TITLE:
Incident Investigation Investigation Guidelines Contents
1. Scope 2. Initiation of an Incident investigation 2.1. Profundity definitions 2.2. Criteria for probability 2.3. Criteria for (potential) Consequences 3. Assignment 3.1. Qualifications 4. Incident Investigation Phases 4.1. Secure the site 4.2. Initial on-site investigation 4.3. Analysis 4.3.1. Bow-tie technique 4.4. Reporting 5. Recommendations 6. Review and approval 7. Follow-up and close-out of the action items 8. Attachments 8.1. Bow-tie 8.2. Checklist Initial Investigation 8.3. Tentative list of immediate causes 8.4. Tentative list of Root Causes
1. SCOPE This guideline provides the details for Incident Investigations as set out in Procedure xxx. It is meant to assist the investigator(s) with the sometimes difficult process of gathering and analyzing the data and to prepare comprehensive reports.
2. INITIATION OF AN INCIDENT INVESTIGATION The (HSE) Management shall carefully consider when and to what profundity detail an Incident should be investigated. A thorough Incident Investigations takes time and requires resources. The management can use the RA Risk Matrix as a guideline to determine this profundity (slightly different criteria).
2.1. Profundity definitions Thorough investigation
An ”in depth” investigation that involves all aspects of the incident and that identifies immediate- and root causes that may be beyond the span of control of the project (HO / Design / Manufacturer / Legislation etc.) Normal investigation A detailed investigation that involves all directly related aspects of the incident and that identifies immediate- and root causes within the span of control of the project. Close-out A standard report that identifies the obvious immediate- and root causes. (statement of facts).
2.2. Criteria for probability Probability for re-occurrence re-occurrence LEVEL DEGREE DESCRIPTION A IMPROBABLE So unlikely that occurrence may not be experienced. B REMOTE Unlikely but possible to occur during project C OCCASIONAL Likely to occur sometime during project D PROBABLY Will occur several times during project E FREQUENT Likely to occur frequently
2.3. Criteria for (potential) Consequences (Potential) Consequences Consequences
LEVEL DEGREE DESCRIPTION 1 Negligible People Slight injury or health effects Assets Slight damage Environment Slight effect Reputation Slight impact 2 Marginal People Minor injury or health effects Assets Minor damage Environment Minor effect Reputation Minor impact 3 Critical People Major injury or health effects Assets Local damage Environment Localized effect Reputation Considerable Impact 4 Severe People Single fatality or permanent total disability Assets Major damage Environment Major effect Reputation National impact 5 Catastrophic People Multiple fatalities or permanent total disabilities Assets Extensive damage Environment Massive effect Reputation International impact
(Potential) Consequences Consequences LEVEL
1
2
3
4
5
DEGREE
Negligible
Marginal
Critical
Severe
Catastrophic
DESCRIPTION People
Slight injury or health effects
Assets
Slight damage
Environment
Slight effect
Reputation
Slight impact
People
Minor injury or health effects
Assets
Minor damage
Environment
Minor effect
Reputation
Minor impact
People
Major injury or health effects
Assets
Local damage
Environment
Localized effect
Reputation
Considerable Impact
People
Single fatality or permanent total disability
Assets
Major damage
Environment
Major effect
Reputation
National impact
People
Multiple fatalities or permanent total disabilities
Assets
Extensive damage
Environment
Massive effect
Reputation
International impact
3. ASSIGNMENT The assignment of Incident investigator or Incident Investigation team shall follow as soon as possible after the initial report of the incident has been assessed as per above criteria. The (HSE) Management shall inform the people in charge of- and/or involved in, the investigation. Most times this will be someone from the HSE department but it is possible that others will be in charge or involved. It is also possible that Client Representatives will participate in the Investigation. The (HSE) Manager shall, if possible, provide the Incident Investigator(s) with a scope or “expected extend” description. In any case he will indicate the type of requires investigation as per above criteria.
3.1. Qualifications Although theoretical models and procedures etc. claim that adequate qualified staff with in-depth knowledge of the processes and years of experience within the field of Incident Investigations, are required, required, it is known that we do not always have that choice. HSE Staff on projects, with all respect, are not always marine or dredging specialists and/or have been involved in a limited number of investigations. Still the client and the management wants quick results with reports of a good quality and analysis of immediate- and root causes and the definitions of adequate recommendations. Also the complexity of the site logistics plays a part in the ass ignment of the incident Investigator(s). Investigator(s). On xxx we might be limited by distances, weather, day/night shift, xxx, leave rotation schedules etc. where and when possible the following aspects shall be c onsidered: Knowledge and experience of the process and subject (e.g. Marine) Experience with Incident Investigations Communication skills
Reporting skills Immediate availability
4. INCIDENT INVESTIGATION PHASES Incident Investigations are performed in different phase which are described in this chapter. Please note that this is only a guideline and Investigator(s) should determine what phases and actions are needed for their ongoing investigation. Also the established profundity of the Investigation influences this process.
