Rapid Risk Assessment Of. Proposed Expansion for the Manufacturing of Pigment and Intermediate Products, Clariant Chemicals (India) Limited

Transcription

1 Rapid Risk Assessment Of Proposed Expansion for the Manufacturing of Pigment and Intermediate Products, Clariant Chemicals (India) Limited Conducted by Cholamandalam MS Risk Services Ltd. (An ISO 9001:2008 Certified Company) Chennai, India April 2017

2 DOCUMENT HISTORY S. No Document Identification No Revision Date Comments / Nature of Changes 1 CCIL/RRA/SR/17-18/ Preparation of original document Prepared By Reviewed By Approved by Venkata Sainath G Nagarajan. R V.S.Bhaskar Cholamandalam Ms Risk Services Limited Clariant Chemicals (India) Limited DISCLAIMER This report is for the sole use by CCIL for the purpose for which they have claimed it is required. We are not responsible to any other person/party for any decision of such person/party based on this report. It is hereby notified that reproduction, copying or otherwise quoting of our report or any part thereof other than the aforementioned purpose, can be done only with our prior permission. The report cannot be used or relied by CCIL for any other purpose or any other third party for any purpose whatsoever and we will not be liable which CCIL may incur on this account including in tort (including but not limited to negligence) arising out of or in connection with this Report. This Report is not intended to identify all hazards which may exist nor is it intended to be an exhaustive review of all possible eventualities. The recommendations for risk improvement contained in the report are purely advisory and the decision and responsibility for implementation rests with CCIL. Page 2 of 32

3 LIST OF ABBREVIATIONS ALARP : As Low As Reasonably Practicable BLEVE : Boiling Liquid Expanding Vapor Explosion CMSRSL: Cholamandalam MS Risk Services Ltd. CR : Catastrophic Rupture EIA : Environment Impact Assessment HSE : Health Safety & Environment IBA : Isobutyl Alcohol IDLH : Immediately Dangerous to Life and Health LFL : Lower flammability limit LOC : Loss of containment MCPA : Mono Chloro Phthalic Acid MOC : Material of Construction NA : Not Applicable NR : Not Reached QRA : Quantitative Risk Assessment RAA : Rapid Risk Assessment TCCPC : Tetra Chloro Copper Phthalocyanine TNT : Tri-Nitro Toluene TPM : Tones per Month UFL : Upper flammability limit UVCE : Unconfined Vapor Cloud Explosion VCE : Vapour cloud explosion Page 3 of 32

4 TABLE OF CONTENTS LIST OF ABBREVIATIONS... 3 EXECUTIVE SUMMARY INTRODUCTION Introduction Scope of the Study Objectives of the Study About the Consultants Software Used FACILITY DESCRIPTION QUANTITATIVE RISK ASSESSMENT METHODOLOGY QRA Methodology Identification of Hazards and Release Scenarios Factors for Identification of Hazards Types of Outcome Events Consequence Calculations Source Strength Parameters Consequential Effects Selection of Damage Criteria Probabilities Failure/ Accident Probabilities Population Probabilities Weather Probabilities Ignition Probabilities Risk Assessment Risk Reduction Measures Modeling Assumptions CONSEQUENCE ANALYSIS Scenarios Inventory Calculations Page 4 of 32

5 4.3 Summary of Consequence Results RISK ANALYSIS Frequency Calculations Individual Risk Societal Risk RISK ACCEPTABILTY CRITERIA EXISTING MEASURES AND RECOMMENDATIONS Existing Control Measures Recommendations REFERENCES Page 5 of 32

6 LIST OF FIGURES Figure 1: Quantitative Risk Assessment methodology Figure 2: Individual risk contour for CCIL Figure 3: Individual risk contours for proposed solvent storage tanks Figure 4: ALARP triangle LIST OF TABLES Table 1: Damage to life due to heat radiation Table 2: Effects due to incident radiation intensity Table 3: Effects due to overpressure Table 4: Approximate population adjacent to facility Table 5: Approximate facility population Table 6: Representative weather classes Table 7: Pasquill stability classes Table 8: Probability of immediate ignition Table 9: Probability of delayed ignition Table 10: Loss of containment scenarios identified for the study Table 11: Process parameters Table 12: Consequence results for CCIL Table 13: Event failure frequency considered for the study Table 14: Risk acceptability criteria Page 6 of 32

