^__ — United States Science Advisory EPA-SAB-EEC-93-004,
*" CDA Environmental Board (A-101) December 1992
Protection Agency
AN SAB REPORT:
REVIEW OF HYDROGEN
FLUORIDE STUDY:
REPORT TO CONGRESS
REVIEW OF THE OSWER/CEPPO
DRAFT HYDROGEN FLUORIDE
STUDY: REPORT TO CONGRESS
.L,J Floor
-------
V UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
| WASHINGTON, D.C. 20460
EPA-SAB-EEC-93-004
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
December 9, 1992
Honorable William K. Reilly
Administrator
U.S. Environmental Protection Agency
401 M Street, SW
Washington, D.C. 20460
Subject: Science Advisory Board report on review of the
OSWER/CEPPO draft document "Hydrogen Fluoride
Study: Report to Congress"
Dear Mr. Reilly:
The Science Advisory Board (SAB) has completed its review of the Office of
Solid Waste and Emergency Response (OSWER), Chemical Emergency
Preparedness and Prevention Office (GEPPO) "Hydrogen Fluoride Study: Report to
Congress" (May 1992 Draft) and is pleased to submit this report. The attached
report stems from a public meeting conducted on July 7 and 8, 1992 by the
Hydrogen Fluoride Review Subcommittee (HFRS) of the Environmental
Engineering Committee (EEC), supplemented with invited experts from academia,
industry and environmental groups. This group reviewed the draft document,
received briefings from the Agency's OSWER/CEPPO staff managers and scientists
who developed the document, debated technical arguments, and offered advice to
the staff on the current draft document.
In contrast to most SAB projects, in this instance we were not asked to
review a completed document. Certain essential elements that were missing from
the draft document are scheduled to be added later. Specifically, we are alerting
you to the fact that the SAB has not reviewed the Findings and Recommendations
that will eventually be included in the final version of the Agency's Report to
Congress. However, we also would like you to be aware that we did in fact accept
this review without the Findings and Recommendations, because of the importance
of making sure that the technical issues are properly understood. Without
reviewing the full document, it is impossible to establish whether conclusions
drawn are supported by the technical details that are presented in the draft
document. The Subcommittee strongly recommends that the full report, including
RtcycM/Ttecyeltblt
PrtnMd on p«p«r thai oonuin*
« toMt 75% mcydad Itotr
-------
the findings and recommendations, be reviewed when it becomes available in order
to ensure scientifically sound interpretation of data.
We commend the Agency's CEPPO staff for collection of a large amount of
background information on the properties, hazards, industrial uses, regulations,
and accidental releases of hydrogen fluoride in the short time dictated by the
Clean Air Act Amendments. We note that the CEPPO staffs approach is basically
sound. Despite the limitations listed below, the draft is well-organized and well-
written, providing good, though limited, background information about the
chemical, its properties, and current practices associated with its production and
use. We also note that the CEPPO staff has properly focused their strategy on
the uses, properties and hazards associated with the anhydrous form of hydrogen
fluoride (HF). Throughout the report, the SAB refers to HF as the anhydrous
form, unless otherwise noted.
Consistent with the draft HF report to Congress, the HFRS identified HF
production and use to be important to a wide range of stakeholders. For example,
HF is used in a wide range of industries including alkylation catalysis in the
production of clean fuels, production of fluorocarbons, nuclear applications,
aluminum production and production of various chemical derivatives. Increased
regulation of HF could potentially have far reaching impacts with extension to
other chemicals. Thus, it is important for EPA, the Office of Management and
Budget (OMB), the Congress and others to be sensitive to this issue and made
aware of the specific nature of those impacts.
In keeping with the tenets of Total Quality Management (TQM), we
encourage the Agency to go through a process of "aligning with its customer" for
development of this document. If Congress intends for the Agency to conduct an
in-depth analysis of the uses of HF in the economy, then considerably more work
needs to be done, since the current draft document is incomplete in this area.
Further, even if Congress has not explicitly asked for it, we believe that the
document should include an assessment of alternatives to HF, since there may be
a Congressional sense that the use of HF should be limited or eliminated —without
adequately considering the consequences; cf., studies conducted by other groups
(e.g., American Petroleum Institute) and agencies (e.g., California, South Coast Air
Quality Management District). We believe that the customer (Congress) should be
made aware of the progress that has been made in preventing HF-related
accidents and in ameliorating the associated harmful effects. In short, the
Agency's document should put the hazard issues associated with HF into a larger
context. For instance, the Committee raises the point that the hazards of HF
appear not to be unique relative to other industrial chemicals, such as chlorine,
fuming sulfuric acid, phosgene, and ammonia.
