Town Gas Quantitative Risk Assessment

QUANTITATIVE RISK ASSESSMENT (QRA) Study For

TOWN GAS COMPANY

[ El-Tebeen (Industrial) Pressure Reduction Station] At

Greater Cairo City

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

TABLE OF CONTENTS 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Executive Summary........................................................................................................................4 Introduction ......................................................................................................................................7 Project Description....................................................................................................................... 11 Technical Definitions ................................................................................................................... 12 Assessment of Risks ................................................................................................................... 13 Methodology ................................................................................................................................. 14 Plan of Work ................................................................................................................................. 15 Operation of the System ............................................................................................................. 17 Town Gas Emergency Plan........................................................................................................ 18 Weather Data ............................................................................................................................ 19 Generic Release Scenarios .................................................................................................... 25 Specific Release Scenarios .................................................................................................... 27 Impairment Criteria ................................................................................................................... 28 Flammability Assessment........................................................................................................ 30

14.1 14.2 14.3 14.4

15 16 17 18

SHELL FRED Version (4.0) Consequence Modelling Software ....................................... 33 Sensitivity Analysis ................................................................................................................... 38 Release Scenarios.................................................................................................................... 41 Ignited Release Scenario ........................................................................................................ 43

18.1 18.2

19

19.1.1 19.1.2 19.1.3

19.2 19.2.1 19.2.2 19.2.3 19.2.4

Fire .........................................................................................................................................47 Flash Fire........................................................................................................................................................................ 48 Unobstructed Jet Fires ............................................................................................................................................... 49 Obstructed Jet Fires ................................................................................................................................................... 50 Pool Fires ....................................................................................................................................................................... 51

High Pressure Release from 100 -mm (4 -Inch) Leak Upstream PRS (1A) ...........................53 High Pressure Release from 25-mm (1-Inch) Leak Upstream PRS (1B).............................56 High Pressure Release from 5-mm (1/4-Inch) Leak Upstream PRS (1C)............................59 Low Pressure Release from 100-mm (4 Inch) Leak Downstream PRS (2A) .......................62 Low Pressure Release from 25-mm (1-Inch) Leak Downstream PRS (2B).........................65 Low Pressure Release from 5 -mm (1/4-Inch) Leak Downstream PRS (2C)........................68 Process Release....................................................................................................................71 Ignition Probability .................................................................................................................72

Risk Assessment ...................................................................................................................... 73

22.1 22.2

23

Gaseous Release........................................................................................................................................................ 44 Liquid Release .............................................................................................................................................................. 45 Toxic Gas release....................................................................................................................................................... 45

Likelihood Data .......................................................................................................................... 71

21.1 21.2

22

Hydrocarbon Releases..........................................................................................................44

Consequence Modelling Results............................................................................................ 52

20.1 20.2 20.3 20.4 20.5 20.6

21

Causes of Release ................................................................................................................43 Causes of Ignition ..................................................................................................................43

Typical Fire Consequence Analysis....................................................................................... 44

19.1

20

General..................................................................................................................................30 Process Hydro carbons ..........................................................................................................31 Electrical Fires .......................................................................................................................32 Conclusions ...........................................................................................................................32

Individual Risks ‘IR’ to Workers.............................................................................................74 Individual Risk to the Public...................................................................................................75

Risk Evaluation.......................................................................................................................... 76

PS-GZT-TG-001 Revision (0) Draft Report

Page 2

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

24 25 26

Risk Reduction Measures (Recommendations) .................................................................. 77 Conclusion ................................................................................................................................. 79 References ................................................................................................................................. 81

PS-GZT-TG-001 Revision (0) Draft Report

Page 3

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

1 Executive Summary Quantitative risk assessment study has been performed for the El-Tebeen PRS (Pressure Reduction Station) for Town Gas Company based on the recommendations from the World Bank. The connecting pipelines and the associated critical crossings are outside the scope of this quantitative risk assessment (QRA) study. El-Tebeen PRS has been considered as the first PRS included for consequence modelling and risk assessment, while other stations shall follow in series. F or the purpose of the analysis it has been assumed that the PRS are within restricted entry open area, which is not normally manned but will be frequently visited by operations and maintenance teams comprising at least two personnel. SHELL FRED version (4 .0) has been selected for the consequence modeling of different types of hazardous consequences as follows: • • •

Flammable gases dispersion (Gas Clouds) Flash fires Jet fires

SHELL FRED version (4.0) is the state of the art Shell’s consequence modeling softwa re kit stands for Fire, Release, Explosion and Dispersion models used to predict the consequences of the accidental release of flammable and toxic materials from different types of process equipment. For the PRS leak scenario, the release rate has been simulated based on 3 -hole sizes 0.25 inch, representing instrument fitting failure [pin hole leak], 1.0 inch representing small pipe leak [minor leak] and 4.0 inches leak representing a 4-inch pipe full bore rupture or 4-inch hole size in a larger pipe diame ter [major leak or catastrophic failure]. This is corresponds to 5-mm, 25-mm and 100-mm leak sizes. Weather conditions have been selected based on wind speed and stability class. The worst case weather conditions for gas dispersion is the stable weather conditions, represented by wind speed of 1 m/s and stability class "F" representing "Very Stable" weather conditions, in order to obtain conservative results. Since the jet fire is originally a high momentum directed jet release, hence the effects of wind direction, wind speed or atmospheric stability on the jet flame are minimal. The PRS comprises two types of pressures, the first is the upstream pressure, which is high pressure ranging from 30 to 70 Bars, while the second pressure is the down stream pressure, which is low pressure ranging from 4 to 7 Bars. For the purpose of the consequence modelling, the maximum of the two types of pressures have been simulated to represent the worst case and mild case respectively (70 Bars as HP and 7 Bars as LP). The jet fire (flame length) and heat radiation distances are measured in meters. The gas dispersion distances have been calculated in meters in concentration terms of Lower Flammability Limits (LFL) and Upper Flammability Limits (UFL) presented by Part Per M illion (PPM)

PS-GZT-TG-001 Revision (0) Draft Report

Page 4

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

concentrations in order to represent the flammability range of the released gas cloud; however the extent of damage is presented by LFL only. The heat radiation from flash fires will not significantly affect humans, equipment or structures due to the short duration of flash fires. Fire consequence analysis has been described in details in fire consequence effects section, which details the hazardous effects from different types of fires. The following table presents the generic extent of damage distances as a result from the consequence modelling simulation analysis performed by FRED. Table 1.1 Generic Extent of Damage Distances from PRS Leaks in Meters High Pressure Low Pressure Case Leak size Leak size Side [70 Bar] Side [7 Bar] Leak type No. in Meters in Inches Jet Gas Jet Gas Flame Cloud Flame Cloud 1.0 Pin Hole 0.005 0.25 6.5 3.5 2.2 1.2 2.0 Minor leak 0.025 1 25 11.2 8.5 5.5 3.0 Major leak 0.1 4.0 70 30 25 11 Notes: The damage distances have been calculated by SHELL's FRED version (4.0) consequence modeling software with the following conditions: 1. 2. 3. 4.

The released gas is Standard Natural Gas. Wind speed of 1 meter/second. Weather stability of "F" representing "Very Stable" weather conditions. Damage distances are based on the maximum damage contours [i.e. the limit of the frustum length or jet flame length and the LFL contour for gas cloud].

As a rule of thumb, the worst case scenario presented by the high pressure side of the PRS shall be considered in the risk assessment study in order to obtain conservative results. From the extent of damage distances calculated, it can be observed that major or catastrophic equipment failure has the maximum potential extent of damage due to increased leak size. Maximum extent of damage is 70 meters in the worst case conditions. The minor leak has a localized extent of damage within the PRS boundary or battery limits due to medium leak size. Proposed extent of damage is 25 meters. While the pin hole leak has the minimum localized extent of damage due to small leak size. Minimum extent of damage is 6.5 meters in the mild case conditions. But on the other hand, the probability of occurrence or failure frequency of major leak or catastrophic equipment failure is deemed to be much lower than a pin hole leak. Process release generic failure frequencies and ignition probabilities have then been identified for the detailed quantitative risk assessment (QRA) purposes from API-581, Lees and E&P Forum.

PS-GZT-TG-001 Revision (0) Draft Report

Page 5

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Quantitative risk assessment (QRA) has been performed to all types of hazardous events developed from the hazardous scenarios development. The risks have been assessed for the industrial workers and general public representing the two types of risk namely the "Individual Risk" and "Societal Risk" with in the PRS area as presented in the following table. No 1.0 2.0

Calculated Risk 4.5 E-05 9.0 E-05

Acceptable Risk 1.0 E-05 1.0 E-05

Area Type Workers Public

Acceptance ALARP ALARP

The risks assessed have been evaluated based on the international risk acceptance criteria in order to demonstrate that risks are within the ALARP limits. From the plot plan and general layout of the PRS area, it can be observed that the PRS is within populated area, with potential for public impacts by potential ignited release scenarios. Hence, the risk from such PRSs shall be within the acceptable limits, if safety precautions have been considered and strictly followed in the design, operation and maintenance of such facilities.

PS-GZT-TG-001 Revision (0) Draft Report

Page 6

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Risk Reduction Measures: Risk reduction measures (recommendations) have been proposed as points of improvement in order to enhance the PRS safety standards. These risk reduction measures (recommendations) are summarized as follows: 1. There is a need to develop a safe system of work, based on risk assessment for dealing with potential gas leaks. 2. Consideration should be given to the remote actuation of isolation and slam-shut valves by Town Gas for different PRS's as well as the transmission and distribution pipelines. 3. There is a need to produce Hazardous Area Classification drawings for all Pressure Reduction Stations. 4. Planned preventive maintenance policy should be in place for the new PRSs. 5. There is a need to produce a 'Station Manual' for each PRS. This manual should include formalized procedures, including precautions and a site scenario specific emergency plan. 6. Site emergency plans must take into account wind direction and stability and should consider interfaces with others, e.g. GASCO as well as the public living nearby. 7. Town Gas needs to consider the security arrangements for all un -manned stations. 8. There is a need that Town Gas should apply risk assessment to all activities and to formalize procedures and permit-to-work systems. 9. The control room inlet door should be located in the upwind direction away from the PRS station (Inlet door should not face the PRS station). 10. Alternatively, the control room should be provided by a secondary means of escape at the back side of the room, which shall be used in case of blockage of the main escape route by jet fires. 11. It is recommended that a jet fire rated passive fire protection system to be applied to all safety critical shutdown valves ESDVs or Solenoid valves in order to maintain small isolatable inventories. (As applicable) 12. It is recommended to have pipeline marking signs indicating in Arabic and in English "Do Not Dig" and "High Pressure Pipeline Underneath" in order to prevent such extreme hazardous situation. 13. It is recommended to include the prevailing wind direction on the PRS site plan. 14. It is reco mmended to have an elevated wind sock installed in the PRS site, which can be seen from distance and from outside the fence to determine the direction of gas migration in case of major gas leak. 15. It is recommended to have a gas detection system within the PRS area to automatically sense the released gases as a percentage of LFL, in order to provide early warnings of gas release. 16. Also, it is recommended to have point gas detectors at the room HVAC intake (if provided) to automatically sense the released gases as a percentage of LFL, in order to provide early warnings of gas release. 17. Investigate a strategy to inform the residential area beside the PRS and the associated pipeline with the risk involved in such accidents as well as the methods required for ann unciating if any leak occurs. 18. The design should fully comply with IGE TD/3 code requirements.

PS-GZT-TG-001 Revision (0) Draft Report

Page 7

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

2 Introduction This report presents the Quantitative Risk Assessment (QRA) carried out by Petrosafe for the Pressure Reduction Station Project at Greater Cairo City for Town Gas Company. Objectives: Town Gas operates a distribution network to supply natural gas to residential and commercial premises as well as car gas filling stations in the Greater Cairo area. Town Gas has set out the main objectives of the risk assessment to include the following: o To identify, assess and quantify risks to people (the general public, Town Gas operations staff, and other associated groups. o To identify, assess and quantify risks arising from the steel pipeline and relevant pressure reduction and odorising stations. o To comprehensively examine the ways in which the identified risks can be eliminated or reduced. o To recommend practical risk reduction measures for consideration. o To determine the final residual risks (assuming adoption/implementation of the recommended risk reduction measures). o To recommend practical residual risk control measures. Terms of Reference : Town Gas identified specific parts of the distribution network for the proposed risk assessment study. These are: 1. El-Tebbeen industrial pressure reduction and odorant station. 2. El-Tebbeen domestic pressure reduction and odorant station. 3. El-Haram pressure reduction and odorant station. 4. El-Mokatam pressure reduction station and odorant station. 5. El-Shorouk pressure reduction station and odorant station, and 6. New Cairo pressure reduction station and odorant station.

