October 29, 2014
The Honorable Mathy Stanislaus, Assistant Administrator
Office of Solid Waste and Emergency Response
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, DC 20460-0001
Re: Comments for Docket Number: EPA-HQ-OEM-2014-0328 - Risk Management
Program Request for Information
Dear Assistant Administrator Stanislaus,
We the undersigned organizations are writing to formally submit comments in response to
Docket Number EPA-HQ-OEM-2014-0328 - Risk Management Program Request for
Information (RFI). The organizations sending this letter represent thousands of facilities across
the nation covered by Risk Management Program (RMP) regulations. In addition, the
International Institute of Ammonia Refrigeration (IIAR) is an ANSI accredited standards writing
body whose standards are used as Recognized and Generally Accepted Good Engineering
Practices (RAGAGEP). The questions posed in the RFI are of great interest to our organizations
and member companies and we appreciate the opportunity to provide comment.
Below are our specific comments on the sections most relevant and applicable to our
organizations and members:
Updating the List of Regulated Substances (page 12)
Lowering the ammonia threshold quantity (10,000 lbs.) under the Risk Management program
would require smaller independent companies to needlessly increase their operating costs and
could inflict financial harm. It is estimated that the initial cost for developing a RMP could be
over $100,000.00, with an annual program maintenance costs greater than $25,000.00.
The Clean Air Act Amendments of 1990, Section 112(r)(1) (General Duty Clause) requires safe
operations of facilities/processes with less than threshold quantities of RMP chemicals through
prevention of accidental releases and minimization of consequences of releases that do occur. .
This requirement has been in place since 1990 and the ammonia refrigeration industry has
suitable, less costly management programs in place that will satisfy the requirements of the GDC
without harshly, unduly requirements by lowering RMP threshold for ammonia.
Additional Risk Management Program Elements (page 43)
Revising the management-system elements raises a number of questions and concerns. Because
RMP is supposed to be performance-based, we are opposed to requiring specifying management-
system metrics required by those subject to the regulation. Requiring facilities to use and share
metrics is more prescriptive than a performance-based regulation should mandate.
In addition, RMP regulations already include management practices in almost all elements. For
example, CCPS – RBPS does not appear to really have 20 separate management systems.
Rather, it appears that they have broken out portions of existing elements and subdivided them.
“Process Safety Culture”, “Process Safety Competency”, and “Stakeholder Outreach” RBPS
elements are already being incorporated into the Employee Participation portion of PSM
programs. For those truly unique elements included in RBPS, not all of them may apply to
facilities with ammonia refrigeration systems.
CCPS primarily focuses on the production of chemicals and the processes used to create
chemicals, thus the CCPS should not be used as an applicable source for all RMP processes. We
believe this would stifle innovation and some industry best practices are not applicable to all
other industries. While we recognize continuous improvement and metrics can be useful tools,
we are concerned that they cannot effectively be a measure of compliance or non-compliance.
Therefore, we do not feel continuous improvement and metrics should be added to the RMP
The IIAR’s PSM and Risk Management Program Guidance contains a sample Management-
System document which provides management system guidance directly applicable to the
ammonia refrigeration industry. Adding the requirement for more complex management
processes has the potential to occupy resources that would otherwise be applied to improving
compliance with existing requirements while adding little value towards safety.
Should such additional management system elements ultimately be required, it would be
important to coordinate the requirements between PSM and RMP. When the EPA was
developing their Accidental Release Prevention Requirements (40 CFR Part 68), the initial drafts
for RMP contained substantial differences with the PSM Standard which had been issued four
years earlier. Facilities were rightfully worried that they would have to maintain two separate,
distinct programs: A PSM program to comply with the OSHA requirements and a Risk
Management Program to comply with the EPA requirements. Numerous comments to the EPA
convinced them to include program elements in the RMP which are essentially identical to the
PSM program elements.
