1、 GUIDANCE NOTES ON THE INVESTIGATION OF MARINE INCIDENTS JUNE 2005 (Updated 1 February 2014 see next page) American Bureau of Shipping Incorporated by Act of Legislature of the State of New York 1862 Copyright 2005 American Bureau of Shipping ABS Plaza 16855 Northchase Drive Houston, TX 77060 USA Up
2、dates February 2014 consolidation includes: June 2005 version plus Corrigenda/Editorials ABSGUIDANCE NOTES ON THE INVESTIGATION OF MARINE INCIDENTS .2005 iii Foreword The marine industry experiences incidents that range from major accidents to near misses. These incidents should be investigated sinc
3、e many flag administration regulations require it; international agreements mandate it (such as the IMO “International Safety Management Code”); and industry initiatives encourage it. Incident investigation is a process that is designed to help organizations learn from past performance and develop s
4、trategies to improve safety. The ABS Marine Root Cause Analysis Technique (MaRCAT) provides an effective and efficient approach for investigating marine incidents of any magnitude. ABS developed the MaRCAT methodology by customizing and combining the best techniques available and by proving and impr
5、oving the overall approach through MaRCATs application during numerous investigations. The ABS MaRCAT approach to incident investigation caters to the unique needs of the marine industry, including human element; machinery and engineering; structural and security concerns. The objectives of the ABS
6、MaRCAT approach are as follows: Provide ABS clients with a technique that will guide incident investigators in the conduct of root cause analyses and in identifying, documenting and trending the causes of accidents and near misses. Assist clients with the investigation of a variety of types (e.g., g
7、roundings, collisions, fires, etc.) and sizes of incidents (minor to major, including near misses) related to their vessels and facilities (ashore and at sea). Allow analysis of losses whether they are related to safety, the environment, human element concerns, security, reliability, quality or busi
8、ness losses. Support Class-related activities such as ABS Safety, Quality and Environment (SQE) notation, as well as the International Safety Management Code (ISM Code) and the International Ship and Port Facility Security (ISPS) Code. Provide a technique that is sufficiently flexible to allow custo
9、mization to a clients own management system, Health, Safety and Environment (HSE) programs or related initiatives. These Guidance Notes provide instructions for the performance of incident investigation activities, including: Incident Investigation Initiation Data Gathering Data Analysis Root Cause
10、Determination Generating Recommendations Reporting and Trending of Incident Investigation Results The ABS Guidance Notes on the Investigation of Marine Incidents provide a structured approach to the investigation of incidents and near-miss events. The information contained can also assist with ident
11、ifying and documenting root causes as required by the ISM Code. These Guidance Notes describe an incident investigation methodology that was expressly developed for the maritime industry, and so it reflects those elements of maritime operations and incident causation particular to the industry. This
12、 Page Intentionally Left Blank ABSGUIDANCE NOTES ON THE INVESTIGATION OF MARINE INCIDENTS .2005 v GUIDANCE NOTES ON THE INVESTIGATION OF MARINE INCIDENTS CONTENTS SECTION 1 Introduction 1 1 Background1 2 The ABS Incident Investigation Model: MaRCAT2 3 Scope of the Guidance Notes2 4 Contents of the G