PS-GZT-TG-001 Revision (0) Draft Report

Page 8

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

FIGURE 2.1: The Planned PRSs in this Project Petrosafe will perform a semi-quantitative and risk assessment to identify the major risk issues and contributors with a “best estimate” of the associated levels of risk for the ‘El-Tebeen Industrial PRS’ as the first urgently required station, while the others shall follow. This work will be designed to perform two main functions: o To provide Town Gas with a clear risk knowledge and awareness such that investment decisions can be well informed.

PS-GZT-TG-001 Revision (0) Draft Report

Page 9

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

o To act as a coherent stage in, and to define the final scope of the full risk assessment. Conduct the remaining work to provide a full risk assessment in compliance with defined and approved Standards and Guidelines. In general the work will cover, but not necessarily be limited to, the following: o Define data to be provided by Town Gas o Review Town Gas data o Perform physical survey of the steel pipeline route to identify possible ‘hot spots’. o Define possible accident scenarios and events o Conduct a full consequence analysis in relation to gas leaks and fire/explosion and toxic release scenarios. o Perform qualitative/quantitative risk assessment o Conduct ALARP risk reduction/mitigation review o Define possible risk elimination/reduction measures o Define/quantify residual risk levels o Propose residual risk control measures

PS-GZT-TG-001 Revision (0) Draft Report

Page 10

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

3 Project Description The objective of the Gas Distribution system at Town Gas is the supply of natural gas to the domestic, commercial and industrial customers in the Greater Cairo Area. The infrastructure consists of a 24” diameter, 57.53 km high pressure steel pipeline from Tibbeen to North Heliopolis stations [to be confirmed]. The line contains several pressure reduction and odorant stations that interface with various supply pipelines. Egyptian Town Gas employs some 2,100 personnel and supplies gas to more than 1.3 million residential and commercial customers. This is the largest gas netwo rk in Egypt for more than 22 years. Documents Reviewed: The risk assessment Technical Proposals are based on the following documents that were received from Town Gas. o Cairo map showing geographical boundaries for the natural gas operating areas. o Transmission steel pipeline route map - Drawing. o Technical Specifications for main steel pipeline. [Typical] o Operating and emergency procedures for pressure reduction and odorising stations in Greater Cairo Area. [Typical] o Odorant- Hazard Data Sheet. [Typical] o Plan layout for pressure reduction and odorant station. [Typical] o P&I Diagram for pressure reduction station. [Typical] Available documents are presented in Appendix-1.

PS-GZT-TG-001 Revision (0) Draft Report

Page 11

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

4 Technical Definitions Confinement

A qualitative or quantitative measure of the enclosure or partial enclosure areas where vapors cloud may be contained.

Congestion

A qualitative or quantitative measure of the physical layout, spacing, and obstructions within a facility that promote development of a vapor cloud explosion.

EERA

Escape, Evacuation and Rescue Assessment

ESD

Emergency Shut Down

FRA

Fire Risk Assessment

Gas cloud dispersion

Gas cloud air dilution naturally reduces the concentration to below the LEL or no longer considered ignitable (typically defined as 50% of the LEL).

Hazard

An inherent physical or chemical characteristic (flammability, toxicity, corrosivity, stored chemical or mechanical energy) or set of conditions that has the potential for causing harm to people, property, or the environment.

Individual risk

The risk to a single person inside a particular building. Maximum individual risk is the risk to the most-exposed person and assumes that the person is exposed.

QRA

Quantitative Risk Assessment

Risk

Relates to the probability of exposure to a hazard, which could result in harm to personnel, the environment or general public. Risk is a measure of potential for human injury or economic loss in terms of both the incident likelihood and the magnitude of the injury or loss.

Risk assessment

The identification and analysis, either qualitative or quantitative, of the likelihood and outcome of specific events or scenarios with judgments of probability and consequences.

Vapor cloud explosion (VCE)

An explosion in air of a flammable material cloud

PS-GZT-TG-001 Revision (0) Draft Report

Page 12

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

5 Assessment of Risks This part of the study would address the identification, analysis and subsequent assessment of major hazards associated with the relevant gas pressure reduction stations in addition to the associated part of the gas pipeline. They are categorised, and makes judgement on the tolerability of risks to personnel associated with these hazards. British Gas criteria for risk tolerability are used to base such judgements. Scenarios that could result in major hazards will be identified and evaluated using semi-quantified and Quantified Risk Assessment ‘QRA’. This technique is used to establish the expected frequency of such incidents occurring on each facility and their consequences. Several commercial software tools are available to P etrosafe are available for consequences modelling of dispersion, fire and explosion will be selected for modelling pipeline gas leaks. Detailed risk profile to different individuals on the facility will be estimated, which in turn becomes an important input in determining the requirements for any remedial work. This section will be linked to the rest of the proposed study in order to tie together the logic of the arguments and bring the findings into better context. It will encompass: o Town Gas Policy, Standards and Criteria, o The sources of hazards, o Hazardous substances, and their inventories, o Events which are capable to cause major accidents, o Analysis of the consequences and their effects on employees, third parties and the public, o Evaluation of individual and societal risks, using BG Risk Tolerability Criterion, o Measures to prevent, control or minimise likely consequences, o Emergency procedures and emergency systems, derived from consideration of the above issues, including evacuation/fire-fighting procedures. From these studies, risk reduction measures are identified, and improvements- to hardware and management systems- are considered.

PS-GZT-TG-001 Revision (0) Draft Report

Page 13

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

6 Methodology Maximum use should be made of all studies and reports produced for the project so far. A list of all-relevant studies and reports relevant to all gas facilities should be identified by The Company and made available to the Consultant. The proposed study will examine the contents of existing documents relating to The Company- HSE Management System. This will cover such documents as: o The Company HSE Policy. o Details of the pipeline design, PFD and PID’s. o Codes and Standards used for the design and construction of the pipeline and relevant stations. o Fire and gas detection/protection systems and procedures of the facilities in question . o Operations manuals, procedures and standing orders relating to the pipeline and stations. o Engineering: 1)

Change control procedures

2)

Relevant stations latest PFDs and PIDs.

3)

Maintenance philosophy & control

4)

Inspections and planned preventive maintenance

5)

Shut-down and blow-down philosophies

6)

Design and locations of sectionalising valves

7)

Procedures for pigging operations.

The proposed study will examine the relationship between relevant documents and management procedures to establish that these are adequate to demonstrate that all reasonable steps have been taken to ensure that the design, construction, operation and maintenance of the facility and its equipment are adequate to provide a safe working environment. Furthermore, the Company should demonstrate that, in the event of an incident, which may escalate and lead to the requirement for personnel to evacuate the facility, such arrangements are in place and are adequate.

PS-GZT-TG-001 Revision (0) Draft Report

Page 14

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

7 Plan of Work The plan of work includes the following: o Review of relevant facilities documentation, which include design drawings, operating and emergency procedures. o To carry out a physical survey of the steel pipeline route in order to identify potential hot spots, where the public may be exposed to possible gas leaks. o To identify parts of the pipeline ‘hot-spots’ for some detailed modelling of release scenarios and fire/explosion. Layout analysis will identify critical sections of the pipeline for further modelling. o To conduct a selective Quantified Risk Assessment ‘QRA’ on the relevant gas distribution stations and to calculate Individual and Societal Risks to Town Gas workers, third parties and the public. Estimation and evaluation of risks is carried out using the se mi-quantitative matrix illustrated in FIGURE 7.1.

Probability B

A Improbable

Severity

People

Asset/ Environ- Perfor Costs ment mance 1 in 100,000 years

5. Catastrophic

Multiple fatalities

4. Severe

Single fatality or Multiple Major injuries

Extensive

Massive

damage

effect

Major

Major

damage

effect

C

Remote 1 in 10,000 years

D

E

Occasional

Probable

Frequent

1 in 1000

1 in 100

1 in 10

years

> 500 0ff > 18 hrs loss of supply

years

years

HIGH RISK A

>5 interruption / customer

Major injury

3. Critical

or health effects

Local

Localised

damage

effect

Minor injury

2. Marginal

or health

Minor

Minor

damage

effect

effects

Slight injury

1. Negligible

or health

Slight

Slight

damage

effect

effects

> 100 mins CML target

(ALARP) B

> 95 off > 3hrs restorations

> 1.1 interruptions per customer

LOW RISK C

FIGURE 7.1 : Criterion for Semi-quantitative Risk Estimation and Evaluation The proposed QRA Framework is shown in FIGURE 7.2 and the QRA Criteria for risk tolerability is shown in FIGURE 7.3.

PS-GZT-TG-001 Revision (0) Draft Report

Page 15

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Failure Case Definition

Pipeline Data

Identify Hazards

Scenario Development

Frequency Analysis

Analysis of Consequences

Impact Assessment

Estimate/ Measure Risks

UGD BG Criteria Criteria

Evaluate Risks

Decide Risk Mitigation Measures

Verify

FIGURE 7.2 QRA Risk Assessment Frame -work

1 in 10,000 1 in 1000

ALARP

ALARP

Region

Region

1 in 100,000 1 in 1 million

Individual Risk to Personnel

Individual Risk to the Public

FIGURE 7.3 British Gas Criteria for the QRA Risk Tolerability SOFTWARE: The Software proposed for consequences modelling and the QRA parts of the project is Shell's FRED (4.0) - Consequences modelling software tools as presented in Appendix-2.

PS-GZT-TG-001 Revision (0) Draft Report

Page 16

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

8 Operation of the System The system of operating all Town Gas Pressure Reduction Stations relies on a high degree of flexibility in terms of supplies that ensures multiple gas inlet sources and effective monitoring system. Town Gas SCADA System: The Company operates a monitoring and control ‘SCADA’ system that monitors all main and substations in Greater Cairo as well as the main gas transmission lines (the 70 Bar and 7 Bar lines) with the following function s: [To be confirmed] o Monitoring of inlet gas pressure for main sources and giving alarms in the cases of high and low gas pressures. o Monitoring of the odorization units in the stations in terms of rate of filling and giving alarms if the storage tanks odorant level and pressure increases or decreases. o Monitoring of all stations outlet gas pressure and flow rates. o Monitoring of the effects of gas output on the transmission lines pressure (70 Bar and 7 Bar). o Controlling critical isolation valves remotely in cases of emergency (as applicable). o The SCADA system should also remotely control the change-over gas reduction streams to auxiliary lines at all stations in cases of emergency or due to failures. However, this function is not implemented due to the theft of spare parts. Filling the Odorant Main Storage Tank: The odorant storage tank is filled when the liquid level drops below the minimum level. The liquid level is monitored via two level measuring systems. [To be confirmed] The tank is pressurised by blanket gas to minimise vaporisation of the odorant. This gas is used to transfer the odorant from its drum into the storage tank. The odorant vaporises at about 1.45 Bar, so that the blanket gas pressure is increased from 1.5 to 1.6 Bar. [To be confirmed] Gas Odorant: The odorant is supplied with a Hazard Data Sheet. This is based on Aliphatic Mercaptan mixtures in clear liquid form that is extremely flammable.

PS-GZT-TG-001 Revision (0) Draft Report

Page 17

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

9 Town Gas Emergency Plan The Central Emergency Room for Town Gas is located at the Head Office. This is connected to the Control Rooms at the Holding Company, GASCO (gas Exporters), Police and Fire Brigade. Communication is via landline telephone as well as wireless network. In addition, speedy communication with the gas network and customers by means of the SCADA control room. Emergencies in Stations: There is an emergency plan for the gas stations. The documents reviewed refer only to Greater Cairo and includes the following hazards scenarios: o Hazards resulting from adding the odorant. o Hazards resulting from gas leaks inside pressure reduction stations and main regulators. o Fire near the odorant storage tank. o Breakage of off-take or outlet line inside odorising station. o Fire in main electricity board . o Fire inside the Control Room. o Fire inside or outside the station fence. o Gas leak during a fire inside a pressure reduction station. Actions in response to the above emergencies are generally restricted to the isolation of valves, reporting the incident and follow up with relevant authorities, e.g. transmission lines, operations and SCADA. Wind direction and stability as well as neighbouring installations / public are not considered in existing emergency plans. Gas Pipeline: The Company has an emergency booklet that covers the main gas transmission line and customers. There is also an emergency room dedicated for such emergencies. Emergencies are prioritised at three levels, and include the following: o Gas leaks (instrumentation and equipment), o External leaks, o House leaks, o Gas exp losion, o Natural events to include earthquakes, heavy rain and external events. Response to these emergencies focus on isolation and reporting for actions.

