When making risk estimations, two variables are involved; the probability that an event will occur and the consequence, the logical and expected result of that event. For unplanned events these are independent of each other. Of the two, more development and investigative time has been spent on determining consequence. Multiple methodologies and approaches are available to reasonably predict the consequences of a chemical release, fire, and/or explosion on surrounding manufacturing equipment, people, the environment, etc. The technology to do so is well developed and is enhanced when new information becomes available. Therefore, consequence estimation is not within the scope of this book.
This book focuses on probability, which is more difficult to quantify objectively or experimentally. With the help of member companies of the Center for Chemical Process Safety (CCPS), research for the book began with a co-op student directed and assisted by a professional librarian and supplemented by a subcommittee composed of volunteers from CCPS member companies.
After approximately two years of research and document reviews, the subcommittee agreed that there was sufficient information and technology available to produce a book that addressed and reasonably estimated the component of the risk equation relating to the probability that a released flammable mass would ignite. Available technology methods could be compiled, codified, and developed into real-world usable tools as part of the book. The CCPS membership approved the proposal to write a book on the subject and a book subcommittee was formed.
The information used to write this book and develop the calculation tool came from the previous research conducted which is not proprietary and is generally available in the public domain. The CCPS subcommittee compiled, codified, clarified, and in some cases further developed the information and data to develop the algorithms in the book and the associated calculation tool. This tool, when populated with appropriate data, yields an estimate of the probability that a defined release of a flammable material will ignite if exposed to an ignition source. This tool can provide users, within reason, information that can be used in risk assessments with a higher degree of confidence than estimates made before.
This book and calculation tool have been developed and are presented with the understanding that users will employ whatever methodology they deem appropriate to estimate the probability that the released flammable mass will reach an ignition source. That estimate is then combined with the probability of ignition determined per the methodology in this book and the associated calculation "tool" to determine the total estimated probability of ignition of a release.
The team agreed that the calculation tool should be conservative, especially when data was scant and/or known to be beyond previously vetted values. The team believed it would be safer and more responsible to estimate a higher probability of ignition than to rely on unconfirmed or proprietary data and methodologies or approaches designed for very specific uses. The desire to err on the side of being conservative was tempered by the committee's recognition that being "conservative" could lead to other issues, e.g., predicting a high ignition probability reduces the probability of a toxic outcome for chemicals that are both flammable and toxic. Additionally, the team recognized that some companies may have proprietary or more detailed data than what is publicly available and could reasonably reach a different estimate as to the probability that a defined release of a flammable material would ignite when exposed to an ignition source. It is critical that users thoroughly read the book to understand the scope and limitations of both the book and calculation tool. The calculation tool is populated with popup caution "flags" to warn the user when the calculation tool is being used at the extremes of its design capabilities as we know them now. However, not all such conditions may be identified at present.
Readers should be aware that the methodologies presented here apply only to process safety incidents, unplanned events where probability and consequence are independent of each other. These methodologies may not be used in evaluating the risk associated with willful acts such as sabotage or terrorism. In such cases, probability and consequence are codependent, i.e., consequence influences probability.
Like many CCPS books, this book breaks new ground. As such, we anticipate that over time these methodologies will improve. For this we would like to enlist your help to improve the precision and utility of the calculation tool. We intend to provide periodic updates to the tool and within a few years determine if a second edition of the book and an updated version of the tool are warranted. The CCPS website provides a way for users to submit information and lessons learned related to usage of the tool and a way for users to download new information and updated versions of the tool should there be any.
Owners of this book can access the calculation tool on the CCPS website:http://vvww.aichc.org/ccps/resources/pubHcations
This "tool" is provided as part of the book purchased and should not be used by anyone other than the owner of this book. The tool should also only be used after the user has read the book, particularly Chapters 1 and 4, and should initially be used in conjunction with the illustrations in Chapter 5.
User Please Note: This tool represents technology that is evolving and is based on the algorithms and logic described in the corresponding CCPS book, "Guidelines for Determining the Probability of Ignition of a Released Flammable Mass." Any such tool cannot anticipate all possible circumstances of usage, and it is expected that the user of this tool has read and understood the scope of the associated book, the tool, and the potential hazards associated with misuse of the tool. Results generated by users who have not read the book may in certain circumstances be incorrect or misinterpreted. Misunderstanding of how to use the tool and/or misuse of the tool may result in inaccurate reported results and possibly inappropriate actions on the user's part.
It is sincerely hoped that the information presented in this tool will lead to an even more impressive safety record for the entire industry; however, the American Institute of Chemical Engineers (AIChE), its consultants, the AIChE's CCPS Technical Steering Committee and the Probability of Ignition subcommittee members, their employers, their employers' officers and directors, and Baker Engineering and Risk Consultants, Inc. (BakerRisk) and its employees, officers, and directors neither warrant nor represent, expressly or by implication, the correctness or accuracy of the content of the information presented in this tool. As between (1) the AIChE, its consultants, the CCPS Technical Steering Committee and Subcommittee members, their employers, their employers' officers and directors, and BakerRisk, and its employees, officers, and directors and (2) the user of this tool, the user accepts any and all legal liability and responsibility whatsoever for the consequence of its use or misuse of this tool.

