

APPENDIX G: CONSEQUENCE RESULTS TABLE OF CONTENTS 1.

INTRODUCTION........................................................................................ 2

2. 2.1. 2.2.

CONSEQUENCES ....................................................................................... 3 Pool Fires ................................................................................................ 3 Full Surface Tank Fires .............................................................................. 9


Page G.1 of 9


1.

ModuSpec Appendix G: Consequence Results

INTRODUCTION

This appendix contains the consequence results. The results are presented in table format. The following list provides a definition for each of the headings presented. Scenario Name Hole Size (mm) Release Rate (kg/s) Duration (s) Frequency (x10-6/y) Distance (m) to Criteria (kW/m2)

Heat


The name of the isolatable section or specific equipment considered as the scenario. The specific hole size considered for the consequence scenario. The rate at which the product is expected to be released. The time taken for the product to be released. Failure frequency for the scenario. The maximum distance at which the consequence is experienced for the given heat flux level from the centre of the pool.

Page G.2 of 9

ModuSpec


2.

CONSEQUENCES

2.1.

Pool Fires

Appendix G: Consequence Results

Table 2.1: Consequence Modelling Results for Pool Fires. Hole size (mm)

Release Rate Used (kg/s)

Duration (s)

Crude Receiving

10

0.57

113352

Crude Receiving Crude Receiving Crude Receiving

10 10 25

0.57 0.57 3.5

114402 128352

Scenario

2 Frequency Distance (m) to Heat Criteria (kW/m ) (x10-6/y) 8.70 9.70 11.70 13.60 16.10

14

13

13

12

12

14 14 28

13 13 27

13 13 26

12 12 25

12 12 24

30711

0.3 0.03 2.25 1.28

Crude Crude Crude Crude Crude

Receiving Receiving Receiving Receiving Receiving

25 25 25 100 100

3.5 3.5 3.5 57 57

31761 45711 45711 8454 9504

0.12 0.02 9.67 0.16 0.02

28 28 28 65 65

27 27 27 62 62

26 26 26 57 57

25 25 25 52 52

24 24 24 46 46

Crude Crude Crude Crude Crude Crude Crude Crude Crude Crude Crude

Receiving Receiving Receiving Receiving Receiving Suction (Line Suction (Line Suction (Line Suction (Line Suction (Line Suction (Line

100 650 650 650 650 25 25 25 25 100 100

57 778 778 778 778 3.5 3.5 3.5 3.5 57 57

16554 1193 2243 2993 2993 21886 22936 36886 36886 7902 8952

0.35 0.55 0.08 0.01 0.11 2.25 0.22 0.03 16.96 1.33 0.21

65 128 128 128 128 28 28 28 28 65 65

62 121 121 121 121 27 27 27 27 62 62

57 110 110 110 110 26 26 26 26 57 57

52 103 103 103 103 25 25 25 25 52 52

46 100 100 100 100 24 24 24 24 46 46

9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250)


Page G.3 of 9

ModuSpec


Scenario Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Crude Suction Tank 632

(Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line (Line

9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250) 9-P962-B4-250) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450) 9-P611-KA1-450)

Tank 632



Hole size (mm)


100 100 250 250 250 250 450 450 450 450 25 25 25 25 100 100 100 100 250 250 250 250 450 450 450 450 25

57 57 78 78 78 78 78 78 78 78 3.54 3.54 3.54 3.54 57 57 57 57 78 78 78 78 78 78 78 78 4.4 4.4

25

2 Frequency Distance (m) to Heat Criteria (kW/m ) Duration (s) (x10-6/y) 8.70 9.70 11.70 13.60 16.10

16002 16002 7539 8589 15639 15639 1191 2241 2991 2991 21886 22936 36886 36886 7902 8952 16002 16002 7539 8589 15639 15639 1191 2241 2991 2991

0.03 2.98 1.04 0.16 0.02 2.33 0.59 0.09 0.01 0.12 4.48 0.44 0.06 33.75 1.29 0.2 0.03 2.89 1.02 0.16 0.02 2.3 0.61 0.09 0.01 0.13 926.11

65 65 128 128 128 128 128 128 128 128 28 28 28 28 65 65 65 65 128 128 128 128 128 128 128 128 30

62 62 121 121 121 121 121 121 121 121 27 27 27 27 62 62 62 62 121 121 121 121 121 121 121 121 29

57 57 110 110 110 110 110 110 110 110 26 26 26 26 57 57 57 57 110 110 110 110 110 110 110 110 28

52 52 103 103 103 103 103 103 103 103 25 25 25 25 52 52 52 52 103 103 103 103 103 103 103 103 27

46 46 100 100 100 100 100 100 100 100 24 24 24 24 46 46 46 46 100 100 100 100 100 100 100 100 26

