187706066 QRA Lyse LNG Base Load Plant Train 1

STAVANGER QRA Lyse LNG Base Load Plant Linde Project No:

Item No:

Lyse Contract No :

2110A11U

Lyse Infra AS

61-10156.05.01

Linde Job Code:

Lyse Project No:

STAVANGER

R100

Linde Doc. No:

Lyse Doc. No:

Page

&AA S-CS 1002

R100-LE-S-RS0003

1 of 133

Quantitative Risk Analysis (QRA) Lyse LNG Base Load Plant Train 1

03

ISSUE

03

25.08.2008

02

ISSUE

02

14.03.2008

01

ISSUE

01

22.02.2008

Revised acc. to Lyse Comments Can Revised acc. to the Comments in Can QRA presentation from 26.02.2008 Revised acc. to the Lyse Comments Can

DRAFT Status

Issue

28.12.2007

Rev (Lyse)

(Linde)

(Linde)

Date

Description

Rath

Buttinger

Rath

Ralph

Rath/Baumgartner

Ralph

Can

Buttinger

Can

Prepared

Reviewed

Approved


2110A11U

Lyse Contract No :

Lyse Infra AS Item No:

61-10156.05.01

Linde Job Code:

Lyse Project No:

STAVANGER

R100

Linde Doc. No:

Lyse Doc. No:

Page

&AA S-CS 1002

R100-LE-S-RS0003

1 of 133

Table of Contents

1.0

Executive Execu tive Summary........... Summary.......................... ............................. ............................. ............................. ................ 3

2.0 2.1 2.2

Introduction Introduct ion ........... ...................... ....................... ....................... ....................... ....................... ....................... ....................... .................... .........9 9 Objective Objective of the Study ...................... ................................. ....................... ....................... ....................... ....................... .................. .......9 9 General General Descriptio Description n of the Approach Approach ...................... ................................. ....................... ....................... .................. .......9 9

3.0

General Gener al Description of Process and Facilities......................... Facilities......................... 13

3.1 3.1.1 3.1.2 3.1.3 3.2 3.2.1 3.2.2 3.3 3.3.1 3.3.2 3.4 3.5 3.6 3.7

Natural Gas Treatment and Gas Liquefaction .......................... ............. .......................... ...................... ......... 13 Feed Gas Reception Reception ...................... ................................. ...................... ....................... ....................... ...................... ................... ........13 13 Natural Gas Pretreatment...............................................................................13 NG Liquefaction..............................................................................................14 Refrigerant Refrigerant System.................. System............................. ...................... ....................... ....................... ...................... ....................... ............... ... 14 Refrigerant Refrigerant Cycle................... Cycle.............................. ....................... ....................... ...................... ....................... ....................... ................ ..... 15 Refrigerant Storage and Make-Up..................................................................15 LNG Storage / LNG Loading ...................... ................................. ...................... ....................... ....................... .................. .......16 16 LNG Storage Storage ...................... .................................. ....................... ...................... ...................... ....................... ....................... ................... ........16 16 LNG Loading Loading ...................... .................................. ....................... ....................... ....................... ....................... ....................... .................. .......16 16 Fuel Gas System System ....................... .................................. ...................... ....................... ....................... ...................... ...................... ............. .. 17 Hot Oil System System ..................... ................................. ....................... ...................... ...................... ....................... ....................... ................. ...... 17 Flare System System ...................... .................................. ....................... ...................... ....................... ....................... ...................... ................... ........18 18 ESD and Blowdown System...........................................................................18

4.0

Study Method Methodology... ology................. ............................ ............................ ............................ ........................ .......... 20

4.1 4.2 4.3 4.4

Risk Analysis Basics.......................................................................................20 Definition and Types of Risk...........................................................................20 Acceptance Criteria ......................... ............ ......................... ......................... .......................... .......................... ......................... ............ 21 Hazard Identification.......................................................................................23

5.0

Data used for the Risk Assessment................................... Assessment.......................................... ....... 25

5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.7.1 5.7.2 5.7.3

Scenarios Scenarios ....................... .................................. ....................... ....................... ....................... ....................... ....................... .......................25 ...........25 Leak Frequencie Frequencies s ....................... .................................. ....................... ....................... ...................... ...................... .......................26 ............26 Release Release Duration............. Duration........................ ....................... ....................... ...................... ....................... ....................... ......................27 ...........27 Atmospheric Conditions................................. Conditions..................... ........................ ......................... ......................... ........................ ............ 28 Population Distribution....................................................................................28 Ignition Sources..............................................................................................29 Consequence Calculations.............................................................................30 Discharge Discharge and Dispersion Dispersion ...................... ................................. ...................... ....................... ....................... ......................30 ...........30 Instantane Instantaneous ous Releases Releases........... ...................... ...................... ....................... ....................... ...................... ...................... .............. ... 31 Continuou Continuous s Releases Releases ...................... ................................. ...................... ....................... ....................... ...................... .................. .......31 31


2110A11U

Lyse Contract No :

Lyse Infra AS Item No:

61-10156.05.01

Linde Job Code:

Lyse Project No:

STAVANGER

R100

Linde Doc. No:

Lyse Doc. No:

Page

&AA S-CS 1002

R100-LE-S-RS0003

2 of 133

5.7.4 5.7.5 5.7.6 5.8

Release Release Duration............. Duration........................ ....................... ....................... ...................... ....................... ....................... ......................31 ...........31 Dispersion Dispersion ...................... .................................. ....................... ....................... ....................... ....................... ....................... ......................31 ...........31 Thermal Radiation and Overpressure.............................................................31 Mitigation Measures taken into Account ......................... ............. ......................... ......................... ................... ....... 32

6.0

Results Result s of the Risk Analysis...................................... Analysis.................................................... ................ .. 34

6.1 6.2 6.3 6.4

Risk 1st and 2nd party........... party ...................... ....................... ....................... ...................... ....................... ....................... .................. .......34 34 rd Risk 3 party ...................... ................................. ....................... ....................... ....................... ....................... ....................... ................... .......39 39 Location Location Specific Specific Risk ....................... .................................. ....................... ....................... ....................... ....................... ............... .... 45 Overpressure Risk..........................................................................................46

7.0

Sensitivity Sensit ivity Evalua Evaluation.................... tion.................................. ............................ ............................ ................. ... 50

7.1 7.1.1 7.1.2 7.2 7.2.1 7.2.2 7.3

Sensitivity 1: Pit on the jetty, LNG Storage Tank and the Pentane Pentane Tank ..................... ................................. ....................... ...................... ...................... ....................... ....................... ................... ........50 50 Discussio Discussion n ...................... ................................. ....................... ....................... ....................... ....................... ....................... .......................50 ...........50 Compariso Comparison n with Criteria Criteria ....................... .................................. ...................... ....................... ....................... ...................... ............. 53 Sensitivity 2: Rock Wall towards towards the public public area on the peninsula ................. .............. ... 53 Discussio Discussion n ...................... ................................. ....................... ....................... ....................... ....................... ....................... .......................53 ...........53 Compariso Comparison n with Criteria Criteria ....................... .................................. ...................... ....................... ....................... ...................... ............. 54 Sensitivity 3: Splitting of process vessels inside the refrigerant cycle into smaller vessels and additional block valves to reduce the volume of inventory loops.........................................................................54

8.0

Conclusions Conclu sions .............. ............................ ............................ ........................... ........................... ......................... ........... 55

9.0

Appendix Appen dix A: Assumption Assumption Sheets ............ .......................... ........................... .................... ....... 56

10.0

Appendix Appen dix B: Hazard Hazard Identification Identification ............. ........................... ............................ ................. ... 90

11.0

Appendix Appen dix C: Equipme Equipment nt Count ............. .......................... ........................... ........................ .......... 95

12.0

Appendix Appen dix D: Results Results of LEAK 3.2 3.2 Calculations Calculations ............ ....................... ........... 99

13.0

Appendix Appen dix E: Individual Risk Ranking Ranking Report............. Report ......................... ............ 104

14.0

Appendix Appen dix F: Details on the Analysis Analysis Procedure Procedure ............. .................... ....... 127

15.0

References Refere nces .............. ............................ ............................ ............................. ............................. ....................... ......... 132

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Introduction

1.0

25.08.2008 Page 3

Executive Summary

Linde Engineering AG (Linde), on the behalf of Skangass AS (Skangass), has conducted a Quantitative Risk Analysis (QRA) of the first train of the new Lyse LNG Base Load Plant, located near Stavanger, Norway. The objective of the study was to determine the level of risk associated with the Lyse LNG Base Load Plant, which is currently being designed, and compare it with the acceptance criteria given by Lyse Infra AS (Lyse).

Approach To achieve this objective, a thorough analysis was made of all hazardous substance inventories and streams within the plant. In particular, all equipment were counted and used as a basis to calculate leakage frequencies. The determination of leakage frequencies was done using the program "LEAK", a proprietary program from Det Norske Veritas (DNV). To achieve this, the whole plant was segmented, four leak size categories were defined, and leakage frequency calculations were performed for the segments based on the categories. Meteorological data as well as population data provided by Lyse were used for study. The data are important for DNV's risk assessment tool PHAST RISK (further SAFETI), which takes into account (when applicable): • • • •

Pool fires, Jet Fires, Flash Fires and Vapour Cloud Explosions.

For the scenarios defined for the Lyse LNG Base Load Plant, the population and the determined ignition source distribution were entered into PHAST RISK and analysed with respect to their contribution to individual risk and to societal risk.

Results PHAST RISK calculates both individual risk and societal risk. The individual risk for 1 st, 2nd and 3rd parties are calculated based on these results which are then compared to the acceptance criteria. As expected, the main contribution to the overall risk is due to vapour cloud explosions and flash fires. Individual Risk, 1st and 2nd party

Individual risk is a measure of risk to which an individual person is exposed. The individual risk criteria are divided in this analysis into Individual Specific Risk (ISR) and Average Individual Risk (AVR). The 1st party risk is defined as a fatality risk for the Lyse LNG Base Load Plant personnel. Maintenance personnel and operators during supervision rounds are considered to be the most exposed personnel. Fatality risk for the LNG Carrier personnel (Truck, Ship Loading and external contractors) have been considered as 2 nd party and are also assumed to be within the most exposed personnel group.


25.08.2008 Page 4

The figures 1 and 2 show the calculated individual risk contour lines for the Lyse LNG Base Load Plant. The calculated risk contours of individual risk for the most exposed person is illustrated in Figure 1. The figure shows the contours of the most exposed person to suffer a fatality every 100 000 years (green line), every 1 000 000 years (dark blue line). The risk is illustrated for the most exposed person present in the process plant area, 20 % of their working time per year. -

10 /yr -6 10 /yr

Figure 1: Most exposed person individual risk contour lines for the Lyse LNG Base Load Plant

The risk contours for the individual risk are also calculated and is illustrated in Figure 2. The figure shows the contours of individual risk for a fatality every 10 000 years (green line), every 100 000 years (dark blue line), etc. The risk is illustrated for 1 person present at any point outside a building in the plant, continuously 8 hours a day, 5 days a week during a whole year (45 weeks).


25.08.2008 Page 5

10- /yr 10-5 /yr 10-6 /yr -7 10 /yr -8 10 /yr

Figure 2: Individual risk contour lines for the Lyse LNG Base Load Plant

The Individual Risk (IR) has been extracted from the PHAST RISK risk report: It is calculated for 1 person and for each worker group present at any point in the plant, continuously 8760 hours per year. The Individual Specific Risk (ISR) for 1st and 2nd party, which considers the individual working hours for each group, is given below in Table 1.

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Introduction st

25.08.2008 Page 6

nd

Table 1: Individual Specific Risk (ISR) for 1 and 2 party

Buildings/Personnel Locations Operator

ISR [year] -4 1.2 X 10

Process Area (Maintenance)

1.2 X 10-4

Truck Loading

6.7 X 10

-5

Ship Loading (Jetty)

2.0 X 10

-5

Ship Bridge

5.0 X 10-5

Ship Deck

4.8 X 10-5

ISR > 1 X 10-3

Not acceptable

1 X 10-3 < ISR < 1 X 10-6

ALARP

ISR < 1 X 10-6

Acceptable

The Average Individual Risk (AVR) of 5.0 X 10-5 per year for all personnel (1 st and 2nd party) is within the ALARP regime, i.e. As Low As Reasonably Practical, which means that the mitigation measures may be applied as long as the respective cost benefit ratio is reasonable. Individual Risk, 3rd party

For the Lyse LNG Base Load Plant such mitigation measures have already been applied (e.g. a rock wall "mound" around the LNG tank, the ESD and Blowdown system). The Individual Specific Risk (ISR) for the 3 rd party risk is given below in Table 2. rd

Table 2: Individual Specific Risk (ISR) for the 3 party

Personnel Locations Peninsula

ISR [year] -8 4.6 X 10

Hiking Track

2.2 X 10

-6

Ferry Terminal_office workers

7.6 X 10

-7

Ferry Terminal_industry workers

3.8 X 10-7

Ferry Terminal_passengers

4.0 X 10

-7

Energiveien+Risavika_office workers

4.6 X 10

-9

Energiveien+Risavika_industry workers

4.6 X 10

-9

Container Area_office workers

3.2 X 10-9

Container Area_industry workers

3.2 X 10-9

Rest Companys_office workers

2.8 X 10-14

Rest Companys_industry workers

2.8 X 10

-14

Living Quarters

3.5 X 10

-10


ISR > 1 X 10-5

Not acceptable

1 X 10-5 < ISR < 1 X 10-7

ALARP

ISR < 1 X 10-7

Acceptable

25.08.2008 Page 7

The Average Individual Risk (AVR) of 1.5 X 10-7 per year for people living, working or staying outside the Lyse LNG base load plant does not exceed the acceptance criteria of 1 X 10 -5 / year and is within the ALARP regime. Societal Risk, 3rd party

Societal risk (or 3 rd party risk) is a measure of the collective risk to which a certain population is subjected as a whole. It is usually depicted in form of a so-called FN curve, which shows the frequency (F), that a given number, N people or more (hence N+) will be exposed to lethal consequences. The societal risk calculated for the Lyse LNG Base Load Plant is shown below in Figure 3.

Figure 3: Societal risk FN curve for the Lyse LNG Base Load Plant


25.08.2008 Page 8

The minimum and maximum risk criteria are shown in Figure 3 as blue and green lines respectively. Calculations of the external societal risk (e.g. Hiking Track, Peninsula, Industry Area and Ferry Terminal) have shown that this risk for the Lyse LNG Base Load Plant falls into the area between the upper and lower limit line, i.e. the ALARP regime.

Conclusions A careful risk analysis of the first train of the Lyse LNG Base Load Plant has been performed, including a very detailed counting of all pieces of equipment (including all pipelines, vessels and compressors etc.). It has been found that the calculated levels of individual risk for the 1 st, 2nd and 3rd parties are in compliance with the criteria set by Lyse. The individual specific risk for 1 st and 2nd party for the most exposed person in each group, maintenance and operators, is lower than 1 X 10 -3 per year and within the acceptance criteria. The average individual risk for personnel is 5.0 X 10-5 per year and therefore clearly below the acceptance criteria of 1 X 10 -4 / year. The individual specific risk for 3 rd party for the most exposed population (e.g. hiking track, ferry terminal industry and office workers) is within the ALARP regime. The average individual risk is 1.5 X 10-7 per year and therefore within the lower region of ALARP, close to acceptable in general. The calculated risk for the Peninsula people is acceptable since the rock wall (mound) is taken into account (refer to the Chapter 7.2).


2.0

25.08.2008 Page 9

Introduction

Linde is currently performing the design of train 1 of the new Lyse LNG Base Load Plant, located near Stavanger, Norway, on behalf of Skangass. The design of the plant shall conform to EN 1473:2007 " Installation and equipment for liquefied natural gas – Design of onshore installations" [1]. To fulfil EN 1473 a hazard assessment shall be carried out during the design of the plant. A part of this hazard assessment is a risk investigation, in this case using Quantitative Risk Analysis (QRA). This document describes in detail the results and methodology used to obtain the results of the QRA.

2.1

Objective of the Study

The objective of the study was to estimate the level of risk by QRA. The performed QRA covered all essential risks of the new Lyse LNG Base Load Plant as far as they are of relevance and have been determined in the Hazard Identification (HAZID) [Appendix B]. The individual personnel risk and the 3 rd party risk are evaluated in this study. The overpressure risk to the plant buildings and equipment (Central Control Room, LNG Tank etc.) and a consequence modelling of worst case scenarios, e.g. hydrocarbon dispersion from the LNG Tank, are included in the calculations.

2.2

General Description of the Approach

QRA is a well established methodology to assess the risks of industrial activities and to compare them with risks of normal activities. Linde has used a QRA methodology as shown in Figure 4. The QRA performed by Linde used the QRA Reports performed by Advantica [2] as a reference. Data Collection

This study is based on the following documents: • • • • • • • • • • •

Process Flow Diagrams (PFDs) Heat and Material Balance Process and Instrument Diagrams (P&IDs) Process Description General Plot Plan Mechanical and Process Data Sheets ESD and Blow-down System Concept Lyse LNG Base Load Plant Site conditions Manning Level Table [3]. Acceptance Risk Criteria for Lyse LNG Base Load Plant Development Area Plan and Information from Lyse


25.08.2008 Page 10

Figure 4: QRA Methodology

Collection and Analysis of Background Data

This was an internal Linde exercise to collect information relevant to the QRA study. The leak frequencies for equipment, valves etc. are based on DNV database and included in DNV's proprietary program “LEAK”. Hazard Identification (HAZID)

The hazard identification process is important for any risk analysis. A HAZID was been performed prior to the QRA by Linde. A HAZOP study for the main plant has been completed.


25.08.2008 Page 11

Frequency Analysis

Failure frequencies were determined for each event in order to perform a probabilistic risk assessment. Generally, a number of techniques are available to determine such frequencies. The approach relies on generic data. This provides failure frequencies for equipment items where data has been obtained from failure reports from a range of facilities. DNV has developed an extensive generic failure frequency database for this purpose, which is compiled in DNV's proprietary LEAK 3.2 software. These leak frequencies are based on the "UK Health & Safety Executive" data for offshore facilities. To reflect the design of the Lyse LNG base load plant, which is a onshore facility and has clean service, new leak frequencies for pipes and process vessels based on the "Purple Book" [6] are implemented in the LEAK Program. The changes are shown in Appendix D. This program was used to determine overall leakage frequencies subsequently subsequently used in the risk assessment. Consequence Analysis

For each hazard scenario PHAST RISK (Software for the Assessment of Flammable, Explosive and Toxic Impacts) and PHAST (Process Hazard Assessment Software Tool) software was used to determine consequence effect zones for each hazard. The different possible outcomes could be: Dispersing of Hydrocarbon Vapour Cloud Explosion Fireball BLEVE Flash Fire Jet Fire Pool Fire.

• • • • • • •

The CO2/H2S (sour gas) in CO 2 wash unit is routed to the regenerative thermal oxidation and then sent to atmosphere at safe location. Dispersion from a leak of CO 2/H2S gas cloud due to low operating pressure is not considered as the contribution to the risk is minor compared to the above mentioned outcomes. The particular outcomes modelled depend on source terms (conditions like fluid, temperature, pressure etc.) and release phenomenology. The current understanding of the mechanisms occurring during and after the release is included in state-of-the-art models in the PHAST RISK and PHAST packages. Risk Calculations

The outcome of the PHAST RISK analysis are risk terms presented in form of risk contours and FN curves, where the former is a form of location specific individual risk measurement while the latter is a measure for societal (group) risk. The individual risk is the risk for a hypothetical individual assumed to be continuously present at a specific location. The individual at that particular location is expected to sustain a given level of harm from the realization of specified hazards. It is usually expressed in risk of death per year. Individual risk is presented in form of risk contours. Societal Risk is the risk posed to a local community or to the society as a whole from the hazardous activity. In particular it is used to measure the risk to every exposed person, even if they are exposed on one brief occasion. It links the relationship between the frequency and the number of people suffering a given level of harm from the realization of a specified hazard. It is usually referred to a risk of death per year.


25.08.2008 Page 12

Risk Criteria

Risk criteria for both individual risk and societal risk have been discussed with Lyse. These criteria are compared to other risk criterion and to the results of the actual risk assessment for the plant. Risk Assessment

Once risks have been determined, they will be assessed against the criteria level and ranked to determine the principal contributors. Ranking enables attention to be focused on the main contributors. This is of particular significance when assessing the viability of different mitigation measures. Risk Mitigation

Risk reduction measures concentrate on the major risk contributors identified during risk ranking. Discussion is made on how different risk reduction measures will affect the overall risk level in relation to the ALARP principle (As Low As Reasonably Practical). Report structure

The safety studies are documented according according to the following report structure: •

•

•

•

•

•

•

Main report The main report summarizes the study data, methodology, the risk results, conclusions and sensitivities Appendix A – Assumptions The main assumptions where the studies are based on are presented in this appendix. Appendix B – HAZID This appendix documents the results of the HAZID workshop in Munich, October 2007. Appendix C – Equipment Count This appendix documents the equipments with their dimensions and inventories used to determine the leak size and – frequency for the risk assessments Appendix D – Result of LEAK 3.2 Calculations Calculations This appendix documents the risk leakage frequencies based on the UK HSE databank [4] and the Dutch Purple Book [ 5] Appendix E – Individual Risk Ranking Report This appendix documents the risk ranking points, for which the individual risk has been calculated Appendix F – Details on the Analysis Procedure This appendix gives details of the actual QRA methodology

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 General Description of Process and Facalities

3.0

25.08.2008 Page 13

General Description of Process and Facilities

The section briefly describes the process and facilities to ensure a common understanding. The description only addresses those parts, which are of relevance to the QRA.