-------
Also, the authors of the draft Report to Congress would do well to study
the February 26, 1992 F. Henry Habicht, II memorandum on Risk
Characterization and the appended Risk Assessment Council (RAC) Guidance for
Risk Assessment. In that document, the Deputy Administrator, and the RAC
provide explicit guidance to Agency personnel on how to develop exposure
scenarios, both "worst-case" and "best estimate." The current draft document
seems uncertain in how to respond to the Congressional request to consider "...a
wide range of events, including worst-case accidental releases." The Risk
Characterization memorandum and RAC attachment explicitly explains how this
can and should be done. The Subcommittee recognizes that the CEPPO staff is
currently working to quantitatively address these hazards, as has been
recommended in this review.
The draft document utilizes a computer model to estimate exposures
associated with releases of HF. We refer the Agency to an earlier report from the
SAB, entitled Resolution on Use of Mathematical Models by EPA for Regulatory
Assessment and Decision-Making (EPA-SAB-EEC-89-012). That report addresses a
number of concerns with the use of models that should also be explicitly
considered in this document as a Report to Congress; e.g., input scenarios and
model validation.
Finally, we would like to stress that the CEPPO staff should continue to
examine the newly-passed (February 24, 1992) Occupational Safety and Health
Administration (OSHA) 1910 "Process Safety Management (PSM) Rule" relative to
accident prevention for highly hazardous chemicals. The Subcommittee
recommends that the CEPPO staff expand its discussion of the PSM standard, and
more importantly, address its role in controlling accidental releases of HF. This
newly enacted rule establishes performance standards for safety management of
hazardous materials, and HF is one of many chemicals covered by this rule.
The SAB has offered a number of broad-ranging and specific
recommendations to improve the quality and usefulness of the current draft
document. In summary, the draft document is a reasonable work-in-progress.
However, limitations identified in the body of this letter need to be addressed
prior to the finalization of this report. In this manner the revised report would
have increased utility as a Report to Congress, and for this body to make decisions
based on sound technical information regarding HF.
-------
We are most pleased to have had the opportunity to be of service in the
preparation of this important document and look forward to your response.
Sincerely,
Dr. William Randall Seeker, Chair
Hydrogen Fluoride Review Subcommittee
Environmental Engineering Committee
Science Advisory Board
Mr. Richard Conway, Chair
Environmental Engineering
Committee
Science Advisory Board
Dr. RaymomTC. Loehr, Chair
Executive Committee
Science Advisory Board
-------
NOTICE
This report has been written as a part of the activities of the Science
Advisory Board, a public advisory group providing extramural scientific
information and advice to the Administrator and other officials of the
Environmental Protection Agency. The Board is structured to provide a balanced,
expert assessment of scientific matters related to problems facing the Agency.
This report has not been reviewed for approval by the Agency; hence, the
comments of this report do not necessarily represent the views and policies of the
Environmental Protection Agency or of other federal agencies. Any mention of
trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
ABSTRACT
The Hydrogen Fluoride Review Subcommittee (HFRS) of the Environmental
Engineering Committee (EEC) of the EPA Science Advisory Board has reviewed
the Office of Solid Waste and Emergency Response (OSWER), Chemical
Emergency Preparedness and Prevention Office (CEPPO) draft Report to Congress
entitled "Hydrogen Fluoride Study: Report to Congress," May 1992 draft and
offered a number of recommendations.
The HFRS agrees that Hydrogen Fluoride (HF) production and use is
important to a wide range of stakeholders. Increased regulation of HF could
potentially have far-reaching impacts if extended with regard to other chemicals.
The Subcommittee suggested that a study using life cycle analysis concepts of
health, environment and safety could be undertaken concerning both the use of
HF and alternatives to the use of HF, noting that Congress should be advised
whether alternatives to HF have substantial risk. The Report to Congress should
indicate the implications of the findings of this study on the evaluation of hazards
associated with other industrial chemicals.