7 EXECUTIVE SUMMARY CCIL intends to conduct a Rapid Risk Assessment (RRA) study for expansion of their facility at Cuddalore as part of EIA. This RRA has being executed for expansion of production of pigment & intermediate products by the addition of other pigments and intermediates at their existing facility which are proposed as part of the expansion project. It approached Cholamandalam MS Risk Services Limited and based on the data shared by CCIL at this stage of the project, potential worst-case loss of containment (LOC) scenarios were identified for the RRA study. DNV GL Phast v 7.21 and Safeti v 7.21 software is used for estimation of consequence and risk values. Below mentioned LOC scenarios are identified for the CCIL facility. S. No Scenarios Proposed new Solvents 1 Leak of Diethylene Glycol storage tank 2 CR of Diethylene Glycol storage tank 3 Leak of Mono ethylene Glycol storage tank 4 CR of Mono ethylene Glycol storage tank 5 Leak of Isobutyl Alcohol (IBA) storage tank 6 CR of Isobutyl Alcohol (IBA) storage tank 7 Leak of 350 P solvent storage tank 8 CR of 350 P solvent storage tank 9 Leak Cyclopentanone drum 10 CR of Cyclopentanone drum Individual Risk: The overall individual risk contour is given the figure below. Page 7 of 32

8 Overall Individual Risk contours for proposed storage tanks, CCIL

9 Societal Risk: Considering the storage capacities, process parameters and population densities provided, it is observed that F-N curve for the proposed storage tanks at CCIL facility is not obtained Recommendations 1. Ensure that periodical inspection and thickness measurement are to be carried out for the proposed storage tanks as specified by OISD129-Inspection of Storage Tanks 2. As per OISD 119, ensure bunds provided at the proposed tank farm have proper drainage system. 3. Ensure that foam Pourers are to be made available for the proposed storage tank area as per OISD Portable monitors/foam hose streams shall be provided for fighting fires in dyked area and spills. 5. Spill control kit and procedure shall be in place to contain any spill, clean them up properly and dispose off any containment waste safely. 6. Wind socks are to be installed for knowing wind direction during emergency. 7. Emergency Response & Control Plan to address the critical scenarios are to be tested by conducting mock drills at regular intervals. With reference to the risk acceptance criteria specified by HSE, UK in IS 15656: Code of Practice on Hazard Identification and Risk Analysis it is observed that the risk levels of proposed new storage tanks in the tank farm area is in ALARP region. The risk levels in ALARP are expected to come down to acceptable limits if all the control measures recommended in this report are implemented in addition to the existing risk control measures. It is highly recommended to conduct full scale Quantitative Risk Assessment during subsequent stages of the project with validated data to ascertain the impact of the units.

10 1 INTRODUCTION 1.1 Introduction Clariant is a globally leading company in specialty chemicals having its head quarters in Muttenz, Switzerland. There are four core business areas, Care Chemicals, Catalysts, Natural Resources and Plastics and coatings. Clariant is represented by more than 110 Group companies and employs around people globally. Clariant Chemicals (India) Limited (CCIL) is a part of global Clariant group. CCIL is now India's leading manufacturer and supplier of Pigments and their intermediates. 1.2 Scope of the Study Scope of the RRA study covers the following Installation of storage tanks for proposed solvents Diethylene Glycol, Mono Ethylene Glycol, Cyclopentanone, Isobutyl Alcohol and 350 P solvent. 1.3 Objectives of the Study The Objective of RRA study is Identification of worst case accidental events Assessment of risk arising from the hazards and consideration of its tolerability to personnel, facility and the environment which includes the following o Calculation of physical effects of accidental scenarios. o Identification and quantification of the risks and contour mapping on the layouts. o Evaluation of risk against the risk acceptable limits. o Risk reduction measures to prevent incidents, to control accidents. 1.4 About the Consultants Cholamandalam MS Risk Services Limited (CMSRSL) is a joint venture between the Murugappa Group and Mitsui Sumitomo Insurance Group of Japan. CMSRSL is an approved HSE consultant of Kuwait Oil Company. CMSRSL offers specialized and innovative risk management solutions to clients in India and rest of Asia. CMSRSL has carried out consultancy and training services to over 2500 units / locations of various organizations belonging to 32 industrial sectors including refineries and petrochemical units. In addition to industrial sector and service sector located in Kuwait, India, Hong Kong and Thailand, the Ministry of Environment and Forests, Government of India and insurance companies located in India, Sri Lanka and Singapore have been engaging the services of CMSRSL to carry out a number of risk analysis and specialized safety studies. CMSRSL also carries out similar studies for companies like Kuwait Oil Company, located outside India. The team members have wide experience in risk management studies and have carried out studies for a number of industrial sectors including refineries located in India and rest of Asia. Page 10 of 32