PS-GZT-TG-001 Revision (0) Draft Report

Page 18

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

10 Weather Data The Weather Data relevant to this study consists of a list of weather conditions in the form of different combinations of wind -speed, temperature, humidity and atmospheric stability. The weather conditions are an important input into the dispersion calculations and results for a single set of conditions could give a misleading picture of the hazard potential. Meteoceanographic data gathered for Greater Cairo over a period of 5 years. This data included wind speed and direction; air temperature and pressure, as well as current speed, direction and wave height. The general climatic conditions at North Cairo are summarised below: Air Temperature oC: o Minimum recorded o Maximum recorded o Yearly average

- 1.1 52.2 28

Relative humidity %: o Average daily minimum o Average daily min imum o Annual average

82 54 78

The recorded annual wind speeds at Cairo are shown in Table 10.1 . Table 10.1 Wind speeds at Cairo (Knots) Month Jan.

Feb.

Mar.

Apr.

May

June

July

Aug.

Sep.

Oct.

Nov.

Dec

Wind speed

5.6

6.3

6.2

5.6

5.2

4.4

3.4

3.6

4.0

3.8

4.4

4.7

In wind Rose figures the radius = 10% Average wind speed =

2.44 m/sec.

Wind Direction : Three permanent high-pressure belts control the wind circulation over Egypt: the Azores, the Indian subtropical and the South Atlantic subtropical. In addition, there is a permanent low-pressure belt ‘the dolorums’ wh ich crosses Africa near the equator. Seasonal high and low pressure systems also alternate over the continental mass, the red sea, the Mediterranean and the Arabian Peninsula.

PS-GZT-TG-001 Revision (0) Draft Report

Page 19

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Table 10.2 Wind Rose for North Cairo

Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec

345/ 014 5.4 9.9 15.8 15.8 21.3 24.4 26.3 29.1 22.8 19.4 16.6 10.6

015/ 044 8.5 14.4 16.3 21.4 24.9 22.2 16.4 16.2 23.9 23.2 17.2 10.8

045/ 074 7.4 8.7 9.3 13.3 13.3 8.3 4.8 3.8 8.1 11.4 6.6 6.8

075/ 104 3.0 3.6 3.6 4.3 3.0 1.3 0.8 0.8 1.4 2.9 1.9 2.4

105/ 134 1.3 1.2 1.4 0.7 0.6 0.1 0.2 0.4 0.2 0.4 0.5 0.8

135/ 164 2.7 1.3 1.6 0.6 0.3 0.1 0.0 0.0 0.2 0.4 0.8 1.9

165/ 194 13.5 8.0 5.0 1.7 0.7 0.2 0.1 0.0 0.7 1.5 3.3 8.8

195/ 224 14.8 12.0 7.5 3.0 1.3 0.7 0.2 0.2 1.2 1.8 5.7 12.0

225/ 254 12.6 10.1 6.4 5.0 2.0 1.2 1.2 0.9 0.8 3.0 6.6 8.3

255/ 284 8.5 7.1 7.9 7.0 4.3 2.9 3.1 2.6 1.3 3.7 6.3 8.2

285/ 314 6.3 6.1 8.2 8.6 7.7 9.3 7.7 6.8 6.3 7.6 7.2 5.9

315/ 344 5.8 8.9 10.1 12.5 14.4 20.1 23.4 21.6 15.4 10.5 9.7 8.8

The prevailing winds are quite parallel to or heading towards the Northwest, mostly from west to north all year, except December and January, when they are from SE. When atmospheric low pressure is passing quite frequently and fast, the wind direction will change ‘anti-clockwise’, normally during a short period of one to two days. After a low pressure has passed, the wind returns to the prevailing direction (W-NW). The mean wind speed at Cairo is 2.44 m/sec. Data on the direction of wind at North Cairo was obtained from the Egyptian Meteorological Office. Table 8 shows the analysis of the 12-months wind distribution data over a period of 10 years. FIGURE 10.1 gives the average wind directions at Cairo throughout the year.

Feb

Jan

May

March

June

Sept

July

Oct

April

August

Nov

Dec

FIGURE 10.1 Average wind directions at Cairo PS-GZT-TG-001 Revision (0) Draft Report

Page 20

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

The overall analysis of the wind data at Cairo is given in what is known as the wind rose. FIGURE 10.2 shows Cairo wind rose, based on data collected during 1992 - 2000. Note that winds blow towards the centre of the rose.

18%

> 22.5 m/sec

20%

20- 22.5 17.5- 20 15- 17.5

15%

7%

12.5- 15 10- 12.5

5%

7.5 -10

6% 5%

4% 3% 4% 8% 5% 2.4

5- 7.5 2.5 - 5

N

0- 2.5

FIGURE 10.2 the Wind Rose at Cairo Stability Categories: The two most significant variables, which would affect the dispersion calculations, are: Wind -speed and atmospheric stability. The stability class is a measure of the atmospheric turbulence caused by thermal gradients. Pasquill Stability identifies six main categories, which are shown in the Table 10.3. Table 10.3 Pasquill Stability Categories A B C Very Unstable Unstable Moderately Unstable

D Neutral

E Moderately Stable

F Stable

Neutral conditions correspond to a vertical temperature gradient of about 1 (oC) per 100m. Cairo weather data for the Geographical area is somewhat limited and do not show seasonal variations over a long time. Therefore, the calculations included in this study have considered alternative stabilities for the average wind speed of 2.4 m/sec. This was done with reasonable accuracy, since the stability is related to the wind speed, and the range of stabilities that is observed for a given wind speed is generally small, as shown in the Table 10.4. As the range is large for a given wind speed, the calculations have initially considered four different combinations of wind speeds and stability classes to include the worst possible conditions. The calculations have also considered atmospheric temperature (30 oC), relative humidity 70% and surface roughness parameter of 0.1 . Table 10 .4 The Relationship between Wind speed and Stability PS-GZT-TG-001 Revision (0) Draft Report

Page 21

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Wind speed (m/s) 6

Day-time Solar Radiation strong medium slight A A-B B C C

A-B B B-C C-D D

B C C D D

thin 3/8 >4/5 D F D E D D D D D

At night, the ground is often cooler than the air if the sky is clear, and this gives rise to the most stable conditions and potentially the greatest effect distances. FIGURE 10.3 shows the criteria used for the selection of weather parameters used for the consequences modelling for this study.

PS-GZT-TG-001 Revision (0) Draft Report

Page 22

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Is the ground covered in frost or snow?

Yes

m/s

Check category against wind speed

F

0- 6 >7

E

No Is it night-time?

Yes

No Inland sites

Yes

Is sky overcast?

D

m/s

No Sky more than half covered

Coastal sites

Y

3

Check category against wind speed

No Check category against wind speed

Sky clear? Wind mainly from from the sea?

Yes

m/s 5

D

No Time within 1 hr before sunset?

Yes

D

No Time within 1 hr after sunrise?

Yes

Sky clear and wind calm/light?

Yes

F

No

D

No Is sky overcast?

Yes

Check category against wind speed

No

Summer only

m/s 8

C

0- 4 >5

D

Select weather type from Hot

Check category against wind speed

m/s

Warm

Cool

Check category against wind speed

A B C

m/s 7

D

A B

Check category against wind speed m/s 0- 4 >5

C D

C D

FIGURE 10.3 Determinations of Modified Pasquill Stability Categories

PS-GZT-TG-001 Revision (0) Draft Report

Page 23

November 2006

F E D F E D

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Category D (neutral) is the most probable at inland sites, and appears to occur for up to 80% of the time at Cairo. To overcome the uncertainty of the accuracy of Cairo weather data results, the following cases were selected in this analysis to study the effects of normal and extreme weather conditions at Cairo. Table 10.5 Sets of weather conditions initially selected for this study: Set 1 Set 2 Set 3 Wind speed Stability Wind speed Stability Wind speed Stability 3 m/s B 1 m/s D 2 m/s D

Set 4 Wind speed Stability 4 m/s F

The wind speed range between 1 to 5 m/s was considered to be reasonable representation of typical conditions at Cairo. This would overcome some of the uncertainty of the meteorological data provided by the meteorological office. Wind speeds in excess of 8 m/s are likely to disperse the cloud over long distances to well below LFL. The weather set 3 was eventually selected to represent the most likely conditions, however the worst case conditions shall be defined by a sensitivity analysis study.

PS-GZT-TG-001 Revision (0) Draft Report

Page 24

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

11 Generic Release Scenarios Events associated with release, dispersion and ignition of flammable releases considered in this study can be summarized in the following figure.

Release

Yes

Ignites?

No

Dispersing cloud

Yes

Ignites?

No

More obstacles Greater confinement Flame acceleration

Jet fire

Pool fire

Cloud fire

Fast flame

Internal

Safe

explosion

dispersion

Impinge?

Yes Structural Failure

BLEVE

Figure 11.1 Hazardous events These events can be more detailed as follows: Jet fires

A jet fire will result from an ignited pressurized hydrocarbon gas release. The consequence of jet fires is directional depending on the on release orientation. Jet fires typically have flame temperature of about 2,200 oF and can produce high intensity thermal radiation. The high temperature poses a hazard from direct effects of heat on humans and also from possibility of escalation. If a jet flame impinges upon a target such as a vessel, pipe or structural member, it can cause failure of the item to fail within several minutes. Jet (spray) fire will also result from ignited continuous releases of pressurized flammable liquid. The momentum of the release carries the material forwards in a plume entraining air to give a flammable mixture as gas is released from the plume.

PS-GZT-TG-001 Revision (0) Draft Report

Page 25

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Flash fires

If flammable gas accumulates in an unconfined area and is ignited, then the result will be a flash fire within the flammable limits of the vapour cloud.

Explosions

Ignition of accumulated gas in semi-confined areas may also be accompanied by an explosion; the overpressure generated will depend on the degree of congestion and confinement of the process area, and the gas cloud size.

Pool fires

If a liquid release is ignited after it has time to form a pool, a pool fire results. Because they are less well aerated, pool fires tend to have lower flame temperatures and produce lower levels of thermal radiation than jet fires. They also produce more smoke. Although a pool fire can still lead to structure failure of items within the flame, this would take longer than in a jet fire. An additional hazard of pool fires is their ability to flow. A burning liquid pool can spread along horizontal surface or run down a vertical surface to give a running fire.

BLEVE (Fire Ball)

BLEVE stands for Boiling Liquid Expanding Vapour Explosion. A fire ball can occur if a vessel containing fuel ruptures in the presence of an ignition source (usually a jet or pool fire). A fraction of the liquefied fuel subsequently released will evaporate immediately and take part in a huge fireball, which has the shape of a hemispherical burning cloud or ball of fire. High degree of turbulent mixing and rapid air entrainment allows large quantities of fuel to be consumed in a short period of time.

Structural failure

Loss of structure integrity due to overheating of structure members. The structure shall collapse under much lower load than the designed due to increased temperature.

Safe dispersion

Dilution of the released gases beyond the lower flammability limits (LFL) leading to safe dispersion situation.

PS-GZT-TG-001 Revision (0) Draft Report

Page 26

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

12 Specific Release Scenarios The following are the different types of hazardous scenarios envisaged in the new facility: •

Leakage of gas from equipment/ pipe flanges, if un-ignited, can result in the formation of a vapour cloud. This cloud is likely to remain partly in the area depending on weather conditions and wind velocity and direction. Subsequently, if this stagnant cloud finds a source of ignition and if the composition of the cloud is within the flammability limits of the gas, flash fire can occur. Fire water protection is not suitable for this scenario, as the flash fire scenario is estimated to be very short in terms of seconds.

•

On the other hand, if this stagnant cloud migrates to a congested area with certain degree of confinement and finds a source of ignition within its flammability limits, an explosion can occur with overpressure waves varies depending on the degree of congestion.

•

Leakage of gas from equipment / pipe flanges, followed by immediate ignition resulting in a gas jet fire. Fighting such jet fires with fire water is not effective. Details of the ESD system needs to be provided, a s if the ESD system will automatically shut-off the gas supply on confirmed fire detection, duration of this scenario will be very short due to small inventory in the PRS station (considered as isolatable section). If ESD system is not provided, the expected duration of such fires is considerably long due to higher inventory from the high pressure pipeline shall feed the fire (considered as un-isolatable section).

•

The over-pressure created by such flash fires is considered negligible due to very low congestion in an unconfined open area.

•

Pool fire due to leakage of liquid hydrocarbon odorant is not considered credible case as the capacity of the odorant storage tank is (1 m 3) only, provided with fire fighting system.