The HFRS recommended that the Agency employ a more rigorous definition
of the concepts of hazards, consequences and worst-case scenarios, and that a
credible worst-case accidental release scenario be developed. The HFRS made
substantial recommendations on the use of dispersion models as they apply to
various accident scenarios, and a number of other specific recommendations
intended to improve the draft report to Congress.
Key Words: Hydrogen Fluoride, Hydrogen Fluoride Study, Hydrofluoric Acid,
Report to Congress
11
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
Science Advisory Board
Environmental Engineering Committee
Hydrogen Fluoride Review Subcommittee
Dr. Wm. K«nd«ll Seeker, Senior Vice President, Energy and Environmental
Research Corp., Irvine, CA
Members &
Dr. Linda M. Abriola, Associate Professor, Department of Civil and Environmental
Engineering, University of Michigan, Ann Arbor, Michigan
Dr. George F. Carpenter, Michigan Department of Natural Resources,
Environmental Response Division, Lansing, Michigan
Mr. Richard A. Conway, Senior Corporate Fellow, Union Carbide Corporation, s.
Charleston, WV
Dr. Wayne M. Kachel, Technical Advisor, Pilko & Associates, Inc., Houston, TX
Dr. Ishwar P. Murarka, Senior Program Manager, Land & Water Quality Studies,
Environmental Division, Electric Power Research Institute, Palo Alto, CA
Dr. Robert B. Pojasek, Vice President, Corporate Environmental Programs, GEI
Consultants, Inc., Winchester, MA
Dr. Paul V. Roberts, Professor of Environmental Engineering, Department of Civil
Engineering, Stanford University, Stanford, CA
Dr. Walter M. Shaub, President, The Corporation on Resource Recovery and the
Environment, Washington, DC
in
-------
Invited Experts
Mr. William J. Hague, Supervisor, Process Engineering, Allied-Signal, Inc.,
Morristown, N.J.
Dr. Jerry Havens, Distinguished Professor, Department of Chemical Engineering,
University of Arkansas, Fayetteville, Arkansas
Dr. Fred Millar, Director of the Toxics Project, Friends of the Earth, Washington,
D.C.
Science Advisory Board Staff
Dr. K. Jack Kooyoomjian, Designated Federal Official, US EPA, Science Advisory
Board (A101-F), 401 M Street, SW., Washington, DC 20460
Mrs. Diana L. Pozun, Staff Secretary
Dr. Donald G. Barnes, Staff Director, Science Advisory Board
IV
-------
TABLE OF CONTENTS
1. EXECUTIVE SUMMARY 1
2. INTRODUCTION 6
3. GENERAL COMMENTS 8
4. MODELING 10
5. WORST-CASE SCENARIOS 13
6. DOSE RESPONSE 15
7. ACCIDENT PREVENTION 15
8. OTHER ISSUES RAISED BY THE HFRS 16
APPENDK A - REFERENCES CITED A-l
APPENDIX B - ALOHA MODEL RUN RESULTS B-l
APPENDIX C - GLOSSARY OF TERMS AND ACRONYMS C-l
-------
1. EXECUTIVE SUMMARY
The Science Advisory Board (SAB) has completed its review of the Office of
Solid Waste and Emergency Response (OSWER), Chemical Emergency
Preparedness and Prevention Office (CEPPO) draft document entitled "Hydrogen
Fluoride Study: Report to Congress," May 1992 Draft, (See Appendix A - reference
#10). At a public meeting on July 7 and 8, 1992, the Hydrogen Fluoride Review
Subcommittee (HFRS), of the Environmental Engineering Committee (EEC),
supplemented with experts from academia, industry and environmental groups,
reviewed the draft document. The Subcommittee also received briefings from the
CEPPO staff managers and scientists who developed the document (See, for
instance, Appendix A-reference #9), received public comments (See, for instance,
Appendix A-reference #3), debated technical arguments, and offered advice to the
OSWER/CEPPO staff on the current draft document.
