Incident Investigation Sample Program
Incident Investigation Sample Program
Incident Investigation Sample Program
(Sample)
Overview
Purpose
This program was developed with the involvement of the organizations management team, technical staff, and hourly employees to ensure that accidents and near misses, particularly those of catastrophic magnitude or potential, are: thoroughly investigated relevant findings are implemented, and results are communicated throughout the organization. The goal of this program is to identify root causes of incidents and address the causes through corrective actions in order to prevent reoccurrence. Note: Assignment of blame to individuals is not productive and should not be a part of the incident investigation process.
Scope
All incidents that result in, or could reasonably have resulted in, the following are investigated: an uncontrolled release of toxic materials, fires, explosions, significant equipment / structural damage, serious personnel injuries, injuries to the public, environmental impacts, and/or a significant impact on reliability, productivity goals, and/or customer satisfaction.
The scope includes injuries to contractor employees, contractors, visitors, and damage to equipment owned by contractors, employees, or visitors. This also includes unexpected shutdowns of equipment, failing to meet chartering requirements, voyage delays, and damage to cargo.
Document Own er
Definitions
Incident
An unplanned sequence of events and/or conditions that results, or could have reasonably resulted, in a loss event.
Accident
An incident with unexpected or undesirable consequences. The consequences may be related to personnel injury or fatality, property loss, environmental impact, business loss, etc., or a combination of these.
An incident or series of incidents that results in: (1) one or more fatalities, (2) multiple serious injuries to personnel, (3) significant property damage, (4) imminent and substantial endangerment to public health, (5) significant environmental damage, (6) a catastrophic financial loss or property damage (>$250,000), or (7) more than 25 similar customer complaints.
An incident, other than a catastrophic accident, that involves: (1) a single serious injury to personnel, (2) serious injuries to an individual, (3) major property damage, (4) minor impact to public health, (5) minor environmental damage, (6) a major financial loss or property damage (>$50,000 but <$250,000), or (7) more than 5 but fewer than 25 similar customer complaints.
Any incident other than a catastrophic or major accident (e.g., an incident that ): (1) does not involve a serious injury, (2) results in a minor financial loss or property damage [>$5,000 but <$50,000] or (3) results in five or fewer similar customer complaints).
Consequences
Definitions, Continued
Near Miss [NM] An incident with no consequences, but could have reasonably resulted in consequences under different conditions. OR An incident that had some consequences that could have reasonably resulted in much more severe consequences under different conditions.
Serious Injury
An injury requiring immediate medical treatment at shore-based facilities (e.g., an emergency room or a doctors office).
Loss Event
Event
A happening caused by humans, automatically operating equipment / components, external events or the result of a natural phenomenon
Condition
A mode or state of being. Note: Includes process states, such as pressure, temperature, composition and level. Also includes the state of training of an employee, the condition of raw material and supplies, and the state of equipment. If negative, then it can be a causal factor, intermediate cause, or root cause.
Causal Factor
Structural/Machinery/Equipment/Outfitting problems, human errors, and external factors that caused an incident, allowed an incident to occur, or allowed the consequences of the incident to be worse than they might have been.
Problem
Structural/Machinery/Equipment/Outfitting performance that deviates from the desired performance of the item.
Human Error
Definitions, Continued
External Fact ors
Issues outside the control of the organization. Examples include uncharted / unknown hazards to navigation, some sea or weather conditions, suicides or homicides, and external events.
Intermediate Cau se
An underlying reason why a causal factor occurred, but it is not deep enough to be a root cause.
Item-of-Note (ION )
A deficiency, error, or failure that is not directly related to the incident sequence that is discovered during the course of the investigation.
Root Cause
Deficiency of a management system that allows the causal factors to occur or exist.
Management Syst em
A system put in place by management to encourage desirable behaviors and discourage undesirable behaviors.
Safeguard
A physical, procedural, or administrative control that prevents or mitigates consequences associated with an incident.
Recommenda -tion
Resolution
An analysis that identifies the causal factors, intermediate causes, and root causes of an incident and develops recommendations to address each level of the analysis.
An analysis that identifies the causal factors for the event and develops recommendations to address them, but does not necessarily identify the root causes of the incident.