3.1

Natural Gas Treatment and Gas Liquefaction

3.1.1

Feed Gas Reception

Feed gas is received via a pipeline pressure let-down station from the Kårstø NG Plant with a pressure of approx. 180 bara. The pressure is controlled at plant inlet to 111 bara. A Feed Gas Fiscal Metering Station 15-XT-101 including a filtration device for the removal of particles is installed. 3.1.2

Natural Gas Pretreatment

CO2 Wash Unit

For CO2 removal from natural gas with the present conditions a chemical wash is the most favourable process. An aqueous amine solution (aMDEA) is utilised as solvent. The CO2 wash unit is a Linde designed unit (contrary to a packaged unit). Material and equipment within the unit are designed and supplied according to Linde Standards and Specifications. The feed gas is first heated in the Feed Gas Heater 20-HA-101 against warm lean solvent and further heated in the Feed Gas Trim Heater 20-HA-103 A/B against warm sweet gas to avoid cold temperatures and to allow for efficient CO2 removal. It enters the Amine Wash Column 20-VE-101 and flows from bottom to top through a random packing. Introduced lean amine flows in the opposite direction extracting the acid gas. The CO2 forms a very weak bond with the alkali. In the top of the column solvent traces are removed by water from the purified gas in some additional trays. The wash water for these trays is recirculated by the Water Circulation Pump 20-PA-101 A/B; a small quantity of water is introduced into the cycle by the Amine Make Up Water Pump 20-PB-102 A/B as fresh water (demin. water) to fulfil the water balance of the amine system. The clean gas exits the wash tower with a CO2 content of max 50 vppm and a temperature of approx. 40°C. It is cooled in heat exchanger 20-HA-103 20-HA-103 A/B against Feed Gas to approx. 25°C and leaves the section at a pressure of approx. 109 bara. The loaded amine solution from 20-VE-101 passes via Amine MP Flash Drum 20-VA-102 through the Solvent Heat Exchanger 20-HB-101, where it is warmed up against regenerated solvent and is further routed to the middle section of the Amine Strip Column 20-VE-102. 20-VE-102. In 20VE-102 the reflux water flows from the top through two packed beds. The CO2 is stripped in hot oil heated Amine Strip Column Reboiler 20-HA-102. The regenerated solvent leaves the column at the bottom via heat exchanger 20-HB-101 and is pumped by the Lean Solvent Pump 20-PA-103 A/B to the top of the Amine Wash Column 20-VE-101 via the Feed Gas Heater 20HA-101 and Lean Solvent Cooler 20-HC-101. Approximately 15 % of the flow is routed through the Cartridge Filter 20-LF-101 to remove particles and then through the Activated Carbon Filter 20-LF-102 for removal of heavy hydrocarbons hydrocarbons to prevent foaming.


25.08.2008 Page 14

The acid gas leaves the top of column 20-VE-102 after having passed the water wash section, installed for reduction of amine vapour in the acid gas fraction. After cooling in the Amine Strip Column Condenser 20-HC-103 the gas is separated from the condensate in the Amine KO Drum 20-VA-101. The Amine Strip Column Reflux Pump 20-PA-102 A/B delivers the condensate back to the top of column 20-VE-102. The Amine KO Drum 20-VA-101 also allows for removal of heavy hydrocarbons. The sour gas is routed to the Regenerative Thermal Oxidation 20-XT-101 and then sent to atmosphere. The Solvent Storage Drum 20-VS-101 is designed to hold the complete liquid inventory of the plant. In case of foaming anti foam agent can be injected into the solvent from the Anti Foam Package 20-XU-101. Dryer Station

The sweet, water oversaturated feed gas from the wash unit is fed to the Feed Gas Water KO Drum 20-VL-111 to remove any free liquid upstream of the driers. The liquid from this vessel is routed back to the Amine MP Flash Drum 20-VA-102 to reduce the water make-up of the CO2 wash unit. The drier station is a two-bed molecular sieve adsorber station with a cycle time of 12 hrs. The natural gas is flowing through one of the Feed Gas Driers 20-VK-111 A/B. The water contained in the natural gas is reduced to a level near to zero where no freezing can occur in the downstream liquefaction section. To reduce the temperature fluctuation of the dry gas, a parallel step of 30 minutes is included, where both drier vessels are on adsorption. The dry feed gas passes the Dry Gas Filter 20-LF-111 to remove mole sieve dust which could affect the performance of the downstream cryogenic process section. During this period the other feed gas drier is heated approx. 9 hrs and then cooled approx. 2 hrs by the regeneration gas stream. Dry feed gas at approx. 106 bara serves as regeneration gas. Heating of the regeneration gas to 210°C is provided in the Regeneration Gas Heater 20HA-111 against hot oil and cooling against ambient air in the Regeneration Gas Cooler 20-HC111, followed by the Regeneration Gas Water KO Drum 20-VL-112 where the water is separated and routed to 20-VE-102. The water saturated regeneration gas is compressed by Regeneration Gas Blower 20-KF-111 and routed back into the feed line upstream of the Feed Gas Driers 20-VK-111 A/B. 3.1.3

NG Liquefaction

After CO2 and water removal the natural gas is routed to the cold part of the process, which consists of three spiral-wound heat exchanger bundles integrated in one shell. Liquefaction and subcooling of the feed gas at high pressure is possible because of absence of heavy hydrocarbon components in the design feed gas. The natural gas from the filter 20-LF-111 is first cooled down to approx. -26°C in the Feed Gas Precooler 25-HX-101. It is then further cooled down in the Feed Gas Liquefier 25-HX-102 and throttled to a subcritical pressure of approx. 20 bara to get pure liquid. Finally the natural gas is subcooled in the Feed Gas Subcooler 25-HX-103 to a temperature of approx. -159°C which is low enough to meet the flow limit of 2000 Sm³/h tank return gas allowed for reinjection into the tailgas pipeline.

3.2

Refrigerant System

The cooling duty required to produce the LNG is provided by a simple but efficient closed mixed refrigerant cycle which consists of nitrogen, ethylene, propane, butane, pentane and a portion of the compressed tank return gas (Linde patent).


25.08.2008 Page 15

A motor driven geared centrifugal compressor is applied to compress the refrigerant. 3.2.1

Refrigerant Cycle

The refrigerant is withdrawn from the shell side of the precooler 25-HX-101 at a temperature of approx. 20°C and a pressure of approx. 4 bara, i.e. approx. 10°C overheated against saturated conditions. The refrigerant passes the Cycle Compressor Suction Drum 25-VL-101 and is then compressed in the first stage of Cycle Compressor 25-KA-101. After cooling to approx. 25°C and partly condensing against air in the Cycle Compressor Intercooler 25-HC-101 the liquid and gas are separated in the Cycle Compressor Interstage Drum 25-VL-102. The gas is further compressed in the 2nd stage of 25-KA-101 and partly condensed in Cycle Compressor Aftercooler 25-HC-102 at a temperature of approx. 25°C. Liquid formed in 25-HC-102 is separated in the Cycle HP Separator 25-VA-101. The liquid from 25-VA-101 is sent to 25-VL-102 which also serves as a buffer for the heavy components of the MRC. The liquid hydrocarbon stream is routed to 25-HX-101 where it is subcooled to approx. –26°C and then, after being expanded in a Joule-Thomson valve, used for the precooling of the natural gas. The cycle gas from the separator 25-VA-101 is cooled in the precooler 25-HX-101 to the same temperature, partly condensed and fed to the Cold MRC Separator 25-VA-102. The liquid from this separator is subcooled in the liquefier 25-HX-102 to a temperature of approx. –114°C and used as refrigerant for 25-HX-102 after expansion in a Joule-Thompson valve. The vapour from this separator is condensed in 25-HX-102 and subcooled in the subcooler 25-HX-103 to a temperature of approx. –159°C and provides the cooling duty for the subcooling of the natural gas after expansion in a Joule-Thomson valve to approx. 4.7 bara. After expansion to shell pressure the cycle gas streams are warmed up in the common shell side of the cryogenic spiral wound heat exchangers and returned jointly to the suction side of the 1st stage of the Cycle Compressor 25-KA-101 via the suction drum 25-VL-101. 3.2.2 •

•

•

•

•

Refrigerant Storage and Make-Up The make-up for the refrigerant system is required mainly due to cycle gas losses via the gas seals of 25-KA-101. The quantities required are adjusted according to the composition readings and the temperatures in the cold part and are provided via flow meters as follows: Pure nitrogen is produced in the Backup Nitrogen Package 61-XT-101 and fed to the make-up header by flow control. The methane rich stream is withdrawn from the discharge of the Tank Return Gas Compressor 59-KB-101 and is fed to the make-up header by flow control. For first start-up, when 59-KB-101 is not in service, the gas is withdrawn downstream of the filter 20-LF-111, expanded and routed to the make-up header. Ethylene is stored in the Liquid Ethylene Tank 58-VS-104. The ethylene is vaporised by the Ethylene Make-Up Heater 58-HE-101. Potential traces of water ant methanol are removed in the Ethylene Drier 58-VK-104. To avoid particles in the refrigerant cycle the ethylene is routed via the Ethylene Filter 58-LF-104 to the make-up header.


•

•

3.3

25.08.2008 Page 16

Commercial propane is stored in the Propane Tank 58-VS-101. To assure dry propane, potential traces of water and methanol are removed in Liquid Propane Drier 58-LF-101. To avoid particles in the refrigerant cycle the propane is routed via the Liquid Propane Filter 58-LF-101 to the make-up header. Commercial butane and commercial pentane are stored in the Butane Tank 58-VS103 and in the Pentane Tank 58-VS-102 respectively. To assure dry butane and dry pentane, potential traces of water and methanol are removed in the Liquid Butane/Pentane Drier 58-VK-102. To avoid particles in the refrigerant cycle the butane is routed via Liquid Butane/Pentane Filter 58-LF-102 to the make-up header.

LNG Storage / LNG Loading

Main Purpose of the LNG Storage (Unit 42) and LNG Loading (Unit 47) is the intermediate storage of LNG prior to loading into LNG Carriers at the Jetty and/or to LNG Trucks at the Truck Loading Bay. The LNG Storage Tank is designed as full containment tank and stores LNG near atmospheric pressure. LNG vapour due to end flash, boil off and cooling of loading lines is routed via the LNG storage tank to the Tank Return Gas Compressors. Warm vapour return from ship and truck loading is routed via the LNG storage tank to the tank return gas compressor to protect the compressor while cold vapour return is sent directly to the compressor. Excess vapours mainly during loading of ships with increased tank temperatures at start of LNG Loading are sent to flare. 3.3.1

LNG Storage

Main Purpose of the LNG Storage (Unit 42) and LNG Loading (Unit 47) is the intermediate storage of LNG prior to loading into LNG Carriers at the Jetty and/or to LNG Trucks at the Truck Loading Bay. The LNG Storage Tank is designed as full containment tank and stores LNG near atmospheric pressure. LNG vapour due to endflash, boil off and cooling of loading lines is routed via the LNG storage tank to the Tank Return Gas Compressors. Warm vapour return from ship and truck loading is routed via the LNG storage tank to the tank return gas compressor to protect the compressor while cold vapour return is sent directly to the compressor. Excess vapours mainly during loading of ships with increased tank temperatures at start of LNG Loading are sent to flare. 3.3.2

LNG Loading

There are two LNG Loading Stations foreseen: One for LNG Ship Loading at the Jetty and one for LNG Truck Loading at the LNG Truck Loading bay. 100 % of the produced LNG can be exported via LNG Carriers and approx. 10 % of the LNG production rate can be exported via LNG Trucks. LNG Ship Loading and Ship Vapour Return

During LNG Ship Loading the LNG is pumped to the LNG Carriers by means of the LNG Ship Loading Pumps 42-PS-101 A/B, which are installed in the LNG Storage Tank 42-TR-101.


25.08.2008 Page 17

The LNG from the LNG Ship Loading Pump is routed via the LNG Ship Loading Line and the LNG Ship Loading Arm 47-MU-101 to the manifold of the LNG Carrier at the Jetty. The normal loading rate of the LNG Ship Loading Pump is 1000 m³/h. The flow rate is controlled by the variable speed of the electric motor. Vapour Return from the LNG Ship will be received at a pressure of approx. 1.1 bara at the presentation flange of the ship's manifold and is routed via the LNG Ship Vapour Arm 47-MV101 and the LNG Vapour Return Line to the LNG Storage Tank 42-TR-101 or to the Tank Return Gas Compressor 59-KB-101 depending on the temperature. Warm Vapour Return is cooled to tank operating temperature by injecting LNG into the Vapour Return Line. During no ship loading operation, the LNG Ship Loading Line is kept cold by continuously circulating LNG by means of one LNG Truck Loading Pump 42-PS-102 A/B via the LNG Recirculation Line and the LNG Loading Line back to the LNG Storage Tank 42-TR-101. This is done to keep the loading system cold and gas free at all times, to allow immediate start up of ship loading after arrival of a LNG Carrier. LNG Truck Loading and Truck Vapour Return

During LNG Truck Loading the LNG is pumped to the LNG Truck by means of the LNG Truck Loading Pumps 42-PS-102 A/B, which are installed in the LNG Storage Tank 42-TR-101. The LNG from the LNG Truck Loading Pumps is routed via the LNG Truck Loading Line and the LNG Truck Loading Hose 47-MU-102 to the LNG Truck at the LNG Truck Loading Bay. During loading of LNG Trucks (normal loading rate per pump: 65 m³/h) both LNG Truck Loading Pumps can be used. Vapour Return from the LNG Trucks will be received at the connection point of the Truck Vapour Return Hose 47-MV-102 and is routed via the Vapour Return Hose and the Vapour Return Line to the LNG Storage Tank 42-TR-101 or to the Tank Return Gas Compressor 59KB-101 depending on the temperature. Warm Vapour Return is cooled to tank operating temperature by injecting LNG into the Vapour Return Line. During no truck loading operation, the LNG Truck Loading Line is kept cold by continuously circulating LNG by means of one LNG Truck Loading Pump 42-PS-102 A/B via the LNG Truck Loading Line and the LNG Recirculation Line back to the LNG Storage Tank 42-TR-101. This is done to keep the loading system cold and gas free at all times, to allow immediate start up of Truck loading after arrival of a Truck.

3.4

Fuel Gas System

LNG vapour due to endflash, heat input, cooling of loading lines, ship loading and truck loading is compressed in the Tank Return Gas Compressor 59-KB-101. Part of the tank return gas is routed to the fired Hot Oil Heater as fuel gas. Approx. 2000 Sm³/h is sent to local grid as Sales Gas. For initial start-up and for backup purpose gas from the grid can be used as fuel gas.

3.5

Hot Oil System

The hot oil system supplies the process heat for the plant at two temperature levels. Two cycles are provided, a medium temperature cycle for regeneration of the amine and a high temperature cycle for the heating of the regeneration gas.


25.08.2008 Page 18

The heat for both cycles is provided by the Fired Hot Oil Heater 52-FA-101, a direct fired heater supplied by fuel gas. In this heater the hot oil is heated to approx. 260°C to supply heat for the Regeneration Gas Heater 20-HA-111. This high temperature cycle is pressurized by the Hot Oil Cycle Pump I 52-PA-101 A/B. The required heat for the regeneration of the amine in the Amine Strip Column Reboiler 20-HA-102 is withdrawn from the high temperature cycle downstream of the Hot Oil Cycle Pump I 52-PA-101 and mixed with the cold hot oil downstream of Hot Oil Cycle Pump II 52-PA-102 A/B to limit the maximum temperature to 190°C to avoid degradation of the amine solvent. The hot oil leaves 20-HA-102 with a temperature of approx. 145°C and approx. A small fraction of the flow is pressurized by the Hot Oil Cycle Pump II 52-PA-102 A/B. Most of the hot oil leaving the Amine Strip Column Reboiler 20-HA-102 enters the first hot oil cycle via the balancing line. The balancing line between the two cycles is also used to provide sufficient suction pressure for the two pumps via the Hot Oil Expansion Drum 52-VL-101. The Hot Oil Surge Drum 52-VS-101 is provided to store the total inventory of the system in case of filling or maintenance, and a small Hot Oil Filling Pump 52-PA-103 serves to ease filling of the system. Blanketing for the Hot Oil Unit will be done with pure nitrogen.

3.6

Flare System

The Plant is equipped with two flare headers: •

warm gas flare header which ties in directly at the Flare Stack 54-FC-101

•

cold gas and liquid flare header including the Blow Down Vessel 54-VD-101 for separation of cold liquid and vapour. The vapour is routed to the bottom of 54-FC-101. The liquid is vaporised in the uninsulated Blow Down Vessel 54-VD-101 by ambient heat. In case a warm liquid remains, this liquid can be discharged manually to a barrel.

In addition the low pressure gas from tank and ship loading is routed to the top of the Flare Stack 54-FC-101.

3.7

ESD and Blowdown System

The Emergency Shutdown, Isolation and Depressuring System is used to prevent escalation and to minimise leakage of flammable fluids in case of major plant malfunctions, emergency conditions or damage. The main purpose is to minimise damage by hazards such as fires, unconfined vapour cloud explosions (UVCE) or a boiling liquid expanding vapour explosion (BLEVE) due to bursting vessels. Those hazards may follow on excessive leakage of flammable fluids. After a leakage or fire is detected and localised by the fire and gas alarm system and indicated in the central control room, the Emergency Shutdown, Isolation and Depressuring System will be activated via push-buttons by the operator from CCR. After activation, the plant will be blocked in automatically by means of remote-actuated valves (e.g. Emergency Shutdown Valves - ESV) and selected rotating equipment (eg cycle compressor) will be shut-down. Subsequently the Emergency blow-down System can be activated by the operator. The Emergency blow-down System is depressurising the whole plant (exclusive of LNG-Tank) to the flare system by remote actuated Blow-down Valves (BDV). The system can be operated from a separate control panel (ESD panel) in the central control room (CCR) and allows remote actions from safe location in case of emergency.


25.08.2008 Page 19

Stop Feed, Energy Input and Export Streams

For the LNG process plant all feed streams and energy inputs into the depressurizing areas can be shut off. Units transferring energy to a safe place are kept in operation for continuous energy removal. All export streams (e.g. tail gas) will be shut-off. Depressurizing Philosophy

According to contract and EN 1473, the isolated sections shall be depressurised to •

50 % of design pressure in 15 minutes or to

•

7 barg in 30 minutes

The higher flow is counting. Units and Equipment without Depressurizing Facilities

Basis for selection of depressurizing sections is the maximum operating or settle out pressure and not the mechanical design pressure, which is for other reasons sometimes well above the maximum operating pressure (compare API RP 521). The following units and equipment have no depressurizing facilities: •

•

•

•

MDEA regeneration; operates at low pressure (appr.1 barg) Feed gas Liquefaction passage in 25-HZ-101 (mass of each passages is below 1000 kg limit, the passage is well protected in the shell, the consequence is considerably low) LNG storage; operates at low pressure (appr. 250 mbarg) LNG Ship, LNG piping and LNG Truck Loading system (subcooled liquid at low pressure)

Basis input to QRA

As basis for the QRA a reaction time from first fire&gas alarm until the operator initiates the ESD and blowdown system is assumed to be 600 seconds. As an average value a depressurisation time of 900 seconds shall be used in the QRA (refer to Assumption Sheet RA-4 in Appendix A).

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Study Methodology

4.0

25.08.2008 Page 20

Study Methodology

Risk Analysis Basics

4.1

Risks are commonly incurred and accepted in everyday life. There are many different types of risk including risk to life and health, risk to the environment and economic risks, which may impair the survival of a company. The risk R is commonly described by its two dimensions, i.e. the consequence of an accidental event C and the frequency of this event (F): R=CxF

The actual risk values can be manifold due to the different types of consequences, which might arise from an accident. It could be a financial loss due to downtime and damage in terms of money per event, a certain number of fatalities or certain damage to the environment, which may also lead to a certain financial loss due to the cost resulting from decontamination etc. The economic loss is very often influenced by the fact that certain accidents will lead to damages in the neighbouring parts of the plant. The frequency of an event usually is a composite magnitude, e.g. for an ignited gas leak the primary leak frequency will be multiplied by the conditional probability of igniting the gas cloud resulting from the leak. Under certain conditions, even more conditional probabilities may factor into this product to yield the total frequency of a certain event, e.g. the probability of in-time detection of a flammable cloud or the conditional probability, that certain isolation measures (e.g. ESD and Blowdown System) work, when required.