LIST OF FIGURES xiii

LIST OF TABLES xv

FOREWORD xvii

ACKNOWLEDGMENTS xxi

GLOSSARY xxiii

1 INTRODUCTION 1

1.1 Objectives 1

1.2 Motivation for this Book 1

      1.2.1 A Brief History of Fire Protection 2

      1.2.2 The Development of Risk-Based Approaches to Flammables Management 3

      1.2.3 Difficulties in Developing Ignition Probability

      Prediction Methods 4

      1.2.4 Missing Variables 5

      1.2.5 Summary of Industry Needs and Path Forward 5

      1.2.6 Applications for This Book 6

      1.2.7 Limitations in Applying the Approaches in This Book 7

1.3 Ignition Probability Overview 8

      1.3.1 Theoretical Basis for Ignition 8

      1.3.2 Key Ignition Factors Related to the Properties of the Fuel and Available Surrogates That Can Be Used for Developing Probability of Ignition Predictions 13

      1.3.3 Key Ignition Factors Related to the Release Source 19

      1.3.4 Key Ignition Factors Related to the External

      Environment After the Release 27

1.4 Control of Ignition Sources 31

      1.4.1 Ignition Source Management 31

      1.4.2 Minimization of Release 33

1.5 Vapor Cloud Explosion Probability Overview 34

1.6 Detonation Overview 35

      1.6.1 Detonation Using a Strong Ignition Source 35

      1.6.2 Deflagration-to-Detonation Transition 35

      1.6.3 Buncefield 36

1.7 Other Ignition Topics—Hydrogen 36

      1.7.1 Ignition Mechanisms 36

      1.7.2 Other Hydrogen Ignition Topics 37

2 ESTIMATION METHODS 40

2.1 Introduction 40

      2.1.1 Event Tree 40

      2.1.2 Failure Frequency Data for Use in Event Trees 42

      2.1.3 Quantification of the Event Tree 42

2.2 Factors Influencing the Probability of Immediate Ignition 42

      2.2.1 Temperature of Release Relative to the Autoignition Temperature 43

      2.2.2 Minimum Ignition Energy (MIE) of Material

      Being Released 43

      2.2.3 Pyrophoricity of Released Material 45

      2.2.4 Pressure/Velocity of Discharge 45

      2.2.5 Droplet Size 46

      2.2.6 Presence of Particulates 47

      2.2.7 Configuration/Orien tation of Equipment Near/At

      the Point of Release 47

      2.2.8 Temperature of Release (As It Relates to Its Effect on MIE) 47

      2.2.9 Phase of Release (API RBI) 48

      2.2.10 Flash Point and Release Rate (TNO) 48

2.3 Factors Influencing the Probability of Delayed Ignition 48

      2.3.1 Strength and Numbers of Ignition Sources 48

      2.3.2 Duration of Exposure 52

      2.3.3 Release Rate/Amount 52

      2.3.4 Material Being Released 54

      2.3.5 Release Phase/Flash Point/Boiling Point 54

      2.3.6 Distance from Point of Release to Ignition Source 55

      2.3.7 Meteorology 55

      2.3.8 Events Originating Indoors 55

2.4 Factors Influencing the Probability of Explosion,

Given Delayed Ignition 57

2.5 Potential Interdependence of Variables 58

2.6 Summary of Variables Used in Each Analysis Level 59

2.7 Basic (Level 1) Probability of Ignition Algorithms 60

      2.7.1 Level 1 Algorithm for Probability of Immediate Ignition 60

      2.7.2 Level 1 Algorithm for Probability of Delayed Ignition 61

2.8 Level 2 Probability of Ignition Algorithms 62

      2.8.1 Level 2 Algorithm for Probability of Immediate Ignition 62

      2.8.2 Level 2 Algorithm for Probability of Delayed Ignition 63

2.9 Advanced (Level 3) Probability of Ignition Algorithms 67

      2.9.1 Level 3 Algorithm for Probability of Immediate Ignition 67

      2.9.2 Level 3 Algorithm for Probability of Delayed Ignition 68

2.10 Developing Inputs When Chemical Properties Are Not Available 70

      2.10.1 Estimating Input Properties of Chemicals Not in the Pick List 70

      2.10.2 Estimating the Properties of Flammable Mixtures 72

2.11 Worked Example 74

      2.11.1 Problem Statement 74

      2.11.2 Level 1 Analysis 74

      2.11.3 Level 2 Analysis 75

      2.11.4 Level 3 Analysis 76