With respect to adding the management system elements of (1) “Stop Work Authority”, (2)
“Ultimate Work Authority”, (3) “Conduct of Operations”, (4) “Process Safety Culture”, and (5)
“Job Safety Analysis”, we feel this is too prescriptive for a performance based regulation. In
addition, many (if not all) of these “elements” are already incorporated into the existing Process
Safety Management / Risk Management Programs through defining of: Operating Limits
(§68.69(a)(2)); assignment of responsibility (§68.69(a)(1)); equipment deficiencies (§68.73(e));
Pre-startup review (§68.77); Job Safety Analysis(29CFR 1910.132(d)(2). Other federal
regulations (i.e. 29 CFR 1910.252) and industry standards (i.e. ANSI/IIAR 5-2013) all provide
additional guidance and regulations that the owner / operator is to be following that adequately
address these types of activities. Additional requirements above what is currently stated and
included in existing regulations and through industry standards would not ultimately improve the
“safety” of the system.
Facilities should be free to choose those management system elements which are applicable to
the complexity of their operations and to their industry; they should not be constrained to use
management-system elements which were developed under circumstances which may not apply
to their operations.
Imposing additional safety requirements on contractor owners and operators since
contractors are increasingly used in a variety of roles at chemical process facilities, (page
We do not believe that additional requirements are necessary for contractor owners and operators
in the ammonia refrigeration industry. Existing federal regulations (EPA and OSHA) are very
clear that the “contract owner or operator” is required to ensure training and understanding,
following of safety rules, etc. Modification or additional requirements are not needed or
required at this time.
Ammonia refrigeration industry contractors are essentially used in the same roles today as when
the RMP regulation was first enacted. The current Contractor section clearly lays out the
requirements for Owners/Operators and Contractors and these requirements have been working
in our industry. Since the roles of contractors have not changed in our industry and since the
overwhelming number of incidents does not involve contractors we do not believe additional
safety requirements are necessary.
Clarifying PHA and hazard review requirements: (page 47)
Specifying the types of failure scenarios and damage mechanisms for PHAs and hazard reviews
could be problematic. The implication of such a policy is that EPA would be essentially stating
that they know more about the processes and potential failure mechanisms than the experts
within each industry. In addition, this would seem to indicate that the scenarios and damage
mechanics for all industries are identical. Identification of failure scenarios and damage
mechanism is best be left to those who have the most experience in each unique process which is
Additionally, providing a list of prescriptive failure scenarios and damage mechanics might
actually result in less thorough PHAs and hazard reviews. If a team were provided with such a
list, they might actually review the list as the only comprehensive list of scenarios and damage
mechanisms to analyze. This may essentially minimize or eliminate the “brainstorming”
approach which should be used to generate scenarios and damage mechanics.
The IIAR has provided clear guidance for its members on the failure scenarios and damage
mechanics which should be considered in PHA studies and hazard reviews in the PSM and Risk
Management Program Guidance developed specifically for our industry. But even this document
recognizes that PHA teams must have to freedom to either customize these guidelines or develop
alternative scenarios and damage mechanisms to fit their unique operations.
It is appropriate for EPA to clearly define when a hazard review or PHA should be updated.
There has been much confusion about this issue throughout industry. This definition should not,
however, require facilities to conduct studies for all changes to the process. The majority of
changes are relatively minor and can clearly be addressed using the existing Management of
Change element. PHA and hazard review updates should be reserved for “major” changes to a
We see no advantage to updating PHA studies more frequently that once every five years or
when major changes to systems or processes are to be implemented. The only exception might
be to require an updated PHA study if an incident requires the five-year accident history to be
updated; in this case the updated PHA should only be conducted on the portion of the process
involved in the incident. As stated above, there are already provisions to update PHA studies
and hazard reviews if there are “major” changes to a process. Beyond incidents and changes, we
do not believe there is justification for changing the frequency of PHA and hazard review
Clearly state what modifications would require a change in process safety information and
thus would require the owner or operator to perform a pre-startup review for new
stationary sources and for modified stationary. (page 47)
The existing verbiage in the regulation (§68.77(a)) is very clear on the requirement to perform
the pre-startup review “new stationary sources and for modified stationary sources when the
modification is significant enough to require a change in the process safety information.”