4.2

Definition and Types of Risk

It has become common in the process industries to quantify risk to people in terms of •

1st party risk, i.e. the risk to onsite personal

•

2nd party risk, i.e. the risk to external contractors

•

3rd party risk, i.e. the risk, to which the site external population is exposed.

Further to this one differentiates individual risk , i.e. the risk, to which a single person is exposed, and societal or group risk , i.e. the risk to which a certain group of people are exposed. Details are given in Table 3. Table 3: Types of Risk

Type of Risk st 1 party individual specific risk

Details Risk to onsite personnel, based on the most exposed person at risk, i.e. operators.

1st party average individual risk

Risk based on the individual specific risk and is calculated as average risk to onsite personnel.

2nd party individual specific risk

Risk to external contractors, based on the most exposed person at risk, i.e. LNG carrier, external maintenance personnel.

nd

2

party average individual risk

Risk based on the individual specific risk and is calculated as average risk to external contractors.


Type of Risk rd 3 party individual specific risk

rd

Societal (3 party) risk

4.3

25.08.2008 Page 21

Details Risk to offsite population expressed as the fatality risk per year. Individual risk is calculated under the assumption that the exposed person is present unprotected at the same location for 24 hours per day over 365 days per year. In case of Individual Specific Risk the actual duration of the presence is taken into account. Risk to a group of people outside of the plant. Societal risk usually is quantified in form of the so-called FN curve, specifying the frequency F (per year), that N or more persons are affected by lethal consequences.

Acceptance Criteria

The risk in this QRA study is discussed in terms of individual risk and societal risk. The Individual Specific Risk for 1 st, 2nd and 3rd party has been defined by Lyse. The 3 rd party risk is also calculated as FN Curve and compared with the societal risk acceptance criteria based on UK HSE Societal Risk Criteria. The acceptance criteria defines for the following personnel categories: •

1st party, i.e. personnel working for the Lyse LNG Base Load Plant facility.

•

2nd party, i.e. LNG Carrier personnel (Truck, Ship Loading and external contractors) can be affected by operation activities.

•

3rd person, i.e. offsite population.

Note: occupational accidents have been not included in the acceptance criteria and therefore are not considered in the QRA. 1st and 2nd party Individual specific risk (ISR) is specified as ISR = Σ (Effective Frequency x Occupancy x Vulnerability),

where "Occupancy" is a factor which relates the time for which a person is exposed to work hazards (in hours) to the total number of hours within a year (8760). For sake of simplicity we assumed, that a typical operator works in 8 hour shifts for 5 of 7 days per week, i.e. his annual working hours are 45 weeks x 5 days x 8 hours = 1800 hours per year. He is 20% of his working time outside. The effective frequency is calculated 0.2 x outdoor frequency + (1-0.2) x indoor frequency. Hence the occupancy factor is 1800 / 8760 = 0.20. For the definition of vulnerability please refer the Appendix F. The acceptance criterion for Individual Specific Risk (ISR) for the most exposed person for 1 st and 2nd party is expressed as the yearly probability for loss of life. The ISR is acceptable for < 1 X 10-6 per year, the risk level above 1 X 10 -3 per year becomes unacceptable. The region in between is the ALARP area. The Average Individual Risk (AVR) is specified as follows: AVR = Σ (ISR x Number of personnel) / Σ Number of personnel

The AVR shall not exceed 1 X 10 -4 per year, the risk level under 1 X 10 -6 per year is acceptable. The region in between is the ALARP area.


25.08.2008 Page 22

If no individual specific risk (ISR) is found to be above 1 X 10 -4 per year, the AVR criteria is fulfilled. 3rd party (Societal Risk) Individual specific risk (ISR) is specified as ISR = Σ (Effective Frequency x Occupancy x Vulnerability),

where "Occupancy" is a factor which relates the time for which a person is exposed to hazards (in hours) to the total number of hours within a year (8760). For sake of simplicity we assumed, that a person on the peninsula stays for 4 hours 2 of the 7 days per week in the summer (4 month) and 2 hours at 2 days per week in the winter, i.e. his annual presence hours are (16 weeks x 2 days x 4 hours) + (32 X 2 days X 2 hours) = 256 hours per year. He is staying 100% outside. The effective frequency is calculated with a location fraction of 1 outdoor frequency. Hence the occupancy factor is 256 / 8760 = 0.03. The acceptance criterion for Individual Specific Risk (ISR) for the most exposed person for 3 rd party is also expressed as the yearly probability for loss of life. The ISR is acceptable for < 1 X 10-7 per year, the risk level above 1 X 10 -5 per year becomes unacceptable. The region in between is the ALARP area. The Average Individual Risk (AVR) is specified as follows: AVR = Σ (ISR x Number of people) / Σ Number of people

The AVR shall not exceed 1 X 10 -5 per year, the risk level under 1 X 10 -7 per year is acceptable. The region in between is the ALARP area.

Societal risk for 3rd party is presented as the probability or frequency of accidents of different extent. The Figure 5 below states the acceptable and not-acceptable range of the yearly frequency (F) – consequence (number of fatalities N or larger) – diagram and shows the acceptance criterion based on UK HSE Societal Risk Criteria. It also indicates an area where the company shall actively seek to reduce the risk based on the ALARP principle.

Figure 5: UK HSE Societal Risk Criteria


4.4

25.08.2008 Page 23

Hazard Identification

The study has been based on identified major inventories of flammable and explosive materials in the LNG Base Load Plant units, together with major lines connecting the inventories. Information on inventories, stream compositions, operating conditions and locations has been based on the available drawings and further information. In addition the results of the hazard identification of Hazard Study (HAZID) (Appendix B) were used. The investigations were verified on the basis of operating procedures, P&IDs and the knowledge provided by LINDE. In the HAZID, only those hazards are identified, which might lead to a leakage of hydrocarbons and a subsequent fire or explosion. Other hazards with operational consequences have been discussed in the normal HAZOP study. The basic results of the HAZID are shown in Table 4 . Table 4: HAZID Summary

Hazard Hydrocarbon (gas / liquid or two phase) leaks outdoors

Hydrocarbon (HC gas / liquid or two phase) leaks in buildings

Non-hydrocarbon fire Non hydrocarbon chemical leak or fire

Loss of power

Treatment in QRA Included in QRA in four event classes of very large, large, medium and small leak at various locations in the individual areas. This hazard covers the majority of flammable leakage scenarios. Not included in the QRA Buildings containing HC: - The buildings are specified with explosion group zone 1; therefore the risk of internal explosion is reduced. - The protective effect of the building is not considered in the SAFETI calculation (conservative consideration). Buildings containing no HC: - Gas entering in a building is presented by adequate gas detection and closing the air-intake. Not included in the QRA as of minor importance. Involved chemicals (e.g. MDEA etc.) have a minor contribution to risk due to quantities; hence they are not of relevance in this QRA. Not included in the QRA since failure leads to fail safe conditions.

Loss of instrument air

Not included in the QRA since failure leads to fail safe conditions.

Loss of safety systems


Loss of control system


Occupational accidents

Not included in the QRA as this is identical to general petrochemical facilities and known to be marginal

Natural environmental impact (extreme weather, earthquake, etc) Pipeline rupture

Not included in the QRA due to low risk contribution. Included in the QRA.

Pipeline exposed/free span

Included as a potential cause for leaks.

Pipeline dented Excessive pipeline expansion

Included as a potential cause for leaks. Included as a potential cause for leaks.


Hazard Reduced pipeline thickness

Treatment in QRA Included as a potential cause for leaks.

25.08.2008 Page 24

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Data used for the Risk Assessment

5.0

25.08.2008 Page 25

Data used for the Risk Assessment

This section informs about the basic data and detailed assumptions which were used for the calculations and the individual steps taken to arrive at the risk picture.

5.1

Scenarios

For the purpose of this QRA the plant was analysed with respect to its hydrocarbon content. Units without relevant hydrocarbon content were excluded from the further analysis. These scenarios consider releases of hydrocarbons from small, medium, large or very large leaks in pipe work or equipments. This leaves the following units for further consideration as shown in Table 5 and Figure 6: Table 5: Units covered in this QRA

Unit

Inventory Loop No. used in fig. 6

Designation

20

IL1

Feedgas Purification

20

IL2A

NG Liquefication Gas

25

IL2B1

NG Liquefication Liquid_103 bar System

25

IL2B2

NG Liquefication Liquid_19 bar System

59

IL3A

LNG Storage Return Gas

42

IL3B

LNG Storage

47

IL4

LNG Truck Loading

47

IL5A

LNG Ship Loading Tank Top

47

IL5B1/2/3

LNG Ship Loading Line

47

IL5C

LNG Ship Loading Jetty

25

IL6A1

Refrigeration Gas System_4 bar System

25

IL6A2


25

IL6A3


25

IL6B1

Refrigeration Liquid 25-HX-101/103 System

25

IL6B2

Refrigeration Liquid 25-VA-101 System

25

IL6B3

Refrigeration Liquid 25-VA-102/25-HX-102 System

25

IL6B4

Refrigeration Liquid 25-VL-102 System

58

IL7

Propane Storage

58

IL8

Pentane Storage

58

IL9

Butane Storage

58

IL10A

Ethylene Storage Gas System

58

IL10B

Ethylene Storage Liquid System

20/52

IL11

Hot Oil System

15

IL12

Feedgas Fiscal Metering

59

IL13

Tailgas Fiscal Metering


25.08.2008 Page 26

The solvent regeneration system has not been taken into account in this QRA due to its comparably small inventories. A leak of MDEA from process equipment or piping leads to a release of CO2 loaded MDEA to dike area and pit, which does not impose a relevant hazard to people. The units can be isolated by ESD and Blowdown system or are directly connected to another area, which can be isolated.

Figure 6: Process Areas defined for this QRA (numbers see Table 5)

For these areas an equipment count was performed (refer to Assumption Sheet FA-1 in Appendix A) and considering: Equipment (vessels, pumps, heat exchangers, compressors etc.) Valves (actuated and non-actuated) Pipelines Small bore fittings, Flange connections (partly, based on the Dutch Purple Book [5])

• • • •

All equipment has been listed with their respective operating characteristics. These data have been used to calculate the overall leak rates for the individual areas. Details are contained Appendix C.

5.2

Leak Frequencies

The leak frequency modelling is based on DNV’s leak frequency database LEAK 3.2 and Purple Book. The leak types and sizes are shown in Table 6:


25.08.2008 Page 27

Table 6: Leak types and sizes

Leak Type Small Medium Large Very Large (Full Bore Rupture)

Leak range [mm] 1 – 10 10-50 50-100 > 100

Leaks with equivalent diameter below 1 mm are not considered as they do not contribute substantially to the overall risk.

5.3

Release Duration

The duration of a release is closely linked to the type of detection and isolation. Table 7 lists typical times involved for various alternatives: Table 7: Typical Duration Times based on DNV database

Description

Duration for Detection and Isolation [s] Gas detector which auto closes ESD/automatic valve (XSFV). 120 Gas detector with isolation by manual valve closure. 960 Gas detector with isolation by remotely operated closure of control valve. 660 Detection by operator and initiation of ESD & Blowdown System 600 Gas detector with isolation by remotely operated closure of ESD. 360 Process trip which auto closes ESD. 360 Process alarm with isolation by manual valve closure. 1200 Process alarm with isolation by remotely operated closure of control valve. 900 Process alarm with isolation of feed by remotely operated closure of control max. 1800 valve. Duration determined by either inventory of material (max 1800s) or valve closure time (900s). Process alarm with isolation of feed by remotely operated closure of ESD. max. 1800 Duration determined by either inventory of material (max 1800s) or valve closure time (600s). Process alarm with isolation by remotely operated closure of ESD. 600 Detection by field operator, remote area, with manual isolation. 2700 Detection by field operator, remote area, with isolation by remotely operated 2400 control valve. Detection by field operator, remote area,, with isolation by remotely operated 2100 ESD. Detection by field operator routine patrol, with manual isolation. 1500 Detection by field operator routine patrol, with isolation by remotely operated 1200 control valve. Detection by field operator routine patrol, with isolation by remotely operated 1200 control valve. Duration determined by either inventory of material (max 1800s) or valve closure time. Detection by field operator on routine patrol with isolation of feed by remotely 900 operated closure of ESD. Duration determined by either inventory of material (max 1800s) or valve closure time. Detection by field operator on routine patrol, with isolation by remotely operated 900 ESD.

The Lyse LNG Base Load Plant is equipped with a fire and gas detection system and remotely operated ESD valves, control valves, compressor and pumps. The reaction time is 600 s for


25.08.2008 Page 28

detection and initiation of ESD & Blowdown System by the operator, e.g. shut-off of main feed and product streams via ESD valves and tripping of main machines. An average blowdown time of 900 s is used in the calculation (refer to Assumption Sheet HC-2 and RA-4 in Appendix A).

5.4

Atmospheric Conditions

Weather data have been taken from the site conditions document [6]. For the wind rose data for Sola, refer to Assumption Sheet MI-2 in Appendix A. Table 8 summarises the results, where an angle of 0 degrees presents a wind originating from the North. Table 8: Weather data for Lyse LNG Base Load Plant

Stability Class Wind [m/s]

F - 1.5 D- 6 D - 12 Sum

Percentage Wind direction [degrees] 292.5337.5 22.567.5337.5 -22.5 67.5 112.5 1.99 0.961 1.012 1.633 14.71 7.09 7.47 12.04 2.79 1.346 1.417 2.293 19.49 9.397 9.899 15.966

112.5157.5 1.335 9.89 1.878 13.103

157.5202.5 0.501 3.69 0.702 4.893

202.5247.5 0.807 5.96 1.13 7.897

247.5292.5 1.977 14.57 2.76 19.307

Wind speed classes have been used ranging from 1.5 m/s to 12 m/s, whereas for atmospheric stability Pasquill classes ranging from D (neutral) to F (stable) have been selected. The atmospheric stability is considered to be neutral during the day and stable during the night. For the calculations 8 wind directions have been used.

5.5

Population Distribution

For the Lyse LNG Base Load Plant facility, a work day is divided into three shifts; a day shift, an afternoon shift and a night shift, each lasting 8 hours (Assumption Sheet MI-3 in Appendix A). The relevant figures listed in Table 9 and Table 10:

st

nd

Table 9: Onsite Population (1 and 2 party)

Buildings / Areas

Personnel / People Day (per Shift) Night Total Number 3 1 7 2 1 5 4 2 10 1 1 3 2 2 6 8 8 24

Administration Building Maintenance Truck Loading Ship Loading (Jetty) Ship Deck Ship Bridge

The personnel in the administration building do the daily operation and supervision of the plant. rd

Table 10: Off-site Population (3 party)

Areas

Peninsula Hiking Track Ferry Terminal_office workers

Personnel / People Day Night 16 (in non-work day) 0 8 (in a non-work day) 0 100 2


Areas

Ferry Terminal_industry workers Ferry Terminal_passengers Energiveien+Risavika_office workers Energiveien+Risavika_industry workers Container Area_office workers Container Area_industry workers Rest Companys_office workers Rest Companys_industry workers Living Quarters

5.6

25.08.2008 Page 29

Personnel / People Day Night 10 0 1500 0 400 5 559 0 10 1 50 0 1139 10 715 0 60 60

Ignition Sources

Release of flammable fluid may have many event outcomes, depending on the timing and type of ignition. For example, a release may ignite immediately at the point of release, or it may ignite after the cloud has been dispersing for two minutes, or after the cloud has been dispersing for five minutes, or it may not ignite at all. If it ignites, it may give either explosion effects or different types of fire effects depending on the type of release (e.g. jet fire, fireball, pool fire or flash fire). Each of the outcomes will have different risk effects because each produces an effect zone of a different size and intensity, at a different location. The risk effects for a flammable release will depend on the timing, location and nature of ignition. For example, if an instantaneous release ignites immediately it will produce a hazard zone at the point of release, whereas if it ignites after the cloud has started to disperse, it will produce a hazard zone at the point of ignition. If the ignition produces a fireball, the intensity of the effects within the zone will be different from those for an ignition which produces a flash fire, or for an ignition which produces an explosion. The different outcomes are presented in the form of event trees (Assumption Sheet RA-1 in Appendix A). Each outcome in an event tree can be assigned a probability, and the program performs the risk calculations for all of the event tree outcomes that are relevant to a particular flammable model. The ignition probability within PHAST RISK is definable according to the respective site knowledge. The immediate ignition probability is directly specified. A default value of 0.3 is used, which would only apply to very large flammable gas releases in a large industrial complex. The delayed ignition probability for any failure case is a calculated value within PHAST RISK, which is based on the defined ignition sources on site, with a unique value for each release case and release direction. The calculation is based on the strength, location and presence factor of all ignition sources specified, and the size and duration of the dispersing flammable vapour cloud. PHAST RISK assumes "diffuse ignition background" (which could be understood as e.g. traffic illumination, cameras etc.), i.e. ignition may occur even if no specific ignition sources are given. Plant specific ignition sources, which have been taken into account are listed in Table 11 and their ignition probability have been discussed in Assumption Sheet RA-2 Appendix A.


25.08.2008 Page 30

Table 11: Ignition Sources in Lyse LNG Base Load Plant

Ignition Source Flare 54-FC-101 Fired Heater for Hot Oil 52-FA-101 Regenerative Thermal Oxidation (RTO) (Incinerator) 20-XT-101 Electrical Substation Traffic (Truck Loading) Parking External Population

To model the conditional probabilities for the ignition resulting into different types of fires and explosion, an event tree of the type shown in Figure 7 has been used:

Figure 7: Event tree used for fire and explosion modelling

For the probabilities in this event tree, standard setting as used normally in PHAST RISK have been applied (most values taken from the Dutch Purple Book [5]).

5.7

Consequence Calculations

The analysis of potential consequences following loss of containment is carried out as the first stage of the risk analysis. Consequence analysis involves the estimation of rates of release in the event of loss of containment and prediction of the potential consequences. 5.7.1

Discharge and Dispersion

Material can be released to the atmosphere because of a failure in the containment system. The magnitude of a release depends primarily on the size of the leak in the system, the phase of the material and the operating pressure. For modelling purposes, releases are usually categorized as either instantaneous or continuous. As the analysis is concerned with major acci-


25.08.2008 Page 31

dent hazards, only releases from equipment and releases from holes giving an excess of a release rate of 0.1 kg/s have been included. 5.7.2

Instantaneous Releases

If a catastrophic failure of the shell of a vessel occurs the contents would be released very quickly (instantaneously). This type of failure has been modelled as a hemispherical cloud centred on the release location. 5.7.3

Continuous Releases

Releases of liquids and gases from pipes or equipment items were estimated using basic release rate calculations assuming a fixed value of discharge coefficient. The value of discharge coefficient used (0.65) is taken from a range of values (typically 0.5 to 0.8) which represent various pipe- and equipment configurations. The release rate calculations were performed using PHAST 6.53.1. The calculated release rate was assumed to be constant throughout the release duration. 5.7.4

Release Duration

The release duration depends on the upstream inventory and the means for detection and subsequent isolation of the release. The release duration has been assumed (refer to Assumption Sheet RA-4 in Appendix A) to be limited by the upstream inventory up to a maximum duration of 1500 s (600 s detection time and 900 s automatic or remotely activated ESD and Blowdown closure time) for small and medium sized leaks. For large and very large leaks an isolation time of 600 s has been used. By all size of leaks, the rest flow of fluids from the upstream system, which will be released before the isolation valves closes (in 600 s), is also considered in the PHAST calculation. 5.7.5

Dispersion

When a vapour cloud is generated, either instantaneously or continuously, there may be a substantial degree of mixing of air with the released material. Dispersion was modelled using PHAST version 6.53.1. To allow for destruction of momentum due to impingement of releases or upwind and downwind releases, 50% of releases were modelled as free-field horizontal releases and 50% were modelled as ‘impinged’ releases. The dimensions of impinged releases were determined assuming that the clouds were cylindrical in shape, but with the same volume as a horizontal release. 5.7.6

Thermal Radiation and Overpressure

On ignition of a flammable cloud, different types of combustion can occur depending on the particular circumstances. It is normal to characterize the combustion in various ways and for the purpose of this analysis, flash fires, jet fires, pool fires, fireballs and explosions have been considered. In the event of a flammable release from containment which is not ignited immediately, a hydrocarbon vapour/air mixture is formed. The concentration of hydrocarbon in the cloud, as progressive dilution with air takes place, is estimated using the dispersion model. The direction and extent of drift of the cloud is influenced by the prevailing weather conditions. The cloud


25.08.2008 Page 32

remains capable of ignition providing the concentration is above the lower flammable limit (LFL). On ignition, a flame front passes at slow speed throughout the flammable cloud and a flame stabilizes near to the point of release as either a jet or pool fire. A flash fire does not produce high levels of overpressure outside the cloud, but inside the cloud there can be isolated regions of overpressure which could lead to equipment or building damage. Levels of thermal radiation which are potentially fatal, are produced within, and for a very small distance outside the LFL envelope. Jet fires are usually the consequence of a momentum dominated release resulting from an immediately ignited release or from a flash fire that burns back to the point of release. This type of fire has been included in the SAFETI calculations.