The Agency study was required pursuant to the Clean Air Act (CAA)
Amendments of 1990, Section 301(n)(6) (See Appendix A-reference #8). It should
be emphasized that Congress did not specify whether the study should address
hydrofluoric acid or anhydrous hydrogen fluoride or both. The Subcommittee
concurs with the CEPPO staffs focus on the uses, properties and hazards
associated with anhydrous hydrogen fluoride. While the CAA references
hydrofluoric acid (HF), the Subcommittee, for the purposes of this report, refers to
HF as anhydrous hydrogen fluoride. Our findings and recommendations are aimed
at improving the current draft document (hereafter referred to as the draft Report
to Congress, the draft document or the HF Study). We commend the CEPPO
staff for collection of a large amount of background information on the properties,
hazards, industrial uses, regulations, and accidental releases of hydrogen fluoride
in the short time dictated by the Clean Air Act Amendments. We found the draft
document to be well organized, well written, concise and representative of the
special properties, production and uses of hydrogen fluoride. However, it is not
clear that the data gathered to date are sufficient to allow EPA to make a
reasonable and comprehensive assessment of the potential risks associated with
the production, storage or uses of Hydrogen Fluoride (HF) to surrounding
communities. We noted that "hazards" are considered intrinsic properties of the
potential to do harm from the use of HF whereas "risks" resulting from exposure
to hazards are what relate extrinsically to the community. The current draft
document lacks the scientific detail requisite for a thorough technical review.
Also, some of the data appear to be out of date, due to the long time required to
gather the data. For example, the future market assessment for use of HF should
be updated in light of acceleration of the adoption of the Montreal protocol for
phase-out of chlorofluorocarbons (CFC's) (See Appendix A-references #5 and #13).
-------
Our major findings and recommendations of the HFRS are summarized
below:
1) The report did not include any definition of the approach to be used
to synthesize the findings and recommendations or to derive the conclusions.
Therefore, a clear assessment of the adequacy of the data base could not be made,
since the use of the data is not defined. The HFRS recommends that, as a
minimum, the final draft report, including the findings and recommendations, and
public comment docket, be peer-reviewed with the same rigorous process that the
preliminary draft has undergone, including stakeholder reviews [stakeholders
include governmental groups, labor, industry, trade associations, public interest
groups, professional societies, and state and federal government agencies (See
Appendix A, reference #10, page 3], The SAB could be requested to review the
final document if the Agency desires.
2) The HFRS identified HF production and use to be relevant to a wide
range of stakeholders. Increased regulation of HF could potentially have far-
reaching impacts with extension to other chemicals.
3) The HFRS suggests that "chemical use trees" should be used to
inform the reader and foster greater awareness of the families of chemicals which
are derived from HF.
4) The HFRS suggests that the Agency should not be so narrow as to
only respond directly to Congress's request, but should attempt to anticipate and
address the questions which their Report to Congress will generate. The most
obvious of such questions is: Is HF so dangerous that its use should singularly be
severely restricted, or is it similar to other materials, all of which may require
special precautions?
5) It is conceivable that alternatives may be as risky or perhaps more
risky than HF. The actual hazards associated with use of substitutes do not
appear to have been adequately considered, especially concerning total process
implications such as materials transport and handling and component
regeneration. A comprehensive study using life cycle analysis concepts of health,
environment and safety should be completed on both the use of HF and its
alternatives. The evaluation of alternatives should be more completely considered
in the report and compared on the basis of all hazards associated with the total
process (i.e., life cycle implementation). If the recommended comparative life cycle
analysis of alternatives cannot be conducted in a timely manner and consistent
with Congressional mandates, then the Report to Congress should recommend this
-------
critical activity as future work. Congress must be aware that the alternatives to
HF do exist and they may have substantial risk. The Subcommittee further notes
that until the life cycle analysis is performed, the above observation remains a
presumption and not an assertion.
6) HF has unique properties and hazards which should be addressed.
However, the section on properties and hazards is almost entirely qualitative,
apart from the table of basic physical properties and the graphical comparison of
guideline exposure levels. The numerous chemical and physical reactions unique to
HF, some of which may create dangerous situations, appear to be only summarily
described. The draft report did not adequately quantify the importance of these
unique properties as they may contribute to accidental releases. The corrosive
nature of the chemical and the heat release associated with reactions of HF with
various caustic materials have contributed to past releases. Nonetheless,
management of HF is not altogether significantly different from a wide range of
commonly used industrial chemicals, such as chlorine, fuming sulfuric acid,
phosgene or ammonia. For this reason, the approach used in this study could
have broader implications for the manner in which hazards associated with other
chemicals are evaluated in the future. The report to Congress should clearly state
the implications of the findings of this study on the evaluation of hazards
associated with other industrial chemicals.
