The Occupational Safety and Health Administration (“OSHA”) published a Request for Information (“RFI”)  on December 9, 2013 concerning possible changes to the Process Safety Management (“PSM”) program codified at 29 C.F.R. 1910.119.  See 78 Fed. Reg. 73756 (Dec. 9, 2013).  Likewise, the Environmental Protection Agency (“EPA”) published an RFI on July 31, 2014 relating to possible changes to the similar Risk Management Program (“RMP”) rules codified at 40 C.F.R. Part 68.  See 79 Fed. Reg. 44604 (July 31, 2014).  At the time of this writing, the respective comment periods have closed and we are waiting to see new proposed regulations. This is the sixth article in a series of articles concerning these potential rulemaking actions.

OSHA and the EPA requested comments concerning revising the mechanical integrity requirements in the PSM and RMP rules to expand the scope to include unspecified safety-critical equipment.  Comments received by OSHA and the EPA requests were similar in nature.  Although many comments addressed safety-critical equipment, this article will compare comments from three organizations:  the American Fuel & Petrochemical Manufactures (“AFPM”), the Mary Kay O’Conner Process Safety Center (“MKOPSC”) at the Texas A&M Engineering Experimental Station, and the U.S. Chemical Safety Board (“CSB”).

Both PSM and RMP rules require that an employer or operator maintain a mechanical integrity program that includes written procedures for inspection and testing pressure vessels, storage tanks, piping systems, relief and vent systems and devices, emergency shutdown systems, controls, and pumps.  See 29 C.F.R. 1910.119(j)(1)-(2) and 40 C.F.R 68.73(a)-(b).  The written procedures and the inspection frequency must be consistent with recognized and generally acceptable good engineering practices “RAGAGEP.” See 29 C.F.R. 1910.119(j)(4) and 40 C.F.R. 68.73(d).

When promulgating the PSM rule in 1992, OSHA recognized that it was untenable to classify individual pieces of equipment as critical (or not) with regulatory certainty.  “Equipment considered critical to a process by one employer may not necessarily be considered critical to a different process by another employer.  As a result, there could be confusion with respect to which equipment is subject to the [mechanical integrity] requirements.” 57 Fed. Reg. 6356, 6389 (Feb. 24, 1992).  As a practical matter, OSHA concluded that “there is certain equipment, critical to process safety that is common to all processes.”  Id. However, OSHA then added “if an employer deems additional equipment to be critical to a particular process, that employer should consider that equipment to be covered by this paragraph and treat it accordingly.”  Id.  That said, it should be noted that the preamble language is permissive (should) and the requirement to inspect and test other safety-critical equipment was not included within the rule.  In publishing the RFI, the agencies are requesting comments as to whether the mechanical integrity section should be revised “to explicitly apply the mechanical-integrity requirements of the PSM standard to all equipment that the employer identifies as critical to process safety.”  78 Fed. Reg. at 2013.

Both agencies requested comment as to how employers or operators identified other equipment that was safety-critical.  78 Fed. Reg. at 73766 and 79 Fed. Reg. at 44617.   Whereas OSHA did not provide any suggestion as to what equipment could be affected, the EPA identified “additional types of equipment and systems that could reasonably be judged to be critical to process safety.”  79 Fed. Reg. at 44617.  Examples provided by the EPA “include computer software systems that interact with process components, electrical power systems, and other utility systems[1] that interact with pumps, valves, or control systems.” Id.

In response to the RFI, the CSB stated that PSM and RMP “should require companies to identify their safety-critical equipment/elements (“SCE”) and demonstrate to the regulator that each SCE has a performance standard that addresses functionality, availability, reliability, survivability, and other interactions with other systems as well as a verification scheme.”   CSB Comment to OSHA, p. 19 (Mar. 31, 2014) and CSB Comments to EPA, p 24 (Oct. 29, 2014).  Unfortunately the CSB does not provide any insight as to how a company should identify safety-critical equipment but instead effectively anticipates regulatory uncertainty by adding “the regulator could provide a list of common suggested SCE and/or criteria for what constitutes a SCE to assist the company in ensuring all SCE are identified and managed.”  Id.  Further, notwithstanding the open-ended scope, the CSB fails to describe how the “performance standard” suggestion differs from following RAGAGEP.[2]

While embracing the use of a list to identify the majority of safety-critical equipment, the MKOPSC believes that any list will likely be incomplete and miss some equipment that is critical to a specific process.  Whereas the MKOPSC believes that the list should be maintained, it asserts that safety-critical equipment should be defined to include “all equipment which [has] been identified through the [Process Hazard Analysis] as potential causes for catastrophic release of [a] hazardous chemical.”  MKOPSC Comments to OSHA at 35 (Mar. 31, 2014).