Additional verbiage and/or changes to this section would not be beneficial at this time. Basically
stated, if a Management of Change is required then a Pre-startup Review is required.
Define and Require Evaluation of Updates to Applicable Recognized and Generally
Accepted Good Engineering Practices (RAGAGEP) (page 50)
We agree that adding a definition for RAGAGEP could be useful to help owners better
understand requirements under the standard. A definition for RAGAGEP may also be helpful in
reducing the instances of EPA inspectors citing standards that are not as applicable to a given
type of facility. For example, there have been occasions where EPA inspectors have applied
American Petroleum Institute (API) standards to ammonia refrigeration facilities. Better
defining RAGAGEP can reduce the misapplication of standards by inspectors and facilitate
better understanding and application by facility owners. A definition of RAGAGEP should
include methods and “whys”, but not go so far as how to do something like inspect, which
becomes a maintenance procedure. A definition of RAGAGEP should not take away the ability
of a facility to identify which RAGAGEP they are applying to their operations.
It is incumbent on covered facilities to identify updates to codes and standards of RAGAGEP as
process changes or improvements must be constructed in accordance with the latest revisions of
these docs. The Management of Change (MOC) and Process Hazards Analysis (PHA) sections
of PSM and RMP are currently sufficient to identify risks without a stand-alone requirement for
evaluation of RAGAGEP. The Formosa Plastics incident used as an example is a situation
where a Process Hazards Analysis Revalidation should adequately identify appropriate
RAGAGEP for incident prevention. Furthermore, it is likely impractical for covered facilities to
update all processes to maintain conformance with current standards without other design
changes which could impose a significant economic burden. In many cases it is impossible to
conform with all RAGAGEP because of possible conflicting requirements. The MOC and PHA
elements coupled with Employee Participation and Pre-Startup Safety Review are adequate for
identification of new hazards created by process changes or to identify hazards based on
incidents since the last PHA Revalidation.
There is also concern about the application of manufacturer’s recommendations within
RAGAGEP. While EPA has a tendency to look to manufactures recommendations for industry
standards for mechanical integrity, it is important to understand that manufactures
recommendations should not automatically be considered industry standards or RAGAGEP
because they can be motivated by factors other than the reductions in injuries, fatalities, and
Extend Mechanical Integrity Requirements to Cover Any Safety-Critical Equipment (page
Conceptually, the proposal to expand the coverage of the Mechanical Integrity element to all
safety-critical equipment seems reasonable. For the ammonia refrigeration industry, covered
facilities already must identify components, controls and PM frequency for them in accordance
with OEM recommendations. In addition, IIAR is currently developing IIAR 6 that will specify
the mechanical integrity requirements for all safety-critical equipment in an ammonia
However, for such a change to be effective, a workable definition of “safety-critical” must be
developed. The determination of what is safety-critical can be subject to broad interpretation.
For example, the loss of any utility within the control of the owner could be construed to
represent a significant risk to the process, even when the process is designed to safely shutdown
on a loss of utilities. However, the concern in this case would be that there would likely be no
special stipulations for the utilities out of control of the owner.
Require Owners and Operators to Manage Organizational Changes (page 55)
There is some merit in expanding the Management of Change requirements to include
organization changes, as long as there is clear guidance on what organizational changes qualify.
We believe that the majority of organizational changes do not rise to the level of requiring
inclusion in Management of Change.
The following “Tip” is contained in the management of change section of the IIAR’s PSM and
RMP Program Guidelines and can provide some useful perspective on what organizational
changes should be included in management of change programs:
Tips: Modify these definitions and add additional definitions applicable to your facility.