Under certain circumstances the flame travelling through a hydrocarbon/air cloud can accelerate and attain a significantly higher flame speed than that associated with a flash fire. This high flame speed also generates an overpressure wave. This phenomenon is referred to as a vapour cloud explosion (VCE). Experimental work and observations on incidents have confirmed that in order for a flame to accelerate from a low speed to a high speed, some form of congestion is necessary, e.g. a gas cloud within a plant area. Flame acceleration does not occur if the cloud is in the open air, e.g. a cloud over open ground, and indeed if a high speed flame exits from a congested region into an open region, flame deceleration occurs. Vapour cloud explosions are characterized by the production of levels of overpressure which can cause damage to equipment and destruction of buildings well beyond the flammable cloud boundary. Although any person within the flammable cloud is likely to be fatally injured, direct human fatalities from blast outside the flammable cloud are unlikely. Most casualties beyond the cloud envelope arise indirectly, i.e. from crush injuries in collapsed buildings or injuries from fragments. PHAST RISK uses a modified version of the TNT equivalent model to describe the consequences of VCE. This model considers a typical congestion. As there is unconfined space between the process area and the administration building, the results for explosion overpressure towards the administration building and installation outside battery limit can be considered conservative.

5.8

Mitigation Measures taken into Account

The present concept takes into account various mitigation measures, which are presented in the Assumption Sheets as indicated in Table 12:

Table 12: Risk Reducing Measures

No.

Risk Reducing Measure

1

Loading frequency consideration

Assumption Sheet No. HC-9

2

Welded Pipes in Feed Gas and LNG service

FA-1

3

Full containment LNG Storage Tank

FA-3

4

Explosion Protection

RA-2

5

Design of the flare stack

RA-6

6

Fire and Gas Detection

RA-3

7

ESD/Blowdown System

RA-4


25.08.2008 Page 33

No.

Risk Reducing Measure

8

Appropriate measures

Assumption Sheet No. RA-2, RA-6

9

Active and Passive Fire Protection

RA-7

10

Escape Ways

RA-8

11

Safe Haven

RA-8

For further reduction of the risk to ALARP additional risk reduction measures are evaluated by means of sensitivity calculations in Chapter 7.0.


6.0

25.08.2008 Page 34

Results of the Risk Analysis

This section presents the results of the risk calculations using PHAST RISK with the assumption specified in the previous sections. Risk to people is described in terms of individual risk for 1st, 2nd, 3rd party and societal risk 3 rd party.

6.1

Risk 1st and 2nd party

Individual Risk, 1st and 2nd party

The subsequent Figure 8 and Figure 9 show the calculated individual risk contour lines for the Lyse LNG Base Load Plant. The figures 8 and 9 show the calculated individual risk contour lines for the Lyse LNG Base Load Plant. The calculated risk contours of individual risk for the most exposed person is illustrated in Figure 8. The figure shows the contours of the most exposed person to suffer a fatality every 100 000 years (green line), every 1 000 000 years (dark blue line). The risk is illustrated for the most exposed person present in the process plant area, 20 % of their working time per year.


25.08.2008 Page 35

-

10 /yr -6 10 /yr

Figure 8: Most exposed person individual risk contour lines for the Lyse LNG Base Load Plant

The risk contours for the individual risk are also calculated and is illustrated in Figure 9. The figure shows the contours of individual risk for a fatality every 10 000 years (green line), every 100 000 years (dark blue line), etc. The risk is illustrated for 1 person present at any point outside a building in the plant, continuously 8 hours a day, 5 days a week during a whole year (45 weeks).


25.08.2008 Page 36

-4

10 /yr 10-5 /yr -6 10 /yr 10-7 /yr -8 10 /yr

Figure 9: Individual risk contour lines for the Lyse LNG Base Load Plant

The figure shows the contours of individual risk for a fatality every 10 000 years (green line), every 100 000 years (dark blue line) etc. The risk is illustrated for 1 person present at any point outside a building in the plant, continuously 8 hours a day, 5 days a week during a whole year (45 weeks). The risk is higher closer to the process area and above 1 fatality every 10 000 years. The other plant areas, e.g. LNG Tank is between the 1 fatality per 10 000 year and 100 000 year risk contour. The frequency of overpressure at the control room is calculated [refer Chapter 6.4]. The Central Control Room is designed for an explosion load of 200 mbar. The risk contributions (pies) for 1 st and 2nd party are illustrated in Figure 10, Figure 11, Figure 12 and Figure 13 and based on calculated individual risk (IR) and include gas and liquid leaks from all leak sizes of the various inventory loops, which are indicated in the risk ranking report (ref. Appendix E) and reflect the indoor (respectively outdoor) risk that the exposed person is present unprotected at the same location for 24 hours per day over 365 days per year.


25.08.2008 Page 37

IRindoor = = 4.9 X 10-8 /yr Other 1%

LNG Ship Loading Jetty 43%

Refrigerant System 56%

Figure 10: Risk contributors to personnel in administration building

Propane Storage 7%

Other 5%

IRoutdoor = 2.9 X 10-3 /yr

Ethylene Storage 6%

Pentane Storage 7% Refrigeration System 47% Butane Storage 6%

Feedgas Purification 22%

Figure 11: Risk contributors to personnel in process area


Butane Storage 3%

Other 6%

25.08.2008 Page 38


Propane Storage 10%


Refrigeration System 23%

Ethylene Storage 11%

Figure 12: Risk contributors to personnel at jetty

IRoutdoor = 8.2 X 10- /yr

Other 9%

LNG Storage Tank Top 4%

Refrigerant System 11%

Feedgas Purification 46% Ethylene Storage 6%

Propane Storage 6%

Truck Loading 18%

Figure 13: Risk contributors to personnel at truck


25.08.2008 Page 39

The main individual risk to onsite personnel comes from the refrigerant system and feedgas purification. Individual Specific Risk, 1st and 2nd party

The Individual Specific Risk (ISR) for 1st and 2nd party, which considers the individual working hours for each group, is given below in Table 13. st

nd

Table 13: Individual Specific Risk (ISR) for 1 and 2 party

Buildings/Personnel Locations Operator

ISR [year] 1.2 X 10 -4

Process Area (Maintenance)

1.2 X 10 -4

Truck Loading

6.7 X 10

-5

Ship Loading (Jetty)

2.0 X 10

-5

Ship Bridge

5.0 X 10

-5

Ship Deck

4.8 X 10-5

Criterion (ISR)

1 X 10

-3

The most exposed 1 st party person will be an operator. He is presumed to be 20% of his time in the process area and exposed to a potential accident when he is at work, i.e. 1800 hours per year. The “indoor” individual risk in administration building is 4.9 X 10 -8 per year. The “outdoor“ individual risk in process area is 2.9 X 10 -3 per year. The Individual Specific Risk is calculated by the following equation, (0.2 x 2.9 X 10 -3) + (0.8 x 4.9 X 10 -8 ) X 1800/8760 = 1.2 X 10 -4 per year. Hence the 1st party individual specific risks for the most exposed person lower than 1 X 10 -3 per year and within the acceptance criteria. The Average Individual Risk (AVR) of 5.0 X 10-5 per year for all personnel (1 st and 2nd party) is within the ALARP regime.

6.2

Risk 3rd party

Individual Risk, 3rd party

The distribution of the external population (3 rd party) is shown in Figure 14:


Residential Area

25.08.2008 Page 40

Living Quarters

E n e r g i v e i e n + R i s a C o v i n t k a A r a i e a n e r

Rest Companys

Ferry Terminal Peninsula

Hiking Track

Figure 14: Representative external population (Peninsula, Hiking Track, Ferry Terminal, Container Area, Energiveien&Risavika, Rest Companys, Residential Area and Living Quarters)

Figure 15 shows the Lyse LNG Base Load Plant location.

LNG lant location

LNG lant location

Figure 15: The LNG Base Load Plant location


25.08.2008 Page 41

The risk contributions (pies) for 3 rd party are illustrated in Figure 16, Figure 17 and Figure 18 and based on calculated individual risk (IR) and include gas and liquid leaks from all leak sizes of the various inventory loops, which are indicated in the risk ranking report (ref. Appendix E) and reflect the indoor (respectively outdoor) risk that the exposed person is present unprotected at the same location for 24 hours per day over 365 days per year.


Propane Storage 38%

Other 6%


Ship Loading Tank Top 22%

Butane Storage 2% Pentane Storage Ship Loading Jetty 4% Feedgas Purification 2% Ethylene Storage 5% 6%

Figure 16: Risk contributors to people at Peninsula and Hiking Track



25.08.2008 Page 42

Other 9%

Truck Loading 6%


Ship Loading Jetty 4%


Ethylene Storage 9% Ship Loading Tank Top 14% Pentane Storage 6% Propane Storage 16%

Figure 17: Risk contributors to other external population (Ferry Terminal, Container Area, Energiveien&Risavika, Rest Companys and Living Quarters)

The main individual risk to offsite personnel comes from the LNG storage tank top (ship loading tank top), refrigerant system and propane storage.


25.08.2008 Page 43


Ship Loading Tank Top 49%


Figure 18: Risk contributors to the Residential Area

Individual risk for people in residential area is negligible (9.7 X 10 -10 per year). Worst case assessment shows, there is no credible scenario that a flammable gas cloud above LFL (lower flammable limit) can reach the residential area. Individual Specific Risk, 3rd party The Individual Specific Risk (ISR) for the 3 rd party risk is given below in Table 14.

Table 14: Individual Specific Risk (ISR) for the 3rd party

Personnel Locations Peninsula

ISR [year] -8 4.6 X 10

Hiking Track

2.2 X 10

Ferry Terminal_office workers

7.6 X 10 -7

Ferry Terminal_industry workers

3.8 X 10

-7

Ferry Terminal_passengers

4.0 X 10

-7

Energiveien+Risavika_office workers

4.6 X 10 -9

Energiveien+Risavika_industry workers

4.6 X 10 -9

Container Area_office workers

3.2 X 10 -9

Container Area_industry workers

3.2 X 10

-9

Rest Companys_office workers

2.8 X 10

-14

Rest Companys_industry workers

2.8 X 10

-14

-6


Personnel Locations Living Quarters

ISR [year] -10 3.5 X 10

Criterion (ISR)

1 X 10

25.08.2008 Page 44

-5

The Average Individual Risk (AVR) of 1.5 X 10-7 per year for people living, working or staying outside the Lyse LNG base load plant does not exceed the acceptance criteria of 1 X 10 -5 / year and is within the ALARP regime, close to acceptable in general. Societal Risk, 3rd party

The societal risk calculated for the Lyse LNG Base Load Plant is shown in Figure 19.

Figure 19: Societal risk FN curve for t he Lyse LNG Base Load Plant

The minimum and maximum risk criteria are shown in Figure 19 as blue and green lines respectively. Calculations of the external societal risk (e.g. Hiking Track, Peninsula, Industry Area and Ferry Terminal) have shown that this risk for the Lyse LNG Base Load Plant falls into the


25.08.2008 Page 45

area between the upper and lower limit line, i.e. the ALARP regime. For the Lyse LNG Base Load Plant such mitigation measures have already been applied (refer to Chapter 5.8 and 7.0).

6.3

Location Specific Risk

It has been suggested by Skangass, that the location of the 1 X 10 -5 per year contour is a suitable measure to use for the outer extent of a “safety zone” around the site. Certain activities, such as smoking, starting open fires and camping would not be allowed in this region. Such a designation is consistent with the use of the 1 X 10 -5 per year contour by the HSE in the UK to mark the extent of the “inner zone” around a site where future residential developments would be prohibited. Figure 20 shows in relation to the safety zone the 1 X 10-5 per year risk contour.

safety zone contour 10 /yr (individual risk contour) -5

Figure 20: Location of safety zone in relation to the calculated individual risk contour of 1 X 10-5 per year


6.4

25.08.2008 Page 46

Overpressure Risk

To assess the effectiveness of mitigation measures, overpressure risk contours are calculated for the potential overpressure design in the Lyse LNG Base Load Plant. In view of the probabilistic nature of a QRA, also peak overpressure reached at a certain location will occur at this location with a certain frequency (per year), i.e. overpressure is of probabilistic nature itself. Therefore overpressure risk calculations have been performed using PHAST RISK. Peak overpressure is of relevance for the design of buildings, in which personnel will be protected against the consequences of pressure waves. Depending on the design and reinforcement of the building it will provide more or less protection for people. In Reference 7 and 8, overpressure fatality probabilities for various peak overpressures have been determined for various building types (so-called BEAST and CIA types). The selection of the respective pressure to be considered is motivated by the vulnerabilities of populations in various building types as shown in Table 15 for some of the building types:

Table 15: Overpressure vulnerabilities in various building types [7, 8] Overpressure Fatality Probabilities Building Type Beast1 Beast2

Beast3

Beast5 Beast7 Beast10

Beast11

Beast12

CIA3

Description of Building Type

30 - 70

70 - 110

110 - 150

150 - 300

300 - 500

> 500

0.0001

0.01

0.065

0.279

0.488

0.488

0.0001

0.01

0.017

0.221

0.668

0.668

0.0001

0.02

0.282

0.282

0.788

0.788

0.00005

0.02

0.02

0.083

0.988

0.988

0.0001

0.0001

0.017

0.171

0.488

0.488

0.02

0.838

0.838

0.838

0.838

0.838

0.0001

0.025

0.322

0.322

0.988

0.988

0.00005

0.00005

0.02

0.322

0.988

0.988

0.010

0.036

0.081

0.267

0.575

0.740

mbar Steel framed structure with metal panels for roof and wall cladding Steel framed structure with metal wall panels and a reinforced concrete roof Steel framed structure with unreinforced masonry (CMU or brick) infill walls (non-load bearing) and a reinforced concrete or metal roof. Steel framed building with reinforced concrete walls panels and a reinforced concrete roof deck Pre-engineered metal structure Unreinforced masonry building with load bearing walls and a reinforced concrete roof. Reinforced concrete frame structure with unreinforced masonry infill walls and a reinforced concrete roof. Reinforced concrete frame structure with reinforced masonry infill walls and a reinforced concrete roof Typical domestic building, 2 storey brick walls timber floor

Whereas 30 mbar represents a threshold for fatalities, overpressures exceeding 500 mbar tend to lead to 50 % and more fatalities in most building types. Table 15 may be used to select the appropriate reinforcement method for a building to reduce the fatality rate and thus achieve risk mitigation where required. Figure 21 and Figure 22 show the various frequency contours for the different overpressure values. Frequency values are shown for 50 and 70 mbar. Overpressure curves above 70 mbar have not been found.


-4

10 / r

Figure 21: Frequency contours for 50 mbar

25.08.2008 Page 47


25.08.2008 Page 48

-4

10 / r

Figure 22: Frequency contours for 70 mbar

These results can be used to specify the Design Accidental Load (DAL) (refer to the Linde Document &AA-S-SD-1002) in conjunction with the following content of Table 16.


25.08.2008 Page 49

Table 16: Building and Plant Effects at Specified Consequence Levels

Consequence Flash Fire

Consequence Level Lower flammable limit

Vapour cloud explosion and subsequent overpressure

500 mbar

300mbar

150 mbar

110mbar

70 mbar

Fireball

30 mbar Within fireball 2 1.333

1000(kW/m ) Pool Fire/Jet Fire

2

20kW/m

12kW/m

2

s

Effects on Building/Plant Ignition of easily ignitable materials which are exposed, e.g. flammable vapour vents, etc., plastics, fabrics etc. Secondary fires are possible but unlikely. Process vessels and pipe work likely to be damaged. Unstrengthened buildings likely to be demolished. Threshold of significant damage to process vessels and pipe work. Unstrengthened buildings likely to be significantly damaged/partly demolished. Plant damage is insignificant except for inherently weak structures e.g., empty atmospheric storage tanks. Structural damage to domestic type buildings could be anticipated. Superficial damage would be expected with failure of unsupported walls and all windows broken. Onset of plant damage for inherently weak structures. Virtually all windows broken. Superficial damage to buildings. Plant damage unlikely and only slight superficial damage to buildings of brick construction. Most windows broken and the glass is likely to cause injury to some people within the buildings. Glass broken, but no fatalities anticipated. Building is likely to be ignited. People in the open air would be killed. The threshold of fatality for people exposed. People inside buildings will not be fatally injured. Carbonaceous material will not ignite spontaneously but could be ignited with a pilot flame. Pain experienced on unprotected skin within 2s. People outdoors will be unlikely to be able t o reach a place of safety No significant damage to buildings of conventional constructions. People should be able to safely remain in a building subject to this level of thermal radiation. Pain experienced on unprotected skin within a few seconds, but workers with protective overalls and able to move in any direction will have a good chance of reaching a place of safety.

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Sensitivity Evaluation

7.0

25.08.2008 Page 50

Sensitivity Evaluation

In order to evaluate the effectiveness of further risk reduction measures the following sensitivity calculations have been carried out. Table 17: Sensitivity Cases

Case No. 1

Pit on the jetty, LNG Storage Tank and the Pentane Tank

Assumption Sheet No. HC-6

2

Rock Wall around the LNG Storage Tank

RA-6

3

Splitting of cooling up medium tanks into smaller sections in order to reduce inventory

7.1

Sensitivity 1: Pit on the jetty, LNG Storage Tank and the Pentane Tank

Sensitivity

In this sensitivity calculation a bound around the “LNG Ship Loading Jetty”, the “LNG Storage Tank” and the “Pentane Storage Tank” is modelled. The design dimensions of pit are implemented as a bound around leaking equipment and shown in the table below: Bound LNG Ship Loading Jetty LNG Storage Pentane Storage Tank

7.1.1

Height [m] 0.25 1.8 0.9

Area [m2] 104 10.2 23.6

Discussion

This assumption affects the Pool Fire risk. This risk mitigation measure reduces the pool spreading and thus the heat radiation. Pit on the Jetty

Early Pool Fire without a bound


Early Pool Fire with a bound

Pit around the LNG Storage Tank

Early Pool Fire without a bund

25.08.2008 Page 51


Early Pool Fire with a bund

Pit around the Pentane Storage Tank

Early Pool Fire without a bund

25.08.2008 Page 52


25.08.2008 Page 53

Early Pool Fire with a bund

7.1.2

Comparison with Criteria

As shown ich chapter 7.1.1 the flame heat radiation impact is significantly reduced by a bound. As the contribution of pool fire from Jetty, LNG Storage- and Pentane Tank is small, this positive effect of the bounds is only marginally reflected by the calculated ISR (Individual Specific Risk). However, in order to reduce the risk to ALARP, it is decided to design the plant with a pit for the “LNG Ship Loading Jetty”, the “LNG Storage Tank” and the “Pentane Storage Tank”. This is considered in the risk results presented in Chapter 6.0. 7.2

Sensitivity 2: Rock Wall towards the public area on the peninsula

In this sensitivity calculation all fire vulnerabilities for the peninsula people have been reduced for 80% due to expected radiation effect zone from fires. 80 % of release sources are at lower than the rock wall. 7.2.1

Discussion

The heat radiation in case of fire at the LNG Tank will be reduced towards the public area on the peninsula, thus reduced the risk.


7.2.2

25.08.2008 Page 54

Comparison with Criteria

Table 18: ISR for people on the Peninsula

Criteria

Base Case

Sensitivity

Base Case Δ

1X10 -7/year
Peninsula people Peninsula people ISR 5.7X10-6 year ISR 4.6X10-8/ year 99%

The positive effect of the rock wall on individual specific risk for peninsula people is significant. It is decided to design the rock wall towards the public area. This is considered in the risk results presented in 6.0. 7.3

Sensitivity 3: Splitting of process vessels inside the refrigerant cycle into smaller vessels and additional block valves to reduce the volume of inventory loops

This measure would mainly lead to a shorter duration of jet fire and smaller pool fire scenarios. The effect therefore is a lower risk to asset, but the personnel risk would not mainly change, as this lethality occurs within the first minutes after a leakage due to good escape possibilities and the fact, that vapor clouds have built up their maximum, static flammable mass a few minutes after start of release. In fact, splitting of vessels and subsequently the addition of valves and flanges would increase the number of potential leak sources and therefore marginaly increase the risk. Furthermore from the process point of view, it is very unfavorable to add additional valves and instrumentation in the mixed refrigerant cylce as this would reduce the availability and operability of the plant. It would required also additional plot space. In summary, the splitting of vessels 1 inside the refrigerant cycle into smaller vessels and additional block valves to reduce the volume is assessed not to be a suggestive risk reduction measure.

1

Splitting of vessels to reduce the risk is only reasonable for larger storage vessels but not for process vessels

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Conclusions

8.0

25.08.2008 Page 55

Conclusions

A careful risk analysis of the first train of the Lyse LNG Base Load Plant has been performed, including a very detailed counting of all pieces of equipment (including all pipelines, vessels and compressors etc.) and sensitivity evaluation for the risk reduction measures (refer to the Chapter 7.0). It has been found that the calculated levels of individual risk for the 1 st, 2nd and 3rd parties are in compliance with the criteria set by Lyse. The individual specific risk for 1 st and 2nd party for the most exposed person in each group, maintenance and operators, is lower than 1 X 10 -3 per year and within the acceptance criteria. The average individual risk for personnel is 5.0 X 10-5 per year and therefore clearly below the acceptance criteria of 1 X 10 -4 / year. The individual specific risk for 3 rd party for the most exposed population (e.g. hiking track, ferry terminal industry and office workers) is within the ALARP regime. The average individual risk is 1.5 X 10-7 per year and therefore within the lower region of ALARP, close to acceptable in general. The calculated risk for the Peninsula people is acceptable since the rock wall (mound) is taken into account (refer to the Chapter 7.2).