PS-GZT-TG-001 Revision (0) Draft Report

Page 27

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

13 Impairment Criteria This section defines the human injury and asset impairment criteria in caring out the consequence analysis of the identified hazardous events scenarios on the proposed facilities. Table 13.1 represents standard human impact criteria as applied in consequence modelling. Table 13.1: Criteria for Assessment of Fire Effects on Humans Event Effect Distance to Effect Jet fire / Pool fire 4.7kW/m2 Will cause pain in 15-20 seconds and injury after 30 seconds exposure. 12.5 kW/m2 Significant chance of fatality for extended exposure and high chance of injury. 2 37.5 kW/m Significant chance of fata lity for people exposed instantaneously. Flash fire LFL Fatal for people in the flammable cloud path Explosion overpressure 0.03 Bar Will cause injuries from flying debris 0.21 Bar 20% chance of fatality to a person in a protected enclosure 0.35 Bar Threshold for eardrum damage, 50% chance of fatality for a person within enclosure, 15% chance of fatality for a person in the open. 0.70 Bar Will cause 100 % fatality for a person within enclosure or in the open. The criteria applied for assessment of the effects of fire on assets are summarised in Table 13.2. Table 13.2: Criteria for Assessment of Fire Effects on Assets Impairment Mechanism Level Effect Thermal Radiation 4.7kW/m2 Impairment of evacuation/embarkation areas 6.3 kW/m 2 Impairment o f escape routes Probit (Pr) Fatality (For personnel protected by Equation clothing) Pr = -37.23 + 2.56.ln(t.I4/3) I: Intensity (W/m 2) t: exposure time (seconds) Thermal Radiation or Flame 500 deg.C Structural Failure. Impingement on Load Bearing Structural Steel Both jet fires and explosions can lead to structure failure of items, though this will take several times longer for jet fires than for explosions. Table 13.3 presents indicative failure times under hydrocarbon fire impact conditions, where times to failure refer to burn through or loss of load bearing capacity.

PS-GZT-TG-001 Revision (0) Draft Report

Page 28

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Table 13.3: Structure Failure times in Fires (Indicative) Time to Failure (Min) Component Jet Fire Pool Fire Unprotected structural steel beam 10 10 Unprotected steel plate 5 10 A-60 firewall 15 60 H-120 firewall 60 120

Table 13.4 reports published information on the explosion overpressure effects. Table 13.4: Explosion Overpressure Effects Explosion Overpressure - Bar(g) 0.02 0.07 0.07-0.14 0.08-0.1 0.15-0.2 0.2 0.3 0.34 1.0 2.0

Damage 50% windows shattering Collapse of tank roof Connection failure of corrugated panelling Minor damage to steel framework Wall of concrete blocks shattered Collapse of steel framework "Reparable damage" cladding blown off. Offshore bridjes and lifeboats impaired Steel walls blown off. Process plant within offshore module rupture, in neighbouring modules damaged. 50% chance for ESD valve closure failing Columns and buoyant deck of semi-sub ruptured Riser wall rupture

The heat rad iation contours, flash fire diameters (the distance to the lower flammability limits) and the explosion overpressures are estimated using SHELL, FRED Version (4.0) consequence modelling software package.

PS-GZT-TG-001 Revision (0) Draft Report

Page 29

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

14 Flammability Assessment 14.1 General An assessment of all flammable and combustible materials present on the facilities is required to determine those materials that can be excluded from further assessment in the QRA due to low flammability and hence the low probability of ignition. In order for a fire to start there must be an ignition source of sufficient heat intensity to cause ignition. However, after a fire has started, the heat necessary to sustain combustion is typically supplied by the combustion process. A flammable gas or vapour burns in air only over a limited concentration range. Below a certain concentration in air, the Lower Flammability Limit (LFL), the mixture is too ‘lean’, and above a certain concentration in air, the Upper Flammability limit (UFL), the mixture is too ‘rich’ to sustain combustion. The concentrations between these limits constitute the flammable range. Flammability limits vary between hydrocarbon gases. For example, methane is flammable between 5 - 15% v/v and propane between 2.1- 9.5% v/v. Process streams consist of a mixture of hydrocarbons and on loss of containment the flammability limits depend on the composition of the gas or vapour that is released to air. The flash point of a flammable liquid is the temperature at which the vapour pressure is sufficient to result in a concentration of vapour in air above the liquid corresponding to the lower flammable limit. On loss of containment or where open to the atmosphere, a hydrocarbon liquid that has a flash point below ambient temperature is readily ignitable. A liquid with a high flash point, could also ignite if raised in temperature above its flash point by an external heat source, if released as a high pressure spray that promotes vaporisation, or if soaked into lagging (insulating materials). Flash point is th e main parameter in the hazard classification of flammable liquids. The flammability of materials has been assessed using the Flammability Hazard Ranking from NFPA 325M under the categories summarised in Table 14.1. Flammable liquid classes referred to in Table 14.1 are explained in IP15.

PS-GZT-TG-001 Revision (0) Draft Report

Page 30

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Flammability Rating 4

3

2

1

0

Table 14.1: NFPA 325M Flammability Rating Description This degree includes flammable gases, liquids and class IA flammable liquids. The preferred method of fire attack is to stop the flow of material or to protect exposures while allowing the fire to burn itself out. This degree includes class IB and IIC flammable liquids and materials that can be easily ignited under almost all normal temperature conditions. Water may be ineffective in controlling or extinguishing fires in such materials. This degree includes materials that must be moderately heated before ignition will occur and includes class II and IIIA combustible liquids and solids and semi-solids that readily give off ignitable vapours. Water spray may be used to extinguish fires in these materials because the materials can be cooled below their flash points. This degree includes materials that must be pre -heated before ignition will occur, such as class IIIB combustible liquids and solids and semisolids whose flash point exceeds 93.4ºC, as well as most ordinary combustible materials. Water may cause frothing if it sinks below the surface of the burning liquid and turns to steam. However, a water fog that is gently applied to the surface of the liquid will cause frothing that will extinguish the fire. This degree includes any material that will not burn.

In general, materials with a flammability rating of 3 and 4 are readily ignited and present a greater fire hazard than materials with flammability rating of 1 or 2 that require pre -heating (e.g. by an existing fire) before ignition can occur. The properties of the various flammable materials present on the facilities are summarised in Table 14.2. The flammability of the various inventories is discussed in further detail in the following sections. Table 14.2 physical properties of Selected Flammable / Combustible materials

14.2 Process Hydrocarbons NFPA Flammability Index = 4 Two major components of the process hydrocarbon inventories are methane and ethane. Propane, butane and pentane are also present. As a result the process hydrocarbon inventories are classified

PS-GZT-TG-001 Revision (0) Draft Report

Page 31

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

as highly flammable, corresponding to an NFPA Flammability Index rating of 4. Based on this rating the fire and explosion hazards associated with the inventories are further assessed in the QRA.

14.3 Electrical Fires NFPA Flammability Index = N/A Electrical faults can cause overheating, sparking and a fire. However, electrical fires are only likely to have minor, localised impact. Smoke/fire detection and active fire protection should be effective in controlling such incidents. Escalation of electrical fires to process areas is highly unlikely, because the electrical equipment are segregated from other areas and hence the risk from electrical fires is assessed as insignificant and does not warrant further assessment within the QRA. However, electrical faults will be considered as a potential cause of ignition of hydrocarbon releases.

14.4 Conclusions Flammable inventories carried forward for further assessment are limited to process hydrocarbon inventories as these are identified as initiating causes of fire/explosion that present significantly. Other inventories will be further assessed in the QRA only in the cases where it is identified that significant risk exists due to the potential for escalation of an initiating fire/explosion (Odorant).

PS-GZT-TG-001 Revision (0) Draft Report

Page 32

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

15 SHELL FRED Version (4.0) Consequence Modelling Software Consequence modelling will be used to simulate the Major Accident Events (MAE) raised from the scenario identification. The consequence modelling will be carried out using SHELL, FRED version (4.0). SHELL, FRED Flyer, including technical capabilities and benefits is presented in Appendix 2. FRED consequence modeling software stands for Fire, Release, Explosion and Dispersion developed by SHELL global solutions. This software calculates in graphical displays and detailed reports the previously mentioned hazardous consequences and presents their extent and degree of danger. FRED determines the heat radiation contours from different fire scenarios depending on the amount of fuel burning, type of fuel and wind direction. It calculates a fluid release flow rate depending on the fluid pressure, the size and location of the hole. Also, It calculates the explosion overpressure contours resulting from the ignition of released gas inside confined space depending on the type of the fuel exploding and the degree of confinement. Finally it performs gas dispersion calculation and calculates the gas concentration contours in fraction of the lower explosive limits depending on the type of gas released, release rate, wind stability, wind speed and surface roughness. SHELL FRED has the following simulation modules (deta iled list): • • • • • • • • • • • • • • • • • • • • • • • •

Tank Top Fire Pool Fire Trench Fire Gas Jet Flame (known rese rvoir pressure) Gas Jet Flame (Known mass flow rate) Shell BLEVE BLEVE (TNO) Temperature Rise Pressurised release (known reservoir pressure) Pressurised release (known mass flow rate) Pressure relief valve Blowdown Two-Phase Blowdown LPG two -phase Explosion CAM Explosion TNO Explosion TNT Dense gas dispersion Gaussian dispersion (in stantaneous) Gaussian dispersion (Continuous) Gaussian dispersion (Non boiling liquid pool) Heat Up Vessel Burst Bubble Plume

This hazardous consequence simulation is normally carried out in order to optimize the design, while on the other hand it will be used in this study to estimate the degree of danger raised from the hazardous events on the facilities under study in order to assess the associated risks. The SHELL, FRED input file is detailed as follows:

PS-GZT-TG-001 Revision (0) Draft Report

Page 33

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Process conditions:

Component

n-Butane Propane Ethane Methane Nitrogen Carbon dioxide • • • • •

Weight Mole Critical Critical Molecular Atmos Freeze Heat of Fraction Fraction Temp °C Pressure Weight BP °C Pt °C Comb norm norm bara kg/kmol kJ/kg 0.0321 0.0100 152.1 37.41 58.12 -0.5001 -138.4 45742.7 0.0487 0.0200 96.7 41.91 44.1 -42.1 -187.7 46383.8 0.0829 0.0500 32.18 48.08 30.07 -88.6 -182.8 47514.8 0.7966 0.9000 -82.6 45.35 16.04 -161.5 -182.5 50043.9 0.0155 0.0100 -146.9 33.56 28.01 -195.8 -210 0 0.0243

0.0100

31.06

72.86

44.01

-86.9

-56.6

Temperature = 30 °C Pressure = 70 bara Pressure downstream of release = 1.013 bara Use standard atmospheric pressure = yes Release source = Vapor space

Hole & release geometry: Hole geometry: • • •

Failure type = Custom Hole diameter = 0.1 / 0.025 / 0.005 m Discharge coefficient = 0.8

Pipe : • • • •

Pipe length = 100 m Pipe diameter = 0.2 m Pipe surface roughness = 4.6e -005 m Sum loss coefficient = 0

Release: • • •

Release height = 1 m Release angle from vertical = 90 deg Release angle, clockwise from North = 90 deg

Weather: • • • • • •

Temperature = 40 °C Relative humidity = 50 % Wind speed = 1 & 10 m/s Direction wind is going to = 180 deg (measured clockwise from North) Atmospheric stability conditions define by = Pasquill class Pasquill class = "E" stable & "B" Unstable

PS-GZT-TG-001 Revision (0) Draft Report

Page 34

November 2006

0

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Thermal radiation: • • •

Radiation contours = 1.5, 2.5, 6.3, 12.5, 32 kW/m² Height at which plan view contours to be plotted = 0 m Cross flame distance at which side view contours to be plotted = 0 m

Dispersion: • • • •

Surface roughness = 0.01 m Contours to plot: Plot type = LFL/UFL Sampling time = Instantaneous

Technical Notes: •

Shell Fred includes two methods of inputs to the discharge modelling, one is “known reservoir pressure” and the second is “known release mass flow rate”. The scenario was selected as “known reservoir pressure” in order to represent the maximum desired flow rate through the hole.

•

The following table provides values of absolute roughness, e, in metres for materials used in the construction of open channels. (Note that where pipes have become corroded, surface roughness can increase 10-fold):

Table 15.1: Values of Absolute Roughness

•

Pipe surface roughness was selected as 4.6e-005, which represents the steel material.

•

Different wind speeds were selected for the gas dispersion and heat radiation modelling, basically 1 m/s and 10 m/s.