For example, you might modify the definition of a “change” to include operating and
maintenance staffing changes such as cutbacks in the staffing of operations or
maintenance personnel or to address labor issues such as work stoppages (e.g. strikes).
If organizational changes are included, “replacements in kind” should be exempted from the
management of change requirements to address normal transitions which occur at a typical
facility provided that actual staffing levels or reporting requirements do not change. For
example, if a Refrigeration Technician retires, the facility should not be required to complete the
management of change form if they plan on hiring a new Refrigeration Technician and training
him to replace the Technician that left. However, the elimination of a Refrigeration Technician
position may warrant inclusion. In the spirit of performance based standards, facility owners
should be given a sufficient level of flexibility to design their own programs to meet the
Require Third-Party Compliance Audits (page 56)
Compliance audits are useful tools for evaluating a facility’s safety. However, we are concerned
about the intended definition of “third-party”. We strongly believe that third-party audit should
not be limited to hiring outside personnel to perform the audit. Outside consultants have their
place, but facilities should have the flexibility to utilize internal safety experts from other
facilities or corporate headquarters to perform audits.
A third party doesn’t necessarily equate to more qualified or independent auditors. Often third
parties are motivated to generate work for themselves as a result of an audit. When the RMP
regulation was first implemented there were many contractors that developed auditing
capabilities as a way of generating business. Often the internal auditor is more familiar with the
process and the inherent risks. Internal audit teams are often more thorough than a third party
and share best practices, company policies and experience from other facilities within the same
company. Using internal auditors develops the auditing experience and expertise in-house where
it is more accessible as opposed to losing it to a third party. Process owners should have the
ability to identify the resources required to conduct audits whether it be by independent internal
resource or a third party.
Requiring facilities to hire outside auditors also has the potential to put small businesses at a
disadvantage. Hiring auditors can be a costly process. If a small business has access to qualified
internal auditors, they should have the flexibility to use them.
Should EPA move forward with this proposal, we strongly encourage the use of a definition
similar to the one used in Section 6.6.4 of IIAR Bulletin #110:
This inspection should be carried out by a person who has the training and knowledge for
this task, for example:
An employee of the owner, competent to perform inspections and who is
independent of the daily operating responsibilities for that installation
An independent organization or individual competent to perform
An inspector from the insurance company who is licensed to write
pressure vessel insurance
A licensed inspector from the jurisdiction where the pressure vessel or
shell-and-tube heat exchanger is located.
We are opposed to the agency defining certification requirements for third-party auditors.
Requirements for the credentials of the auditors has some merit, but due to the breath of the RMP
standard, a single standard is not recommended (e.g. an oil refinery is much different than a
refrigeration system). The performance based concept in RMP should be applied here, where the
employer sets the standard for the “qualifications” of the auditors.
In addition, certification is not required for PHA Team Leaders. The PHA process and the audit
process are equivalent and equally important activities, therefore there should not be added
qualifications to perform an audit. The following requirements are consistent with those for
PHA team leaders and should be applicable to auditors:
“One member of the audit team must be knowledgeable in the Risk Management Program
requirements and in auditing techniques.”
The RMP regulation already requires owners/operators to address each finding for a compliance
audit and to document how they address each requirement in writing. Therefore we feel no
additional changes are required. We have concerns with establishing a procedure to appeal the
findings of the third-party auditor presupposes that facilities must implement each finding
identified by a third-party auditor. This would establish a dangerous precedent that we believe it
is against the intent of RMP. As stated earlier, facilities should always have relative freedom to
address, but not necessarily implement each finding, whether the finding was generated by
facility personnel or a third-party auditor.
We do not agree with requiring facilities to perform full compliance audits if they experience an
incident or near miss. These situations should trigger an incident investigation and the incident
investigation should get to the root cause, including which sections if any of PSM and/or RMP
need improvement to prevent another incident and/or near miss. However, making a
requirement of a full compliance audit could possibly influence facilities to report/document and
investigate less incident investigations and near misses because of the concern about the burden
of the requirement to perform a compliance audit.
