Revisions to the EPA’s Risk Management Plan Involve Accident Investigation & Root Cause Analysis
President Obama issued Executive Order 13650 that directed agencies to improve chemical safety performance. In response, the EPA is proposing changes to the RMP (Risk Management Plan) regulation. A preliminary copy of the changes have been published HERE (they have not yet been published in the Federal Register).
For readers interested in root cause analysis, the main changes start on page 28 in the Incident Investigation and Accident History Requirements section.
The revision to the regulation actually mentions “causal factors” and “root causes” that were not mentioned in the previous regulation. On page 33 the revision states:
Thus EPA is proposing to require a root cause analysis to ensure that facilities determine
the underlying causes of an incident to reduce or eliminate the potential for additional accidents
resulting from deficiencies of the same process safety management system.
The EPA document uses the following definition of a root cause:
Root cause means a fundamental, underlying, system-related reason
why an incident occurred that identifies a correctable failure(s) in management systems.
The revision document gives examples of poor investigations of near-miss accidents that did not get to root causes so that a future accident that included a fatality or severe injuries occurred. These examples include and explosion and fire at a Tosco refinery, an explosion at a Georgia-Pacific Resins facility, an explosion an fire at a Shell olefins plant, and a runaway reaction at a Morton International chemical plant. In each case, root causes of issues were not identified and fixed and this allowed a more serious accident to eventually occur.
Of course, I have said many times that I’ve never seen a major accident that didn’t have precursor incidents (call them near-misses if you must). Performing adequate root cause analysis of smaller incidents has always been one of the goals that we have suggested to TapRooT® Users and now even more fully support with the new Using the Essential TapRooT® Techniques to Investigate Low-to-Medium Risk Incidents book.
The document asks for comments on the proposed revision to the regulation (page 41):
- EPA seeks comment on whether a root cause analysis is appropriate for every RMP reportable accident and near miss.
- Should EPA eliminate the root cause analysis, or revise to limit or increase the scope or applicability of the root cause analysis requirement?
- If so, how should EPA revise the scope or applicability of this proposed requirement?
- EPA also seeks comment on proposed amendments to require consideration of incident investigation findings, in the hazard review (§ 68.50) and PHA (§ 68.67) requirements.
- Finally, EPA seeks comment on the proposed additional requirement in § 68.60 to require personnel with appropriate knowledge of the facility process and knowledge and experience in incident investigation techniques to participate on an incident investigation team.
In the document, there is extensive discussion about defining and investigating near-misses. The section ends with …
- EPA seeks comment on the guidance and examples provided of a near miss.
- Is further clarification needed in this instance?
- Should EPA consider limiting root cause analyses only for incidents that resulted in a catastrophic release?
The document also discusses time frames for completing investigations. Should it be 30 days, 60 days, six months? It’s interesting to note that many investigations of process safety incidents by the US Chemical Safety Board takes years. The EPA is suggesting that a one year time limitation (with the possibility of a written extension granted by the EPA) be the specified time limit.
The EPA is asking for feedback on this time limit:
- EPA seeks comment on whether to add this condition to the incident investigation requirements or whether there are other options to ensure that unsafe conditions that led to the incident are addressed before a process is re-started.
- EPA also seeks comment on whether the different root cause analysis timeframes specified under the MACT and NSPS and proposed herein will cause any difficulties for sources covered under both rules, and if so, what approach EPA should take to resolve this issue.
The document also discusses reporting of root cause information to the EPA and suggests that common “categories” of root causes be reported to the EPA. The document even references an old (1996) version of the TapRooT® Root Cause Tree® and a potential list of root cause categories, They then request comments:
- EPA seeks comment on the appropriateness of requiring root cause reporting as part of the accident history requirements of § 68.42, as well as the categories that should be considered and the timeframe within which the root cause information must be submitted.
Although I am flattered to be the “father” of this idea that root causes should be reported so that they may be learned from, I’m also concerned that people may think that simply selecting from a list of root causes is root cause analysis. Also, I’ve seen many lists of root causes that had bad categorization. The main problem is what I would call “blame” categorization. I’m not sure if the EPA would recognize the importance of the structure and limits that need to be enforced to have a good categorization system. (Many consultants don’t understand this, why should the EPA?)
As everyone who reads the Root Cause Analysis Blog knows, I am always preaching the enhanced use of root cause analysis to improve safety, process safety, patient safety, quality, equipment reliability, and operations. But I am hesitant to jump aboard a bandwagon to write federal regulations that require good management. Yes, I understand that lives are at stake. But every time a government regulation is written, it seems to cement a certain protocol and discourages progress. Imagine all the improvements we have made to TapRooT® since 1996. Would that progress be halted because the EPA cements the “categorization” of root causes in 1996? Or even worse… what if the EPA’s categories include “blame” categories and managers all over the chemical industry start telling investigators to stop looking for other system causes and find blame related root causes? It could happen.
I would suggest that readers watch for the publication of EPA’s revision of the RMP in the Federal Register and get their comments in on the topics listed above. You can’t blame the EPA for making bad regulations if you don’t take the opportunity to comment when the comments are requested.