13、uidance Notes3 5 Terminology .4 5.1 Incident .5 5.2 Consequences 5 5.3 Loss Event 5 5.4 Accident 6 5.5 Near Miss6 5.6 Event.6 5.7 Condition.6 5.8 Causal Factor7 5.9 Structural/Machinery/Equipment/Outfitting Problems7 5.10 Human Errors7 5.11 External Factors7 5.12 Intermediate Cause.7 5.13 Item-of-No
14、te (ION) 8 5.14 Root Cause .8 5.15 Management System (MS) .8 5.16 Safeguard .8 5.17 Recommendation8 5.18 Resolution .8 5.19 Root Cause Analysis (RCA)9 5.20 Apparent Cause Analysis (ACA) .9 FIGURE 1 ABS Incident Investigation Model 2 FIGURE 2 Relationship of Incident Investigation Terms.4 vi ABSGUIDA
15、NCE NOTES ON THE INVESTIGATION OF MARINE INCIDENTS .2005 SECTION 2 Basics of Incidents and Investigations 11 1 The Need for Incident Investigation.11 1.1 Rationale for Taking a Structured Approach to Incident Investigation 11 1.2 Depths of Analyses .12 1.3 Structured Analysis Process14 2 Selecting I
16、ncidents to Investigate 14 3 The Investigation Thought Process .15 3.1 Differences Between Traditional Problem Solving and Structured Incident Investigation .15 3.2 An Incident Investigation Approach to the Analysis.15 4 Incident Investigation within a Business Context.18 5 The Elements of an Incide
17、nt 19 6 The Goal of the Incident Investigation Process .19 7 The MaRCAT Marine Incident Investigation Process20 7.1 Step 1: Should the Incident Be Analyzed Now? 21 7.2 Step 2: Initiating the Investigation21 7.3 Step 3: Gathering and Preserving Data.21 7.4 Step 4: Analyzing Data21 7.5 Step 5: Identif
18、ying Root Causes 21 7.6 Step 6: Developing Recommendations .21 7.7 Step 7: Completing the Investigation.22 7.8 Step 8: Selecting Problems for Analysis22 7.9 Step 9: Trending22 7.10 Step 10: Following up an Investigation22 7.11 Developing an Overall Incident Investigation Program Management Issues 22
19、 8 Levels of the Analysis: Root Cause Analysis and Apparent Cause Analysis 22 9 Summary24 FIGURE 1 Task Triangle Showing Possible Depths of Analyses.13 FIGURE 2 Overlap of Multiple Task Triangles 13 FIGURE 3 Differences Between Traditional Problem Solving and Structured Root Cause Analysis.16 FIGURE
20、 4 Relationship Among Proactive Analysis, Reactive Analysis and Management Systems18 FIGURE 5 The MaRCAT Marine Incident Investigation Process 20 FIGURE 6 Levels of Analysis 23 FIGURE 7 Connection Between Causal Factors and Root Causes.23 ABSGUIDANCE NOTES ON THE INVESTIGATION OF MARINE INCIDENTS .2
21、005 vii SECTION 3 Initiating Investigations.25 1 Initiating the Investigation 25 2 Notification .26 3 Emergency Response Activities 26 4 Immediate Response Activities26 5 Beginning the Investigation27 6 Corrective Action Requests .27 6.1 Reasons to Generate a CAR 28 6.2 Typical Information Contained
22、 in a CAR .28 6.3 Using the CAR in the Incident Investigation Process 29 7 Incident Classification 29 8 Investigation Management Tasks30 9 Assembling the Team 31 10 Restart Criteria.31 11 Gathering Investigation Resources32 12 Summary32 TABLE 1 Incident Classification Criteria30 FIGURE 1 Initiating
23、Investigations within the Context of the Overall Incident Investigation Process.25 SECTION 4 Gathering and Preserving Data 33 1 Introduction 33 1.1 Importance of Data Gathering.34 1.2 Overall Types of Data .34 1.3 Prioritizing Data-gathering Efforts .34 1.4 Types of People Data35 1.5 Types of Electr
24、onic Data .36 1.6 Types of Position Data 36 1.7 Types of Physical Data .37 1.8 Types of Paper Data .37 2 Gathering Data.37 3 Gathering Data from People 37 3.1 Initial Witness Statements.39 3.2 The Interview Process.39 4 Physical Data.43 4.1 Sources of Data 43 4.2 Types and Nature of Questions.44 4.3
25、 Basic Steps in Failure Analysis .44 4.4 Use of Test Plans47 4.5 Chain-of-Custody 48 4.6 Use of Outside Experts .48 5 Paper Data.49 6 Electronic Data.49 viii ABSGUIDANCE NOTES ON THE INVESTIGATION OF MARINE INCIDENTS .2005 7 Position Data50 7.1 Unique Aspects .50 7.2 Data Collection 50 7.3 Documenta