Classification of Incidents
Introduction
The organization applies appropriate resources to adequately investigate catastrophic, major, and minor incidents, as well as near misses. Because of the varying levels of risk and the desire to focus investigation resources to manage the most significant risks, the company uses different types of investigation teams, as well as different levels of investigation/documentation for each category of incident .
Role of Vessel The vessel safety officer (or appropriate shore-based personnel for incidents that Safe occur at shore-based facilities) classifies the event to determine the appropriate investigation protocol. ty Offic er
The organizations Incident Investigation Manager will review this classification and adjust the classification (if necessary) of the reported incident.
Company management may choose to modify the classification of an incident based on extenuating circumstances.
Incident Reporting
Initial Notif Personnel are to immediately notify the vessels safety officer of incidents. icati on
Role of Vessel Safe The safety officer determines the appropriate classification for the investigation. ty Offic er
Any incident involving personal injury must be reported immediately to the organizations Safety Manager.
The notification process must not hinder the immediate dispatch of an emergency response team to the incident site when necessary.
The Emergency Response Plan controls immediate notifications required to organization management and outside agencies.
Acute losses are usually reported by personnel in the field. However, chronic losses must usually be identified by examining incident data.
Investigation of chronic events should concentrate on those types of risks that contribute the most to the overall risk of the organization. This means that events that occur at high frequencies and/or have significant consequences should be the highest priorities. These events should be the highest priorities because they represent the greatest potential opportunities to reduce the overall risk levels of the organization .
To identify candidate chronic events for incident investigations, incidents are grouped to determine the dominant factors that are contributing to risk.
This investigation can be performed using a variety of techniques such as Pareto investigation, failure modes and effects analysis, and fault tree analysis.
Intent
The intent is to identify the characteristics of the dominant loss events. Once the dominant failure types have been identified, incident investigation can be used to determine the causes of the events.
Because the identification methods are standard risk analysis techniques, the details required for the investigation should not be covered in the incident investigation program but within the organizations procedures and training programs.
Investigation Team
Team Req uire men ts
Although the size and composition of an incident investigation team vary based on the incidents classification, each incident investigation team should meet the following composition requirements: At least one person knowledgeable in the process or activity involved A team leader and/or others with appropriate knowledge and skills to thoroughly investigate and analyze the event.
Team Leaders
Investigation team leaders must have received basic training in the requirements of this incident investigation program and in basic investigation techniques.
For personnel who may lead investigations of catastrophic incidents, additional training in more advanced investigation approaches may be necessary, at the discretion of the Incident Investigation Manager.
Table 2, Typical Investigation Team Structure for Each Incident Classification, describes the typical investigation team structure for each incident classification.
Reporting incidents as described in Section 6 table. Individuals shall notify the safety officer that an incident has occurred. Completing Incident Investigation Initial Witness Statements. These forms should be completed by all individuals involved in or witnessing an incident. Assisting the incident investigation team in investigating the event.
The Vessel Safety Officer will appoint the team members and ensure that the investigation is begun within 24 hours of discovery. The Vessel Safety Officer will ensure that the area is secured to prevent further injuries and equipment losses. The investigation must be completed as soon as possible, and results must be documented and sent on to the appropriate member(s) of management. Upon review of these results, management determines and initiates further investigation if necessary
For
The Vessel Safety Officer immediately notifies the Corporate Safety Manager, Cata who appoints an appropriate incident investigation team and leader. stro The Corporate Safety Manager will determine the scope of each investigation. phic The Corporate Safety Manager will ensure that the area is secured to prevent or further injuries and equipment losses and to ensure proper emergency response for Majo the incident. r Acci dent s
Investigation The investigation team follows the basic investigation procedure outlined in the Tea organizations incident investigation program. The investigation team is responsible m Resp for the following: onsi- Beginning the investigation within 24 hours whenever possible and no later than 48 hours biliti es Completing the investigation as soon as possible Documenting the results, including recommendations Submitting the report to the Safety Manager for subsequent review, distribution, and communication.
Team Leader
The investigation team leader is responsible for communicating additional resource needs (e.g., expertise) to management when necessary to properly conduct the investigation.