Quantitative Risk Analysis for the LNG Plant Stavanger Appendix A: Assumption Sheets

9.0

25.08.2008 Page 56

Appendix A: Assumption Sheets

Appendix A documents the basis for the QRA of the Stavanger LNG Base Load Plant. The actual Assumption Sheets are presented in Table 19. Table 19: List of assumption sheets

Subject Area

Hazard Identification / Consequence Analysis

Assumption Sheet No.

HC-1

Main area definitions

HC-2

Release Rates

HC-3

Hydrocarbon Releases

HC-4

Gas Dispersion

HC-5

Gas Fire Modelling

HC-6

Liquid Fire Modelling

HC-7 HC-8

Vapour Clod Explosion Modelling Non-Process

HC-9

Loading Frequency

HC-10

Ship Transport Accidents Occupational Risk

HC-11 Frequency Analysis

Risk Assessment

Assumption Sheet Topic

FA-1

Inventory Count

FA-2

Frequency Database

FA-3

Leak Frequency (LNG Storage Tank) Event Tree Probabilities Ignition Sources – Probabilities Fire and Gas Detection

RA-1 RA-2 RA-3 RA-4 RA-5 RA-6 RA-7 RA-8

ESD/Blowdown System-Duration Time Vulnerabilities Heat Radiation Mitigation Active and Passive Fire Protection Escape Ways and Safe Haven

Revision No. / Revision Date

Page No.

Rev0 / 2008.05.26 Rev1 / 2008.07.31 Rev1 / 2008.07.31 Rev0 / 2008.05.26 Rev0 / 2008.05.26 Rev1 / 2008.07.31 Rev0 / 2008.05.26 Rev0 / 2008.05.26 Rev0 / 2008.05.26 Rev1 / 2008.07.31 Rev0 / 2008.05.26 Rev1 / 2008.06.02 Rev0 / 2008.05.26 Rev1/ 2008.07.31 Rev0 / 2008.06.02 Rev0 / 2008.06.02 Rev1 / 2008.07.31 Rev0 / 2008.06.02 Rev0 / 2008.06.02 Rev0 / 2008.06.02 Rev1 / 2008.07.31 Rev2 / 2008.08.25

58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 74 76 77 79 80 82 83

Remarks


Miscellaneous

MI-1

Acceptance Criteria

MI-2

Meteorological Data / Parameters Manning Levels / Population Distribution

MI-3

25.08.2008 Page 57

Rev0 / 2008.07.31 Rev0 / 2008.07.31 Rev1 / 2008.08.25

84 86 88


25.08.2008 Page 58

Stavanger LNG Base Load Plant

Date: 2008.05.26

Assumption No.: Subject Area: Topic:

Revision: 0

HC-1

Hazard Identification/Consequence Analysis Main area definitions

Assumption/Rule Set

The Stavanger LNG Base Load Plant has been divided into the following main areas: Main area definition Inventory Loop 1 Inventory Loop 2A Inventory Loop 2B1 Inventory Loop 2B2 Inventory Loop 3A Inventory Loop 3B Inventory Loop 4 Inventory Loop 5A Inventory Loop 5B1/2/3 Inventory Loop 5C Inventory Loop 6A1 Inventory Loop 6A2 Inventory Loop 6A3 Inventory Loop 6B1 Inventory Loop 6B2 Inventory Loop 6B3 Inventory Loop 6B4 Inventory Loop 7 Inventory Loop 8 Inventory Loop 9 Inventory Loop 10A Inventory Loop 10B Inventory Loop 11 Inventory Loop 12 Inventory Loop 13

Description Feedgas Purification NG Liquefication Gas NG Liquefication Liquid_103 bar System NG Liquefication Liquid_19 bar System LNG Storage Return Gas LNG Storage LNG Truck Loading LNG Ship Loading Tank Top LNG Ship Loading Line LNG Ship Loading Jetty Refrigeration Gas System_4 bar System Refrigeration Gas System_18 bar System Refrigeration Gas System_40 bar System Refrigeration Liquid 25-HX-101/103 System Refrigeration Liquid 25-VA-101 System Refrigeration Liquid 25-VA-102/25-HX-102 System Refrigeration Liquid 25-VL-102 System Propane Storage Pentane Storage Butane Storage Ethylene Storage Gas System Ethylene Storage Liquid System Hot Oil System Feedgas Fiscal Metering Tailgas Fiscal Metering

Implication of assumption: Impact on safety function impairment. Reference: Prepared by: Internal Verification:

Sign: CAN Sign: BAUMGARTNER/RATH

Date: 2008.05.26 Date: 2008.05.26

Sign:

Date:

Comment from Lyse: Approved by Lyse:


Stavanger LNG Base Load Plant Assumption No.: Subject Area: Topic:

25.08.2008 Page 59

Date: 2008.07.31 Revision: 1

HC-2

Hazard Identification/Consequence Analysis Release Rates

Assumption/Rule Set

Only releases of hydrocarbons are considered. A release of MDEA in the Solvent Regeneration system is not considered, as it is used in a not flammable aqueous solution. CO2/H2S leaks from pipelines to Regenerative Thermal Oxidation (RTO) due to a low operating pressure are not considered, but discussed qualitatively (see Appendix B: Hazard Identification). Releases have been analysed in terms of four characteristic hole sizes: Small: 1-10 mm Medium: 10-50 mm Large: 50-100 mm Very Large Large (Full (Full bore): taken to be hole hole equivalent equivalent to the largest diameter pipework in that inventory ( > 100 mm) Releases from pipelines, flanges, pumps etc. are modelled as liquid, gas, or two-phase releases. Where an inventory comprises significant liquid and gas sections, e.g. in a vessel, then both are modelled. The representative release height for all cases is taken 1 m; except for the LNG Tank, where 30 m are applied, since the leak sources (flanges) by the LNG Tank are expected on the tank top. Release rates are assumed to be constant throughout the release duration time and calculated with isolation (ESD System), and with blowdown (see Assumption Sheet RA-4). According to EN 1473, the isolated sections shall be depressurised to 50 % of design pressure in 15 minutes or to 7 barg in 30 minutes. Based on this, the calculated time to detect and initiate is 600 s. An average blowdown time of 900 s is used in the calculation. Release rates of gas systems with small gas volume are limited by flow controlled gas supply. Liquid release rates are limited by pump rates. However, the times to detect will vary, depending on leak size, release rate, location of release, etc. In practice, some releases may be isolated much quicker, but it is assumed that this represents a “realistic worst case” value. • • • •

Implication of assumption: Releases of hydrocarbons affect the fire and explosion risk. Reference: Prepared by: Internal Verification:


Date: 2008.07.31 Date: 2008.07.31

Sign:

Date:



25.08.2008 Page 60


Date: 2008.07.313


Revision: 1

HC-3

Hazard Identification/Consequence Analysis Hydrocarbon Releases

Assumption/Rule Set

Outdoor Releases of hydrocarbons (gas/liquid or two phases) are considered from the counting equipment (see Assumption Sheet FA-1). Hole sizes are defined in Assumption Sheet HC-2. Release duration time is based on the fire and gas detection and ESD&Blowdown System (see Assumption Sheets RA-3 and RA-4). Hydrocarbon leaks in buildings, which contain Hydrocarbons, are defined as explosion group zone 1 and are assumed to have a minor contribution to risk compared to outdoor releases due to forced ventilation. Hydrocarbon entering in a building is prevented by adequate gas detection and closing the air-intake. Therefore Hydrocarbon leaks in buildings are not analysed, but are discussed qualitatively qualitatively (see Appendix B: Hazard Identification). Implication of assumption: Outdoor hydrocarbon releases affect the fire and explosion risk. The buildings are specified with explosion group zone 1. Gas entering in a building is prevented by adequate gas detection and closing the air-intake. Therefore the risk of internal explosion is not considered. considered. Reference: Prepared by: Internal Verification:


Date: 2008.07.31 Date: 2008.07.31

Sign:

Date:



25.08.2008 Page 61


Date: 2008.05.26


Revision: 0

HC-4

Hazard Identification/Consequence Analysis Gas Dispersion

Assumption/Rule Set

The gas dispersion is calculated by the UDM model implemented in the PHAST / PHAST RISK software. This model considers only free field dispersion, so that any local air stream effects at equipment/ buildings are not included in the dispersion calculation. Dispersion generally is modelled as horizontal releases. A representative gas cloud size to 50% of lower flammable limit (LFL fraction) has been used to determine the magnitude / extent of flash fires / explosions. Implication of assumption: Gas dispersion affects the consequence calculations associated with the fire and explosion risk. Reference: Prepared by: Internal Verification:


Date: 2008.05.26 Date: 2008.05.26

Sign:

Date:



25.08.2008 Page 62


Date: 2008.05.26


Revision: 0

HC-5

Hazard Identification/Consequence Analysis Gas Fire Modelling

Assumption/Rule Set

Gas fires resulting from ignited hydrocarbon releases are modelled as jet fire, flash fire and fire ball for each release scenario. For unimpinged gas releases the jet fire is calculated using the Shell model. The original Shell model uses the Chamberlain correlation for calculation of the flame length as function of the release rate, which was developed for near-vertical vapour-phase releases. This correlation was modified by Cook et al. to describe the shape of jets that contain liquid. Therefore the option DNV Recommended has been used, that means the PHAST / PHAST RISK program will use the correlation that is most appropriate for the release-conditions. For impinged releases the fireball diameter is calculated from the release rate using the correlation given in Dutch Yellow book. For delayed ignition the flash fire limit is the distance to ½ LFL. Implication of assumption: This assumption affects the fire risk. See also the assumption sheet RA-6. Reference: Methods for the calculation of physical effects (Yellow Book), CPR14E Prepared by: Sign: CAN Date: 2008.05.26 Internal Verification: Sign: Date: 2008.05.26 BAUMGARTNER/RATH Comment from Lyse: Approved by Lyse:

Sign:

Date:



25.08.2008 Page 63


HC-6

Hazard Identification/Consequence Analysis Liquid Fire Modelling

Assumption/Rule Set

Fires resulting from ignited liquid releases are modelled as a pool or a jet fire. Pool fire dimensions are modelled using the spill rate to compute pool development with allowance for burning (if ignited) or boil off. The maximum pool sizes are defined either by hitting a dike wall or by reaching a minimum thickness. The minimum thickness depends on surface and is set by the PHAST / PHAST RISK program to 5 mm for a concrete surface. For pool fires the effects are calculated for an early and late ignition. The late pool fire is assumed to occur when the pool reaches its maximum radius. For early pool fires pool size evolution is based on ignition occurring at 10 sec. Jet flame lengths and radiation effects distances are calculated as per gas fires (refer to HC5). The bund around the pentane tank is implemented in the calculation. The effect of an LNG pit at the storage tank and at the jetty is implemented in the LNG PLANT QRA calculations as a bund around the tank (Inventory Loop 3B) and around the jetty (Inventory Loop 5C), since the PHAST RISK program can not directly simulate such a pit. The bunds limit the pool spreading. Implication of assumption: This assumption affects the fire risk. The LNG pool fires around storage tank and around the jetty loading are limited by a pit in each case (modelled as bund). This reduces the pool spreading and thus the heat radiation. Reference: Prepared by: Internal Verification:


Date: 2008.07.31 Date: 2008.07.31

Sign:

Date:




25.08.2008 Page 64


HC-7

Hazard Identification/Consequence Analysis Vapour Cloud Explosion Modelling

Assumption/Rule Set

The TNT model is used to calculate vapour cloud explosion effects. The explosion efficiency is set to 10 %. For gases lighter than air an air burst is assumed. For gases heavier than air a ground burst is taken into account. Then the PHAST / PHAST RISK program multiplies the explosion efficiency by factor two, to account for the effects of reflection on the overpressure. The flammable mass is calculated as mass between LFL and UFL. The explosion location criterion is the cloud front (1/2 LFL fraction). Vapour cloud explosion effects are calculated if the minimum explosion energy of 5 x 10^6 kJ (DNV default value) is exceeded. Implication of assumption: This assumption affects the explosion risk. Reference: Prepared by: Internal Verification:


Date: 2008.05.26 Date: 2008.05.26

Sign:

Date:




25.08.2008 Page 65


HC-8

Hazard Identification/Consequence Analysis Non-Process

Assumption/Rule Set

Non-process events include: loss of utilities (failure leads to fail safe conditons), utilities releases, non-hydrocarbon fires (e.g. transformer fire in electrical/instrument room). They are not included in the LNG PLANT QRA due to their low frequency and low consequence and active and passive fire protection , but discussed qualitatively (see Appendix B: Hazard Identification). Implication of assumption: This assumption has none impact on fire and explosion risk in the LNG PLANT QRA. Reference: Prepared by: Internal Verification:


Date: 2008.05.26 Date: 2008.05.26

Sign:

Date:




25.08.2008 Page 66


HC-9

Hazard Identification/Consequence Analysis Loading Frequency

Assumption/Rule Set

Loading operations are assumed to be 1 cargo ship loading every 5th day (filling time 6h) and truck loading 10 times in a day (filling time 1.2 h). Implication of assumption: This assumption reduces the release and ignition probabilities. Reference: Prepared by: Internal Verification:


Date: 2008.05.26 Date: 2008.05.26

Sign:

Date:




25.08.2008 Page 67


HC-10

Hazard Identification/Consequence Analysis Ship Transport Accidents

Assumption/Rule Set

Types of accidents are not a part of LNG PLANT QRA. A ship collision risk assessment is recommended (important risk). As it has an impact on third party population risk. A ship collision with jetty could be significant with respect to 1 st risk. Implication of assumption: This assumption could have impact on fire and explosion risk in the LNG Plant. Collision incidents per port visit - while mooted at jetties, berths etc or within locks, enclosed harbours etc. is 3.7 X 10-5 [LMIS database]. Therefore, such accidents can be neglected. Reference: Lyod’s Maritime Information Services (LMIS). Prepared by: Sign: CAN Date: 2008.07.31 Internal Verification: Sign: Date: 2008.07.31 BAUMGARTNER/RATH Comment from Lyse: Approved by Lyse:

Sign:

Date:



25.08.2008 Page 68


HC-11

Hazard Identification/Consequence Analysis Occupational Risk

Assumption/Rule Set

The occupational accidents have been not included in the acceptance criterion, and are therefore not considered in the LNG Plant QRA. Implication of assumption: This assumption has none impact on harm/death risk in the LNG Plant QRA. Reference: OGP, Safety Performance Indicators – 2006 data, Report no. 391, June 2007 Section 2.2 & 4.1 Prepared by: Sign: CAN Date: 2008.05.26 Internal Verification: Sign: Date: 2008.05.26 BAUMGARTNER/RATH Comment from Lyse: Approved by Lyse:

Sign:

Date:



25.08.2008 Page 69


FA-1 Frequency Analysis Inventory Count

Assumption/Rule Set

For each inventory the leak frequencies are estimated using a full parts count of the equipment shown on the P&ID. Typically this includes: Flanges (not to consider in welded pipelines) Valves Small bore fittings Pipelines Pressure vessels Heat exchangers Pumps Compressors Atmospheric Tanks Equipment counts assumptions are detailed below: Drums and other vessels that are primarily gas (e.g. cycle compressor interstage drum) or liquid (e.g. cold MRC separator) are conservatively treated as 100% gas or liquid, respectively Relief valves to flare and blow down valves are counted as normal valves and assumed to be closed in normal operation. Therefore downstream equipment are not considered Flanges and small bore fittings in pipelines are not counted since the failure of flanges is included in the failure frequency of the pipeline [Purple Book] For jetty, a double flange per valve connections and associated flanges are counted Flanges and small bore fittings at vessels and at the LNG Tank are not counted since their failure frequencies are included in the failure frequency of the vessels and tanks [Purple Book] • • • • • • • • •

•

•

•

•

•

Further details are given in Appendix C: Equipment Count. Implication of assumption:

The amount of inventories as leakage sources affects the release frequency. Reference: Guidelines Risk calculations (Purple Book) BEVI Module C, Version 3.0 Date 1 January 2008: Modelling specific BEVI categories. BEVI is the abbreviation of the decree implementing the SEVESO directive. Prepared by: Sign: CAN Date: 2008.06.02 Internal Verification: Sign: Date: 2008.06.02 BAUMGARTNER/RATH Comment from Lyse: Approved by Lyse:

Sign:

Date:



25.08.2008 Page 70


FA-2 Frequencies Analysis Frequency Database

Assumption/Rule Set

As Linde AG has not received a database from the Client, the leak sizes and- frequencies, are calculated with DNV Software Leak 3.2. The generic failure data used as the basis of the frequency analysis is the UK HSE Offshore Hydrocarbon Release Statistic 1992-2006, or HCRD [Ref. A]. This is a DNV recommended database for Hydrocarbon releases. To reflect the LNG plant, which is considered a clean service and an onshore facility, leak frequencies for pipelines, vessels and the LNG Storage Tank are applied as given in the Purple Book [Ref. B]. Accordingly, failures of flanges in pipelines or at vessels are included in the failure frequency of the pipeline or of the vessel (see Assumption Sheet FA-1). Further details are given in Appendix D: Results of Leak 3.2. Calculations. Implication of assumption: Key influence on the risks (i.e. risk is directly proportional to frequency). Reference: A: HSE, 2000. Offshore Hydrocarbon Release Statistics, 1999 , Offshore Technology Report OTO 079, HSE Offshore Safety Division (OSD), January 2000. B: Guidelines Risk calculations (Purple Book) BEVI Module C, Version 3.0 Date 1 January 2008: Modelling specific BEVI categories. BEVI is the abbreviation of the decree implementing the SEVESO directive. Prepared by: Sign: CAN Date: 2008.05.26 Internal Verification: Sign: Date: 2008.05.26 BAUMGARTNER/RATH Comment from Lyse: Approved by Lyse:

Sign:

Date:



25.08.2008 Page 71


FA-3 Frequencies Analysis Leak Frequency (LNG Storage Tank)

Assumption/Rule Set

Acc. to EN 1473 roof collapse/tank collapse is considered negligible for full containment tanks. Therefore, a very large leak (full rupture) associated with the full containment LNG tank is not considered in Inventory Loop 3A “LNG Storage Return Gas” (refer to the Assumption Sheet HC-1). Further details are given in Appendix D: Results of Leak 3.2. Calculations. Implication of assumption: The leak frequency is directly proportional to risk. Reference: Prepared by: Internal Verification:


Date: 2008.07.31 Date: 2008.07.31

Sign:

Date:



25.08.2008 Page 72

Date: 2008.06.02


Revision: 0

RA-1 Risk Assessment Event Tree Probabilities

Assumption/Rule Set

The development of a release is largely defined by the stage at which ignition occurs, where the immediate and delayed ignition may give an explosion, or a flash fire, or a fireball. These different developments are represented in a diagram called an “event tree”, and the probabilities for the developments are known as “event tree probabilities” or “event tree parameters”. The sum of the probabilities for Fireball, Flash Fire, Explosion and Pool Fire alone is usually 100%. An example risk model event tree for a continuous release with rainout (with probability of a pool fire) is shown in Figure 23. Figure 23: Example Risk Model Event Tree Structure

1.0

0.6

1.0 0.3* 1.0

0.4

0.6

0.15

* The default probability of immediate ignition (0.3) has not been used to account the effects of fluid properties (e.g. reactivity) and source strength on the ignition probability. If no immediate ignition occurs, the program models the dispersion of the cloud through a succession of time steps until it has diluted below a hazardous concentration. At each time step the program models the effects of delayed ignition of the cloud, calculating the probability of delayed ignition by considering the ignition sources (see the Assumption Sheet RA-2) within the flammable area of the cloud during that time-step.