•

Dispersion sampling time was selected to be “Instantaneous”, which represents the worst-case scenario (stricter than 10 minutes sampling).

•

Typical values of the incident radiation intensity if the target is impinged by flame are:

PS-GZT-TG-001 Revision (0) Draft Report

Page 35

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Table 15.2: Values of incident radiation intensity

•

Different Pasquill stability classes were selected for the gas dispersion and heat radiation modelling, basically B (Unstable) and E (Stable).

Table 15.3: Pasquill Stability Classes

Numbe r 1. 2. 3. 4. 5. 6. •

Class A B C D E F

Description Very Unstable Unstable Slightly Unstable Neutral Stable Very Stable

The following table contains the criteria for the Pasquill stability classes, taking time of day, wind speed and cloud cover influences into account.

Table 15.4: Pasquill Stability Classes

•

For the flammable gas dispersion modelling, the upper flammability limits (UFL) and lower flammability limits (LFL), have been selected as dispersion contours for representing different gas composition cloud contours.

•

For the toxic gas dispersion modelling, modelling values in the terms of parts per million (ppm) shall be selected as concentrations of interest for toxic gases dispersion. (As highly toxic gases, shall represent a critical safety factor in designing such types of facilities).

•

Obstacles on the level over which the plume is dispersing will have a tendency to break up the plume. This effect is quantified in the gas dispersion models by a surface roughness. Typical

PS-GZT-TG-001 Revision (0) Draft Report

Page 36

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

surface roughness lengths, as used in many models, are given in the following table. (The roughness values are NOT the actual size of the obstacles on the ground ). Table 15.5: Values of Surface Roughness

•

Failure Types supported by Shell FRED is listed in the following table, indicating the typical hole diameter resulting from the failure.

Table 15.6: Typical Hole D iameters

•

Fire characteristics can be summarized in the following table.

Table 15.7: Fire Characteristics

PS-GZT-TG-001 Revision (0) Draft Report

Page 37

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

16 Sensitivity Analysis The sensitivity analysis shall be performed in order to determine the worst case parameters (or the combination of the worst case parameters), which shall be utilized in the consequence modelling. The sensitivity analysis shall include the following parameters: 1.0

Release flow rate:

The release flow rate depends on the size of the hole assumed to leak, which can be summarized as follows: •

Major leak or full bore rupture (presents maximum release rate).

•

Pin hole leak (presents minimum release rate).

2.0

Release pressure :

The release pressure depends on the process design and operating pressure of the released materials, which can be summarized as follows: •

The proposed maximum equipment design pressure.

•

The proposed minimum equipment design pressure.

3.0

Release temperature:

The release temperature depends on the process design and operating temperature of the released materials, which can be summarized as follows: •

The proposed maximum equipment design temperature.

•

The proposed minimum equipment design temperature.

4.0

Ambient temperature:

Ambient temperature varies from high ambient temperatures in the summer to low ambient temperatures in the winter, which can be summarized as follows:

5.0

•

The proposed maximum ambient temperature in the summer is 40 (oC)

•

The proposed minimum ambient temperature in the winter is 5 (oC) Relative humidity:

Relative humidity varies from high relative humidity in the summer to low relative humidity in the winter, which can be summarized as follows:

PS-GZT-TG-001 Revision (0) Draft Report

Page 38

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

•

The proposed maximum relative humidity is 90 %.

•

The proposed minimum relative humidity is 50 %.

6.0

Wind speed:

Wind speed varies from high wind speed to low wind speed depending on the weather conditions, which can be summarized as follows: •

The proposed maximum wind speed is 10 m/s (presents very unstable weather conditions).

•

The proposed minimum wind speed is 1 m/s (presents very stable weather conditions).

7.0

Wind stability:

Wind stability presented as Pasquill stability classes varies from very unstable weather to very stable weather depending on the weather conditions, which can be summarized as follows: •

The proposed very unstable weather is (A).

•

The proposed very stable weather is (F).

Different Pasquill stability classes are represented in the following table: Number 1. 2. 3. 4. 5. 6.

Class A B C D E F

Descrip tion Very Unstable Unstable Slightly Unstable Neutral Stable Very Stable

Each of the previously mentioned parameters shall be checked with all other parameters are constant. (I.e. these parameters shall be checked one by one, and for each case all other parameters shall remain unchanged in order to determine the worst case scenario for each parameter). From the sensitivity analysis for the gas dispersion, it can be concluded that: 1. The gas dispersion distances shall be increased b y higher release flow rate. 2. The gas dispersion distances shall be increased by higher release pressures. 3. The gas dispersion distances shall be increased by lower release temperatures. 4. The gas dispersion distances shall be increased by higher ambient temperatures. 5. The gas dispersion distances shall be increased by lower relative humidity. 6. The gas dispersion distances shall be increased by lower wind speeds. 7. The gas dispersion distances shall be increased by higher weather stability.

PS-GZT-TG-001 Revision (0) Draft Report

Page 39

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

From the sensitivity analysis for the heat radiation, it can be concluded that: 1. The jet flame heat radiation distances shall be increased by higher release flow rate. 2. The jet flame heat radiation distances shall be increased by higher release pressures. 3. The jet flame heat radiation distances shall be increased by higher release temperatures. 4. The jet flame heat radiation distances shall not be affected by ambient temperatures. 5. The jet flame heat radiation distances shall not be affected by relative humidity. 6. The jet flame heat radiation distances shall be increased by higher wind speeds. 7. The jet flame heat radiation distances shall not be affected by weather stability. From the sensitivity analysis performed, it has been concluded that there is a combination of set of parameters that gives the worst case scenarios for the gas dispersion and heat radiation, while on the opposite side; there is a combination of set of parameters that gives the mild case scenarios for the gas dispersion and heat radiation. Both cases (the worst cases and mild cases) can be simulated using FRED consequence modelling software, however only the worst case scenarios for gas dispersion and heat radiation shall be governing in this report in order to present a conservative approach leading to conservative QRA results. From the sensitivity analysis, the following parameters have been selected to represent the worst case scenario parameters and shall be utilized in the consequence modelling analysis: •

The proposed maximum ambient temperature in the summer is 40 (oC),

•

The proposed minimum relative humidity in the winter is 50 %,

•

The proposed minimum wind speed is 1 m/s (presents very stable weather conditions),

•

The proposed very stable weather stability class is (F).

PS-GZT-TG-001 Revision (0) Draft Report

Page 40

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

17 Release Scenarios For the purposes of the analysis a number of representative release cases are defined in Table 9.1. Release rates have been calculated with SHELL FRED (4.0) assuming release materials to be a mixture of Methane, Ethane and Propane. Calculated release rates a re based on initial flow conditions. Release rates will diminish with time due to a reduction in pressure at the breach. Factors causing the pressure reduction include resistance to flow through the inventory system and depletion of the inventory. Emergency shutdown will initiate isolation of the inventory in case of provided (manually or automatically). Release durations depend on inventory size and the rate of inventory depletion. An indication of release durations based on simplifying assumptions is provided in consequence modelling simulation cases (Appendix 3). Case No. 1

2

Systems Included Release from the HP side of the PRS Release from the LP side of the PRS

Table 17.1: Representative Release Cases Locations Released Press . Temp . Hole Size Material (Bar) (°C) (mm) Upstream the Standard 5 PRS Natural 70 30 25 Gas 100 Downstream Standard 5 the PRS Natural 7 30 25 Gas 100

Simulation Case Name 1A 1B 1C 2A 2B 2C

For the PRS, two conditions are present described as follows: 1. High pressure gases upstream the PRS. 2. Low pressure gases downstream the PRS. The high pressure stream conditions have been selected to present the release scenario from the PRS, as it presents the worst case scenario (conservative design approach ). The following are the summary of the representative release cases: 1

High Pressure Release from 100-mm (4-Inch) Leak Upstream PRS (Case -1A),

2

High Pressure Release from 25 -mm (1 -Inch) Leak Upstream PRS (Case-1B),

3

High Pressure Release from 5-mm (1/4 -Inch) Leak Upstream PRS (Case-1C),

4

Low Pressure Release from 100-mm (4 Inch) Leak Downstream PRS (Case-2A),

5

Low Pressure Release from 25-mm (1-Inch) Leak Downstream PRS (Case-2B),

6

Low Pressure Release from 5 -mm (1/4 -Inch) Leak Downstream PRS (Case-2C).

The effects of fire on personnel and asset will vary with fire type. From the listing of release scenarios, the locations of potential fire types are identified and summarised in Table 11.1.

PS-GZT-TG-001 Revision (0) Draft Report

Page 41

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Table 17.2: Summary of Potential Fire Source Locations by Fire Type Fire Potential? (Yes / No) Gas Cloud Flash Fire Jet Fire Pool Fire Explosion PRS system Yes Yes Yes - (2) No (3) (1) (2) Pipeline Yes Yes Yes No (3) Notes: Location

1. Pipeline point release is similar to the PRS piping release; hence the simulation cases have been done once for both of them. 2. Pool fire has not been simulated as the type and composition of the Odorant has not been specified and advised yet, while there is no possibility of pool fire from the process stream as it is a standard Natural Gas streams. 3. Explosion is not considered in the analysis, as the PRS area is considered as an open area with no confinement or congestion. The listing of flammable material release scenarios contained in Appendix 3 provides details of: • • • •

Inventory size; Release location; Release rate; Potential fire types.

For the purpose of the analysis it is assumed that the facilities is normally unmanned but will be frequently visited by operations and maintenance teams comprising at least two personnel.

PS-GZT-TG-001 Revision (0) Draft Report

Page 42

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

18 Ignited Release Scenario 18.1 Generic Causes of Release Historically, incidents involving the loss of containment of flammable/combustible material in the oil and gas facilities have occurred due to causes including: • • • • • • • • • • •

Human error associated with operations and maintenance activities; Corrosion Erosion Fatigue/vibration/vortex shedding; Brittle fracture (e.g. due to low temperatures embrittlement); Impact (e.g. due to dropped object, projectile, impacts…etc.); Creep; Natural causes (e.g. storm, earthquake…etc); Operation beyond design envelope; Inappropriate choice of materials; and Inadequate design.

In relation to the new facilities, thorough design and the implementation by Company of an appropriate Safety Management System will ensure many of the causes listed above are either avoided or significantly reduced in potential.

18.2 Generic Causes of Ignition Historically, the causes of ignition of released flammable/combustible material in the oil and gas facilities have included: • • • • • • • • • • • • • •

F lames/direct heat; Hot surfaces; Hot work (e.g. welding, flame cutting, grinding); Mechanical sparks; Electrical equipment not classified for hazardous areas; Faulty electrical equipment; Lightning; Engines; Distressed equipment (e.g. overheated bearings); Impact energy (e.g. tools, dropped objects, projectiles); Chemical energy; Static electricity; Illicit smoking; and Hot soot particles.

Similar to causes of release, the above listed causes of ignition on the new facilities will be either avoided or significantly reduced in potential through thorough design and the implementation by Company of an appropriate Safety Management System.

PS-GZT-TG-001 Revision (0) Draft Report

Page 43

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

19 Typical Fire Consequence Analysis 19.1 Hydrocarbon Releases Hydrocarbon releases in the industry are either gaseous, mists or liquids and are either atmospheric releases or pressurized. Gas and mist releases are considered more significant since they are readily ignitable since they are in the gas state and due to the generation of vapour clouds which if ignited are instantly destructive in a widespread nature versus liquid fires that may be less prone to ignition, generally localized and relatively controllable. The cause of a release can be external or internal corrosion, internal erosion, equipment wear, metallurgical defects, operator errors third party damage or for operational requirements. Generally releases are categorized as: 1

Catastrophic Failure: A vessel or tank opens completely immediately releasing its contents.

2

The mount of release is dependent of the size of the container.

3

Long Rupture: A se ction of pipe is removed leading to two sources of gas. Each section being vented in an opening whose cross sectional areas are equal to the cross sectional area of the pipe (e.g., pipeline external impact and a section is removed).

4

Open Pipe: The end of a pipe is fully opened exposing the cross sectional area of the pipe.

5

Short Rupture: A split occurs on the side of the pipe or hose. The cross sectional area of the opening will typically be equal to the cross sectional area of the pipe or hose (e.g., pipe seam split).

6

Leak: Leaks are typically developed from valve or pump seal packing failures, localized corrosion or erosion effects and are typically "small" to "pin -hole" sized (e.g., corrosion or erosion leakages).

7

Vents, Drains, Sample Ports Failures: Small diameter piping or valves may be opened or fail which release vapours or liquids to the environment unexpectedly.

8

Normal Operational Releases: Process storage or sewer vents, relief valve outlets, tank seals, which are considered normal and acceptable practices that release to the atmosphere.