11 1.5 Software Used Phast v7.21 and Safeti v7.21 The software developed by DNV is used for risk assessment studies involving flammable and toxic hazards where individual and societal risks are also to be identified. It enables the user to assess the physical effects of accidental releases of toxic or flammable chemicals. Phast v7.21 is used for consequence calculations and Safeti v7.21 is used for assessing risk. The software contains a series of up to date models that allow detailed modeling and quantitative assessment of release rate pool evaporation, atmospheric dispersion, vapor cloud explosion, combustion, heat radiation effects from fires etc. The software is designed to facilitate compliance with regulatory requirements of many countries, with tailor-made specifications incorporated into the program. The software is developed based on the hazard model given in TNO Yellow Book as well as various incidents that had occurred over past 25 years. CMSRS has used the latest version of DNV software for developing the consequences for each model. 2 FACILITY DESCRIPTION The existing unit was started during the year 1977 in the name of M/s. Vanavil Dyes and Chemicals Ltd. as a public limited company. Until In 1983-M/s Colour-Chem Ltd joined as a co-promoter and in 2006, it was merged with Colour-chem Ltd. Later the company name was changed as Clariant Chemicals (India) Ltd. The facility was involved in the manufacture of Naphthol, Fast Color base and intermediates with a total production capacity of 75 TPM until August The existing facilities at CCIL have capacity of manufacturing 275 TPM (Tones per Month) of Blue pigments and 115 TPM of Intermediates in the existing plant located at Cuddalore. The proposed project involves the production of pigments product such as Violet pigments, Tetra Chloro Copper Phthalocyanine (TCCPC) and one intermediate Mono Chloro Phthalic Acid (MCPA). The proposed expansion will be carried out within the existing production facility of Clariant Chemicals (India) Limited which is located in SIPCOT Industrial Complex, Cuddalore. Violet Pigment is used for pigmentation in various coating industries like automobile industry, printing ink, plastic industries. Due to its specific properties like more stability in solvent, fastness to heat and light, it is strongly recommended for above industries. Tetra Chloro Copper Phthalo cyanine is useful in various areas of Automobile coating and ink industry etc., The facilities considered to conduct RRA are the proposed storage tanks Diethylene Glycol, Mono Ethylene Glycol, Cyclopentanone, Isobutyl Alcohol and 350 P solvent. Page 11 of 32

12 3 QUANTITATIVE RISK ASSESSMENT METHODOLOGY 3.1 QRA Methodology Define the Goal (Statutory, Emergency Planning, Consequence, Etc. Location, Layout, Process Parameters Hazard Identification Quantification of Hazard Select most Credible Scenario Select Worst Case Scenario Frequency Estimation Estimate Consequence Estimate Effect of Damage Estimate Frequency of Occurrence Estimate Risk Is Risk Acceptable? Yes End No Prioritize & Recommend Risk Measure Figure 1: Quantitative Risk Assessment methodology Page 12 of 32

13 3.2 Identification of Hazards and Release Scenarios A hazard is generally realized as a loss of containment of a hazardous material. The routes for such loss of containment can include release from pipe fittings containing liquid or gas, releases from vents, relief and releases from vessel rupture. The containment is defined as one or several devices, any parts which are permanently in open contact with one another, and which are intended to contain one or multiple substances. The following data were collected to envisage scenarios: Composition of materials stored in storage tanks Inventory of materials stored in storage tanks; Storage tanks process conditions (phase, temperature, pressure); Accidental release of flammable liquids/gases can result in severe consequences. Delayed ignition of flammable gases can result in blast overpressures covering large areas. This may lead to extensive loss of life and property. In contrast, fires have localized consequences. Fires can be put out or contained in most cases; there are few mitigating actions one can take once a flammable gas or a vapor cloud gets released. Major accident hazards arise, therefore, consequent upon the release of flammable gases. 3.3 Factors for Identification of Hazards In any installation, main hazard arises due to loss of containment during handling of flammable chemicals. To formulate a structured approach to identification of hazards, an understanding of contributory factors is essential. Inventory Inventory analysis is commonly used in understanding the relative hazards and short listing of release scenarios. Inventory plays an important role in regard to the potential hazard. Larger the inventory of a vessel or a system, larger is the quantity of potential release. A practice commonly used to generate an incident list is to consider potential leaks and major releases from fractures of pipelines and vessels/tanks containing sizable inventories. Parameters Potential vapor release for the same material depends significantly on the operating conditions. This operating range is enough to release a large amount of vapour in case of a leak/rupture, therefore the storage tank leaks and ruptures need to be considered in the risk assessment calculations. Blast overpressures depend upon the reactivity class of material and the amount of gas between two explosive limits. For example, LPG once released and not ignited immediately is expected to give rise to a vapour cloud. These vapours in general have medium reactivity and in case of confinement of the gas cloud, on delayed ignition may result in an explosion and overpressures. Page 13 of 32