4.1. Secure the site If Necessary, the site shall be secured by the site supervisor s upervisor or a HSE officer on site, even before the investigation is assigned or defined. Securing the site might be necessary to ensure that a p roper investigation can take place. This may mean shutting down any work in progress and blocking access to materials and equipment that may have been involved in the incident. The goal is to prevent tampering with evidence and exposing workers to additional hazards .
4.2. Initial on-site investigation The assigned investigator(s) will start gathering as much data about the incident as possible. It‟s critical that this process begin immediately, before witnesses begin to forget details and before regular work compromises any evidence. It is possible that Management support is required in order to ensure cooperation of supervisory personnel. Depending on the profundity of the Investigation it is possible that major plant has to shut down. Usually this requires approval of the Management. The information-gathering process starts with questioning the people involved and witnesses. It is recommended to go through the timeline with them and to draft the statement with them. It is important that they sign this statement. Less reliable but sometimes more practical (e.g. for close -out report only) the witnesses can be asked to write their statement and to send it to the investigator(s). All findings should be documented, even if a worker says that he didn‟t see the inc ident, because that information will be helpful if stories change down the road. The purpose of questioning is not to determine who deserves any blame; it‟s to collect as much information as possible about exactly what took place. Investigators should also gather any documentation that would be helpful, from equipment logs to photos and diagrams of the accident scene. As the investigator looks over what he has been able to c ollect, it‟s important to identify any gaps in the information and attempt to fill those gaps through additional investigation. A checklist for initial investigation is attached to this guideline.
4.3. Analysis The investigator(s) shall in this next phase analyze the information that has been gathered. Usually that consists of two separate steps. 1) Is the step where a timeline is prepared and where the investigator(s) gain understanding of what happened? And, 2) Is the s tep where the events or „triggers‟ are linked in a logical way to show how they relate to each other. In this guideline the Bow-tie technique is explained. This technique can be replaced by other techniques if the investigators are more acquainted with such a technique. E.g. Fault tree analysis Event tree analysis Fish bone diagram The “5 why” method
Black-box analysis Etc. During the step where the investigator(s) create the timeline and gain understanding of what happened, it is possible that information need to be verified. This can e.g. include: Maintenance records
Instructions given Toolbox meetings held Training records Standards (Inter)national (Inter)national rules and regulations Handover documents Certificates Docking information Sub-contractor assessments Etc. The Investigator(s) shall seek adequate evidence of their assumptions and arguments in the final report. Note: It is important to know how people respond to investigation results that involve them; “By nature” people will feel attacked (even while this cannot be the purpose of the report) and will start to “defend” themselves. If there is o ne mistake or uncertainty in the report people tend to focus on that issue, leading the attention away from their worries. The report becomes doubtfully and with that also the recommendations in it. Therefore the Investigator(s) shall only incorporate facts that are or can be proven or clarify the status of his assumption clearly.
4.3.1. Bow-tie technique After that the Investigator(s) have completed the time line and understand what happened, they shall create the root cause analysis. Attached to this guideline is a example of the Bow-tie methodology. The Bow-tie technique has been developed by Shell and is used for both, Incident Investigations and Risk Assessments. Shell and its consultants have developed de veloped software for the detailed use for the technique (Thesis) which is nowadays used in oil-and gas, chemical, shipping and airline industries. (In our need we only use the top layer of what it does for these industries, but is suits our purpose well). The name of the bow-tie is only because of its usual shape, nothing else. On the left hand side the triggers are shown. The Top-event is in the middle and on the right hand side the (potential) consequences (further explained in attached template). Building the bow-tie (making the analysis) requires some exercise. Make it on paper p aper first. Using pencil and eraser still works the best and you can concentrate on the process and events instead of the links in Word or o r any other software program. That comes later. Start with the unwanted event in the middle. Then put the triggers for the incident on the left hand side in a logical sequence top-down. Incorporate escalation factors if relevant. Keep on changing that until the process is logical and you think that others can follow the process as well. Then start to include the barriers. Think logical (and discuss) and incorporate the facts that you learned about in the timeline and initial investigation. In case of doubt…leave it out. Than the consequence side and the recovery measures same process in reverse. For (potential) consequences it is important that we choose the correct wordings. It is easy and maybe interesting to exaggerate but it will not lead to correct understanding of the incident by report readers whom weren‟t there. Avoid the use of strengthening words such as: “A huge hole”, “An enormous fire”, “Major damage” etc. It does not bring factual information and works distracting. A hole, A fire, damage will do. If you need to explain the severity or extend of a consequence, make sure you use the facts only. o nly. Be careful with potential worse case consequences (e.g. required in client report). You will rapidly