19.1.1 Gaseous Release There are a number of factors that determine the release rate and initial geometry of a hydrocarbon gas release. The most significant is whether the gas is under pressure or released at atmospheric conditions.

PS-GZT-TG-001 Revision (0) Draft Report

Page 44

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Depending on the release source the escaping gas can last from several minutes or hours, until the supply is isolated, depleted or fully depressurized. Common long duration sources are massive storage equipment, or long pipelines without intermediate isolation capabilities. If released under atmospheric conditions the gas will either rise or fall depending on its vapour density and will be directed in the path of the prevailing wind. In the absence of a wind, heavier gases will collect in low points in the terrain. Normally atmospheric gas releases are dispersed within relatively close distances to their point source, usually about 3 meters (10 ft.) These atmospheric releases, if ignited, will burn relatively close to the source point, normally in a vertical position with flames of short length. For gases released under pressure, there are a number of determining factors that influence the release rates and initial geometry of the escaping gases. The pressurized gas is released as gas jet and depending on the nature of the failure may be directed at any direction. All or part of a gas jet may be deflected by surrounding structures or equipment. If adequate isolation capabilities are available and employed, the initial release will be characterized by high flow and momentum which decreases as isolation is applied or supplied are exhausted. Within a few pipe diameters of the release point, the pressure of released gases decreases. Escaping gases are normally very turbulent and air will immediately be drawn into the mixture. The mixing of air will also reduce the velocity of the escaping gas jet. Obstacles such overhead platforms or structures will disrupt momentum forces of any pressurized release. These releases will generally produce a vapour cloud, which if not ignited will eventually disperse in the atmosphere. Where turbulent dispersion processes are prevalent (e.g., high pressure flow, winds, congestion, etc.), the gas will spread in both horizontal and vertical dimensions while continuing mixing with available oxygen in the air. Initially escaping gases are above the UEL but with dispersion and turbulence effects they rapidly pass into the flammable limits. If not ignited and given an adequate distance they will eventually disperse below the LEL. Various computer software programs are currently available that can calculate the turbulent jet dispersion, downwind explosive atmospheric locations, and volumes for any given flammable commodity, release rates and atmospheric data input. Generally most gases have a low vapour density and will rise. In any event, the height of a gas plume will mostly be limited by the ambient atmospheric stability and wind speed. If the gases are ignited, the height of the plume will rise due to the increased buoyancy of the high temperature gases from the combustion process.

19.1.2 Liquid Release When a liquid is released from process equipment, several things may happen, as shown in the Figure. If the liquid is stored under pressure at a temperature above its normal boiling point (superheated), it will flash partially to vapour when released to atmospheric pressure. The vapour produced may entrain a significant quantity of liquid as droplets. Some of this liquid may rainout onto the ground, and some may remain suspended as an aerosol with subsequent possible evaporation. The liquid remaining behind is likely to form a boiling pool which will continue to evaporate, resulting in additional vapour loading into the air. An example of a superheated release is a release of liquid ammonia from a pressurized container stored at ambient temperature.

19.1.3 Toxic Gas release The inhalation of toxic gases can give rise to effects, which range in severity from mild irritation of the respiratory system to death. Lethal effects of inhalation depend on the concentration of the gas to which people are exposed and on the duration of exposure. Mostly this dependence is non linear; as the concentration increases, the time required to produce a specific injury decreases rapidly. PS-GZT-TG-001 Revision (0) Draft Report

Page 45

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Immediately dangerous to life and health (IDLH) is defined as a condition that poses immediate danger to life or health, or a condition that poses a threat of severe exposure. Two factors are considered when establishing the IDHL limits: •

Personnel must be able to escape such an environment without suffering permanent health damage,

•

Personnel must be able to escape without severe eye or respiratory tract irritation or other condition that might impair their escape.

Immediately Dangerous to Life and Health: (IDLH) is an atmospheric concentration of any toxic, corrosive, or asphyxiate substance that poses an immediate threat to life or would cause irreversible or delayed adverse health effects or would interfere with an individual’s ability to escape from a dangerous atmosphere.

PS-GZT-TG-001 Revision (0) Draft Report

Page 46

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

19.2 Fire The combustion process: Fire, or combustion, is a chemical reaction in which a substance combines with oxygen and heat is released. Usually fire occurs when a source of heat comes into contact with a combustible material. If a combustible liquid or solid is heated it evolves vapour, and if the concentration of vapour is high enough it forms a flammable mixture with the oxygen of the air. If this flammable mixture is then heated further to its ignition point, combustion starts. Similarly, a combustible gas or vapour mixture burns if it is heated to a sufficiently high temperature. Thus there are three conditions essential for a fire: (1) fuel, (2) oxygen, and (3) heat. These three conditions are often represented as the fire triangle. If one of the conditions is missing, fire does not occur and if one of them is removed, fire is extinguished. Normally the heat required is initially supplied by an external source and then provided by the combustion process itself. The amount of heat needed to cause ignition depends on the form of the substance. A gas or vapour may be ignited by a spark or small flame . Ignition of a combustible gas or vapour mixture may occur in two ways. In the first the energy for ignition is supplied by a local source such as a spark or small flame at a point within the mixture. In the second the bulk gas mixture is heated up to its ignition temperature. The three conditions of the fire triangle indicate how fires may be fought. The first method is to cut off the fuel. Th is is particularly relevant for fires caused by leaks on process plant. The second method is to remove heat. This is usually done by putting water on the fire. The third method is to stop the supply of oxygen. This may be affected in various ways, including the use of foam or inert gas. Fire is sustained only if there is a net release of heat. The heat comes from the combustion of fuel. If this fuel is liquid or solid, it must first be vaporized. With liquids or solids fire usually involves a process of positive feedback. The heat evolved by the fire causes the vaporization of an increasing amount of fuel and the fire spreads. Fire growth and spread : Fire normally grows and spreads by direct burning, which results from impingement of the flame on combustible materials, by heat transfer or by travel of the burning material. The three main modes of heat transfer are (1) conduction, (2) convection and (3) radiation. All these modes are significant in heat transfer from fires. Conduction is important particularly in allowing heat to pass through a solid barrier and ignite material on the other side. Most of the heat transfer from fires, however, is by convection and radiation. It is estimated that in most fires some 75% of the heat emanates by convection. O n open plant much of the heat is dissipated into the atmosphere, but in steel structures it is transferred to the steel supports. PS-GZT-TG-001 Revision (0) Draft Report

Page 47

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Radiation is the other main mode of heat transfer. Although it usually accounts for a smaller proportion of the heat issuing from the fire, radiated heat is transferred directly to nearby objects, does not go preferentially upwards and crosses open spaces. For these reasons it is generally the most significant mode of transfer on open plant. Combustion of a flammable gas/air mixture occurs if the composition of the mixture lies in the flammable range and if the conditions exist for ignition. As already mentioned, ignition may result from either (1) bulk gas temperature rise or (2) local ignition. The combustion of the mixture occurs if the bulk gas is heated up to its auto-ignition temperature. Alternatively, combustion occurs if there is applied to the mixture a source of ignition which has sufficient energy to ignite it. Flammability limits: A flammable gas burns in air only over a limited range of composition. Below a certain concentration of the flammable gas, the lower flammability limit, the mixture is too `lean', while above a certain concentration, the upper flammability limit; it is too `rich'. The concentrations between these limits constitute the flammable range. The lower and upper flammability limits (LFL and UFL) are also sometimes called, respectively, the lower and upper explosive limits (LEL and UEL). They are distinct from the detonability limits. Flammability limits are affected by pressure, temperature, direction of flame propagation and surroundings.

19.2.1 Flash Fire A flash fire would result if a flammable vapour cloud builds up and engulfs a source of ignition, or an ignition source is introduced. The volume of the combustion products are approximately 8 times the volume of the vapour cloud, hence a flash fire would be much larger than the initial un-ignited vapour cloud. Although a flash fire can cause fatalities by flame impingement, it would be of insufficient duration to cause escalation unless it develops significant overpressure. It would then be termed a vapour cloud explosion. Due to the short duration of a flash fire, fatalities are considered to occur only within the flame itself. The size of the vapour cloud depends on: • • •

Release rate; Composition; Wind conditions.

Dispersion calculations should be performed to estimate the maximum gas cloud sizes within the LFL. These have been based upon horizontal releases into open air in the same directionas the wind for various wind speeds. The results of the gas dispersion calculations shall be represented graphically. These results will be used to assess the potential for an ignition source to be engulfed in a vapour cloud, the extent of potential flash fires and the potential for explosion.

PS-GZT-TG-001 Revision (0) Draft Report

Page 48

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

The dispersion calculations are valid for open area releases. Releases in congested areas will not disperse so readily and this will be taken into account in the assessment of effects on personnel and asset. The conclusions from the dispersion calculations are: • •

the heavier gases, propane and butane, produce similar size gas clouds for the same releases rate; methane gas tends to rise more rapidly due to buoyancy, particularly in light wind conditions;

The larger the gas cloud, the greater the size of the flash fire and potential explosion overpressure upon ignition.

19.2.2 Unobstructed Jet Fires Gas or vapour releases from holes in high-pressure hydrocarbon inventories give rise to turbulent jet fires if ignited. With this fire type pure fuel is released through an orifice and the air required for combustion is entrained from the surrounding atmosphere. At high release rates, the jet becomes highly turbulent, entrains more air and burns hotter. The jet lengths have been modelled using SHELL FRED (4.0). FRED uses the ‘Chamberlain’ model developed by the Shell Oil Company to derive gas jet flame lengths. Releases from the liquid phase of a process vessel (e.g. separator) will typically be driven by the vapour pressure of the liquid. Once the gas/liquid interface falls below the level of the leak a gas jet fire release will ensue driven by the pressure of the gas in the system. High-pressure condensate releases will atomise due to the momentum of release and vaporise due to the heat from the fire and burn as a self sustaining jet, some heavier fractions can drop out when the pressure drops to below approximately 5 bar(a), resulting in surface pool fire forming below the jet fire. Thermal radiation isopleths are propo rtional to the size of the jet fire. The dimensions of 1.5, 4.7, 6.3, 12.5 and 37.5 kW/m2 isopleths shall be calculated and included on the graph to facilitate assessment of effects on personnel and impairment of safety critical systems. The jet flame length (metres) for methane releases may be approximated from the mass release rate, m (kg/s) using a power law curve as follows: Jet Length = 13.5 m0.45 (based on BP Cirrus modelling results) Jet flame lengths for propane, butane and condensate are approximately 15% longer than for methane. The unobstructed jet fires will only occur from ignited releases originating from inventories at the edge of the process area and orientated outboard. These are less likely to cause damage or fatalities. Due to the congestion presented by the equipment and pipe work, the majority of potential process fires on the process area will be obstructed. These obstructed jet fires will result in a fireball type of fire, instead of a jet fire. For jet fires, the fire fighting systems (firewater or other fire fighting agents) are not efficient to fight such types of fires due to the high momentum release initiating such jet fires. Hence, the only way to control jet fires is to limit the isolatable inventory feeding the jet flame. PS-GZT-TG-001 Revision (0) Draft Report

Page 49

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

The jet fire will deplete by time due to the decrease in driving force across the release point "the hole", consequently the jet flame is expected to be reduced by time. For jet fires, it is essential to predict the approximate jet fire time duration in order to assess the extent of the hazardous consequences. Based on the isolatable section inventory within the system and the assumption that all operators are aware and trained to deal with such emergency situations, the approximate jet fire duration can be estimated as short duration fires. If the ESD system shall operate effectively in such cases; hence the approximate jet fire durations can be estimated as too short to cause fatality, injury or massive damage to equipment.

19.2.3 Obstructed Jet Fires Most jet fires will be obstructed due to the relatively congested layouts. These will burn as a continuous fireball. The diameter of these fireballs and the associated thermal radiation isopleths are calculated by considering the thermal radiation levels surrounding the fire. For fires above single grade level, the radiation isopleths are in the shape of a hemisphere. The heat radiated through the hemispherical skin is assumed to be equal to the heat generated by the burning as follows: Surface area of a hemisphere, A = 2?r2 Hence Q.(2.?.r 2) = m.H.p And

r = v(m.H.p/2.?.Q)

Where Q = Heat flux (kW/m2) p = Proportion of heat radiated (typically 20%) H = Heat of Combustion (kJ/kg) m = Burning Rate (kg/s) (equivalent to release rate) r = Radius (m) The actual fireball radius is estimated based on setting Q at 150 kW/m 2, which gives a conservative fire size. Curves are also calculated for the 1.5, 4.7, 6.3, 12.5 and 37.5 kW/m 2 isopleths. For fires between multiple levels structure, the radiation isopleth is assumed to be in the shape of a cylinder, the height of which is the distance between decks. The equilibrium equation for this case is calculated as follows: Surface area of a cylinder (excluding ends), A = 2.?.r Hence Q.(2.?.h.r) = m.H.p And

r = (m.H.p/2.?.h.Q)

Where h = Height (or length) of cylinder (m) For instance, a fireball in the centre of the deck level associated with a release rate greater than approximately 5 kg/s would produce fatal radiation levels to a distance about 20m from the fire PS-GZT-TG-001 Revision (0) Draft Report

Page 50

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

source. In reality, the fire would soon become ventilation limited and would tend to fill the area with flames lapping out around the perimeter.