26、tion of Data Collection .50 7.4 Alternative Sources of Position Data .51 8 Overall Data Collection Plan51 9 Application to Apparent Cause Analyses and Root Cause Analyses.52 10 Summary52 TABLE 1 Forms of Fragility .35 TABLE 2 Application of Data Collection Methods .53 FIGURE 1 Gathering Data within
27、the Context of the Overall Incident Investigation Process .33 FIGURE 2 Overall Types of Data Resources34 FIGURE 3 Flowchart of Typical Interview Sequence 40 FIGURE 4 Basic Steps in Failure Analysis45 SECTION 5 Analyzing Data . 55 1 Introduction 55 2 Overview of Primary Techniques.56 3 Fault Tree Ana
28、lysis .57 4 The 5-Whys Technique58 5 Causal Factor Charts.60 6 Using Causal Factor Charts and Fault (or 5-Whys) Trees Together during an Investigation .64 7 Application to Apparent Cause Analyses and Root Cause Analyses.64 8 Summary64 TABLE 1 Applicability of Analysis Techniques56 TABLE 2 Guidance o
29、n Using Causal Factor Charts and Fault Trees 65 FIGURE 1 Analyzing Data within the Context of the Overall Incident Investigation .55 FIGURE 2 Tank Spill Example Fault Tree.57 FIGURE 3 Sandblasting Fault Tree Example58 FIGURE 4 5-Whys Technique Example60 FIGURE 5 Sandblasting Causal Factor Chart Exam
30、ple62 ABSGUIDANCE NOTES ON THE INVESTIGATION OF MARINE INCIDENTS .2005 ix SECTION 6 Identifying Root Causes67 1 Introduction 67 2 Root Cause Analysis Traps .68 2.1 Trap 1 Hardware Problems 68 2.2 Trap 2 Personnel Problems .68 2.3 Trap 3 External Event Problems 69 3 Procedure for Identifying Root Cau
31、ses69 4 Using the ABS Marine Root Cause Analysis Map.69 5 Observations About the Structure of ABSs Marine Root Cause Analysis Map 69 6 The ABS Marine Root Cause Analysis Map70 6.1 Multiple Coding .70 6.2 Using the ABS Guidance Notes on the Investigation of Marine Incidents71 6.3 Typical Problems Enc
32、ountered When Using the ABS Marine Root Cause Analysis Map.71 6.4 Advantages and Disadvantage of Using the ABS Marine Root Cause Analysis Map.73 7 Documenting the Root Cause Analysis Process .73 8 Application to Apparent Cause Analyses and Root Cause Analyses 74 9 Summary74 TABLE 1 First Example of
33、a Root Cause Summary Table .75 TABLE 2 Second Example of a Root Cause Summary Table 76 TABLE 3 Third Example of a Root Cause Summary Table 77 FIGURE 1 Identifying Root Causes with the Context of the Overall Incident Investigation Process.67 FIGURE 2 Document Hierarchy 71 SECTION 7 Developing Recomme
34、ndations.79 1 Introduction 79 2 Timing of Recommendations .80 3 Levels of Recommendations .81 3.1 Level 1 Address the Causal Factor 81 3.2 Level 2 Address the Intermediate Causes of the Specific Problem .81 3.3 Level 3 Fix Similar Problems81 3.4 Level 4 Correct the Process that Creates These Problem
35、s.82 4 Types of Recommendations 82 4.1 Eliminate the Hazard.82 4.2 Make the System Inherently Safer/More Reliable .82 4.3 Prevent the Occurrence of the Incident.82 4.4 Detect and Mitigate the Loss.82 x ABSGUIDANCE NOTES ON THE INVESTIGATION OF MARINE INCIDENTS .2005 5 Suggested Format for Recommenda
36、tions.83 6 Special Recommendation Areas .83 7 Management Responsibilities83 8 Examples of Reasons to Reject Recommendations .84 9 Benefit-Cost Ratios85 9.1 Estimating the Benefit of a Recommendation85 9.2 Estimate the Costs of Implementing a Recommendation by Considering the Total Life-cycle Costs o
37、f the Change85 9.3 Cost-Benefit Ratio .85 10 Assessing Recommendation Effectiveness.86 11 Application to Apparent Cause Analyses and Root Cause Analyses.87 12 Summary87 TABLE 1 Effectiveness of Various Shift Turnover Alternatives 87 TABLE 2 Recommendations for Apparent Cause Analyses and Root Cause