25.08.2008 Page 73

Implication of assumption: The event tree is a key aspect of the QRA model and affects of fire and explosion risk depending on the timing and type of ignition. Reference: Prepared by: Internal Verification:


Date: 2008.06.02 Date: 2008.06.02

Sign:

Date:



25.08.2008 Page 74

Date: 2008.06.02


Revision: 0

RA-2 Risk Assessment Ignition Sources – Probabilities

Assumption/Rule Set

The ignition sources are defined by ignition probability and time period. The ignition probability is the probability that the ignition source will ignite a flammable cloud if the cloud is exposed to the source for the specified time period, which is assumed 10 s (default value in PHAST RISK). The expected ignition sources and their probabilities are listed in following table: Ignition Source Speed

Ignition Probability

Flare 54-FC-101 Fired Heater for Hot Oil 52-FA-101 H2S Conventer (Incinerator) 20-XT-101 Electrical Substation Traffic (Truck Loading) Maintenance Traffic Parking Area Traffic

[Fraction] 0.5 0.1 0.1 0.1 0.1 0.1 0.1

Traffic Density

Average

[day]

[kph]

10 1 20

30 30 20

It is assumed that all electrical equipment will be EX-safe (Explosion Protection). Due to fire and gas detection, the Regenerative Thermal Oxidation (RTO) Incinerator and all the not exsafe units will be isolated by closing the damper at the air inlet. Therefore the ignition probabilities have been reduced to 0.1 expect for the flare. The flare pilots will burn with a pilot flame continuously. Due to the fact that the flare pilots are installed 70 m above ground and the assumption that only 50 % of flammable gas clouds will disperse to this height, the flare ignition probability has been reduced from 1 to 0.5. The reduced ignition fractions may reduce the delayed ignition probabilities on the event trees, which are depends on the ignition fractions (e.g. Assumption Sheet RA-1). The default value assigned within PHAST RISK for the ignition source associated with people corresponds to 1.68 x 10 -4 per person per second of cloud exposure. This value has been derived to account for the probability of ignition associated with people in general, and includes an allowance for smoking and general human behaviour associated with residential areas. The ignition probability of personal within the LNG Plant would be zero (except the truck and maintenance traffic), with no smoking. Immediate ignition for material properties of the released material have been taken into account and shown in table below: Class (Material) immediate

source-continuous

Average/high reactivity <10 kg/s e.g. Ethane,Propane/Hydrogen 10-100 kg/s >100 kg/s Low Reactivity <10 kg/s e.g. Methane 10-100 kg/s >100 kg/s

source-instantaneous

<1000 kg/s 1000-10,000 kg/s >10,000 kg/s <1000 kg/s 1000-10,000 kg/s >10,000 kg/s

probability of ignition 0.2 0.5 0.7 0.02 0.04 0.09


25.08.2008 Page 75

Implication of assumption: This assumption affects of fire and explosion risk depending on the timing and type of ignition. The ignition probabilities due to ex-safe design and fire and gas detection reduce the fire and explosion risk. Reference: Prepared by: Internal Verification:


Date: 2008.06.02 Date: 2008.06.02

Sign:

Date:




25.08.2008 Page 76


RA-3 Risk Assessment Fire and Gas Detection

Assumption/Rule Set

The fire and gas detection depends on the location and magnitude of the event, the number, location of detectors and their PFD (probability of failure on demand). However, the basis design of the LNG Plant considers a sufficient fire and gas detection (refer to the Linde Document &AA S-PC-1004 Fire Protection Concept and &AA S-ZA-1003 Fire & Gas Detection Plan). The F&G System is considered in the QRA by prevention / mitigation of Hazards either automatic or manual actions activated upon gas detection: Automatic shutdown of the following ignition sources (refer to Assumption Sheet RA-2) : Regenerative Thermal Oxidation (RTO) (Incinerator) 20-XT-101 o fired heater (Hot-oil unit) 50-XT-101 o Manual activation of the Emergency Shutdown, Isolation and Depressuring System via push-buttons by the operator in the CCR. The F&G detection system is the basis of the ESD- and Blowdown duration time (see the Assumption Sheet RA-4). •

•

Implication of assumption: This assumption affects the release duration. Reference: Prepared by: Internal Verification:


Date: 2008.07.31 Date: 2008.07.31

Sign:

Date:




25.08.2008 Page 77


RA-4 Risk Assessment ESD/Blowdown System-Duration Time

Assumption/Rule Set

The initial release rate [in kg/s] is calculated within the PHAST RISK discharge model and set constant during the representative release duration time. In reality, the internal pressure is reduced and this reduces the release rate. If the initial release rate is very large, the release duration time will be short. Low release rates will compensate with the representative release duration time, which is typically of 1500 s and shown in Figure 2. Figure 24: Refrigerant Depressurizing Calculation pressure

P1

A1

A1 ~ A2

A2

time

900 s

Time used in QRA model as max. release duration at constant release rate after initiation of ESD and Blowdown

According to EN 1473, the isolated sections shall be depressurised to 50 % of design pressure in 15 minutes or to 7 barg in 30 minutes. Based on this, the calculated time to detect and initiate is 600 s followed by an average blowdown time of 900 s. Implication of assumption: This assumption affects release the release rates and duration times. Reference:


Prepared by: Internal Verification:

25.08.2008 Page 78


Date: 2008.06.02 Date: 2008.06.02

Sign:

Date:



25.08.2008 Page 79


Date: 2008.06.02


Revision: 0

RA-5 Risk Assessment Vulnerabilities

Assumption/Rule Set

A given accident may lead to fatalities both indoors and outdoors depending on each hazardous effect. The lethality factors are given for individual and socitial risk calculations and shown in Figure 3. Figure 25: Default values based on Purple Book set in General Risk Parameters in PHAST RISK

Some of this default values have been changed based on the mitigation measures by the assumption sheets RA-6 and RA-8. Implication of assumption: The risks are directly influenced by the impact and fatality assumptions, which quantify the severity of the consequences. Reference: Guidelines Risk calculations (Purple Book) BEVI Module C, Version 3.0 Date 1 January 2008: Modelling specific BEVI categories. BEVI is the abbreviation of the decree implementing the SEVESO directive. Prepared by: Sign: CAN Date: 2008.00.02 Internal Verification: Sign: Date: 2008.06.02 BAUMGARTNER/RATH Comment from Lyse: Approved by Lyse:

Sign:

Date:



RA-6

25.08.2008 Page 80


Risk Assessment Heat Radiation Mitigation

Assumption/Rule Set

The impact of heat radiation from fires to the peninsula depends on the height of the release sources. 80% of all release sources (vessels, pipework and the LNG Storage Tank) are at a lower level than the rock wall. The heat radiation effect zones are simplified shown in figure below. The heat radiation effects are expected at the rock wall top. It is confirmed by the Client that no people will enter the rock wall top, which will be ensured by appropriate measures.

Therefore all fire vulnerabilities for the peninsula people have been reduced in the General Risk Parameters (default vulnerabilities in Assumption Sheet RA-5) for 80% (i.e. outdoor vulnerability factor for fire is reduced from 1 to 0.2).


25.08.2008 Page 81

For a more detailed evaluation of this effect, CFD calculations would be required. Implication of assumption: This assumption reduces the risk to people on the peninsula. (Refer to the Table 2 “Individual Specific Risk (ISR) for the 3 rd party”) Reference: Prepared by: Internal Verification:


Date: 2008.06.02 Date: 2008.06.02

Sign:

Date:




25.08.2008 Page 82


RA-7 Risk Assessment Active and Passive Fire Protection

Assumption/Rule Set Active Fire Protection: All vessels and piping are protected by the water application like monitors, hydrants. Passive Fire Protection: Fireproof insulation for supporting steel structures shall be provided to protect the plant and fire fighting personnel against the effect of support failure in case of fire. Fireproof insulation is needed within fire hazardous areas. For details, refer to the Linde Document &AA S-PC-1004 Fire Protection Concept.

To consider the Stavanger LNG Base Load Plant being an onshore plant with clean service etc. the leak frequencies for piping, the LNG storage tank and pressure vessels given in the Dutch Purple Book have been applied. But additive to that no credit for active and passive fire protection can be taken, as leakage failure rates given in the HSE database as well as failure rates given in the purple book consider only initial leakages, but not secondary leakages due to heat ingress caused by fire radiation. Implication of assumption: This assumption prevents the escalation risk. Reference: Guidelines Risk calculations (Purple Book) BEVI Module C, Version 3.0 Date 1 January 2008: Modelling specific BEVI categories. BEVI is the abbreviation of the decree implementing the SEVESO directive. Prepared by: Sign: CAN Date: 2008.07.31 Internal Verification: Sign: Date: 2008.07.31 BAUMGARTNER/RATH Comment from Lyse: Approved by Lyse:

Sign:

Date:



25.08.2008 Page 83


RA-8 Risk Assessment Escape Ways and Safe Haven

Assumption/Rule Set Escape Ways: Escape ways are considered to be state of the art in hydrocarbon processing plants Therefore no credit is taken to account. Safe Haven: The probability of death within the building will depend on the vulnerabilty of the building in the first instance and given a level of damage how vulnerable the occupants are to the combined hazards. The control room is designed for dynamic resistance pressure of 200 mbar and may verify a overpressure vulnerabilty factor of 0.322. Therefore, the Heavy Explosion Damage value in the General Risk Parameters (default vulnerabilities) has been changed from 1 to 0.322 for the administrative personal (refer to the Assumption Sheet RA-5 and the table in the QRA Issue 02). Implication of assumption: This assumption affects the fire and explosion risk. Reference: Prepared by: Internal Verification:


Date: 2008.07.31 Date: 2008.07.31

Sign:

Date:




25.08.2008 Page 84


MI-1 Miscellaneous Acceptance Criteria

Assumption/Rule Set

The below tables and figures summarize the Acceptance Risk Criteria for Stavanger LNG Base Load Plant given by the Client.

Acceptance Risk Criterion for Stavanger LNG Base Load Plant Individual Risk (Pr År) st nd 1 and 2 party risk 3rd party risk

Individual specific risk (ISR) for most exposed person Avarage individual risk (AVR)

.

<1x10-3

<1x10-5

<1x10-4 (Note 1)

<1x10-5

Acceptance Risk Criterion for Stavanger LNG Base Load Plant Individual Risk (Pr År) st nd 1 and 2 party risk 3rd party risk (ISR/AVR)

Ikke akseptabelt ALARP Akseptabelt

>1x10-3 / >1x10-4 1x10-6 to 1x10-3 / 1x10-6 to 1x10-4 (Note 1) <1x10-6

1x10-5 1x10-7 to 1x10-5 <1x10-7

Note 1: If no individual risk (ISR ) is found to be above 1 > 10-4 the average individual risk (AIR) criteria is fulfilled. Societal Fatal Risk

Figure above shows the Stavanger LNG Base Load Plant F-N Curve Societal Risk Criteria (same std. as UK HSE F-N curve)


25.08.2008 Page 85

Implication of assumption: Impact on safety function impairment. Reference: Prepared by: Internal Verification:


Date: 2008.07.31 Date: 2008.07.31

Sign:

Date:



25.08.2008 Page 86


Date: 2008.07.31


Revision: 0

MI-2 Miscellaneous Meteorological Data/Parameters

Assumption/Rule Set

Data on distribution of wind speed and wind direction in the surroundings of Sola is presented in the form of a wind-rose in Figure 26. In combining the wind rose and stability data the eight representative weather categories have been rationalised to three weather categories. The probability of each weather category (stability and speed) and each direction is shown in Table 20. Figure 26: Wind Rose Data for Sola

Table 20: Rationalised Representative Weather Categories for Sola Stability Class

Wind direction [degrees]

Wind [m/s]

292.5 337.5

337.5 22.5

22.5 67.5

67.5 112.5

112.5 157.5

157.5 202.5

202.5 247.5

247.5 292.5

DAY: D - 1.5 NIGHT: F-1.5 D- 6 D - 12

1.99 14.71 2.79

0.961 7.09 1.346

1.012 7.47 1.417

1.633 12.04 2.293

1.335 9.89 1.878

0.501 3.69 0.702

0.807 5.96 1.13

1.977 14.57 2.76

SUM

19.49

9.397

9.899

15.966

13.103

4.893

7.897

19.2307

In addition to the weather categories, certain meteorological constants are defined as inputs to the consequence modelling. These values are summarised below.


25.08.2008 Page 87

Implication of assumption: The weather conditions have a key influence on flammable cloud dispersion and hence the consequences associated with any release. The influence of any specific weather category and direction will vary for each and every release, where on balance the resulting influence of any changes in the meteorological assumptions will have a negligible influence on the risk results. The dispersion and consequences associated with LNG and other dense gas releases are relatively sensitive to assumptions affecting the heat transfer to the cloud. Hence, the above values are relatively conservative representative conditions, but will not necessarily correspond to the worst-case dispersion conditions that may occur. Representative conditions used – relevant to consequences, with relatively minor influence on subsequent risks. Reference: Prepared by: Internal Verification:


Date: 2008.07.31 Date: 2008.07.31

Sign:

Date:



25.08.2008 Page 88

Date: 2008.08.25


Revision: 0

MI-3

Miscellaneous Manning Levels/Population Distribution

Assumption/Rule Set Manning levels are defined for the following time periods:

‘Daytime’: Normal working day is divided into 2 shifts; a morning shift (08:00 to 16:00) and an afternoon shift (16:00 to 00:00).

•

‘Night’: Night shift between 00:00 and 08:00. The risk analysis is based on the onsite population (1 st and 2nd party) and off-site population (3rd party). It is assumed that 70% of release events relate to some form of manual intervention, which is therefore assumed to apply only to the daytime shifts. The remaining 30% of events apply to the night shift. The loading operations (truck and jetty loading), which are not to be present 24 hours a day, all days throughout the year, are taken into account. Loading operations are planned to be 1 cargo ship loading every 5th day by a ship filling time of 10 h and truck loading 10 times in a day by a truck loading time of 1.2 h. A total number of 55 workers (1 st and 2nd party) are present at the plant temporarily within 24 hours. This figure is used to calculate the average individual risk for onsite population. The QRA calculation of the average individual risk is based on 4567 off-site people by day and 78 by night. The distribution of onsite- and off-site people is summarised in Table 21 and Table 22. •

Table 21: Onsite Population (1st and 2nd party)

Buildings / Areas Administration Building Maintenance Truck Loading Ship Loading (Jetty) Ship Deck Ship Bridige

Day (per 3 2 4 1 2 8

Personnel / People Shift) Night Total 1 1 2 1 2 8

Number 7 5 10 3 6 24


25.08.2008 Page 89

rd

Table 22: Off-site Population (3 party)

Areas Peninsula Hiking Track Ferry Terminal_office workers Ferry Terminal_industry workers Ferry Terminal_passengers Energiveien+Risavika_office workers Energiveien+Risavika_industry workers Container Area_office workers Container Area_industry workers Rest Companys_office workers Rest Companys_industry workers Living Quarters

Personnel / People Day Night 16 (in non-work day) 0 8 (in a non-work day) 0 100 2 10 0 1500 0 400 5 559 0 10 1 50 0 1139 10 715 0 60 60

Implication of assumption: Onsite and off-site are directly influenced by the numbers of people exposed to hazardous events and hence the individual specific risk and the average individual risk are sensitive to the manning assumptions. Reference: Prepared by: Sign: CAN Date: 2008.07.31 Internal Verification: Sign: Date: 2008.07.31 BAUMGARTNER/RATH Comment from Lyse: Approved by Lyse:

Sign:

Date:

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Appendix B: Hazard Identification

10.0

25 08 2008 Page 90

Appendix B: Hazard Identification

HAZID – Lyse LNG Base Load Plant, Munich 30th of October 2007 #

Hazard

Causes

Consequences

Safeguards

Further follow up/Comments

1

NG Leak from Process Equipment or piping

Corrosion Erosion Gasket Failure etc.

Regular Maintenance Fire and Gas Detection ESD System Emergency Blow Down System Active/Passive Fire Protection

considered in Linde QRA

2

LNG Leak from Process Equipment or piping




3

LNG Leak from Tank


Regular Maintenance Fire and Gas Detection ESD System Emergency Blow Down System Active/Passive Fire Protection Impoundment Pit


4

Mixed Refrigerant Gas Leak from Process Equipment or piping


Atmospheric Dispersion of Gas Toxic Consequences Potential Explosion Flash Fire Jet Fire Flashing and Evaporation of LNG Atmospheric Dispersion of Gas Potential Explosion Flash Fire Jet Fire Pool Fire Flashing and Evaporation of LNG Atmospheric Dispersion of Gas Potential Explosion Flash Fire Jet Fire Pool Fire Atmospheric Dispersion of Gas Potential Explosion Flash Fire Jet Fire




5

Mixed Refrigerant Liquid Leak from Process Equipment or piping


6

Mixed Refrigerant Gas Leak from Compressor or piping in Compressor Noise Hood

Corrosion Erosion Gasket Failure Seal Failure etc.

7

LNG Leak at Truck Loading

Corrosion Erosion Mechanical Damage Gasket Failure etc.

8

LNG Leak at Jetty Loading (Ship Loading)

Corrosion Erosion Mechanical Damage Gasket Failure etc.

9

Transformer Oil Fire in Electrical/Instrument Room

Gasket Failure Oil Leakage

25 08 2008 Page 91

Flashing and Evaporation of LNG Atmospheric Dispersion of Gas Potential Explosion Flash Fire Jet Fire Pool Fire Potential Explosion Flash Fire Jet Fire


Considered in Linde QRA

Regular Maintenance Fire and Gas Detection ESD System Emergency Blow Down System Active fire Protection


Flashing and Evaporation of LNG Atmospheric Dispersion of Gas Potential Explosion Flash Fire Jet Fire Pool Fire Flashing and Evaporation of LNG Atmospheric Dispersion of Gas Potential Explosion Flash Fire Jet Fire Pool Fire





Oil Pool Fire


Not considered in Linde QRA (minor contribution to risk) Considered in Linde Document &AA S-PC-1004 Fire Protection Concept


10

MDEA Leak from Process Equipment or piping


Release of CO2 loaded MDEA to dike area and pit

11

CO2 leak from piping to Regenerative Thermal Oxidation (RTO) Incinerator


12

NG leakage in shell side of 25-HZ-101

13

Failure of LNG Ship Loading

Corrosion Erosion Mechanical Stress etc. Corrosion Erosion Mechanical Damage Operator Failure Gasket Failure Coupling Failure etc.

Release of CO2. The CO2/H2S (sour gas) in CO2 wash unit is routed to the regenerative thermal oxidation and then sent to atmosphere at safe location. Dispersion from a leak of CO2/H2S gas cloud due to low operating pressure is not considered as the contribution to the risk is minor compared to the above mentioned outcomes. NG to refrigerant system

Flashing and Evaporation of LNG Atmospheric Dispersion of Gas Potential Explosion Flash Fire Jet Fire Pool Fire

25 08 2008 Page 92

Dike Area and Pit

Not considered in Linde QRA (minor contribution to risk) Not a flammable medium

Not considered in Linde QRA (no flammable Gas) Refer to the Linde Document S-RA 1001 Air Dispersion Study of CO2 off gas

Safety Relief Valve to Flare

Not considered in Linde QRA (only flaring) 25 SV 1002 is designed for this case

Regular Maintenance Fire and Gas Detection ESD System Active/Passive Fire Protection



25 08 2008 Page 93

14

Failure of Truck Loading

Corrosion Erosion Mechanical Damage Operator Failure Failure of Coupling Failure Gasket Failure etc.

Flashing and Evaporation of LNG Atmospheric Dispersion of Gas Potential Explosion Flash Fire Jet Fire Pool Fire

Regular Maintenance Fire and Gas Detection ESD System Active/Passive Fire Protection


15

Cable Fire in Electrical/Instrument Room

Overload Mechanical Damage etc.

Local Fire

Regular Maintenance Fire Detection Active Fire Protection

16

Fuel Gas Leak from Boil-off Compressor or piping in Compressor Noise Hood

Corrosion Erosion Gasket Failure Seal Failure etc.

Potential Explosion Flash Fire Jet Fire

Regular Maintenance Fire and Gas Detection ESD System Emergency Blow Down System Active fire Protection

Not considered in Linde QRA (minor contribution to risk) Refer to the Linde Document &AA S-PC-1004 Fire Protection Concept Considered in Linde QRA

17

Hot Oil Fire in 20-HA102 and 20-HA-111

Gasket Failure Oil Leakage

Oil Pool Fire




18

Hydrocarbon (HC gas / liquid or two phase) leaks in buildings

19

LNG release from PSV’s on the LNG tank to the atmosphere

25 08 2008 Page 94

Buildings containing HC: - The buildings are specified with explosion group zone 1; therefore the risk of internal explosion is reduced. - The protective effect of the building is not considered in the PHAST RISK calculation (conservative consideration). Buildings containing no HC: - Gas entering in a building is presented by adequate gas detection and closing the air-intake.