There should not be a mandated timeframe for completing compliance audit items. Facilities
should have a documented plan with target completion dates and action being shown on each
item to completion. However, some items might require capital funding, which can take time
due to the process and/or lengthy construction time. As long as engineering and/or
administrative controls are in place (if necessary) and the facility can show action on an item
until completion, a mandated timeframe is not needed.
We are opposed to the requirement to mandate specific time frames to respond to issues
identified during an audit since no two audits and no two recommendations are identical. Some
recommendations can be implemented before the audit is completed, others may take years, and
significant capital investments, to fully implement. Having a written action plan and showing
action on items should be acceptable.
We believe that requiring audits every three years is an appropriate timeframe. Auditing more
frequently than every three years is an extra burden on the process owner since there is time and
expense associated with the auditing process whether it is internal or external. Audits are
disruptive to the normal operation of the process since the operating staff must also prepare and
participate in formal audits. This is particularly true of smaller entities whose staffs handle
multiple responsibilities. Increasing the frequency of auditing to annual or bi-annual is no more
likely to improve compliance than the current three year requirement.
Effects of OSHA PSM Coverage on RMP Applicability (page 59)
Non-refrigeration ammonia users and/or chlorine users typically have greater inventories of
identical chemicals, yet are not held to the same standards as the refrigeration industry.
Locations of many exempted chemical users are in close proximity to rural refrigeration systems.
Enhanced educational programs and compliance assistance support offered by EPA for these
small entities would likely enhance safety where prevention practices are not being used at the
The Program 2 elements do not have a place for ammonia users. Facilities are either Programs 1
or Program 3. Most, if not all, are Program 3 and as such must comply with OSHA PSM
Standards. As such, the EPA should consider adopting or mirroring the OSHA PSM Standard
and simply having a separate element for the Offsite Consequence Analysis, Management
System and Risk Management Plan requirements.
Safer Technology and Alternatives Analysis (page 63)
We are very concerned with mandating inherently safer technology reviews. The regulatory
burden of requiring costly IST reviews tends to stifle innovation. For those companies who are
already looking to improve safety by implementing IST options, a formal IST review would add
costs to a process by forcing them to document the activities they are already performing. Small
operations might not have the manpower or expertise to do this and lack the resources to hire it
out cost effectively.
For those companies who do not implement IST options, the IST review becomes a “paper
exercise” where they document why it is “infeasible” to implement these options. If RMP
facilities are required to perform safer alternative options analyses and implementation plans,
EPA should not require that the analyses and/or implementation plans be submitted to the
agency. Likewise, EPA should not have any role in analyzing or approving such plans.
We are strongly opposed to requiring the involvement of local communities in IST analysis.
While local communities have good intentions, they have virtually no expertise in safety and risk
management as it applies to a specific industry, and often little knowledge of why processes and
systems are necessary for economic and material well being of the community and the country.
For example, in industrial refrigeration, there is no substitute refrigerant for ammonia that is
available on the scale that is required to produce food for human consumption. A public demand
to replace ammonia would, if even feasible, would drive up consumer prices and could possibly
force the closure of facilities. In addition, the public would tend to feel that a company has
“deep pockets” and thus should be able to implement all IST options regardless of feasibility.
Also, there are security issues related to revealing issues related to IST options and sensitive
information could be compromised by involving local communities.
Emergency Drills to Test a Source’s Emergency Response Program or Plan (page 71)
Requirements for tabletop drills and/or functional emergency response exercises on alternating
years will likely enhance a chemical user’s awareness and preparedness for emergencies.
Recommendation or suggestion of inclusion of local emergency responders would be reasonable
to enhance effectiveness of coordination requirements already in effect in the Risk Management
Program Rules. Requirements for inclusion of offsite emergency responders, however is not
recommended as these resources are not controlled by covered facilities. Staffing and funding
limitations may inhibit some responders’ exercise and/or coordination abilities. Covered
facilities must not be penalized if offsite responders are not able to participate.
