PT Notes
EPA RFI - Incident Investigation and Accident History Requirements
PT Notes is a series of topical technical notes provided periodically by Primatech for your benefit. Please feel free to provide feedback.
On July 24, 2014, the U.S. Environmental Protection Agency (EPA) announced a Request for Information (RFI) seeking comments on potential revisions to its Risk Management Program (RMP) regulations and related programs to modernize its regulations. Multiple issues were addressed in the RFI.
One of the issues addresses incident investigation and accident history reporting which can provide valuable information about potential hazards and the steps needed to prevent future events. The RMP rule's incident investigation requirements closely track those established in the Occupational Safety and Health Administration's (OSHA's) Process Safety Management (PSM) standard. Likewise, EPA's hazard assessment requirements include a five-year accidental release history, which has some overlap with similar process hazard analysis (PHA) requirements in OSHA's PSM standard.
RMP facilities must investigate each significant incident which resulted in, or could reasonably have resulted in, a catastrophic release. However, the accident history requirement does not include near-miss accidents or accidents with only the potential for consequences.
EPA's experiences with RMP facility inspections and incident investigations show there have been incidents that were not investigated, even though under slightly different circumstances, the incident could have resulted in a catastrophic release. Additionally, there have been some cases where the facility chose not to conduct an investigation because the owner/operator elected to decommission the process involved, or because the process was destroyed in the incident. While an investigation would have no impact on a decommissioned or destroyed process, other similar processes or operations at the facility, or at similar facilities, could potentially benefit from its findings.
In other instances, facilities have failed to investigate serious releases because they determined there were no actual or potential off-site consequences. Investigating these types of incidents and including them as part of the RMP accident history report could provide facilities with important information on what problems could lead to an incident, and allow for the facility to address them before a catastrophic release occurs. Furthermore, including some of these incidents as part of the accident history could also improve process safety at facilities with similar processes, where operators could learn from the shared information.
Furthermore, EPA has discovered situations where incident investigations by regulated facilities have been delayed indefinitely. Conducting investigations as soon as possible after an incident may yield better quality data and information. Delays could result in an increased risk of incident recurrence as root causes and the appropriate corrective actions are not necessarily identified or implemented promptly. EPA did amend the RMP accident history requirements to require facilities that have had a qualifying accident to update their RMP accident history to include the new information within six months of the date of the accident.
EPA is soliciting comments on whether broadening the incident investigation and accident history requirements to include clear requirements to investigate near misses and determine root causes of accidents, near misses, and process upsets, and establishing specific time frames for incident investigations to be completed, would promote increased safety.
Questions posed by EPA on this issue include:
- Are the RMP incident investigation requirements too narrowly focused? Would identifying a broader range of incidents requiring investigation (e.g. near misses) help prevent additional accidental releases? Please provide specific examples where possible. EPA requests information on alternative definitions or incident classifications that could be included within the RMP rule's incident investigation requirements.
- Are there any data or information on process upsets, near misses or other incidents that were not required to be investigated, but where an investigation and resulting changes in management systems might prevent accidental releases?
- Does your facility routinely investigate incidents not required to be investigated under the RMP rule? If so, please describe the types of incidents investigated, and the effects these investigations have had on facility operations.
- Would a specific time frame for incident investigations to be completed benefit overall safety? What should be the basis for establishing an appropriate timeframe requirement for an incident investigation to be completed? What are the challenges and limitations to completing an incident investigation within a specified timeframe?
- Are there benefits from requiring that investigations must be performed even in cases where the owner/operator elects to decommission the process involved, the process is destroyed in the incident, or a facility determines there were no actual or potential off-site consequences? Would such a requirement provide a disincentive to decommission potentially risky processes?
- Would a modification of the definition of "catastrophic release" assist in addressing the concerns regarding the appropriate scope of incidents that require investigation?
- Would a modification of the accident history reporting requirements to reflect a broader range of incidents being investigated assist in disseminating lessons learned across industry?
- Should EPA require facilities that have incidents or near misses to conduct a full compliance audit?
- Is it appropriate for facilities to share the results of accident investigations with the local community, or, alternatively, a summary of the accident, and its root causes? Is there an appropriate role for the local community in conducting investigations?
- What would be the economic impact of broadening the RMP incident investigation requirements to require root cause investigations of near misses? Are there any special circumstances involving small entities that EPA should consider? Would small businesses have the capacity to investigate near miss incidents?