12442307 12443657

127.34

30

29

28

27

26

Page G.4 of 9

ModuSpec



Hole size (mm)


Tank 632

25

4.4

Tank 632 Tank 632 Tank 632

25 100 100

4.4

Tank 632

100

Tank 632 Tank 632 Tank 632

Scenario

2 Frequency Distance (m) to Heat Criteria (kW/m ) Duration (s) (x10-6/y) 8.70 9.70 11.70 13.60 16.10

104.19 4208.88 13.23 2.43

30

29

28

27

26

70.1 70.1

12486407 12486407 779050 780400

30 70 70

29 67 67

28 61 61

27 55 55

26 48 48

70.1

783550

1.98

70

67

61

55

48

100

70.1

38.56 0.14 2.63

67

61

55

48

64,041 64,041

783550 953 954

70

350 350

177 177

168 168

155 155

150 150

149 149

632 632 632 632 632

350 350 Rupture Rupture Rupture

60,891 59,541 N/A N/A N/A

963 967

0.17 0.90

Instantaneous Instantaneous Instantaneous

0.10 0.12 0.13

177 177 209 209 209

168 168 198 198 198

155 155 176 176 176

150 150 176 176 176

149 149 176 176 176

Tank 632

Rupture

N/A

Instantaneous

0.50

209

198

176

176

176

Tank Tank Tank Tank Tank


Page G.5 of 9

ModuSpec



Table 2.2: Consequence Modelling Results for the 4.7 kW/m2 Heat Criteria for Pool Fires

Scenario

Hole size (mm)

Crude Receiving

PLT 2 Suction



Duration (s)

Frequency (x10-6 / y)

Distance (m) to 4.7kW/m2 Heat Criteria

10

0.6

113352

0.30

16

10 10 10

0.6 0.6 0.6

114402 114402 128352

0.03 0.00 2.25

16 16 16

25 25 25 25

3.5 3.5 3.6 3.6

30711 31761 45711 45711

1.28 0.12 0.02 9.67

32 32 32 32

100 100 100 100 650

56.7 56.7 56.7 56.7 778

8453 9503 16553 16553 1193

0.16 0.02 0.00 0.35 0.55

78 78 78 78 159

650 650 650 25 25

778 778 778 3.5 3.5

2243 2993 2993 21885 22935

0.08 0.01 0.11 2.25 0.22

159 159 159 32 32

25 25 100

3.6 3.6 56.7

36885 36885 7902

0.03 16.96 1.33

32 32 78

100

56.7

8952

0.21

78

100

56.7

16002

0.03

78

Page G.6 of 9

ModuSpec


Scenario


Hole size (mm)


Duration (s)



100

56.7

16002

2.98

78

250 250 250 250

78 78 78 78

7539 8589 15639 15639

1.04 0.16 0.02 2.33

159 159 159 159

450 450 450 450

78 78 78 78

1191 2241 2991 2991

0.59 0.09 0.01 0.12

159 159 159 159

25 25 25 25 100

3.5 3.5 3.6 3.6 56.7

21885 22935 36885 36885 7902

4.48 0.44 0.06 33.75 1.29

32 32 32 32 78

100 100 100 250 250

56.7 56.7 56.7 78 78

8952 16002 16002 7539 8589

0.20 0.03 2.89 1.02 0.16

78 78 78 159 159

250 250 450

78 78 78

15639 15639 1191

0.02 2.30 0.61

159 159 159

450

78

2241

0.09

159

450

78

2991

0.01

159

PLT 45 Suction


Page G.7 of 9

ModuSpec


Scenario


Hole size (mm)


Duration (s)



450

78

2991

0.13

159

25 25 25 25

4.4 4.4 4.4 4.4

12442307 12443657 12486407 12486407

926.11 127.34 104.19 4,208.88

34 34 34 34

100 100 100 100

70.1 70.1 70.1 70.1

779050 780400 783550 783550

13.23 2.43 1.98 38.56

85 85 85 85

350 350 350 350

64,042 64,042 60,892 59,542 N/A

0.14 2.63 0.17 0.90 0.10

224 224 224 224

Rupture

953 953 963 967 Instantaneous

Rupture Rupture

N/A N/A

Instantaneous Instantaneous

0.12 0.13

258 258

Rupture

N/A

Instantaneous

0.50

258

Tank 632


258

Page G.8 of 9

ModuSpec


2.2.


Full Surface Tank Fires

Table 2.3: Consequence Modelling Results for Full Surface Tank Fires Scenario

Tank 632 Full Surface Tank Fire


Frequency (x10-6/y)

120

Downwind Distance (m) to Heat Criteria (kW/m2) 4.7

8.7

9.7

11.7

13.6

16.1

81

81

81

81

81

81

Page G.9 of 9

Quantitative Risk Assesment - 06_0160_appendix_d

Recommend Documents