14 Initiating Events The range of initiating events and incidents covered affects both complexity of study and number of incident outcome cases. This not only reflects the inclusion of accidents and/or non-accident-initiated events, but also the size of those events. For instance, studies may evaluate one or more of the following: Catastrophic failure of storage tanks, tank trucks and process vessels Large hole (large continuous release) Small hole (continuous release) Leaks at fittings or valves (small continuous release) In general, quantitative studies do not include very small continuous releases or short duration small releases if past experience or preliminary consequence modeling shows that such releases do not contribute to the overall risk levels. MCLS (Maximum Credible Loss Scenario) is described as the worst credible accident or as an accident with a maximum damage distance, which is still believed to be possible is taken into consideration in calculating risk. Selection of Initiating Events and Incidents The selection of initiating events and incidents should take into account the goals or objectives of the study and the data requirements. The data requirements increase significantly when nonaccident initiated events are included and when the number of release size increase. While the potential range of release sizes is tremendous, groupings are both appropriate and necessitated by data restrictions. The main reasons for including release sizes other than the catastrophic rupture are to reduce the conservatism in an analysis and to better understand the relative contributions to risk of small versus large releases. As per Reference Manual Bevi Risk assessment version 3.2, only the Loss of Containment (LOC), which is, basically the release scenarios contributing to the individual and/or societal risk are included in the QRA. There may be number of accidents that may occur quite frequently, but due to proper control measures or fewer quantities of hydrocarbons released, they are controlled effectively. A few examples are a leak from a gasket, pump or valve, release of a chemical from a vent or relief valve, and fire in a pump due to overheating. These accidents generally are controlled before they escalate by using control systems and monitoring devices. 3.4 Types of Outcome Events Considering the present case the outcome possibilities are: Flammable Gas Dispersion Jet fires Pool Fire Vapor Cloud Explosion (VCE) Page 14 of 32

15 Flammable Gas Dispersion (Flash fire) Flammable vapors, after loss of containment, will normally spread in the direction of the wind. If it finds an ignition source before being dispersed to below its Lower Flammability Limit (LFL), a flash fire is likely to result and the flame may travel back to source of the release. Any person caught in a flash fire is likely to suffer fatal burn injuries. Typically the burning zone is defined as ½ LFL and LFL limit back to the release point, even though the vapour concentration might be above UFL. Jet fire Jet fire occurs when a pressurized release (of a flammable gas or vapor) is ignited by any source. They tend to be localized in effect and are mainly of concern in establishing the potential for domino effects and employee safety zones rather than for community risks. The jet fire model is based on the radiant fraction of total combustion energy, which is assumed to arise from a point slowly along the jet flame path. The jet dispersion model gives the jet flame length. Pool fire This represents a situation when flammable liquid spillage forms a pool over a liquid or solid surface and gets ignited. Flammable liquids can be involved in pool fires where they are stored and transported in bulk quantities. Early pool fire is caused when the steady state is reached between the outflow of flammable material from the container and complete combustion of the flammable material when the ignition source is available. Late pool fires are associated with the difference between the release of material and the complete combustion of the material simultaneously. Late pool fires are common when large quantity of flammable material is released within short time. Vapor Cloud Explosion (VCE) Vapor cloud explosion is the result of flammable materials in the atmosphere, a subsequent dispersion phase, and after some delay an ignition of the vapor cloud. Turbulence is the governing factor in blast generation, which could intensify combustion to the level that will result in an explosion. Obstacles in the path of vapor cloud or when the cloud finds a confined area, e.g. as under the bullets, often create turbulence. The VCE will result in overpressures. 3.5 Consequence Calculations In consequence analysis, use is made of a number of calculation models to estimate the physical effects of an accident (spill of hazardous material) and to predict the damage (lethality, injury, material destruction) of the effects. The calculations can roughly be divided in three major groups: a. Determination of the source strength parameters; Page 15 of 32

16 b. Determination of the consequential effects; c. Determination of the damage or damage distances. The basic physical effect models consist of the following Source Strength Parameters Calculation of the outflow of liquid vapors out of a vessel/tank, in case of rupture. Also two-phase outflow can be calculated. Calculation, in case of liquid outflow, of the instantaneous flash evaporation and of the dimensions of the remaining liquid pool. Calculation of the evaporation rate, as a function of volatility of the material, pool dimensions and wind velocity. Source strength equals pump capacities, etc. in some cases Consequential Effects Dispersion of gaseous material in the atmosphere as a function of source strength, relative density of the gas, weather conditions and topographical situation of the surrounding area. Intensity of heat radiation [in kw/ m 2 ] due to a fire or a BLEVE, as a function of the distance to the source. Energy of vapor cloud explosions [in N/m 2 ], as a function of the distance to the distance of the exploding cloud. Concentration of gaseous material in the atmosphere, due to the dispersion of evaporated chemical which may lead to explosion Selection of Damage Criteria The damage criteria give the relation between the extents of the physical effects (exposure) and the effect of consequences. For assessing the effects on human beings consequences are expressed in terms of injuries and the effects on equipment / property in terms of monetary loss. In consequence analysis studies, in principle three types of exposure to hazardous effects are distinguished: 1. Heat radiation due to fires - in this study, the concern is that of Jet fires and pool fires 2. Explosions The knowledge about these relations depends strongly on the nature of the exposure. Following are the criteria selected for damage estimation Heat Radiation The effect of fire on a human being is in the form of burns. There are three categories of burn such as first degree, second degree and third degree burns. The consequences caused by exposure to heat radiation are a function of: Page 16 of 32