Not considered in Linde QRA (minor contribution to risk) Refer to the Linde Document SPC-1009 Explosion Protection Concept and S-ZA-1002 Hazardous Area Classification Plan

Not considered in Linde QRA (minor contribution to risk) Refer to the Linde documents SRS Specific Part - SIL Allocation &AA S-SD 1003 and Dispersion Calculation Safety Valves LNG Tank &AA S-RA 1002

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Appendix C: Equipment Count

11.0

25.08.2008 Page 95

Appendix C: Equipment Count

To determine the leak size and -frequency, an equipment count was performed using the P&IDs. The equipment count for pipelines, valves and number all individual vessels, heat exchangers, pumps, filters etc. have been compiled and transferred into the LEAK frequency calculation. The resulting data is used in DNV's LEAK 3.2 software, to determine overall leak frequencies and leak sizes. Details are indicated in the subsequent tables. Unit

Eq-ID

Name

Number

Diameter inch

Diameter Length mm

mm

Operating Operating Volumen pressure tempeature bara °C m³

Stream content

Inventory

Remarks

kg

Feedgas Purifiction (IL1) IL1

20-HA-101

Tube Bundel Heat Exchanger

IL1

20-HA-103

Tube Bundel Heat Exchanger

IL1

20-VE-101

Amine Wash Column

IL1 IL1

20-VL-111 20-VK-111

Feed Gas Water KO Drum Feed Gas Drier

IL1

20-LF-111

Dry Gas Filter

IL1

20-HA-102

Amine Strip Column Reboiler

IL1

20-HA-111

Regeneration Gas Heater

IL1 IL1

20 HC-111 20-VL-112

Regeneration Gas Cooler Regeneration Gas KO Drum

IL1

20-KF-111

1

500

5000

110

-10.44

0.98

Feed Gas Feed Gas

1

520

5500

109.28

39.9

1 1 2

1150 800 1400

23800 1930 4225

110 107.8 109

32 25.29 25.25

1

500

107.8

28

1

1400

4000

109

25

1

381.4

5500

108.2

28

2 1

2000 350

1400

107.1 105.6

180 25

Regeneration Gas Blower

1

3000

105

24.97

IL1 IL1 IL1

Fittings Flanges Flanges

38 41 5

2" 1-6" 14-28"

150 600

109 109 109

60 60 60

Feed Gas Feed Gas Feed Gas

IL1 IL1

Pipelines <=3" Pipelines 4-8"

50 200

<=3" 4-8"

80 150

109 109

60 60

Feed Gas Feed Gas

IL1 IL1

Valves non-actuated Values actuated

75 25

1-6" 1-6"

150 150

109 109

60 60

Feed Gas Feed Gas

24.72 0.97 6.5

Feed Gas Feed Gas Feed Gas

103 small 2365.5 93 1243

Feed Gas

small

Feed Gas

small

Regeneration Gas 0.13

R eg en er at io n G as R egen er at ion Ga s

sm al l 1 2. 2

R egen er at ion Ga s

1 2. 2

in IL11 as leak source counted in IL11 as leak source counted

Total 4000

NG Liquefication Gas (IL2A) IL2A

25-HX-101

Feed Gas Precooler

1

2100

8000

1.21

Feed Gas

107.19

28

107 107

-27 -27

Precooled Feed Gas Precooled Feed Gas

IL2A IL2A

Fittings Flanges

1 16

2" 1-6"

IL2A

Pipelines 4-8"

50

4-8"

150

107

-27

Precooled Feed Gas

IL2A

Valves non-actuated

2

1-6"

150

107

-27

Precooled Feed Gas

190

in IL6A1 as leak source counted

Total 190

NG Liquefication Liquid System (IL2B) IL2B

25-HX-102

Feed Gas Liquefier

-113.92 1

IL2B

25-HX-103

1900

16000

1098 -113.92

1200

10000

Liquid Feed Gas

102.5

Feed Gas Subcooler 1

2.8

0.9

Liquid Feed Gas

102.5

406

IL2B IL2B

Fittings Flanges

2 4

2" 1-6"

150

103 103

-114 -114

Liquid Feed Gas Liquid Feed Gas

IL2B IL2B


4 1

1-6" 1-6"

150 150

103 103

-114 -114

Liquid Feed Gas Liquid Feed Gas

IL2B

Pipelines 4-8"

110

4-8"

150

103

-112

Liquid Feed Gas Total 1504

in I6A1 as leak source counted (refer remarks 25-HX-101) in I6A1 as leak source counted (refer remarks 25-HX-101)

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Appendix C: Equipment Count Unit

Eq-ID

Name

Number

Diameter

Diameter Length

inch

mm

mm

25.08.2008 Page 96 Operating Operating Volumen pressure tempeature bara °C m³

Stream content

Inventory

Remarks

kg

LNG Storage (IL3B) IL3B

16 bara LNG Rundown Line

IL3B

Fittings

IL3B

Pipelines 4-8"

IL3B IL3B

Valves non-actuated Values actated

IL3B

1 bara Flanges

150 1

200000 16

-160

LNG

2"

50

16

-160

LNG

200

4-8"

150

16

-160

LNG

9 1

1-6" 1-6"

150 150

16 16

-160 -160

LNG LNG

20

1-6"

150

1

-160

LNG

1600

Total 1600

LNG Truck Loading (IL4) 5 bara IL4

LNG Truck Loading Line

100

250000 5

-161.5

IL4 IL4

Fittings Flanges

5 24

IL4 IL4


IL4 IL4

1.96

LNG

2" 1-6"

50 150

5 5

-161.5 -161.5

LNG LNG

150 100

2" 4-8"

50 100

5 5

LNG LNG


36 9

1-6" 1-6"

150 150

5 5

-161.5 -161.5 -161.5 -161.5 -161.5

IL4

1 bara Pipelines <=3"

10

<=3"

50

1

-160

LNG

IL4

Valves non-actuated

1

1-6"

150

1

-160

LNG

900

LNG LNG

Total 900

LNG Ship Loading Tank Top (IL5A) IL5A

LNG Ship Loading Tank Top

IL5A IL5A IL5A


2 20 10

2" 1-6" 8-12"

IL5A

Valves non-actuated Valves non-actuated Valves actuated Valves actuated

27 4 6 2

1-6"

IL5A IL5A IL5A

8-12" 1-6" 8-12"

5

-161.5

LNG

50 150 300

5 5 5

-161.5 -161.5 -161.5

LNG LNG LNG

150 300 150 300

5 5 5 5

-161.5

LNG

-161.5 -161.5 -161.5

LNG LNG LNG

1500

Total 1500

LNG Ship Loading Line (IL5B1/B2/B3) IL5B1/B2/B3

LNG Ship Loading Line

IL5B1/B2/B3 IL5B1/B2/B3


110 110

3" 9-12"

325

340000 5

-161.5

80 300

5 5

-161.5 -161.5

28.21

LNG

127800

LNG LNG Total 127800

LNG Ship Loading Jetty (IL5C) IL5C

LNG Ship Loading Jetty

IL5C IL5C

Fittings Flanges

2 10

2" 8-12"

IL5C IL5C


2 1

8-12" 8-12"

5

-161.5

LNG

50 300

5 5

-161.5 -161.5

LNG LNG

300 300

5 5

-161.5 -161.5

LNG LNG

127800

Total 127800

Refrigeration Gas_4 bar System (IL6A1) I L6A 1

IL6A1

2 5- VL- 101

25-HX-101

C ycl e co mp ressor S uc ti on D rum

1

2600

7960

3.79

19.42

53.7

Feedgas Precooler

Mixed LP Refrigerant 1128

Mixed LP Refrigerant 1

2100

8000

4

20

27.71

158

IL6A1 IL6A1 IL6A1 IL6A1

Fittings Flanges Flanges Flanges

2 15 7 10

2" 1-6" 1-6" 30"

50 50 150 750

4

20

Mixed LP Refrigerant

4 4

20 20

Mixed LP Refrigerant Mixed LP Refrigerant

IL6A1 IL6A1

Pipelines 18-28" Pipelines 30"

20 40

20" 30"

500 750

4 4

20 20


IL6A1 IL6A1 IL6A1 IL6A1 IL6A1

Valves non-actuated Valves non-actuated Valves non-actuated Valves non-actuated Values actuated

4 2 1 2 2

1-6" 1-6" 8-12" 30" 1-6"

50 150 300 750 150

4 4 4 4 4

20 20 20 20 20

Mixed LP Refrigerant Mixed LP Refrigerant Mixed LP Refrigerant Mixed LP Refrigerant Mixed LP Refrigerant Total 1286

25-HX101/102/103 in this inventory loop as Vessel in leak counted


Eq-ID

Name

Number

Diameter

Diameter Length

inch

mm

mm


Stream content

Inventory

Remarks

kg

Refrigeration Gas_18 bar System (IL6A2) IL6A2 IL6A2

25-VL-102 Cycle Compressor Interstage Drum 1 2 25-KA-101 Cycle Compressor

I L6A 2

25- HC -1 01

C yc le Co mpr esso r I nt er coo le r

2600 3500

4

7320

17.6 17.6

20.95 20.95

38.86

Mixed LP Refrigerant 933 Mixed LP Refrigerant 1128

17.6

4000

small

IL6A2 IL6A2 IL6A2 IL6A2 IL6A2

Fittings Flanges Flanges Flanges Flanges

2 12 9 5 9

2" 1-6" 8-12" 14-28" 14-28"

50 150 300 400 500

18 18 18 18 18

20 20 20 20 20

IL6A2 IL6A2

Pipelines 13-18" Pipelines 18-28"

20 50

16" 20"

400 500

18 18

20 20


IL6A2 IL6A2 IL6A2 IL6A2 IL6A2 IL6A2

Valves non-actuated Valves non-actuated Valves non-actuated Values actuated Values actuated Values actuated

4 1 2 2 1 1

1-6" 14-28" 14-28" 1-6" 8-12" 14-28"

150 400 500 150 300 450

18 18 18 18 18 18

20 20 20 20 20 20

Mixed LP Refrigerant Mixed LP Refrigerant Mixed LP Refrigerant Mixed LP Refrigerant Mixed LP Refrigerant Mixed LP Refrigerant

Mixed LP Refrigerant Mixed LP Refrigerant Mixed LP Refrigerant

Total 2061

Refrigeration Gas_40 bar System (IL6A3) I L6A 3

25- HC -1 02

C yc le Com pr esso r A ft er coo ler

IL6A3

25-VA-101

Cycle HP Separator

IL6A3

25-VA-102

Cold MRC Separator

2

6000

1

2000

1

2000

IL6A3 IL6A3

Fittings Flanges

3 18

2" 8-12"

50

IL6A3 IL6A3

Pipelines 4-8" Pipelines 9-12"

20 60

8" 12"

200 300

IL6A3 IL6A3


4 2

8-12" 8-12"

41.03 40.53

3860 8470

12"

12" 12"

76.56 25 -26.5

40

Light HP Refrigerant

small

1 2. 13

L ig ht H P R ef ri ge ra nt

7 03 .5

12.1

Light HP Refrigerant

40 40

-26.5 -26.5

Light HP Refrigerant Light HP Refrigerant

40 40

-26.5 -26.5


40 40

-26.5 -26.5


642

in IL6B as leak source counted

Total 1345.5

Refrigeration Liquid 25-HX-101/25-HX-103 System (IL6B1) IL6B1

25-HX-101

Feed Gas Precooler

40.2 1

IL6B1

25-HX-103

2100

-25.6

33.2 1200

-25.6

5.5

3086

in I6A1 as leak source counted (refer remarks 25-HX-101)

Mixed Refrigerant

10000

IL6B1 IL6B1 IL6B1


3 14 4

2" 1-6" 8-12"

50 150 300

40.2 40.2 40.2

-25.6 -25.6 -25.6

Mixed Refrigerant Mixed Refrigerant Mixed Refrigerant

IL6B1 IL6B1 IL6B1

Pipelines 4-8" Pipelines 9-12" Pipelines 9-12"

25 10 5

4-8" 9-12" 9-12"

150 250 300

40.2 40.2 40.2

-25.6 -25.6 -25.6

Mixed Refrigerant Mixed Refrigerant Mixed Refrigerant

IL6B1 IL6B1


2 1

1-6" 1-6"

150 150

40.2 40.2

-25.6 -25.6

Mixed Refrigerant Mixed Refrigerant

10

962

in I6A1 as leak source counted (refer remarks 25-HX-101)

Mixed Refrigerant

8000

Feed Gas Subcooler 1

1.6

Total 4050

Refrigeration Liquid 25-VA-101 System (IL6B2) IL6B2

25-VA-101

Cycle HP Separator

40.53 1

2000

Flanges

7

Propane Tank Liquid Propane Drier Liquid Propane Filter

1 1 1

IL7 IL7 IL7


7 36 1

2" 1-6" 14-28"

IL7

Pipelines <=3"

15

IL7 IL7


12 3

IL6B2

3860

100

1-6"

1 2. 13

M ix ed R ef ri ge ra nt

25 40.2

5700

25

Propane Storage (IL7) IL7 IL7 IL7

58-VS-101 58-VK-101 58-LF-101

2100 390 500

7500 1240

8 8 8

15 15 15

26

50 50 600

8 8 8

15 15 15

Propane Propane Propane

2"

50

8

15

Propane

1-6" 1-6"

50 50

8 8

15 15

Propane Propane

0.15 0.2

Propane Propane Propane

18395 106 100

Total 18600

Pentane Storage (IL8) Pentane Tank Liquid B utane/Pentane D rier Liquid Butane/Pentane Fi lter

1 1 1

IL8 IL8 IL8


11 74 1

2" 1-6" 14-28"

IL8

Pipelines <=3"

15


20 4

IL8 IL8 I L8

IL8 IL8

58-VS-102 58-VK-102 58-LF -102

1800 390 500

5400 1240

8 8 8

14.84 14.84 14.84

13.74 0.15

50 150 600

8

14.84

Pentane Pentane Pentane

3"

80

8

14.84

Pentane

1-6" 1-6"

150 150

8 8

14.84 14.84

Pentane Pentane

Pentane Pentane Pentane

8656 87 small

Total 8743

in IL6A1 as leak source counted


Eq-ID

Name

Number

Diameter

Diameter Length

inch

mm

mm


Stream content

Inventory

Remarks

kg

Butane Storage (IL9) IL9

58-VS-103

Butane Tank

IL9 IL9

58-VK-103 58-LF-103

Liquid Batane Drier Liquid Batane Filter

1 1 1

IL9 IL9 IL9


4 45 1

2" 1-6" 14-28"

IL9

Pipelines <=3"

15

IL9 IL9


15 3

1750 390 500

9.5 0.15

Butane

8 8 8

14.84 14.84 14.84

150 600

8

14.84

Butane Butane Butane

3"

80

8

14.84

Butane

1-6" 1-6"

150 150

8 8

14.84 14.84

Butane Butane

1240

Butane

5482 86 small

Total 5568

Ethylene Storage Gas System (IL10A) IL10A IL10A

58-VK-104 58-LF-104

Liquid Ethylene Drier Liquid Ethylene Filter

IL10A

58-VS-104

Ethylene Tank

1 1

390 500

1240 1000

8 8

1

4900

2800

8

-3 -3

0.15 0.2 -58.4

Ethylene Ethylene Ethylene

3

IL10A IL10A

Fittings Flanges

2 16

2" 1-6"

50 150

8 8

-58.4 -58.4

Ethylene Ethylene

IL10A

Pipelines <=3"

30

3"

80

8

-58.4

Ethylene

IL10A

Valves non-actuated

5

1-6"

150

8

-58.4

Ethylene

1 1 16

Total 18

Ethylene Storage Liquid System (IL10B) Ethylene Tank

1

IL10B IL10B IL10B


10 55 1

2" 1-6" 14-28"

IL10B

Pipelines <=3"

50

IL10B IL10B


19 4

IL10B

58-VS-104

4900

2800

8

-58.4

25

50 150 600

8 8 8

-58.4 -58.4 -58.4

Ethylene Ethylene Ethylene

3"

80

8

-58.4

Ethylene

1-6" 1-6"

150 150

8 8

-58.4 -58.4

Ethylene Ethylene

Ethylene

8350

Total 8350

Hot Oil Systemes (IL11) Am ine Strip Col umn Reboiler Regenaration Gas Heater

1 1

IL11 IL11

Fittings Flanges

10 60

2" 1-6"

IL11


100 100

3" 6"

IL11 IL11


17 4

1-6"

IL11 IL11

20-HA-102 20-HA-111

4 4

185 185

50 150

6 6

192 192

Hot Oil Hot Oil

80 150

6 6

192 192

Hot Oil Hot Oil

6 6

192 192

Hot Oil Hot Oil

381.4

150 6"

5500

6.2 0.63

Hot Oil Hot Oil

2060 446

Total 2500

Feedgas Fiscal Metering (IL12) 110

20

Feed Gas

50 200

110 110

20 20

Feed Gas Feed Gas

8"

200

110

20

Feed Gas

8"

200

110

20

Feed Gas

IL12

Feedgas Fiscal Metering Package

IL12 IL12

Fittings Flanges

5 8

2" 8"

IL12

Pipelines 4-8"

40

IL12

Valves non-actuated

8

150

Total 150

Tailgas Fiscal Metering (IL13) 10

20

Tailgas

50

10

20

Tailgas

3"

80

10

20

Tailgas

3"

80

10

20

Tailgas

IL13

Tailgas Fiscal Metering Package

IL13

Fittings

3

2"

IL13

Pipelines <=3"

20

IL13

Valves non-actuated

8

2

Total 2

in IL10B as leak source counted

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Appendix D: Results of LEAK 3.2 Calculations

12.0

Appendix D: Results of LEAK 3.2 Calculations

The resulting leakage frequencies shown in the subsequent tables:

25.08.2008 Page 99


25.08.2008 Page 100

IL1: Feedgas Purification IL6B1:Refrigerant Liquid 25-HX-101/103 System Total Leaks Repr. Size Total Leaks Repr. Size Category Gas [ / year] Liquid [ / year] [mm] Category Gas [ / year] Liquid [ / year] [mm] small 0.0429422 0.000E+00 3.432 small 0.000E+00 7.756E-04 3.378 medium 6.211E-03 0.000E+00 23.376 medium 0.000E+00 1.094E-04 23.147 large 1.082E-03 0.000E+00 62.175 large 0.000E+00 1.837E-05 59.031 very large (Rupture) 2.038E-03 0.000E+00 206.271 very large (Rupture) 0.000E+00 5.000E-05 133.517 Total 0.05227254 Total 9.535E-04 IL2A: NG Liquefication Gas IL6B2:Refrigerant Liquid 25-VA-101 System Total Leaks Repr. Size Total Leaks Repr. Size Category Gas [ / year] Liquid [ / year] [mm] Category Gas [ / year] Liquid [ / year] [mm] small 2.985E-04 0.000E+00 3.375 small 0.000E+00 7.594E-04 3.376 medium 4.878E-05 0.000E+00 23.152 medium 0.000E+00 1.059E-04 23.159 large 8.937E-06 0.000E+00 62.775 large 0.000E+00 1.763E-05 71.205 very large (Rupture) 3.343E-05 0.000E+00 123.736 very large (Rupture) 0.000E+00 4.516E-05 120.198 Total 3.896E-04 Total 9.281E-04 IL2B1: NG Liquefication Liquid_103 bar System IL6B3:Refrigerant Liquid 25-VA-102/25-HX-102 System Total Leaks Repr. Size Category Total Leaks Repr. Size Category Gas [ / year] Liquid [ / year] [mm] Gas [ / year] Liquid [ / year] [mm] small 0.000E+00 1.134E-03 3.381 small 0.000E+00 5.999E-05 3.336 medium 0.000E+00 1.658E-04 23.138 medium 0.000E+00 8.342E-06 22.981 large 0.000E+00 2.848E-05 53.878 large 0.000E+00 1.404E-06 71.086 very large (Rupture) 0.000E+00 9.344E-05 123.720 very large (Rupture) 0.000E+00 1.161E-05 135.931 Total 1.422E-03 Total 8.135E-05 IL3A: LNG Storage Return Gas IL6B4:Refrigerant Liquid 25-VL-102 System Category Total Leaks Repr. Size Category Total Leaks Repr. Size Gas [ / year] Liquid [ / year] [mm] Gas [ / year] Liquid [ / year] [mm] small 1.373E-02 0.000E+00 3.356 small 0.000E+00 1.660E-03 3.386 medium 1.686E-03 0.000E+00 23.095 medium 0.000E+00 2.401E-04 23.200 large 3.108E-04 0.000E+00 62.622 large 0.000E+00 7.110E-05 57.373 very large (Rupture) ---very large (Rupture) 0.000E+00 7.833E-05 124.105 Total 1.600E-02 Total 2.049E-03 IL3B: LNG Storage IL7_: Propane Storage Category Total Leaks Repr. Size Category Total Leaks Repr. Size Gas [ / year] Liquid [ / year] [mm] Gas [ / year] Liquid [ / year] [mm] small 0.000E+00 2.312E-03 3.414 small 0.000E+00 5.668E-03 3.448 medium 0.000E+00 3.441E-04 23.320 medium 0.000E+00 9.057E-04 23.430 large 0.000E+00 7.302E-05 66.635 large 0.000E+00 3.172E-04 53.278 very large (Rupture) 0.000E+00 1.648E-04 123.796 very large (Rupture) 0.000E+00 1.632E-04 242.147 Total 2.894E-03 Total 7.054E-03 IL4: LNG Truck Loading IL8: Pentane Storage Category Total Leaks Repr. Size Category Total Leaks Repr. Size Gas [ / year] Liquid [ / year] [mm] Gas [ / year] Liquid [ / year] [mm] small 0.000E+00 9.266E-03 3.441 small 0.000E+00 7.112E-03 3.426 medium 0.000E+00 1.379E-03 23.409 medium 0.000E+00 1.154E-03 23.354 large 0.000E+00 3.043E-04 62.462 large 0.000E+00 2.156E-04 62.170 very large (Rupture) 0.000E+00 5.388E-04 122.702 very large (Rupture) 0.000E+00 4.152E-04 145.496 Total 1.149E-02 Total 8.897E-03 IL5A: LNG Ship Loading Tank Top IL9: Butane Storage Category Total Leaks Repr. Size Category Total Leaks Repr. Size Gas [ / year] Liquid [ / year] [mm] Gas [ / year] Liquid [ / year] [mm] small 0.000E+00 7.257E-03 3,438 small 0.000E+00 6.094E-03 3.447 medium 0.000E+00 1.097E-03 23,405 medium 0.000E+00 9.974E-04 23.440 large 0.000E+00 1.923E-04 68025.000 large 0.000E+00 1.879E-04 65.711 very large (Rupture) 0.000E+00 4.363E-04 137,182 very large (Rupture) 0.000E+00 3.522E-04 154.000 Total 8.983E-03 Total 7.632E-03 IL5B1/2/3: LNG Ship Loading Line IL10A: Ethylene Storage Gas System Category Total Leaks Repr. Size Category Total Leaks Repr. Size Gas [ / year] Liquid [ / year] [mm] Gas [ / year] Liquid [ / year] [mm] small 0.000E+00 3.558E-04 3.383 small 3.576E-03 0.000E+00 3.497 medium 0.000E+00 4.512E-05 23.142 medium 6.222E-04 0.000E+00 23.604 large 0.000E+00 4.773E-05 62.063 large 1.282E-04 0.000E+00 65.953 very large (Rupture) 0.000E+00 1.222E-05 177.546 very large (Rupture) 2.018E-04 0.000E+00 183.126 Total 4.609E-04 Total 4.528E-03