19.2.4 Pool Fires Liquid pool fires may occur on ground floors or elevated plated levels or plated decks or in bunded areas due to: • • •

Rain-out from ignited liquid releases at pressures under approximately 3 to 5 bar(g); Delayed ignition of higher pressure liquid releases; or Ignition of low pressure liquid releases (e.g. from storage tanks at atmospheric pressure).

Liquid rain out from high pressure jet fires will be relatively small and will not be assessed as the resulting pool fire would have no additional consequences to the coincident jet fire. A high pressure liquid release would spray over a wide area and create a large flammable vapour cloud. Upon ignition, the fire would rapidly flash back to the source of the leak and burn as a liquid jet fire until the fuel is exhausted. A low pressure release of flammable liquid will drain into the drip pans located under vessels and tanks containing liquid inventories. These will be drained to the hazardous open drains system. Ignition would result in a pool fire in the drip pan under the vessel. The risk from pool fires to personnel or the asset is much less significant when compared to the jet fire or explosion hazards present on the facility and can be reasonably screened out from further assessment. Since, liquid hold up volumes is small compared to the gas volumes, liquid fires can be reasonably screened out from further assessment.

PS-GZT-TG-001 Revision (0) Draft Report

Page 51

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

20 Consequence Modelling Results General: The release scenarios shall be simulated based on 3-hole sizes 0.25 inch, representing instrument fitting failure, 1.0 inch rep resenting small pipe leak and 4.0 inches leak representing a 4-inch pipe full bore rupture or 4-inch hole size in a larger pipe diameter (corresponds to 5-mm, 25-mm and 100mm), with wind speed of 1 m/s and stability class "F" representing "Very Stable" weather conditions and with wind speed of 10 m/s and stability class "A" representing " Very Unstable" weather conditions. The released gases shall form a dispersing plume which tends to rise upwards to the lighter-than-air characteristics of the Natural Gas. Since the main function of the pressure reduction station is to reduce the high pressure gas to low pressure gas by throttling the gas, there are basically two categories of pressures mainly: 1. High pressure gases upstream the pressure reduction station (Maximum of 70 Bars). 2. Low pressure gases downstream the pressure reduction station (Maximum of 7 Bars). The release scenarios shall be performed for both HP and LP cases for the 3 -hole sizes, in order to give a full picture of the released gas characteristics. If the released gases between the LFL and the UFL inside the plume ignite, shall for a flash fire, which has relatively short time duration. The jet frustum fire (flame length) and heat radiation distances are measured in meters. Since the jet fire is originally a high momentum directed jet release, hence the effects of wind direction, wind speed or atmospheric stability on the jet flame are minimal. For jet fires, the fire fighting systems (firewater or other fire fighting agents) are not efficient to fight such types of fires due to the high momentum release initiating such jet fires. Hence, the only way to control jet fires is to limit the isolatable inventory feeding the jet fire. The jet fire will deplete by time due to the decrease in driving force across the release point "the hole", consequently the jet flame is expected to be reduced by time. For jet fires, it is essential to predict the approximate jet fire time duration in order to assess the extent of the hazardous consequences.

PS-GZT-TG-001 Revision (0) Draft Report

Page 52

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

20.1 High Pressure Release from 100-mm (4-Inch) Leak Upstream PRS (1A) General: This case model considers the scenario of a full bore rupture of 4 inch piping or 4 inch hole in a larger diameter piping, which represents the worst case scenario as the release source is a high pressure major leak. Gas Cloud / Flash Fire: The proposed hazardous area resulting from the gas cloud shall be limited by the upper flammability limit (UFL) and the lower flammability limit (LFL), which if ignited results in flash fire. The following figure represents the side view of the gas cloud (Plume) as a graphical display illustrating the LFL/UFL limits and the maximum plume height.

The following table represents the LFL and UFL limits and heights of the gas cloud (Plume) in figures. Concentration

LFL

UFL

Contour value (ppm) Downwind distance (m)

47226.4 148597.8 32.28 10.12

Height above ground (m) 0.8994 0 The following figure represents the plan view of the gas cloud (Plume) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 53

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Jet Fire: The following figure represents the side view of the jet fire (Torch Flame) as a graphical display illustrating the heat radiation levels.

PS-GZT-TG-001 Revision (0) Draft Report

Page 54

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

The following figure represents the plan view of the jet fire (Torch Flame) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 55

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

20.2 High Pressure Release from 25-mm (1 -Inch) Leak Upstream PRS (1B) General: This case model considers the scenario of a leak of 1 inch hole size in a larger diameter piping, which represents the medium case scenario as the release source is a high pressure minor leak. Gas Cloud / Flash Fire: The proposed hazardous area resulting from the gas cloud shall be limited by the upper flammability limit (UFL) and the lower flammability limit (LFL), which if ignited results in flash fire. The following figure represents the side view of the gas cloud (Plume) as a graphical display illustrating the LFL/UFL limits and the maximum plume height.

The following table represents the LFL and UFL limits and heights of the gas cloud (Plume) in figures. Concentration Contour value (ppm) Downwind distance (m)

LFL UFL 47226.4 148597.8 12.08 5.5

Height above ground (m) 0

0.7974

The following figure represents the plan view of the gas cloud (Plume) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 56

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Jet Fire: The following figure represents the side view of the jet fire (Torch Flame) as a graphical display illustrating the heat radiation levels.

PS-GZT-TG-001 Revision (0) Draft Report

Page 57

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

The following figure represents the plan view of the jet fire (Torch Flame) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 58

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

20.3 High Pressure Release from 5-mm (1/4-Inch) Leak Upstream PRS (1C) General: This case model considers the scenario of a pin hole leak of 1/4 inch hole size in a larger diameter piping, which represents the mildest case scenario as the release source is from a pin hole leak. Gas Cloud / Flash Fire: The proposed hazardous area resulting from the gas cloud shall be limited by the upper flammability limit (UFL) and the lower flammability limit (LFL), which if ignited results in flash fire. The following figure represents the side view of the gas cloud (Plume) as a graphical display illustrating the LFL/UFL limits and the maximum plume height.

The following table represents the LFL and UFL limits and heights of the gas cloud (Plume) in figures. Concentration

LFL

UFL

Contour va lue (ppm) Downwind distance (m)

47226.4 148597.8 4 1.2

Height above ground (m) 0.9443 0.9764 The following figure represents the plan view of the gas cloud (Plume) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 59

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Jet Fire: The following figure represents the side view of the jet fire (Torch Flame) as a graphical display illustrating the heat radiation levels.

PS-GZT-TG-001 Revision (0) Draft Report

Page 60

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

The following figure represents the plan view of the jet fire (Torch Flame) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 61

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

20.4 Low Pressure Release from 100-mm (4 Inch) Leak Downstream PRS (2A) General: This case model considers the scenario of a full bore rupture of 4 inch piping or 4 inch hole in a larger diameter piping, which represents the worst case scenario of the low pressure case downstream the PRS, as the release is from major leak. Gas Cloud / Flash Fire: The proposed hazardous area resulting from the gas cloud shall be limited by the upper flammability limit (UFL) and the lower flammability limit (LFL), which if ignited results in flash fire. The following figure represents the side view of the gas cloud (Plume) as a graphical display illustrating the LFL/UFL limits and the maximum plume height.

The following table represents the LFL and UFL limits and heights of the gas cloud (Plume) in figures. Concentration

LFL

Contour value (ppm) Downwind distance (m)

47226.4 148597.8 12.06 6.001

Height above ground (m) 0

UFL

0.675

The following figure represents the plan view of the gas cloud (Plume) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 62

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Jet Fire: The following figure represents the side view of the jet fire (Torch Flame) as a graphical display illustrating the heat radiation levels.

PS-GZT-TG-001 Revision (0) Draft Report

Page 63

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

The following figure represents the plan view of the jet fire (Torch Flame) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 64

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

20.5 Low Pressure Release from 25-mm (1-Inch) Leak Downstream PRS (2B) General: This case model considers the scenario of a minor leak from 1 inch hole size in a larger diameter piping, which represents the medium case scenario as the release source is from a minor leak. Gas Cloud / Flash Fire: The proposed hazardous area resulting from the gas cloud shall be limited by the upper flammability limit (UFL) and the lower flammability limit (LFL), which if ign ited results in flash fire. The following figure represents the side view of the gas cloud (Plume) as a graphical display illustrating the LFL/UFL limits and the maximum plume height.

The following table represents the LFL and UFL limits and heights of the gas cloud (Plume) in figures. Concentration

LFL

UFL

Contour value (ppm) Downwind distance (m)

47226.4 148597.8 5.5 1.8

Height above ground (m) 0.8138 0.9689 The following figure represents the plan view of the gas cloud (Plume) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 65

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Jet Fire: The following figure represents the side view of the jet fire (Torch Flame) as a graphical display illustrating the heat radiation levels.

PS-GZT-TG-001 Revision (0) Draft Report

Page 66

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

The following figure represents the plan view of the jet fire (Torch Flame) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 67

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

20.6 Low Pressure Release from 5 -mm (1/4 -Inch) Leak Downstream PRS (2C) General: This case model considers the scenario of a pin hole leak of 1/4 inch hole size in a larger diameter piping, which represents the mildest case scenario as the release source is from a pin hole leak. Gas Cloud / Flash Fire: The proposed hazardous area resulting from the gas cloud shall be limited by the upper flammability limit (UFL) and the lower flammability limit (LFL), which if ignited results in flash fire. The following figure represents the side view of the gas cloud (Plume) as a graphical display illustrating the LFL/UFL limits and the maximum plume height.

The following table represents the LFL and UFL limits and heights of the gas cloud (Plume) in figures. Concentration

LFL

UFL

Contour value (ppm) Downwind distance (m)

47226.4 148597.8 1.2 0.3

Height above ground (m) 0.9803 0.984 The following figure represents the plan view of the gas cloud (Plume) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 68

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

Jet Fire: The following figure represents the side view of the jet fire (Torch Flame) as a graphical display illustrating the heat radiation levels.

PS-GZT-TG-001 Revision (0) Draft Report

Page 69

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

The following figure represents the plan view of the jet fire (Torch Flame) as a graphical display on the actual plot plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 70

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

21 Likelihood Data 21.1 Process Release The detailed Generic values for equipment failure frequencies are summarized in the following table (based on API-581 - Table 8.1): Table 21.1 Generic Equipment Failure Frequencies

The hole size distribution base on the following areas: • •

Pressure reduction Station (PRS). Under ground pipeline.

Based on the above mentioned hole size distribution and data indicated by the API-581, the failure data for various equipment are indicated in the following tables. The failure data for the pipeline are summarized in the following table: No. 1 2 3

Table 21.2 Pipeline Failure Rate Data Pipe Size Failure rate / ft / year Failure rate / year Pin hole of 0.25 inch 4 X 10 -7 1.3 X 10 -4 -7 Pipe leak of 1 inch 4 X 10 1.3 X 10 -4 -8 Pipe rupture of 4 inch 8 X 10 2.6 X 10 -5

PS-GZT-TG-001 Revision (0) Draft Report

Page 71

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

The pipeline leak frequency shall be calculated from the process piping of diameter of 4 inches failure rate with overall length of approximately (100 meters) as the worst case scenario of the highest failure rate of (1.3 X 10 -4). The failure data for the PRS is summarized in the following table:

No. 1 2 3

Table 21.3 PRS and Process Piping Failure Rate Data Pipe Size Failure rate / ft / year Failure rate / year Pin hole of 0.25 inch 9 X 10 -7 1.5 X 10 -4 -7 Pipe leak of 1 inch 6 X 10 1.0 X 10 -4 -8 Pipe rupture of 4 inch 7 X 10 1.2 X 10 -5

The PRS leak frequency shall be calculated from the process piping of diameter ranges from 1 inch to 4 inches failure rate with overall length of approximately (50 meters) as the worst case scenario of the highest failure rate of (1.5 X 10-4).