17 The radiation energy onto the human body [kw/m 2 ]; The exposure duration [sec]; The protection of the skin tissue (clothed or naked body). The limits for 1% of the exposed people to be killed due to heat radiation, and for seconddegree burns are given in the table below: Table 1: Damage to life due to heat radiation Exposure Duration Radiation energy (1% lethality), kw/m 2 Radiation energy (2 nd degree burns), kw/m 2 Radiation energy (1st degree burns), kw/m 2 10 sec sec Table 2: Effects due to incident radiation intensity Incident Radiation (kw/m 2 ) Type of Damage 4 Causes pain if unable to reach cover within 20 s 12.5 Minimum energy required for melting of plastic 37.5 Sufficient to cause damage to equipment Reference: IS 15656:2006 Hazard identification and risk analysis- Code of practice. The actual results would be less severe due to the various assumptions made in the models arising out of the flame geometry, emissivity, angle of incidence, view factor and others. The radiation output of the flame would be dependent upon the fire size, extent of mixing with air and the flame temperature. Some fraction of the radiation is absorbed by carbon dioxide and water vapor in the intervening atmosphere. Finally the incident flux at an observer location would depend upon the radiation view factor, which is a function of the distance from the flame surface, the observer s orientation and the flame geometry. As per the guidelines of CPR 18 E Purple book: The lethality of a jet fire and pool fire is assumed to be 100% for the people who are caught in the flame. Outside the flame area, the lethality depends on the heat radiation distances. For the flash fires, lethality taken as 100% for all the people caught outdoors and for 10% who are indoors within the flammable cloud. No fatality has been assumed outside the flash fire area. Page 17 of 32

18 Overpressure more than 0.3 bar corresponds approximately with 50% lethality. An overpressure above 0.2 bar would result in 10% fatalities. An overpressure less than 0.1 bar would not cause any fatalities to the public. 100% lethality is assumed for all people who are present within the cloud proper. Table 3: Effects due to overpressure Peak Overpressure Description 0.3 bar Major damage to plant equipment structure 0.1 bar Repairable damage to plant equipment and structure 3.6 Probabilities 0.03 bar Crack in glass Failure/ Accident Probabilities The failure data is taken from CPR 18E Guidelines for Quantitative Risk Assessment, developed by the Committee for the Prevention of Disasters, Netherlands. The impacts due to internal domino effects are not explicitly covered in QRA. An internal domino needs to be considered only in case of a situation in which the failure of one component clearly leads to the failure of another component. As the biggest vessel/ tank are considered for instantaneous failure the impact due to internal domino effects are assumed to get covered in the analysis Population Probabilities It is necessary to know the population exposure in order to estimate the consequences and the risk resulting from an incident outcome. The exposed population is often defined using a population density. Population densities are an important part of a QRA for several reasons. The most notable is that the density is typically used to determine the number of people affected by a given incident with a specific hazard area. Sometimes, population data are available in sketchy forms. In the absence of specific population data default categories can be used. The population density can be averaged over the whole area that may be affected or the area can be subdivided into any number of segments with a separate population density for each individual segment. In this study, based on the data provided by CCIL on nearby facilities On site population details are as follows. Page 18 of 32

19 Table 4: Approximate population adjacent to facility S. No Location Industry Name Population 1 Plant North Side (Outside Population) Tanfac Plant South Side (Outside Stride Shasun & Population) Kudikadu village Plant East Side (Outside Population) Main road 50 4 Plant West Side (Outside Population) Back water 5 Total Outside Population 755 The facility population provided for general shift is Table 5: Approximate facility population S. No Location Population 1 Plant 25 2 Site Weather Probabilities Total Inside Population 175 As per CPR 18E guidelines 6 representative weather classes have been used: Table 6: Representative weather classes Stability Class B D D D E F Low wind speed corresponds with 1-2 m/s Medium wind speed corresponds with 3-5 m/s High wind speed corresponds with 8-9 m/s Wind Speed Medium Low Medium High Medium Low Observations in the Pasquill stability classes C, C/D and D are allocated to stability class D. Wind speeds below 2.5 m/s, between 2.5 m/s and 6 m/s and above 6 m/s are allocated to the wind speed categories low, medium and high respectively. Table 7: Pasquill stability classes Wind Speed A B B/C C C/D D E F <2.5 m s -1 D Low F Low B Medium m s -1 D Medium E Medium Page 19 of 32