Automated Detection and Monitoring for Releases of Regulated Substances (page 75)
RMP should remain performance based, and adding specific requirements for detection and
monitoring would be counter to performance based. RAGAGEP should be used to determine
if/when detection and monitoring devices should be used. The IIAR has clear guidance on the
requirements for detection and monitoring devices in the latest version of IIAR 2 and we feel that
industry experts should be used to determine when/where these devices should be used, not
government regulators. We feel that EPA should allow facilities to install these devices based on
industry standard and guidelines, not regulatory restrictions made by personnel who may have
little to no understanding of the nuances of each industrial sector.
Below are responses to specific points raised under this section:
Ammonia Refrigeration facilities are required to install monitoring equipment and
sensors per our industry codes and standards.
This is something that should ultimately be required by industry specific
organizations and incorporated into that industry’s codes and standards.
We have strong concerns that this opens the door for the detection devices,
sensors, panels, etc. to be potentially required to meet requirements by a testing
laboratory, UL, or similar.
These monitoring systems should only contact the Owner / Operator which would
then initiate the Emergency Response plan and make necessary notifications.
This is something that should ultimately be required by industry specific
organizations and incorporated into that industry’s codes and standards.
g. There should be no requirement for alarm/alert activation to responders and the
public (see “c” above). It is very rare that facilities are able to perform emergency
response drills with local responders due to time and availability restraints on the
local responders’ parts. It is an unreasonable request to expect the LEPCs and
SERCs to participate, review, and have involvement in the development process
of every location’s emergency response plan.
Monitoring equipment / sensor testing requirements should be developed and
provided by the manufacture. Any specific time limit / requirement included in
the regulation would not be beneficial (or reasonable).
With the vast difference in “emissions” versus the ammonia refrigeration
industry, there really is no valid or reasonable method that would allow this
similar type of program to be successful.
Should not be required other than through industry specific codes and standards.
Additional Stationary Source Location Requirements (page 79)
The current regulatory climate already dictates that facilities consider buffer and setback zones
and occupancies when siting new facilities. The issues related to occupancy are already
addressed by industry applicable standards such as ASHRAE 15 and is currently being
incorporated into IIAR 2. However, providing specific requirements is, as stated numerous
times, against the performance based nature of the RMP regulation.
Introducing siting restrictions such as distance from the covered process to occupancies could
become a problem in situations where a facility might want to upgrade to a new and safer engine
room but the best location is not compliant with siting restrictions. If adopted, this policy may
ultimately encourage facilities to keep running older (potentially less safe) facilities rather than
construct newer, modern (and potentially safer) facilities due to the burden involved with the
specific requirements which would be imposed by the EPA.
Compliance with Emergency Response Program Requirements in Coordination with
Local Responders (page 82)
Coordination with local emergency planning and response authorities is an important aspect of
safety. As a result, such coordination is already specifically required in the RMP Rules.
Enforcement of this issue is linked through Hazard Communication, Emergency Action and
HAZWOPER Standards. The coordination with local agencies (e.g. LEPC, Fire Department,
Police, etc.) is required by the EPA’s Chemical Accident Prevention Provisions (40 CFR Part
68.95(c)). An argument can be made that the issue of coordination is already well covered and
that adding requirements would be redundant.
It should also be noted that some facilities may not be in an area with a LEPC that could/would
assist in an event which would also impose and unnecessary burden on facilities. Additionally,
there may be circumstances wherein facility personnel are more trained and qualified through
HAZWOPER training and practice than the LEPC. While we recognize inclusion of the LEPC
for a covered process facility is mostly beneficial, it may not always be the most beneficial
course of action for a facility depending on experience and training.