end up with Multiple Fatalities, Catastrophe, and Disaster etc. if you get carried away. Understandwhat they want to know; the potential po tential worse case under the specific circumstances of that incident. Anyhow, real fixed guideline isn‟t possible here; use your common sense again. Often “standard” terminology can be used for Immediate and Root causes, followed by and explanation. The Standard terminology is needed for the database and statistical trend analysis, an alysis, the explanation for the understanding by the readers of the report. E.g. Material not suitable. 2 ½“ mooring ropes are not sufficient for the mooring this barge. Inspections inadequate. The mooring ropes were in poor condition due to wear and tear. Equipment not available There was no equipment available to retrieve control over the barge. A tentative list with Immediate- and root causes is attached to this guideline
4.4. Reporting Develop the draft report. The report could include: Scope and extend of the Investigation Summary of what happened General information with regard to the incident Equipment involved Timeline Consequences Immediate causes Root Causes (reference to Bow-tie) Post event actions (e.g. Emergency Response) Related aspects (e.g. finding that did not caused the incident) Responsibility (Functions, names & signatures of investigator(s) Recommendations Attachments
5. RECOMMENDATIONS Based on the findings of the incident Investigations the investigator(s) shall now define recommendations for improvement. These action items are meant to reduce the change of reoccurrences of the incident and shall be “SMART” ( Specific, Measurable, Attainable, Relevant and Time –bound). Recommendations shall solve or reduce the root causes of the incident. There is no need for cliché‟s such as: “Good Communication at all times” or “Always focus on the tasks”. In these cases it would be better to put: Standing orders shall be made by the PM that emphasis the importance of Pre -Start meetings, TBT and/or Take 5. Works managers to assign responsible persons for mooring inspections and to establish an procedure that includes criteria for inspection, frequencies and register.
6. REVIEW AND APPROVAL Before the report is submitted to the client and / or Deme and Medco it shall be reviewed by the HSE Manager (or someone else if the report was made by the HSE Manager). The report shall be approved after it is found in line with the procedures and this guideline. Although time pressure may exists for Incident reports, this shall not affect the quality of the reporting.
7. FOLLOW-UP AND CLOSE-OUT OF THE ACTION ITEMS It is obvious that follow-up and close out of the action items that are defined is essential as all efforts of the investigation will be lost if not closed out properly. A system for follow-up and close out of action items is established on the project and shall be maintained at all times.
8. ATTACHMENTS 8.1. Bow-tie The editable template can be found on the xxx Server in the xxx folder
8.2. Checklist Initial Investigation o o o o o o o o
Overview of the site (sketch / photos / drawings etc.) Identification of equipment involved Identification of people involved Identification of witnesses Position(s) in timeline of equipment involved Final position of equipment / material / people Distances Speed
o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o
Time Weather conditions Sea state conditions (wave / swell / current) (force and direction) Illumination conditions Conditions that could have limited judgments Material People PPE Depth Draft List Height Size Numbers Quantities Diameter Surface Damage Failure Logbook Records Sea chart Update status (Sea charts / DGPS / Pilots / Certificates) Calibration status (measuring equipment / critical equipment) Voice Data Vessel tracking Forms (used) Position of switches / handles / buttons Position of safety devices Doors / hatches open / closed / locked Replacements / changes made after the event
8.3. Tentative list of immediate causes Collision Cut / punch Dust Edge Collapse Electrical Failure Excessive forces Falling object Flooding Grounding Hit by an object Horseplay Hose burst Human error Hydraulic Failure Inadequate Hygiene Inadequate Response Loss of Control Lack of oxygen Management Failure Mechanical failure Not applicable Operator Failure Others Over speeding Slip, trip or fall Slippery surface Snapping wire
Systems Failure Unacceptable behavior Wear & Tear Weather
8.4. Tentative list of Root Causes Alcohol Certification failure Communication insufficient Cooperation failed Current Design Error Drugs Emergency Response failure Equipment not available Equipment not suitable Experience inadequate Familiarization inadequate Fatigue Handover not done / insufficient High Tension Housekeeping inadequate Incompatible goals Induction insufficient Inspections not adequate Instructions not followed JSA not available Lack of enforcement Lack of Visibility Lifting gear / SWL i ncorrect Maintenance Inadequate Material not suitable Motivation Not applicable Obstructions Others Personnel inadequate Plan not implemented Planning failure PPE not used / insufficient Preparations inadequate Procedures not followed Qualifications inadequate RA inadequate Short cut / unplanned action Signage insufficient Smoking Standards not followed / available Subcontractor failure Subcontractor selection failure Supervision inadequate Third party Time Pressure / Rush Culture Training inadequate / not provided Underground slippery / unstable Wind / Waves / S well