7) The HFRS suggests follow-up visits to sites where chemical safety
audits were previously undertaken as one possibility to more completely determine
how current practice has changed since the audits.
8) The Subcommittee had substantial concerns regarding the dispersion
modeling used in the assessment of worst-case accident scenarios. The detailed
nature of the model in relation to the role of the dispersion modeling was not
clearly stated in the draft report and, therefore, the HFRS was not able to
comment on the adequacy of the modeling for evaluations to be performed.
Nonetheless, the HFRS can make several recommendations on the use of these
models in the context of regulatory decision-making. The HFRS refers the
CEPPO investigators to the SAB Resolution on Use of Mathematical Models by
EPA for Regulatory Assessment and Decision-Making (EPA-SAB-EEC-89-012) (See
Appendix A - reference #15). This resolution addressed a number of concerns
with the use of models that should be considered in this study. These include
such concerns as input scenarios and model validation. In the draft report, little
or no justification was given pertaining to the selected model inputs.
Furthermore, the presented simulations are clearly inadequate to fully explore the
consequences of postulated scenarios. A number of specific recommendations
pertaining to improving modeling efforts are contained in the body of this report.
-------
9) The definition of hazards, consequences and worst-case scenarios are
central to the request from Congress for this study. However, no clear definitions
were provided in the draft report, and definitions discussed during the briefing
were rather arbitrary. It was not made clear to the HFRS whether the
Congressional intent was to evaluate a truly worst-case scenario, even if the
scenario had a very low probability of occurrence, or a credible worst case scenario
based upon a scientifically defensible probability of occurrence. The HFRS
recommends that the Agency (i.e, CEPPO) staff develop a more rigorous definition
of these concepts in order to provide Congress with a useful assessment. The
rationale for the selection of the "worst-case" HF accidental release scenarios is
one of the most important aspects of the assessment of hazards and should be
better established and defended. A scientifically defensible, systematic approach to
the definition of the credible worst-case accidental release scenario must be
developed.
10) The dose response analysis of the worst-case scenario appears to be
particularly weak in the draft Report to Congress. The guideline exposure levels
such as Immediately Dangerous to Life and Health (IDLH), Emergency Exposure
Guideline (EEGL), and Emergency Response Planning Guideline (ERPG) are
based upon times of exposure which are typically of 30 and-60 minute durations
(See Appendix A-reference #14). It is inappropriate to compare peak
concentrations determined from the dispersion analysis without the consideration
of exposure time. The HFRS did not evaluate the adequacy of the various
guideline exposure levels.
11) Regarding significant releases of HF, no information was available
that would suggest such accidents are inevitable, (i.e., lie beyond the prospect of
mitigation by human intervention). This suggests that HF major accidental
releases can be prevented or mitigated. Prudence dictates a need for increased
attention and diligence to workforce training, equipment inspection and improved
monitoring, maintenance and mitigation activities. However, the CEPPO staff
have made no major effort to evaluate or rank the effectiveness of various
prevention and mitigation measures. At a minimum, the Report to Congress
should provide all available information regarding opportunities to prevent or
mitigate accidental releases, consistent with Congressional intent to prevent
pollution. In addition, the standard operating procedures guidance recently
proposed by the American Petroleum institute (API) for HF alkylation units (See
Appendix A-reference #2) should be evaluated as to its adequacy for accident
prevention, and the quantitative reductions in exposure from mitigation
approaches like water quench and remote-operated valves should be determined.
-------
12) The HF industry appears to be proactive in research associated with
both new uses of HF as well as assessment and mitigation of accidental releases.
Much of the current knowledge base concerning fate and transport modeling of HF
after accidental releases is attributable largely to existing industry research which
is not in the open peer-reviewed literature, as well as the Department of Energy
(DOE) programs. The Report to Congress should address how uncertainty in the
regulation of HF ought to impact and focus on research and development relative
to the chemical and to its alternatives. Case histories and scenarios should be
utilized to illustrate the effects on research and development of the current
regulatory climate.
13) The issue of vulnerability analysis was not addressed in the draft
report. This is one of the major concerns to the Local Emergency Planning
Committees (LEPC's) set up under SARA Title III. These groups should be
surveyed to determine if any of them have conducted Vulnerability Analyses for
HF in their specific communities.