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Appendix D: Results of LEAK 3.2 Calculations IL5C: LNG Ship Loading Jetty Total Leaks Gas [ / year] Liquid [ / year] small 0.000E+00 3.082E-03 medium 0.000E+00 3.952E-04 large 0.000E+00 8.889E-05 very large (Rupture) 0.000E+00 1.196E-04 Total 3.686E-03 IL6A1:Refrigerant Gas_4 bar System Category Total Leaks Gas [ / year] Liquid [ / year] small 2.031E-03 0.000E+00 medium 3.115E-04 0.000E+00 large 7.922E-05 0.000E+00 very large (Rupture) 1.064E-04 0.000E+00 Total 2.528E-03 IL6A2:Refrigerant Gas_18 bar System Category Total Leaks Gas [ / year] Liquid [ / year] small 2.750E-02 0.000E+00 medium 3.394E-03 0.000E+00 large 5.271E-04 0.000E+00 very large (Rupture) 7.534E-04 0.000E+00 Total 3.218E-02 IL6A3:Refrigerant Gas_40 bar System Total Leaks Category Gas [ / year] Liquid [ / year] small 4.531E-03 0.000E+00 medium 7.034E-04 0.000E+00 large 1.241E-04 0.000E+00 very large (Rupture) 2.251E-04 0.000E+00 Total 5.583E-03 Category

Repr. Size [mm] 3.386 23.186 60.189 177.616

Repr. Size [mm] 3.389 23.216 57.637 218.910

Repr. Size [mm] 3.355 23.104 69.693 489.209

Repr. Size [mm] 3.440 23.390 65.120 494.431

25.08.2008 Page 101

IL10B: Ethyleme Storage Liquid System Category Total Leaks Repr. Size Gas [ / year] Liquid [ / year] [mm] small 0.000E+00 4.118E-03 3.402 medium 0.000E+00 6.274E-04 23.253 large 0.000E+00 1.290E-04 60.076 very large (Rupture) 0.000E+00 2.452E-04 127.720 Total 5.120E-03 IL11: Hot Oil System Category Total Leaks Repr. Size Gas [ / year] Liquid [ / year] [mm] small 0.000E+00 3.982E-03 3.398 medium 0.000E+00 6.000E-04 23.236 large 0.000E+00 1.050E-04 60.093 very large (Rupture) 0.000E+00 2.655E-04 124.629 Total 4.953E-03 IL12: Feedgas Fiscal Metering Category Total Leaks Repr. Size Gas [ / year] Liquid [ / year] [mm] small 3.089E-03 0.000E+00 3.431 medium 4.402E-04 0.000E+00 23.365 large 7.439E-05 0.000E+00 56.344 very large (Rupture) 3.244E-04 0.000E+00 143.970 Total 3.928E-03 IL13: Tailgas Fiscal Metering Total Leaks Repr. Size Category Gas [ / year] Liquid [ / year] [mm] small 2.285E-03 0.000E+00 3.435 medium 3.239E-04 0.000E+00 23.382 large 2.496E-04 0.000E+00 58.160 very large (Rupture) ---Total 2.859E-03

Category These names resemble the respective hole size categories are listed in Assumption Sheet HC-2. Total Leak Freq. The leak frequency calculated for the category. This is the frequency of leaks in the hole size range defined for the category. Repr. Size This column will gives the weighted mean value for the range of hole sizes in the category.

The database in LEAK is used for unclean services and offshore facilities which result in high leak frequencies. To reflect the Lyse LNG Base Load Plant, which is considered a clean service and an onshore facility, leak frequencies for pipelines, vessels and the LNG Storage Tank have been changed to the leak frequencies in given in the Purple Book. Accordingly, failures of flanges are assumed to be included in the failure frequency of the pipeline. A pressure vessel (VESSEL_PROC) leak frequency for small leaks (Ø10mm) of 1E-5 / year consists of the vessel wall and the welded nozzles, mounting plates and instrumentation pipes. This reduces the leak frequency by deleting the flanges on pipelines and vessels (tanks). A comparison between the LEAK and the Purple Book leak frequencies for pipes and process vessel are included below in Table 23.


25.08.2008 Page 102

Table 23: LEAK Database used for the QRA DNV LEAK Program Database

LINDE LEAK Program Modification [Purple Book]

Equipment

Descripti on

Ø 1 mm Failure Rate [ /year]

Full Rupture Failure Rate [ /year]

COMP_CENTR

Centrifugal compressor.

0.0106109

COMP_RECIPR

Reciprocating compressor.

FILTER

Description

Remarks



0

0.0106109

0

No changes

0.0710059

0

0.0710059

0

No changes

Solid from liquid filter.

0.0037615

0

0.0037615

0

No changes

FLANGE

Flanged joint, all diameters.

5.15E-5

5E-6

5.15E-5

5E-6

No changes

HEATEX_FINFAN

Heat exchanger finfan cooler.

0.00187091

0

0.00187091

0

No changes

HEATEX_PLATE

Heat exchanger plate.

0.0104712

0

0.0104712

0

No changes

HEATEX_SH&T_ HCinshell

Heat exchanger shell and tube, hydrocarbon in shell.

0.00413223

0

0.00413223

0

No changes

HEATEX_SH&T_ HCintube

Heat exchanger shell and tube, hydrocarbon in tube.

0.00340633

4E-5

0.00340633

4E-5

No changes

PIGTRAP

Pig trap (launcher and receiver).

0.00701052

0

0.00701052

0

No changes

PIPE_PROC

Process pipe (1m), all diameters.

5.95E-5

5E-6

Process pipe (1m), Ø < 75 mm, outflow is from a leak with an effective Ø of 10% of the nominal Ø, a max. of 50 mm

5E-6

1E-6

to PIPE_PROC_LINDE <75 mm changed

Process pipe (1m), 75 mm <= Ø <= 150 mm, outflow is from a leak with an effective Ø of 10% of the nominal Ø, a max. of 50 mm

2E-6

3E-7

to PIPE_PROC_LINDE 75mm
Process pipe (1m), Ø > 150 mm, outflow is from a leak with an effective Ø of 10% of the nominal Ø, a max. of 50 mm

5E-7

1E-7

to PIPE_PROC_LINDE >150 mm changed

PUMP_CENTR

Centrifugal pump, singel and double seal.

0.00755287

0

0.00755287

0

No changes

PUMP_RECIPR

Reciprocating pump, singel and double seal.

0.0071644

0

0.0071644

0

No changes

SMALL_BORE_F IT

Small bore fitting.

5.9E-4

0

5.9E-4

0

No changes

VALVE_MAN

Non-actuated valve, all diameters.

2E-8

2E-6

2E-8

2E-6

No changes


DNV LEAK Program Database

25.08.2008 Page 103

LINDE LEAK Program Modification [Purple Book]

Remarks

Equipment

Descripti on



Description



VESSEL_PROC

Process pressure vessel incl. reactors and columns.

0.00211501

4E-5

Pressure vessel, Frup: instantaneous release of the complete inventory

1E-5

5E-7

to VESSEL_PRES_LINDE changed

VALVE_ACT_NO Actuated nonN_P/L pipeline valve, all diameters.

0.00109165

1E-5

0.00109165

1E-5

No changes

VESSEL_STOR_ Atmospheric ATM storage vessel.

0.00535818

0

1E-4

5E-6

to VESSEL_STOR_ATM_ LINDE changed

A single-containment atmospheric tank consists of a primary container for the liquid. An outer shell is either present, or not, but when present, primarily for the retention and protection of insulation.

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Appendix E: Individual Risk Ranking Report

13.0

25.08.2008 Page 104

Appendix E: Individual Risk Ranking Report

This is the risk ranking report produced by PHAST RISK for Individual Risk and illustrates the risk for 1 person, present at 1 point continously, 24 hours a day, 7 days a week during a whole year (52 weeks). It provides the contribution of the individual scenarios. These scenarios are specified as ILXXYZ, where "IL" stands for "Inventory Loop", XX for the respective inventory loop number (1-13) and Y for gas or liquid and Z, if there, a divided part in each inventory loop, respectively. For example " IL5B1" is for ship loading inventory (IL5) a liquid line (B), which is divided in 3 parts (1 part describes a 110 m). The risk ranking points, for which the individual risk has been calculated, are shown in the subsequent figures (blue dots).

Figure 27: Individual Risk Ranking Points (Administration Building, Peninsula, Hiking Track, LNG Ship Bridge, Process Area, Substation, Offsite Ferry, Offsite Companys, Offsite Living Quarters)


25.08.2008 Page 105


25.08.2008 Page 106


25.08.2008 Page 107


25.08.2008 Page 108


25.08.2008 Page 109


25.08.2008 Page 110


25.08.2008 Page 111


25.08.2008 Page 112


25.08.2008 Page 113


25.08.2008 Page 114


25.08.2008 Page 115


25.08.2008 Page 116


25.08.2008 Page 117


25.08.2008 Page 118


25.08.2008 Page 119


25.08.2008 Page 120


25.08.2008 Page 121


25.08.2008 Page 122


25.08.2008 Page 123


25.08.2008 Page 124


25.08.2008 Page 125


25.08.2008 Page 126

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Appendix F: Details on the Analysis Procedure

14.0

25.08.2008 Page 127

Appendix F: Details on the Analysis Procedure

Until now, the discussion on the QRA methodology has been very generic. The following discussion is intended to provide more insight into the actual QRA methodology. An electronic database of approximately 1200 materials is available to the PHAST / PHAST RISK software, with the material properties regularly reviewed. The PHAST RISK risk modelling software requires the following inputs to be able to produce risk results: •

•

Import an electronic map of the study area, on which individual fatality risk contour results may be produced The electronic map may be programmed in PHAST RISK to: −

−

−

−

•

•

•

•

•

Superimpose all on-site and off-site populations within the study area by location, and specifying the day / night number of people for each location Superimpose all potential ignition sources within the study area, which may cause delayed ignition of a flammable release Delayed ignition sources may be specified as point sources (e.g. flares and fired heaters etc.), area sources (e.g. parking area etc.) or line sources (e.g. traffic roads etc.). Each ignition source carries additional specification in terms of presence factor and ignition source strength (probability of ignition per unit time, when in contact with a flammable vapour cloud between LFL and UFL. The actual delayed ignition probability of any release is calculated by PHAST RISK, based on the dispersion modelling results and event duration The immediate ignition probability associated with each failure case of flammable fluid is a risk analyst programmed value, based on historical ignition data, which varies with leak size and release phase (gas / liquid / 2-phase) (the larger the leak vapour flow rate, the higher the ignition probability, typically varying from 1% to 30%, unless above auto ignition, then 100%)

Prepare and import weather class, wind speed and wind direction probability data for the study area. For day and night due to lack of data the same weather class, wind speed and wind directional probability are used in PHAST RISK Enter all identified failure cases, which are defined in terms of: Location, Material released, Quantity released (or release duration), Temperature, Pressure, Leak size, Leak direction (e.g. horizontal, vertical), Leak elevation, Leak frequency and Immediate ignition probability Each failure case calculation in PHAST RISK starts with discharge modelling. Based on release duration and release phase (gas, liquid, 2-phase), PHAST RISK directs the dispersion and consequence calculations to one of 4 alternate, built-in consequence outcome event trees (continuous vapour release, continuous release with rain-out, instantaneous vapour release, instantaneous release with rain-out), where each event tree branch probability carries default values, which may be re-programmed by the risk analyst. PHAST RISK will then calculate all alternate consequence outcomes (e.g. jet fire, explosion) of the event tree selected, in terms of hazard range and event duration (where applicable), for each weather class / wind speed combination So far the calculations performed in PHAST RISK only relate to the alternate consequence outcomes and the consequence hazard ranges, for each specified failure case. To produce risk results, PHAST RISK will perform impact frequency calculations, using


25.08.2008 Page 128

the failure case specified leak frequency as starting point. Frequency aspects of the risk calculations relate to the: −

−

−

−

−

−

•

•

•

•

•

•

•

•

•

•

Defined failure case leak frequency Weather class, wind speed and wind directional probability, for each of the 8 wind directions Specified immediate ignition probability and PHAST RISK calculated delayed ignition probability. The delayed ignition probability calculation is based on the strength and location of all specified ignition sources and the failure case dispersion hazard range, combined with vapour cloud persistence (duration). Selected event tree branch probabilities, for each alternate consequence out come. Impact probability for each alternate consequence outcome. This is based on the PHAST RISK calculated magnitude of each consequence and the PHAST RISK default impact probability criteria or risk analyst specified impact criteria for that type of consequence. Location and number of people (or equipment) within hazard area for societal risk results, with separate calculations for day and night, indoors and outdoors.

PHAST RISK performs its individual and societal risk calculations based on a 200 x 200 grid (40,000 points), with the grid point spacing automatically varied, based on the consequence hazard range results. For each release case, PHAST RISK takes the failure case release frequency as initial input, multiplies this by the first weather class / wind speed probability, for the first of 8 wind directions. PHAST RISK takes this result and multiplies it by the immediate ignition probability, while also separately multiplying this result by the PHAST RISK calculated delayed ignition probability. These 2 results are multiplied by the first of the event tree consequence branch probabilities, relating to immediate or delayed ignition branch path. PHAST RISK takes the PHAST calculated consequence hazard range and verifies which grid points are within the consequence hazard area. For each grid point within range PHAST RISK then calculates the magnitude of the consequence at each grid point (e.g. explosion overpressure at a particular grid point may be 3 psi (200 mbar)). The calculated consequence magnitude at each grid point is then compared to the PHAST RISK programmed impact criteria level, and the likelihood of fatality or damage calculated, based on the impact probability criteria specified in PHAST RISK, for the type of consequence and the magnitude of the consequence. This calculation is repeated for each event tree alternate consequence outcome at each grid point, for that weather class / wind speed and wind direction, and the result added to the previous risk level, at each grid point. The above calculations are then repeated for each of the 8 wind directions, cumulatively adding to the risk level at each grid point. The above calculations are repeated for all day / night weather classes, wind speeds and wind directions, cumulatively adding these risk results at each grid point. Once all risk calculations at these grid points have been completed for the first failure case, the next failure case will be calculated, again adding all results cumulatively at each grid point. This is repeated until all failure cases have been calculated, while


25.08.2008 Page 129

PHAST RISK also tracks the risk contribution made by each failure case at each grid point. •

•

Once completed, PHAST RISK produces individual risk contour results by linking points of equal risk, based on the pre-specified individual fatality risk (or equipment damage) criteria levels, and using linear interpolation between relevant grid points. The risk contour results are super imposed on the electronic site map, entered in the PHAST RISK software. PHAST RISK can also produce societal risk results by comparing the calculated level of individual risk at all 40,000 grid points, and combining this with the number of people indoors and outdoors, entered by the risk analyst by location.

The above discussion demonstrates that the meteorological data, ignition data and population data entered into the PHAST RISK software are critical to the risk results. Note that by default the risk modelling within PHAST RISK aims to produce offsite fatality risk results. This is achieved by the build-in but programmable parameter settings, which include: •

•

•

Indoor & outdoor people fatality impact criteria levels, for each alternate consequence outcome. For flammable releases the alternate consequences would be spill fires, fire balls, jet fires, flash fires and vapour cloud explosions (VCEs), each with pre-defined values for the impact levels that will affect people. For fire ball exposure this is based on the Eisenberg Probit equation. For spill fires and jet fires a single criterion flammable dose of 250 KJ is used, with the radiation impact level set at 9.8 kW/m 2, corresponding to 1 % lethality in 20 seconds. For flash fires the 0.5 LFL envelope is used and for VCE overpressure two impact criteria levels are used, 1.5 psi (0.1 barg) and 5 psi (0.34 barg). 4 built-in event trees (Continuous No Rain Out; Continuous With Rain Out; Instantaneous No Rain Out; Instantaneous With Rain Out) that are automatically selected based on the type of material and the release conditions. Each event-tree assigns a ‘split’ between alternate consequence outcomes (spill fires, fire balls, jet fires, flash fires, VCEs and no hazard), based on the immediate ignition, delayed ignition and no ignition probabilities. People vulnerability criteria, which pre-determines the fraction of fatalities resulting indoor & outdoor from being exposed to specific consequence outcomes for a specified duration, or to one or more specified criteria levels. Most of these values go back to data provided in the Dutch "Purple Book" [5].

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Appendix E: Details on the Analysis Procedure

25.08.2008 Page 130

Figure 28: SAFETI Default Vulnerability Parameters

By default PHAST RISK uses (programmable) blast overpressure levels of 1.5 psi (110 mbar) and 5 psi (approx. 345 mbar) for assessing indoor and outdoor offsite fatality impact. These overpressure levels correspond to light damage and total destruction of ordinary brick (residential) housing. The default blast criteria levels need to be modified for assessing the impact of overpressure on prefabs, brick and concrete buildings. In particular, it is known: •

Wood or corrugated asbestos panels shatter at 1 – 2 psi (0.07 – 0.14 barg) overpressure,

•

Partial collapse brick housing at 2 psi (0.14 barg) overpressure,

•

Concrete walls (non-reinforced) shatter at 2 – 3 psi (0.14 – 0.21 barg),

•

Complete destruction of houses at 5 – 7 psi (0.34 – 0.48 barg),

•

Blast proof concrete control buildings fail at 10 psi (0.68 barg) overpressure.

Consistent with these data, Linde (expert knowledge from DNV) set the PHAST RISK explosion parameters to achieve the following overpressure risk results: •

1.5 psi (0.1 barg) - Prefabs severely damaged by overpressure,

•

3 psi (0.2 barg) – Brick buildings severely damaged by overpressure,

•

5 psi (0.34 barg) – Reinforced concrete buildings (non blast proof) severely damaged by overpressure.

The default PHAST RISK hydrocarbon vapour cloud explosion efficiency is set to 10 percent, consistent with the objective to produce conservative offsite fatality risk estimates. This is an overestimation compared to historical data. The Dutch Government Coloured Book Risk Assessment Guidelines reports historical Hydrocarbon Explosion Efficiencies ranging from 2 percent to 5 percent. To be less conservative than the PHAST RISK defaults, DNV has used the upper estimate of 5 percent explosion efficiency in the PHAST RISK Explosion Parameters.

Quantitative Risk Analysis, Lyse LNG Base Load Plant, Train 1 Appendix E: Details on the Analysis Procedure

25.08.2008 Page 131

Early ignition vapour cloud explosions are modelled in PHAST RISK at the centre of release. For delayed ignition the PHAST RISK explosion modelling results are (conservatively) based on locating the centre of the explosion at the maximum (theoretical) displacement distance. For delayed ignition the PHAST RISK explosion modelling takes into account the defined ignition sources. PHAST RISK will first perform discharge calculations and dispersion modelling. Where a dispersing vapour cloud contacts one or more of the defined ignition sources, PHAST RISK will then calculate the time varying ignition probability at that time. The explosion mass is calculated based on the mass of vapour between UFL and LFL at that time. PHAST RISK allows to choose the location for the epicentre of a delayed explosion, where the location is defined in relation to the dimensions of the cloud. If Cloud Front (LFL Fraction) is chosen, the epicentre will be located at the furthest location downwind at which the concentration is equal to the LFL fraction to finish, set in PHAST RISK's Flammable Parameters. If one chooses Cloud Centroid , the epicentre will be located at the centre of the cloud for an instantaneous release, and at the cloud centroid for a continuous release, where the downwind ( x) location of the centroid is obtained by taking a weighted average of the centre point of each release segment. If one chooses Cloud Front (LFL) , the epicentre will be located at the furthest location downwind at which the concentration is equal to the LFL. All three options are very conservative when compared to historical explosion incidents, where explosion damage is most often reported around the area of the release. If PHAST RISK default options are used, the explosion centres would (most often) be modelled as located outside the process units in open areas, which are very unlikely to give rise to explosions. To obtain more accurate and realistic explosion modelling results for this QRA, ignition location was set to the cloud centroid.

187706066 QRA Lyse LNG Base Load Plant Train 1

Recommend Documents