2.12 Application of the Models to a Study with Multiple Ignition Sources 77

3 TECHNICAL BACKGROUND AND

DATA SOURCES 80

3.1 Introduction and Summary 80

3.2 Government-driven studies 85

      3.2.1 Rewetal. 85

      3.2.2 Bevi Risk Assessment Manual (TNO Purple Book) 94

      3.2.3 HSE/Crossthwaite et al. 98

      3.2.4 HSE/Thyer 98

      3.2.5 HSE/Gummer and Hawksworth—Hydrogen 100

      3.2.6 Cawley/U.S. Bureau of Mines 101

      3.2.7 Canvey 102

      3.2.8 Witcofski (NASA) Liquid Hydrogen 103

3.3 Information Developed by Industry Groups 103

      3.3.1 Cox/Lees/Ang 103

      3.3.2 E&P Forum 106

      3.3.3 API RBI 106

      3.3.4 API RP 2216 111

      3.3.5 IEEE 112

      3.3.6 UK Energy Institute 113

3.4 Information Developed in Academia 116

      3.4.1 Ronzaetal. 116

      3.4.2 Offshore Explosions (Loughborough) 119

      3.4.3 Srekl and Golob 119

      3.4.4 Duarteetal. 120

      3.4.5 Swain—Ignition of Hydrogen 121

      3.4.6 Dryer et al.—Hydrogen and Light Hydrocarbons 121

      3.4.7 Britton—Silanes and Chlorosilanes 122

      3.4.8 Peseeetal. 123

3.5 Information Developed by Individual Companies 124

      3.5.1 Spouge 124

      3.5.2 Moosemiller 125

      3.5.3 Johnson—Humans as Electrostatic Ignition Sources 126

      3.5.4 Jallais—Hydrogen 128

      3.5.5 Zalosh—Hydrogen 128

      3.5.6 Smith—Pipelines 130

3.6 Studies Specific to Ignition of Sprays 131

      3.6.1 Leeetal. 131

      3.6.2 Babrauskas 133

3.7 Case Histories 134

      3.7.1 Britton—External Ignition Events 134

      3.7.2 Pratt—Gas Well and Pipeline Blowouts 135

      3.7.3 Gummer and Hawks worth—Hydrogen Events 136

4 ADDITIONAL EXAMPLES 140

4.1 Introduction to Examples and Potential "Lessons Learned" 140

      4.1.1 "Reality" vs. Predictions 140

      4.1.2 "Conservatism"—Does It Exist? 141

      4.1.3 Cases Where the Model May Not Be Appropriate

      or the Results Misinterpreted 142

      4.1.4 Summary of Worked Examples 143

4.2 Worked Examples (based on other CCPS books) 144

      4.2.1 Vapor Cloud Explosion Hazard Assessment of a Storage Site 144

      4.2.2 Op en Fi eld Release of Propan e 149

      4.2.3 Release from Pipeline 153

4.3 Worked Examples (Chemical and Petrochemical Plants) 156

      4.3.1 Ethylene Tubing Failure 156

      4.3.2 Benzene Pipe Rupture 158

      4.3.3 Spill from Methyl Ethyl Ketone Tank 159

      4.3.4 Indoor Puncture of MEK Tote 163

      4.3.5 Elevated Release 166

4.4 Worked Examples (oil refineries) 169

      4.4.1 Gasoline Release from a Sight Glass 169

      4.4.2 Overfilling a Gasoline Storage Tank 173

      4.4.3 Overfilling a Propane Bullet 175

      4.4.4 Hydrogen Release from a Sight Glass 177

4.5 Worked Examples (Unusual Cases) 179

      4.5.1 Indoor Acid Spill—Ventilation Model 179

      4.5.2 Release of Ammonia 184

4.6 Worked Examples ("Out of Scope" Cases) 185

      4.6.1 Release of Gas from an Offshore Platform Separator 185

      4.6.2 Dust Ignition 189

4.7 Worked Examples of the Benefits of Plant Modifications and

Design Changes 193

      4.7.1 Ignition by Hot Surfaces 193

      4.7.2 Release Prevention 196

      4.7.3 Duration of Exposure 196

      4.7.4 Benefit of Improved Ventilation of Indoor Releases— Continuation of "Indoor Acid Spill" Example 198

5 SOFTWARE ILLUSTRATION 200

5.1 Explanation and Instructions for Software Tool 200

5.2 Opening the Software Tool 200

5.3 General Inputs and Outputs 201

5.4 Level 1 Inputs 203

5.5 Level 2 Analyses 205

5.6 Level 3 Analyses 207

5.7 Explosion Probability 207

5.8 Illustrations of Software Use 208

      5.8.1 Vapor Cloud Explosion Hazard Assessment of a Storage Site (Example from Section 4.2.1) 208

      5.8.2 Open Field Release of Propane

      (Example from Section 4.2.2) 211

APPENDIX A. CHEMICAL PROPERTY DATA 214

APPENDIX B. OTHER MODELS FOR

CONSIDERATION 220

REFERENCES 226

INDEX 232

中国化工信息中心