38、Analyses .88 FIGURE 1 Developing Recommendations within the Context of Overall Incident Investigation Process.79 SECTION 8 Completing the Investigation 89 1 Introduction 89 2 Writing Investigation Reports.90 2.1 Typical Items to Be Included in an Investigation Report90 2.2 Tips for Writing Reports.9
39、1 3 Communicating Investigation Results94 3.1 Decide to Whom94 3.2 Decide How.94 3.3 Document the Communication 95 4 Resolving Recommendations and Communicating Resolutions 95 4.1 Tracking Recommendations95 4.2 Resolution Report Phase and Closure of Files95 5 Addressing Final Issues.97 5.1 Enter Tre
40、nding Data 97 5.2 Evaluate the Investigation Process .97 6 Application to Apparent Cause Analyses and Root Cause Analyses.98 7 Summary98 TABLE 1 Typical Items to Include in Reports90 TABLE 2 Investigation Completion Activities for Apparent Cause Analyses and Incident Investigations .98 ABSGUIDANCE N
41、OTES ON THE INVESTIGATION OF MARINE INCIDENTS .2005 xi FIGURE 1 Completing the Investigation within the Context of Overall Incident Investigation Process.89 FIGURE 2 Tracking Recommendations 96 SECTION 9 Selecting Incidents for Analysis.99 1 Introduction 99 2 Why Be Careful when Selecting Incidents
42、for Investigation?.100 3 Some General Guidance .101 3.1 Incidents to Investigate (High Potential Learning Value) 102 3.2 Incidents to Trend (Moderate to Low Potential Learning Value). 102 3.3 No Investigation Behavior-based Risk Management (BBRM) (Low Potential Learning Value) . 102 4 Performing the
43、 Investigation 102 4.1 Incidents to Investigate Immediately (Acute Incidents) 102 4.2 Incidents to Trend (Potentially Chronic Incidents). 102 5 Near Misses .103 5.1 Factors to Consider When Defining Near Misses 103 5.2 Reasons Why Near Misses Should Be Investigated . 103 5.3 Barriers to Getting Near
44、 Misses Reported 103 5.4 Overcoming the Barriers . 105 5.5 Acute Analysis versus Chronic Analysis . 105 6 Identifying the Chronic Incidents that Should Be Analyzed 105 6.1 Pareto Analysis . 105 6.2 Examples of Pareto Analysis 106 6.3 Weaknesses of Pareto Analysis 107 6.4 Other Data Analysis Tools 10
45、7 7 Application to Apparent Cause Analyses and Root Cause Analyses.108 8 Summary108 TABLE 1 Learning Potential from Incidents 101 FIGURE 1 Selecting Incidents for Analysis Within the Context of the Overall Incident Investigation Process.99 FIGURE 2 Investigation Cycle if Too Many Investigations Are
46、Performed 100 FIGURE 3 Pareto Charts Developed Using Two Different Attributes106 xii ABSGUIDANCE NOTES ON THE INVESTIGATION OF MARINE INCIDENTS .2005 SECTION 10 Results Trending 109 1 Introduction 109 2 Benefits of a Trending Program.110 3 Determining the Data to Collect.110 3.1 Deciding What Data t
47、o Collect.111 3.2 Defining the Data to Collect.111 3.3 Other Data Collection Guidance112 4 Data Analysis .112 4.1 Interpreting Data Trends113 5 Application to Apparent Cause Analyses and Root Cause Analyses .114 6 Summary114 FIGURE 1 Results Trending Within the Context of the Overall Incident Invest
48、igation Process .109 SECTION 11 Developing Incident Investigation Programs 115 1 Introduction 115 2 Program Implementation Process .116 2.1 Design the Program 116 2.2 Develop the Program 118 2.3 Implement the Program.119 2.4 Monitor the Programs Performance119 3 Key Considerations120 3.1 Legal Consi
49、derations.120 3.2 General Legal Guidelines120 3.3 Media Considerations122 3.4 Some Regulatory Requirements and Industry Standards122 4 Management Influence on the Program 124 5 Typical Reasons Why an Incident Investigation Program May NOT Work 124 5.1 There Is No Business Driver to Change 124 5.2 There Is No Organizational Champion for the Program.124 5.3 The Organization Never Leaves the Reactive Mode .125 5.4 The Organization Must Find an Individual to Blame 125 5.5 You Are Unwilling to Critique Management Systems.125 5.6 The Organization Tries