21.2 Ignition Probability The probability of ignition depends on the availability of a flammable mixture, the flammable mixture reaching an ignition source and the type of ignition source (energy etc.). The ignition sources on the facilities include: • • • • • • •

Hot work Faults in electrical equipment Faults in rotating equipment Ignition caused by combustion engines or hot surfaces Automatic ignition in the event of a fracture or rupture Static electricity Flare / open flame

Generic ignition probabilities have been taken from Lees. Ignition probability data are provided for both gas and oil releases based on mass release rate. Typical ignition probab ility data are given in Table 21.4. Table 21.4 Ignition Probability Data Mass Release Category

Mass Release Rate (Kg / Sec)

Minor Major Massive

PS-GZT-TG-001 Revision (0) Draft Report

50

Page 72

Ignition Probability Gas Oil 0.01 0.01 0.07 0.03 0.3 0.08

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

22 Risk Assessment Risk shall be determined for both workers and public using international risk management guidelines as a reference. The risk will be compared with international risk acceptance criteria. Risk assessment will comprise the following items: • • • •

Failure rate. Consequence rate. Ignition probability. Vulnerability assumptions.

The risk equation shall be as follows: Risk = Consequence Frequency X Occupancy X Vulnerability Consequence Frequency = Leak Frequency X Ignition Probability Where: Consequence Frequency: is the likelihood of a hazardous event occurring. Leak Frequency: is the equipment failure frequency. Ignition Probability: is the likelihood of a release to become a fire or explosion . Occupancy: is the personnel presence in the area. Vulnerability: is the likelihood that the specific person will be fatally injured by the effect of the event (determined from the consequence modelling software).

PS-GZT-TG-001 Revision (0) Draft Report

Page 73

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

22.1 Individual Risks ‘IR’ to Workers In order to calculate Individual Risk ‘IR’ for workers, there is a need to identify who is exposed to the fire and explosion hazards from all hazards at the PRS station, not just as a result of gas leaks. The proportion of time individuals is exposed to the hazards and their vulnerability should be considered in estimating this risk. Vulne rability is the probability that exposure to the fire/explosion hazards will result in fatality. The following calculations relate to the most vulnerable individuals on site, identified to be workers involved in fire-fighting. ‘IR’ is calculated using the following model: IR

=

IR (Workers) =

S (frequency of fires/explosions) X Vulnerability 1.5 E-04 X 0.3 X 1.0 = 4.5 E-05 per year

The major contributory factor for the increased level of ‘IR’ is the potential gas vapour cloud explosion due to the confined conditions of the pressure reduction streams and the Odomatic system. Evaluation of Individual Risks as shown in FIGURE 22.1 indicates that individual risk to workers at the PRS to be within the ALARP region. This should be reduced to a level that is as low as reasonably practicable, taking cost into account.

1.0E 1.2E

- 03/year

1.0E-03 per person/yr

- 03/yr

ALARP ALARP

2.2E - 05 per yr 4.5E 1.0E

- 05/year

1.0E-05 per person/yr

FIGURE 22.1 Evaluation of IR to Town Gas Workers

PS-GZT-TG-001 Revision (0) Draft Report

Page 74

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

22.2 Individual Risk to the Public The general public exposed to major hazards as a result of the PRS activities are road users around the site and residents in buildings nearby. Modelling of the consequences identified gas/odorant releases to affect the public outside the station. The station is surrounded by busy roads, as well as the public buildings. ‘IR’ is calcula ted using the following model: IR

=

S (frequency of fires/explosions) X Occupancy X Vulnerability

IR (Public)

=

1.5 E-04 X 2 X 0.3 X 1.0 = 9.0E-05 per year

Evaluation of IR to the public is shown in FIGURE 22.2 .

1.0E 1.2E

- 03/year

1.0E-03 per person/yr

- 03/yr

ALARP ALARP

2.2E - 05 per yr 9.0E 1.0E

- 05/year

1.0E-05 per person/yr

FIGURE 22.2 Evaluation of IR to the public It is therefore concluded that, Individual Risk to the public is also within the ALARP region and should be reduced to a level as low as reasonably practicable. These risks shall be evaluated against the international risk acceptance criteria.

PS-GZT-TG-001 Revision (0) Draft Report

Page 75

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

23 Risk Evaluation The risks assessed shall be evaluated based on the international risk acceptance criteria. The ALARP principle has been adopted for risk evaluation. The ALARP region is that point at which the time, effort difficulty and cost of further risk reduction become out of proportion compared with the amount of risk reduction achieved. Risks lower than the ALARP region risks will be considered minor risk and consequently they will not be considered. Risks higher than the ALARP region risks will be considered major risk and consequently they will be not acceptable and further reduction measures are required. The international risk acceptance criteria are presented in the following figure .

UNACCEPTABLE REGION

Workers

Public

Maximum tolerable limit

1 in 1000 per year A L A R P Benchmark existing installations 1 in 5,000 per year

Maximum tolerable limit

1 in 10,000 per year

ALARP OR TOLERABILITY REGION

ALARP OR TOLERABILITY REGION

ALARP Benchmark new installations 1 in 50,000 per year

(Risk must be demonstrated to have been reduced to a level which is practicable with a view to cost/benefit)

Minimum tolerable limit

1 in 100,000 per year

Minimum tolerable limit

ACCEPTABLE REGION

1 in 1 million per year ACCEPTABLE REGION

INDIVIDUAL RISK TO WORKERS (including contractor employees)

INDIVIDUAL RISK TO THE PUBLIC (all those not directly involved with company activities)

From the risk assessment and the international risk acceptance criteria the conclusion is presented in the following table. No 1.0 2.0

Calculated Risk 4.5 E-05 9.0 E-05

Acceptable Risk 1.0 E-05 1.0 E-05

PS-GZT-TG-001 Revision (0) Draft Report

Page 76

Area Type Workers Public

Acceptance ALARP ALARP

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

24 Risk Reduction Measures (Recommendations) Risk reduction measures (Recommendations) may include reducing the risk by several technically feasible methods, generally are as follows: •

Measures to eliminate the risk.

•

Measures to reduce the exposure of personnel to the hazards.

•

Measures to reduce the frequency of occurrence.

•

Measures to mitigate the consequences if the event does occur.

•

Measures to improve evacuation in case of emergency (event occurs).

It has been concluded that the risk falls within the ALARP region for the individual risk to workers and public within the industrial area (PRS). This is due to population density of the residential areas, as well as the population is present most of the time, while in the industrial areas the population is relatively low as well as because the existence of the safety precautions and procedures and the protection measures. Since the calculated risk (ALARP region) is close to the acceptable region (1 X 10-5), hence there are some minor risk reduction measures required to reduce the calculated risk beyond the acceptance border. These risk reduction measures (recommendations) are summarized as follows: •

It is strongly recommended to install an automatic fire detection system to activate ESD valve (or solenoid valve – depend on the PRS design) at the PRS station inlet in order to stop feeding the flame in case of fire

•

The control room inlet door should be located in the upwind direction away from the PRS station (Inlet door should not face the PRS station).

•

Alternatively, the control room should be provided by a secondary means of escape at the back side of the room, which shall be used in case of blockage of the main escape route by jet fires.

•

Consider jet fire rated passive fire protection system be applied to all safety critical shutdown valves ESDVs or Solenoid valves in order to maintain small isolatable inventories. (As applicable)

•

It is strongly recommended that the block isolation valve at the off-take point from the coming pipeline, to be placed in a safe place protected fro m jet fire for personnel intervention in case of emergency in order to isolate the pipeline as soon as possible.

•

It is strongly recommended to have pipeline marking signs indicating in Arabic and in English "Do Not Dig" and "High Pressure Pipeline Underneath" in order to prevent such extreme hazardous situation.

•

It is recommended to include the prevailing wind direction on the PRS site plan.

PS-GZT-TG-001 Revision (0) Draft Report

Page 77

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

•

It is recommended to have an elevated wind sock installed in the PRS site, which can be seen from the control room and from outside the fence to determine the direction of gas migration in case of major gas leak.

•

It is recommended to have a gas detection system within the PRS area to automatically sense the released gases as a percentage of LFL, in order to provide early warnings of gas release.

•

Also, it is recommended to have point gas detectors at the control room HVAC intake to automatically sense the released gases as a percentage of LFL, in order to provide early warnings of gas release, if provided .

•

Investigate a strategy to inform the residential area beside the PRS with the risk associated with the activities as well as the methods required for annunciating if any leak occurs.

•

The design should fully comply with IGE TD/3 code requirements.

•

There is a need to develop a safe system of work, based on risk assessment for dealing with potential gas leaks.

•

Consideration should be given to the remote actuation of isolation and slam-shut valves by Town Gas SCADA System for PRS’s as well as the transmission and distribution pipelines.

•

There is a need to produce Hazardous Area Classification drawings for all Pressure Reduction Stations.

•

Review planned preventive maintenance policy and implementation.

•

There is a need to produce a ‘Station Manual’ for each PRS. This manual should include formalized procedures, including precautions and a site scenario specific emergency plan.

•

Site emergency plans must take into account wind direction and stability and should consider interfaces with other adjacent parties as well as the public living nearby.

•

Town Gas needs to consider the security arrangements for all PRS’s.

•

Consider formalizing procedures for filling the odorant storage tank, to include necessary precautions in the event of possible leaks.

•

There is a need that Town Gas should apply risk assessment to all activities and to formalize procedures and permit-to-work systems.

PS-GZT-TG-001 Revision (0) Draft Report

Page 78

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

25 Conclusion Quantitative risk assessment study has been performed for the new facilities (Pressure Reduction Station and only the associated portion of the connecting pipeline) by Town Gas Company. For the purpose of the analysis it has been assumed that the PRS open area is not normally manned but will be frequently visited by operations and maintenance teams comprising at least two personnel. SHELL FRED version (4.0) has been selected for the consequence modeling of different types of hazardous consequences as follows: • • •

Flammable gas clouds Flash fires Jet fires

SHELL FRED version (4.0) is Shell’s suite of Fire, Release, Explosion and Dispersion models used to predict the consequences of the accidental or design release of product from process, storage or distribution operations. For the PRS release scenario, the leak have been simulated based on 3 -hole sizes 0.25 inch, representing instrument fitting failure, 1.0 inch representing small pipe leak and 4.0 inches leak representing a 4-inch pipe full bore rupture or 4-inch hole size in a larger pipe diameter (corresponds to 5 -mm, 25-mm and 100 -mm), with wind speed of 1 m/s and stability class "F" representing "Very Stable" weather conditions and with wind speed of 10 m/s and stability class "A" representing " Very Unstable" weather conditions. For the PRS gas dispersion scenario, the flammable gas dispersion distances (flash fires) have been simulated based on 3-hole sizes (5mm, 25mm and 100mm) and 2-wind cases (1F and 10A). The gas dispersion distances are calculated in terms of Lower Flammability Limits (LFL) and Upper Flammability Limits (UFL). The heat radiation from flash fires will not significantly affect equipment and structure due to the short duration of flash fires. Flash fires are represented by the extent of the flammability limits of the released gases. As a conclusion, flash fires are predicted to emanate from the PRS isolatable section. The pipeline flash fire presents the worst case scenario due to the relatively higher inventory and long duration of release potential. The jet frustum fire (flame length) and heat radiation distances are measured in meters. Process release failure frequencies and ignition probabilities have been identified for the detailed quantitative risk assessment (QRA) purposes. Quantitative risk assessment (QRA) has been performed to all hazardous events developed from the scenario development section. The risks have been assessed for the industrial workers and general public in different areas. The risks assessed have been evaluated based on the international risk acceptance criteria.

PS-GZT-TG-001 Revision (0) Draft Report

Page 79

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

It has been concluded that the risk falls within the ALARP region for the individual risk to workers and public within the industrial area (PRS). Finally, risk reduction measures (recommendations) have been proposed to reduce the risk and improve the facilities safety standards.

PS-GZT-TG-001 Revision (0) Draft Report

Page 80

November 2006

TOWN GAS COMPANY EL-TEBEEN PRESSURE REDUCTION STATION QUANTITATIVE RISK ASSESSMENT STUDY

26 References 1. NFPA 325M, 2. SHELL FRED Version (4.0) documentation, 3. Frank P. Lees, Loss Prevention in the Process Industries, 2001, 4. API-581, Risk Based Inspection recommended practice , 5. IGE Codes, Institution of Gas Engineers and Managers, 6. Town Gas Project Documents.

PS-GZT-TG-001 Revision (0) Draft Report

Page 81

November 2006