A clearly defined reasonable level of coordination with planning and response authorities
presents a significant challenge. If facilities are “required” to coordinate their response
activities, EPA must recognize that despite the best efforts of facilities sometimes the
coordination is a “one-way” street. It is very difficult at times to get the local responders to visit
each PSM-covered facility regardless of the urgency and the number of the invitations provided.
If a facility can demonstrate that they provided appropriate information to the off-site responders
and contacted them on a regular basis that should be sufficient to demonstrate that they have
attempted to encourage coordination with these responders.
In addition, there needs to be recognition of the ability of an industrial response team that deals
with fire, bomb threats, boiler alarms, etc. A trained operator should be allowed to engage the
plant’s emergency shut-down procedure while working with the same PPE that they would be
using when opening the system for repairs (APR, full skin protection, ammonia monitor, and
The following information from the emergency planning and response section of the IIAR’s
PSM and RM Program Guidelines provides a useful perspective on the issue of coordination:
10. Coordination with Off-Site Responders
The following elements of the emergency plan are coordinated with off-site responders:
(1) Procedures that will be followed to inform the local emergency response agencies
about emergencies at the facility
(2) Procedures that will be followed to conduct search and rescue operations
(3) First aid and emergency medical procedures that will be followed to treat accidental
(4) Procedures that will be followed to respond to accidental ammonia releases
(5) Amount and type of emergency equipment that is available on-site and through
(6) Training and drills that will be conducted with outside agencies
(7) Decontamination and clean-up procedures
(8) (Insert Additional Elements if Applicable)
To address the above, the following procedures will be followed to verify that our
emergency plan is coordinated with off-site responders:
(1) Copies of the Emergency Plan will be sent to each of the off-site responders for their
review and comments whenever changes are made to the plan.
(2) Representatives from each of the off-site responders will be invited to tour the facility
and talk with facility personnel to review the Emergency Plan on an annual basis.
(3) Off-site responders will be invited to participate in any related training and drills
with facility personnel.
(4) (Insert Additional Procedures if Applicable)
Tips: Expand and/or modify this section to reflect the procedures followed to
coordinate the emergency action and/or response plan with off-site responders.
For example, insert additional procedures if applicable.
All facilities are encouraged to develop written or oral arrangements of
responsibilities regarding emergency response activities with their local emergency response agencies. These arrangements should address the following
names and/or titles of primary and secondary contacts;
telephone numbers and pager numbers for 24-hour contact; and
a brief description of major responsibilities of each party.
Incident Investigation and Accident History Requirements (page 86)
Incident investigation requirements as currently written are sufficient for determining the cause
of accidents. OSHA PSM 29CFR 1910.119 Appendix C(12) Incident Investigations, addresses
details and provides compliance guidance on the intent of incident investigation and the
applicability of incidents that require investigations to include near misses.
There is value in having a good definition of what constitutes a “near miss” and a “process
upset”, but proper development of the definitions will be critical. Root cause analysis is a
common practice of investigating incidents but it is not the only method used. For example the
“Five-Why” approach is in use by many companies. Facilities should be free to determine the
exact technique used provided that the major causes and all contributing factors are identified as
part of the incident investigating technique.
We believe it would be difficult, if not impossible, to specify time frames for investigations to be
finished. Investigations of relatively straight-forward incidents can be completed in a matter of
days. But serious incidents which involve, for example, expertise by specialists and examinations
by outside laboratories could take a year or more.
Modification to 5-year accident history reporting requirements would potentially lead
to the requirement for locations to meet the requirements of a different “Program”
level. The benefit outlined in the RFI for “disseminating lessons learned across
industry” is not a valid basis for expanding this requirement. The information that is
reported through the RMP reporting (CDX) process on a 5-year accident does not
currently include any information that could lead to the industry fully understanding
(i.e. the factors leading up to the release, release response, root cause(s), remediation
activities taken, etc.) and being able to extrapolate ways to make the process “safer”.