20 >6 m s -1 D high The wind speed and direction considered for the study are taken from Cuddalore data provided in Climatological Normals, The two predominant weather conditions considered for the study are 1.5 F (Where F denotes Stable Condition night with moderate clouds and light moderate winds; 1.5 denotes wind velocity in m /sec) 5.0 D (where D denotes neutral condition little sun and high wind or over cast / windy night; Temperature and Relative Humidity: Based on climatological data from the Indian Meteorological Department, an average temperature of 30 C and relative humidity of 70% is found pre-dominant in Cuddalore Ignition Probabilities Immediate Ignition Probability Immediate ignition can be considered as the situation where the fluid ignites immediately on release either through auto-ignition or because of availability of ignition source. Immediate ignition probability is assumed based on the Reference Manual BEVI Risk Assessments version 3.2, developed by the National Institute of Public Health and the Environment (RIVM), Centre for External Safety, Netherlands. Substance category Category 0 average/ high reactivity Category 0 low reactivity Table 8: Probability of immediate ignition Source term Continuous Source term Instantaneous Probability of direct ignition < 10 kg/s < 1,000 kg kg/s ,000 kg 0.5 > 100 kg/s > 10,000 kg 0.7 < 10 kg/s < 1,000 kg kg/s ,000 kg 0.04 > 100 kg/s > 10,000 kg 0.09 Category 1 All flow rates All quantities Category 2 All flow rates All quantities 0.01 Category 3, 4 All flow rates All quantities 0 Category 1(Highly flammable) Liquid substances and preparations with a flash point below 21 C, which are not, however, extremely flammable Category 2(Flammable) Page 20 of 32

21 Liquid substances and preparations with a flash point greater than, equal to 21 C and less than, or equal to 55 C. Delayed Ignition Probability Delayed ignition is the result of the build-up of a flammable vapour cloud which is ignited by a source remote from the release point. It is assumed to result in flash fires or explosions, and also to burn back to the source of the leak resulting in a jet fire and/or a pool fire. Delayed ignition probability is assumed based on the National Institute of Public Health and the Environment (RIVM), Centre for External Safety, Netherlands. Table 9: Probability of delayed ignition Source type Ignition source Probability of ignition Point source Adjacent process installation Flare Oven (outside) Oven (inside) Boiler (outside) Boiler (inside) Line source high-voltage cable (per 100 m) Ship Population source Households (per person) Offices (per person) This involves identifying opportunities to reduce the likelihood and/or consequence of an accident where deemed necessary. 3.7 Risk Assessment Risk assessment combines the consequences and likelihood of all incident outcomes from all selected incidents to provide a measure of risk. The risk of all selected incidents are individually estimated and summed to give an overall measure of risk. Risk is presented as individual and societal risks. Individual Risk is the probability of death occurring as a result of accidents at a plant, installation or a transport route expressed as a function of the distance from such an activity. It is the frequency at which an individual or an individual within a group may be expected to sustain a given level of harm (typically death) from the realization of specific hazards. Such a risk actually exists only when a person is permanently at that spot (out of doors). Societal Risk involves the concept of the summation of risk from events involving many fatalities within specific population groups. This focused on the risk to society rather than to a specific individual and is termed societal risk. In relation to the process operations we can identify specific groups of people who work on or live close to the installation; for example communities living or working close to the plant. Page 21 of 32

22 Risks arising from the hazards are evaluated for its tolerability to personnel, the facility and the environment. The acceptability of the estimated risk must then be judged based on IS criteria appropriate to the particular situation. Acceptability of Risk is provided as follows Unacceptable risk: Risk greater than 1.00E-04 per average year ALARP: Between 1.00E-04 and 1.00E-06 per average year Acceptable risk: Risk less than 1.00E-06 per average year) 3.8 Risk Reduction Measures Risk-reduction measures include those to prevent incidents (i.e. reduce the likelihood of occurrence) to control incidents (i.e. limit the extent and duration of a hazardous event) and to mitigate the effects (i.e. reduce the consequences). Preventive measures, such as using inherently safer designs and ensuring asset integrity, should be used wherever practicable. In many cases, the measures to control and mitigate hazards and risks are simple and obvious and involve modifications to conform to standard practice. The general hierarchy of risk reducing measures is: Elimination & Substitution (by inherent safer design) Reduction & Isolation (by distance or inherent safer design) Detection (E.g. fire and gas, Leak detection); Control (E.g. emergency shutdown and controlled depressurization); Mitigation (E.g. firefighting and passive fire protection); Emergency response (In case safety barriers fail). 3.9 Modeling Assumptions In addition to the methods and assumptions in the modeling as noted above, the following additional assumptions are used: For the Phast modeling the horizontal option is selected for release orientation, this provides the maximum horizontal distances. Jet fires in Phast have been modeled using the un-impinged jet model. This leads to conservative, longer jet fire lengths as the model assumes that there are no obstacles to reduce jet momentum and therefore jet length and distances to radiation levels. TNT explosion model is used in the study. Page 22 of 32