This process involves gathering information related to the event(s) in order to understand what occurred. Note that the level of effort should be greater for events with greater actual or potential losses.[All accidents]
Step
Inspect the scene and the structures/ machinery / equipment/ outfitting involved Obtain on-the-spot information from eyewitnesses, if possible Schedule interviews with those directly involved as soon as possible
Action
Stabilize the vessel / equipment / process in a safe condition Once stable, secure the area to preserve physical data so it is not disturbed Have witnesses complete an Initial Witness Statement from the MaRCAT toolkit Interview those who were injured (if any) and others whose input might be useful Interview those directly involved in the incident as soon after the incident as possible Conduct interviews privately and individually so that the comments of one witness will not influence the responses of others Document the results of these interviews Photographs Field sketches Structures Equipment components Outfitting items Videos Missile maps (for projectiles) Chemicals Product samples Other
Prepare visual aids of the affected physical data for the investigation Determine the physical data that are relevant to the investigation
Obtain samples of unknown spills, vapors, residues, etc. Develop test plans for the analysis of each item of physical data, including chemical samples Perform the analysis of the equipment components and samples, following the test plan for each Review all sources of potentially useful documentation / information
Note conditions that may have affected the samples Have other interested parties agree to the test plan before physical data are examined When a preliminary analysis reveals that an item / sample may have failed to operate correctly, was damaged, etc. make arrangements to either preserve the items or carefully document any subsequent repairs or modifications. Computer logs Drawings Customer records Written logs Charts Previous incident reports Manuals MOC* records Safety, hazard, engineering analyses Safety Procedures Test records QA records Training and performance records of those involved Maintenance Procedures
Examine the applicable written procedures Determine which incident-related items should be preserved, and establish chain-of-custody to control these items / samples Carefully document the sources of information contained in the incident report * - MOC = Management of Change
Operating procedures
Access to these items should be controlled Note: This will be valuable should it subsequently be determined that further study of the incident is necessary
Develop an understanding of the causes of the event using a simplified fault tree, a causal factor chart, or other appropriate methodology to structure each investigation.
The description of the incident facts (events and conditions) will include timing information to the extent practical.
The causal factor chart is typically the primary investigation tool for incidents involving timing and people actions. A causal factor chart is constructed by working backwards from the end result (the ultimate consequence of the incident) and by letting the questions generated by each step backwards drive the data-collection efforts. For each step taken backwards, the sufficiency of the facts should be tested to ensure the completeness of the chart. This questioning will lead the investigators to collect the data necessary to determine any conditions that must have existed or events that must have occurred.
Fault trees (or why trees) are typically the primary investigation technique for equipment/outfitting/structural issues and chronic problems. The fault tree should be developed level by level, identifying the potential causes of the event above. The tree that is developed should be as small as possible by truncating branches as soon as possible. Branches should be trimmed when the past experience indicates the risk associated with the branch is low or when data or information indicates that the branch is not possible or likely.
Suppositions included on the data analysis charts/trees are clearly distinguished from facts (such as by using dashed lines under or around suppositions). All data sources should be pursued to convert the supposition into a fact.
Continued on next page
The focus of charting the incident should be to direct the data collection process to determine what happened, how it happened, when and where it happened, what actions were taken or not taken, and who was involved. While the organization understands that nearly all incidents result from human error (except natural disasters), the organization also understands that placing blame on individuals is inappropriate in nearly all cases. The facts will be established, including human errors committed, and then the root causes of the errors will be determined as described later.
List
List alternative scenarios when the precise scenario cannot be definitively Alter established because of missing or contradictory information. In some cases it may nati not be ECONOMICALLY feasible to collect data even though it is TECHNICALLY ve feasible. Sce nari os
[CA, MaA]
Identify Cau sal Fact ors [All] Identify all the causal factors.
Identify potential management system weaknesses that explain why the causal factors either occurred or existed.
Step 2
Determining root causes often requires more data collection, but focus the data collection on the management systems that were in place to control the human activities and equipment integrity/reliability.
Step 3
Use the Marine Root Cause Analysis Map to provide structure and consistency to the results.
Step 4
Document the paths through the Marine Root Cause Analysis Map.
Direct
Develop recommendations that are directly related to a causal factor in the incident.
Four Levels
Recommendations should address all of the following four levels: Level 1: Recommendations to address the causal factor Level 2: Recommendations to correct the intermediate causes discovered as part of this investigation Level 3: Recommendations to correct other similar problems that exist on the vessel or in other areas of the organization (other vessel and/or shore facilities) Level 4: Recommendations to either improve or augment existing management systems or reduce the likelihood or consequence of incidents by adding or improving safeguards (which in turn require sufficient management systems to ensure that the features remain sufficiently reliable).