The AFPM expressed concern about having an open-ended mechanical integrity scope as “there is no general consensus among our members as to what safety-critical equipment consists of.”  AFPM Comments to EPA, Section IV (Oct. 29, 2014). Perhaps of most concern, without a functional definition of safety-critical that is neither vague nor ambiguous, “the phrase ‘safety-critical’ will be misused in the enforcement process to justify post-incident notice of violation.”  Id.  As a result, the AFPM commented that “if the EPA is aware of an existing hazard with certain pieces of equipment the Agency has deemed safety-critical, it should specify the equipment of concern and seek public comment on adding specific equipment to Sec. 68.73.” Id.

Interesting, the EPA separately notes that the mechanical integrity requirements apply specifically to emergency shutdown systems, “however, the regulation does not explicitly require that all covered sources install emergency shutdown systems.”  79 Fed. Reg. at 44617.  As a result the EPA requested comments as to whether the mechanical integrity section should be supplemented or clarified, presumably to require emergency shutdown systems.[3]  In response, the AFPM commented that “our members are very concerned that EPA is moving away from performance-based toward prescriptive regulations.”  AFPM at VI.  The CSB responded by encouraging the “EPA to specify situations where covered sources should be required to install emergency shutdown systems (and maintain them per the existing requirements in §68.73).”  CSB Comments to EPA at 24.  According to the CSB, in 2002, a release of 48,000 pounds of chlorine occurred due to a ruptured transfer hose that could have been minimized had the unloading facility been equipped with an emergency shutdown system. Id. at 24-25.  That said, in a 2007 “Safety Bulletin,” the CSB noted that the Chlorine Institute requires that its members install “remotely operated or automatically actuated emergency shutoff valves systems in place which can safety isolate both ends of the transfer hose/flexible piping.”[4]  The CSB also notes that “the Chlorine Institute members produce 98 percent of chlorine manufactured in the United States and Canada.”  Id.  No information was provided with the CSB comments as to the magnitude of any residual risk.[5]

In conclusion, both OSHA and the EPA requested comments regarding the scope of equipment that should be covered under the mechanical integrity regulations.  Although both agencies suggested an open-ended approach by seeking to cover all safety critical equipment, OSHA, the EPA, and the CSB (in their comments) did not provide any definitions or guidance to assist in identifying safety-critical equipment.  The EPA identified a few examples of equipment that could be safety-critical.  The AFPM commented that any additions to the list should be proposed and go through public comment.  Finally, the MKOPC suggested using the Process Hazard Analysis process to identify additional safety-critical equipment (beyond the enumerated list).


[1] Although silent within the RFI, OSHA has separately provided guidance concerning the applicability of PSM to utility systems:

It is OSHA’s position that if an employer determines that a utility system or any aspect or part of a process which does not contain an HHC [highly hazardous chemical] but can affect or cause a release of HHC or interfere in the mitigation of the consequences of a release, then, relevant elements of PSM could apply to these aspects. OSHA’s position is that any engineering control, including utility systems, which meets the above criteria must be, at a minimum, evaluated, designed, installed, operated (training and procedures), changed, and inspected/tested/maintained per OSHA PSM requirements. 

OSHA Guidance letter, Richard Fairfax, Director, to Howard Feldman (Jan. 31, 2008), found at (last visited Apr. 9, 2015).

[2] The CSB suggests that safety-critical equipment should be tested and inspected in accordance with a performance standard.  This suggested performance standard addresses the functionality, availability, reliability, and survivability of the equipment.  Under the current rules, employers and operators test and inspect in accordance with procedures that follow RAGAGEP.  Typically, the RAGAGEP is based on an industry code and standard.  Presumably, these codes and standards are developed considering the same sort of considerations: functionality, reliability, etc.  Other codes and standards, such as construction RAGAGEP, may address design issues like survivability (i.e., specifying the amount and type of insulation required to survive a pool fire).  As such, the CSB’s comment may actually address another request in the RFI, i.e., requiring use of updated codes and standards in the design of existing equipment.

[3] It is curious that the EPA would use mechanical integrity requirements to require a process be constructed to a code or standard.  Such would be more consistent with process safety information requirements that equipment comply with RAGAGEP.  See 40 C.F.R. 68.65(d)(vi)and (viii) and (d)(2).  As such, the question posed by the EPA appears to be a back-door attempt to define RAGAGEP, a subject separately included in the RFI.  79 Fed. Reg. at 44617.

[4]U.S. CSB, Safety Bulletin, No. 2005-06-I-LA, Emergency Shutdown Systems for Chlorine Transfer (June 2007), found at (last visited Apr. 10, 2015).

[5] The CSB specifically references The Chlorine Institute Pamphlets 85 as describing practices that are mandatory to its members as well as their customers.  As such, these should be considered a RAGAGEP.  Residual risk would either result from the failure of Chlorine Institute members (and their customers) to follow the cited industry practice or from non-members that fail to take the same precautions.