23 4 CONSEQUENCE ANALYSIS 4.1 Scenarios This section documents the consequence-distance calculations, which have been computed for the accident release scenarios considered. Following are the potential Loss of Containment scenarios envisaged for proposed storage tanks at CCIL facility Table 10: Loss of containment scenarios identified for the study S. No Scenarios Proposed new Solvents 1 Leak of Diethylene Glycol storage tank 2 CR of Diethylene Glycol storage tank 3 Leak of Mono ethylene Glycol storage tank 4 CR of Mono ethylene Glycol storage tank 5 Leak of Isobutyl Alcohol (IBA) storage tank 6 CR of Isobutyl Alcohol (IBA) storage tank 7 Leak of 350 P solvent storage tank 8 CR of 350 P solvent storage tank 9 Leak Cyclopentanone drum 10 CR of Cyclopentanone drum 4.2 Inventory Calculations Sudden release of hydrocarbon can result in a number of accident situations. As large number of failure cases can lead to the same type of consequences, representative failure cases are selected for this analysis. The failure cases are based on conservative assumptions and engineering judgment. Typically, failure models are considered for 100% catastrophic rupture of vessels and 10mm leak size for vessels/ storage tanks, based on the guidelines of CPR 18 E. Table 11: Process parameters S. No List of Chemicals Inventory (m 3 ) Pressure (bar) Temperature Proposed new solvents 1 Diethylene Glycol 20 Atm Atm 2 Mono ethylene Glycol 20 Atm Atm 3 Isobutyl Alcohol (IBA) 20 Atm Atm P 20 Atm Atm 5 Cyclopentanone 200 lit drum Atm Atm 4.3 Summary of Consequence Results The summary of consequence results are tabulated in Table 12 and subsequently analysis is provided. Page 23 of 32

24 S. No Scenario Leak of Diethylene Glycol storage tank CR of Diethylene Glycol storage tank Leak of Mono ethylene Glycol storage tank CR of Mono ethylene Glycol storage tank Leak of Isobutyl Alcohol (IBA) storage tank CR of Isobutyl Alcohol (IBA) storage tank Leak of 350 P storage tank CR of 350 P storage tank Leak Cyclopentanon e Drum CR of Cyclopentanon e Drum Flammable Gas Dispersion (Distance to concentration results in m) 1.5F Weather Condition 5D Weather Condition UFL LFL 0.5LFL UFL LFL 0.5LFL Table 12: Consequence results for CCIL Jet fire (Distance downwind to intensity level in m) 1.5F Weather 5D Weather Condition Condition kw/ kw/ kw/ kw/ kw/ kw/ m 2 m 2 m 2 m 2 m 2 m 2 Proposed new solvents Pool fire (Distance downwind to intensity level in m) 1.5F Weather 5D Weather Condition Condition kw/ kw/ kw/ kw/ kw/ m 2 m 2 m 2 m 2 m kw/ m 2 bar VCE (Distance downwind to overpressure in m) 1.5F Weather 5D Weather Condition Condition NR NR NR NR NR NR NR NR NR NR NR NR NR NR NA NA NA NA NA NA NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NA NA NA NA NA NA NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NA NA NA NA NA NA NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NA NA NA NA NA NA NR NR NR NR NR NR NR NR NR NR NR NR NA NA NA NA NA NA NR NR NR NR NR NR Legend: NR: Not Reached; NA: Not Applicable 0.1 bar 0.3 bar 0.03 bar 0.1 bar 0.3 bar Page 24 of 32

25 Flammable gas Dispersion Analysis Of the five atmospheric storage tanks, in the event of catastrophic rupture of 350 P storage tank at a weather condition of 1.5F, will cause maximum damage due to flammable gas dispersion. The UFL will each up to 8.14 m, LFL will reach up to m. The 350 P flammable cloud covers a distance from 8.14 m (UFL) to m (LFL). Jet fire Analysis Considering the tank head of storage tanks, it is observed that the jet fire results are not envisaged in any scenario. Pool fire Analysis Of the five atmospheric storage tanks, in the event of catastrophic rupture of Monoethylene glycol storage tank at a weather condition of 5D, will cause maximum damage due to pool fire. The pool fire radiation of 4 KW/m 2 will reach up to a distance of m, causing first degree burns likely; 12.5 KW/m 2 will reach up to a distance of m, where in the energy is sufficient for pilot ignition of wood, melting plastic tubing etc KW/m 2 will reach up to a distance of m, where in the energy is sufficient to cause damage to equipment. Vapor Cloud Explosion Analysis Of the five atmospheric storage tanks, in the event of catastrophic rupture of 350 P storage tank at a weather condition of 1.5 F, will cause maximum damage due to overpressure. The overpressure damage distances for 0.03 bar will reach up to a distance of m which can cause significant damage (Shattering of glass); 0.1 bar will reach up to a distance of m, which can cause Moderate damage (Reparable damage to plant equipment and structure); 0.3 bar will reach up to a distance of m, which can cause heavy damage (Major damage to plant equipment, structure). Page 25 of 32