Practical
Recommendations should be practical, feasible, and achievable, and should reduce the risk of future incidents to acceptable levels.
Flexible
Recommendations may (and many times should) allow for a variety of resolutions.
Items-of-Note
Recommendations related to Items of Note should be documented in a report/memo to management, separate from the investigation report.
When an accident or near miss is discovered, it is an opportunity to examine the potential consequences of the incident, in addition to the actual consequences. By doing this, the potential risk associated with the incident are examined. In other words, if the incident had happened under slightly different circumstances, could the result have been catastrophic, or is this as bad as it can be? By estimating the potential outcomes, the proper level of response to the incident can be assessed.
Generally qualitative estimates of the potential outcomes for the incident are used. It is not practical to develop quantitative estimates of the potential consequences for each incident. Therefore, the incident investigation team will often use a loss potential matrix to estimate potential consequences. Although this is a very subjective estimate, it will provide the guidance needed to develop effective corrective actions and to perform incident trending.
To estimate the loss potential for an incident, the investigation team must estimate the probability of recurrence and the potential severity. The following two tables provide the categories to estimate these two parameters.
Probability of Recurrence
Category Frequency 1 Less than once in 10 years 2 Once in 10 years 3 Once a year 4 Once a month or more
Potential Consequences
Category Personal Consequences Equipment / Property Damage A First Aid Injury $ 1,000 $ 10,000 B Medical Treatment Injury > $10,000 $ 100,000 C Permanent / Disabling Injury > $ 100,000 $ 1,000,000 D Fatal Injury > $1,000,000
The probability of recurrence should estimate the probability that the incident occurs again, assuming that no corrective actions are taken. When estimating the probability of recurrence, the following factors should be considered: (1) the number of people and the number of components/equipment/vessels/etc., and (2) the number of times the activity is performed. For example: If a failure of each pump is expected to occur once a year and there are 12 pumps on board, the expected probability of recurrence is 1/month (Category 4). A procedure that is used once per year contains an error. When the procedure is performed as written, a small amount of hazardous material is dumped on to the deck. The probability of recurrence is once per year because the procedure is only performed at this frequency (this assumes there is only one piece of equipment that uses this procedure).
When estimating the potential consequences, consider what other events could reasonably occur, not the worst possible event that could occur. For example, a fire in a trash can in the lunch room could result in sinking a vessel. However, it is much more likely that the worst potential consequences of this incident would be the destruction of a small portion of the vessel, some personnel injuries, and a minor effect on the schedule.
Reporting Requirements
Team Leader The team leader is responsible for ensuring that, at the conclusion of the Res investigation, the Incident Summary form and supporting documentation are pon sibili prepared. ties
The purpose of the report is to help others understand the incident and the corrective actions that are recommended to prevent recurrence of the same incident and other similar incidents.
The report, regardless of the type of incident, will contain as a minimum: Date and time of the incident Date and time the investigation started A description of the incident Identification of causal (contributing) factors Identification of root causes Recommendations from the investigation List of investigation team members and their roles.
The level of detail required will be related to the actual and/or potential risks associated with the incident(s). Additional supporting documentation may include the following: Parts testing/examination reports Witness statements Causal factor chart Fault tree Incident investigation forms Test plans Photographs or videotapes Maps and diagrams.
Each recommendation should be coupled with a brief description of the rationale so that people not involved in the investigation (e.g., management) can understand the recommendation.
The Safety Manager is responsible for retaining the approved report for at least 5 years.
Report
The reports should be available for use during the next proactive analysis of the Avai systems/equipment/process/vessel involved in the incident, training sessions, safety labili meetings, and subsequent investigations. ty
Report
The completed reports and documented resolutions of the recommendations will be Distr distributed to the vessels so that they can communicate these to personnel who work ibuti in the affected area and/or perform job tasks relevant to the investigation findings. Contract employees are included in these reviews when applicable (e.g., a contract on worker was involved in the incident, a contract employee performed an activity related to the incident, or a contract employee was injured).
This review is accomplished by routing a copy of the approved report to potentially affected personnel and by discussing the incident in a safety meeting.