It is unclear as to what the benefit of a plant conducting a “Compliance Audit” on all
program elements due to having an “incident or near miss” would benefit the plant or
There is no clear benefit for improving safety of the process by sharing incident
investigation results or including the local community in the conducting of the
investigations. Requiring this in the regulation is not going to benefit anyone but will
lead to fewer incidents and near misses being investigated (locations will not want to
“share” that they had or could have had a catastrophic release). Also, this could lead to unfavorable public perception of the plant and company – a potentially huge
economic impact to business and local economy.
Worst Case Release Scenario Quantity Requirements for Processes Involving
Numerous Small Vessels Stored Together (page 91)
In an ammonia refrigeration system, it is virtually impossible to release the entire ammonia
inventory in any single event. Therefore to require the entire system inventory for the process to
be used for the worst-case scenario would be require facilities to analyze a scenario which is
virtually impossible and thus create an unrealistic picture of risk at the facility. Nearly all
incidents, including very serious incidents, at ammonia refrigeration systems involve releases
from a single vessel.
Accident history for the ammonia refrigeration industry shows that the Worst Case Scenario
stipulations for RMP is not practical and is unreasonable for emergency planning in the
refrigeration industry. Ammonia vessels do not explode like process reactors in petrochemical
facilities. Fires involving ammonia equipment are not made worse by ammonia leaks. Typically
ammonia is found to remain in pressure vessels after the fires have been extinguished. A
problem with a vessel is not likely to result in a release from an adjoining or collocated vessel in
an ammonia refrigeration system.
The current definition is clearly understood and unambiguous and it is not recommended to
change worst case scenario requirements to evaluate all vessels together. The changes suggested
could increase confusion due to their ambiguous nature of the requirements and thus lead to a
lack of uniformity in worst-case scenarios and would not present a true representative of risk in
Public Disclosure of Information to Promote Regulatory Compliance and Improve
Community Understanding of Chemical Risks (page 94)
The information contained in the RMP which is submitted to EPA is readily available on-line
and there are already requirements for the distribution of Tier II reports to appropriate local
authorities. Allowing more information to be disclosed (audits, PHA, etc.) would have the effect
of compromising public safety and security by allowing sensitive information be too readily
available. Due to the confidential nature of this information, we should be taking steps to avoid
disclosure, not increasing its availability.
Threshold Quantities and Off-site Consequence Analysis Endpoints for Regulated
Substances Based on Acute Exposure Guideline Level Toxicity Values (page 96)
Based on the current AEGL-2 levels for anhydrous ammonia, we do not oppose using AEGL-2
to recalculate RMP reporting thresholds and toxic endpoints for off-site consequence analyses.
Program 3 NAICS Codes Based on RMP Accident History Data (page 105)
With the exception of some facilities located in California and one facility in New Jersey, all
ammonia refrigeration facilities in the United States are Program 3 facilities since the OSHA
threshold for ammonia is identical to the EPA threshold for ammonia (10,000 pounds). Therefore
this change should have minimal impact on our industry.
The “Safety Case” Regulatory Model (page 109)
We are strongly opposed to shifting to the “Safety Case” regulatory model. Facilities have
invested a lot of time and resources and in many cases are justifiably proud of their PSM and RM
Programs. To suddenly move away from these programs in favor of a method not commonly
practiced in the United States seems highly inappropriate.
Streamlining RMP Requirements (page 112)
None of the issues listed in this RFI would make it easier for regulated sources to comply with
RMP. Therefore, minimizing the number of changes to the existing regulation would have the
greatest impact on streamlining.
Thank you for the opportunity to provide comment on the future direction of the CFATS
program. Please let us know if you have any questions about our submission or if we can be of
any assistance as the rulemaking process moves forward.
American Frozen Food Institute
American Meat Institute
Global Cold Chain Alliance
International Institute of Ammonia Refrigeration
International Association of Refrigerated Warehouses
Refrigerating Engineers and Technicians Association
U.S. Poultry and Egg Association