26 5 RISK ANALYSIS 5.1 Frequency Calculations Event Frequency: A measure of the expected probability or frequency of occurrence of an event. This may be expressed as events/year. For the scenarios, event frequency values are obtained from CPR 18E (Purple book) and tabulated below. Table 13: Event failure frequency considered for the study S. No Scenario Event failure frequency from CPR 18E Proposed new solvents 1 Leak of Diethylene Glycol storage tank 1.00E-04 2 CR of Diethylene Glycol storage tank 5.00E-06 3 Leak of Mono ethylene Glycol storage tank 1.00E-04 4 CR of Mono ethylene Glycol storage tank 5.00E-06 5 Leak of Isobutyl Alcohol (IBA) storage tank 1.00E-04 6 CR of Isobutyl Alcohol (IBA) storage tank 5.00E-06 7 Leak of 350 P storage tank 1.00E-04 8 CR of 350 P storage tank 5.00E-06 9 Leak Cyclopentanone drum 1.00E CR of Cyclopentanone drum 5.00E Individual Risk Individual risk is a measure of the risk (expressed as a frequency of fatality per year) for an individual exposed to a single hazard or the combined effects of several hazards. This measure of risk is important for a small group of people, particularly plant operators who will be exposed to higher levels of risk than other groups, due to their proximity. The individual risk calculation is done using specific locations of the known sources at the establishment. The calculated risks are typically presented as risk contours, which provide an easily understood graphical presentation of the risks. Individual risk contours are indicative of the potential magnitude or intensity of the risk. The individual risk is presented as contour lines on a topographic map. Overall individual risk contour for the facility is given in the Figure 2.Individual contours for the proposed solvent storage tanks is given in Figure 3. Page 26 of 32

27 Figure 2: Individual risk contour for CCIL Page 27 of 32

28 Figure 3: Individual risk contours for proposed solvent storage tanks Page 28 of 32

29 5.3 Societal Risk Societal risk considers the risk of hazardous events with the potential to give rise to large numbers of casualties; in particular, it is used to assess the risk to the public and other groups around the site who are not voluntarily exposed to the risks. It is also expressed as a frequency of fatality per year, but for a defined level of severity. The acceptability of the risk will depend on the number of fatalities (N) and the frequency (F). For this reason societal risk and the acceptability criteria are often given using an FN curves. Considering the storage capacities, process parameters and population densities provided, it is observed that F-N curve for the proposed storage tanks at CCIL facility is not obtained. Page 29 of 32

30 6 RISK ACCEPTABILTY CRITERIA In India, there are no defined criteria for risk acceptance. However, in IS Code of Practice for Hazard Identification and Risk Analysis, Annexure E summarizes the risk criteria adopted in some countries. Extracts from the same are presented below: Table 14: Risk acceptability criteria Authority and Application Maximum Tolerable Risk Negligible Risk (Per Year) (Per Year) VROM, The Netherlands (New) 1.0E-6 1.0E-8 VROM, The Netherlands (Existing) 1.0E-5 1.0E-8 HSE, UK (Existing Hazardous Industry) 1.0E-4 1.0E-6 HSE, UK (New Industries) 1.0E-5 1.0E-6 HSE, UK (Substance Transport) 1.0E-4 1.0E-6 HSE, UK (New Housing Near Plants) 3 x 1.0E-6 3 x 1.0E-7 Hong Kong Government (New Plants) 1.00E-5 Not Used To achieve the above risk acceptance criteria, ALARP principle was followed while suggesting risk reduction recommendations Unacceptable region 1E-04 per annum The ALARP or tolerability region (risk is undertaken only if a benefit is desired) Risk cannot be justified Tolerable only if further risk reduction is impractical, or the cost is not proportionate 1E-06 per annum Broadly acceptable Region Negligible risk Risks closer to the unacceptable region merit a closer examination of potential risk reduction measures Figure 4: ALARP triangle Page 30 of 32

31 7 EXISTING MEASURES AND RECOMMENDATIONS 7.1 Existing Control Measures The existing measures in the proposed solvent tank farm area are MOC for Diethylene Glycol, Monoethylene Glycol, Isobutyl Alcohol storage tanks is SS304, MOC for 350 P is SS316. Hydrant system, three number of portable foam monitors; dry powder, foam and CO 2 fire extinguishers are provided in the proposed tank farm area. 7.2 Recommendations 1. Ensure that periodical inspection and thickness measurement are to be carried out for the proposed storage tanks as specified by OISD129-Inspection of Storage Tanks 2. As per OISD 119, ensure bunds provided at the proposed tank farm have proper drainage system. 3. Ensure that foam Pourers are to be made available for the proposed storage tank area as per OISD Portable monitors/foam hose streams shall be provided for fighting fires in dyked area and spills. 5. Spill control kit and procedure shall be in place to contain any spill, clean them up properly and dispose off any containment waste safely. 6. Wind socks are to be installed for knowing wind direction during emergency. 7. Emergency Response & Control Plan to address the critical scenarios are to be tested by conducting mock drills at regular intervals. Page 31 of 32

32 8 REFERENCES IS 15656:2006: Hazard identification and risk analysis - Code of Practice Guidelines for Quantitative Risk Assessment CPR 18E (Purple book), Committee for the Prevention of Disasters, Netherlands (Edition: PGS 3, 2005) Guidelines for Hazard Evaluation Procedures - Centre for Chemical Process Safety, American Institute of Chemical Engineers, New York, Page 32 of 32