Safety
The Safety Manager is responsible for sending out copies of the report and collecting Man and retaining completed (i.e., signed) routing forms or safety meeting agendas and ager attendance lists. Res pon sibili ties
Each recommendation is assigned by the Safety Manager or the Assistant Vice President Operations to a responsible person who prepares a recommendation tracking form and issues it to the personnel assigned to implement the recommendation.
Designated personnel respond to each assigned recommendation by either resolving the recommendation or documenting the rationale for modifying or rejecting the recommendation.
Typical reasons for rejecting a recommendation are: Implementation of the recommendation would increase the overall risk of operations The recommendation is no longer valid Implementation of other team recommendations adequately address this recommendation The risk reduction associated with this item can be accomplished by a more effective (less costly, less complicated, or greater risk reduction) action The recommendation is not necessary to protect the health and safety of personnel or the environment, and/or The recommendation is infeasible.
Personnel assigned responsibility for resolving recommendations provide periodic updates on the status of recommendations to the Safety Manager.
Quarterly
Upd ates
The Safety Manger retains the final (complete) recommendation tracking summary (and completed recommendation rejection forms, if applicable) in an incident file, and documentation of the final resolutions are transmitted to the vessels to allow communication to the affected employees.
Trending
The Safety Manager will trend the results of the incident investigations. This will consist of collecting and analyzing information related to incidents.
Requirements Incident information that will be included in the incident investigation database for Data include: base Date and time of the incident Date and time the investigation started The process/equipment/items/vessels involved in the incident Environmental conditions at the time of the incident Identification of causal (contributing) factor types and numbers Identification of root causes codes from the Marine Root Cause Analysis Map. tm Recommendations from the investigation Groups responsible for the implementation of recommendations.
The Safety Manager will periodically analyze the information contained in the database to determine the effectiveness of the incident investigation program.
Training Requirements
Training Poli cy
All employees receive instruction in identifying incidents requiring investigation. All contract employees receive this instruction from their own supervisors through required contractor safety orientations.
The Safety Manager ensures that training programs for employees and contractors include criteria and examples for identifying incidents requiring investigation.
Team leaders receive a minimum of 3 days of formal training in investigation methodology, including: (1) Effective methods for gathering data and data control, (2) Causal factor charting method, fault tree analysis, or the 5-Whys technique ( or any combination of these) for analyzing the data that are gathered, (3) Marine Root Cause Analysis Map tm methodology, and (4) Guidance for writing effective recommendations and reports.
One of the challenges we face is to continue our efforts to improve [safety/ reliability/ quality] performance. In order to achieve our goal of [an accident-free workplace/improved reliability/improved quality], we need to eliminate not only the [incidents /loss events] themselves, but also the underlying conditions that create the potential for them to occur. If we are going to be successful in accomplishing this, it is critical that we determine the root causes of these [incidents/loss events]. We must go beyond addressing the symptoms to address the underlying root causes of these [incidents/loss events]. Unless we are certain that the root causes are identified and actions are taken to eliminate them, we cannot ensure that the incidents will not occur again. We have begun taking steps to improve the process we use for investigating [incidents/loss events]. Recently, we provided training to XX individuals in incident investigation methods. The method of incident investigation that we are training our personnel to use provides a structured process for gathering information and identifying root causes. This new process is used not only for [incidents involving injury/significant losses], but also for near misses. Near misses are incidents in which [no one is seriously injured/there are no significant losses] but there is a potential for [serious injury/serious losses]. It is important for everyone to understand that the intent of this process is not to find fault or place blame. It is, by design, a process for identifying failures or weaknesses associated with a [safety/reliability/quality] management system. Once the root causes are identified, we will develop recommendations to eliminate the root causes and set individuals up to succeed in future operations. Punishment of employees involved in investigations will NOT occur unless they are involved in illegal activities such as use of drugs, stealing, or sabotage. We have already started performing incident investigation using the personnel we have recently trained. This requires that those individuals be released from their normal duties to collect information, conduct interviews, analyze the incidents, determine the root causes, and develop recommendations. As a result, other people will need to fill in for those conducting the investigations or, in some cases, work may get delayed. Preventing someone else from getting hurt far outweighs the temporary inconvenience resulting from the persons participation in the investigation process. As people conduct more investigations, the time required will decrease. We, as members of the [company/division/organization] leadership team, support this investigation process and ask that employees support the efforts of their co-workers when they are asked to participate. Signed, The Management Team