14) We would like to stress that the CEPPO staff should continue to
examine the newly-passed (February 24, 1992) Occupational Safety and Health
Administration (OSHA) 1910 "Process Safety Management (PSM) Rule" relative to
accident prevention for highly hazardous chemicals (See Appendix A, reference
#6). The HFRS recommends that the CEPPO staff expand its discussion of the
PSM standard, and more importantly, address its role in controlling accidental
releases of HF. This newly enacted rule establishes performance standards for
safety management of hazardous materials, and HF is one of many chemicals
covered by this rule.
A number of other broad-ranging and specific recommendations are made in
this SAB report with the aim to improve the quality and usefulness of the current
draft document as a Report to Congress. Once the CEPPO staff addresses
substantitively the limitations identified in this document, then the revised report
would have increased utility as a Report to Congress, and for this body to make
decisions based on sound technical information regarding HF.
-------
2. INTRODUCTION
On July 7 and 8, 1992, the Hydrogen Fluoride Review Subcommittee
(HFRS) of the Environmental Engineering Committee (EEC) and consultants of
the U. S. Environmental Protection Agency's Science Advisory Board reviewed the
preliminary draft of the "Hydrogen Fluoride Study: Report to Congress." The
EEC's HFRS was supplemented by three invited technical experts representing
academic, environmental and industry perspectives. This Report to Congress was
mandated by Section 301(n)(6) of the Clean Air Act (CAA) Amendments of 1990
(See Appendix A-reference #80), which required the Agency to complete a study
of:
"the industrial and commercial applications of hydrofluoric acid (HF) and
examine the potential hazards of hydrofluoric acid in industrial and
commercial applications to public health and the environment considering a
range of events including worst-case accidental releases and shall make
recommendations to the Congress for the reduction of such hazards, if
appropriate"
The Agency has correctly distinguished between anhydrous hydrogen
fluoride (HF) and hydrofluoric acid. The requirements of the CAA, Section
301(n)(6) Amendments of 1990 did not specify whether the study should focus on
the anhydrous form of hydrofluoric acid. The CEPPO staff and the SAB have
focused their critique on the properties and uses associated with hazards of
anhydrous hydrogen fluoride, which, for the purpose of the SAB report, is referred
to as HF. In normal terminology, the term HF refers to hydrofluoric acid, and
not anhydrous hydrogen fluoride.
The main text of the draft document presents background information on
the properties and hazards of HF, characterization of the HF industry, current
regulations and initiatives, HF industry process descriptions, hazards and industry
practices for processes involving HF, industry practice to detect and mitigate HF
releases, characterization of HF accidents, research efforts and future actions, and
community and facility emergency preparedness and planning. This preliminary
draft report did not include the executive summary or findings and
recommendations of the report, and these sections could therefore not be reviewed
by the Subcommittee. This SAB report can only be viewed as a review of the
quality of the supporting data per se, and not as a review of whether these data
support particular findings as is typical in most SAB reviews. Any reference by
the Agency to the SAB review should include this caveat.
-------
Since the CEPPO staff are inviting comment from a number of sectors,
including the public and the SAB, and since the time requirement in the
Congressional mandate calls for a report by mid-November of 1992, the CEPPO
staff have labored to assemble the basic data, and have not had the luxury of the
time needed to summarize the findings and recommendations at the time that the
SAB conducted its review. Nonetheless, the SAB review should be viewed as
technical, and adding value to make sure that the basic building-block data are
correct.
A prompt informal review was requested by the EPA. Therefore, the
Subcommittee provided individual comments to the Agency and the public at the
time of the meeting, but opted to later prepare a consensus report in its usual
manner. The charge given to the HFRS was to review the report and to focus on
the following questions:
a) Does the technical information related to chemical and process
hazards of HF appear to be complete?
b) Is the industry manufacturing, processing and use information
properly characterized?
c) Do the industry practices sections appear to be complete and
adequately characterized?
d) Does the technical information on accident history and accident
scenarios appear to be adequately characterized and complete?
The Subcommittee elected to answer the charge in a general manner, and
not focus this report explicitly around the charge, especially since the draft
document does not include the critical findings and recommendations.
Presentations made by representatives of the EPA's CEPPO summarized the
salient features of the HF report and highlighted some of the changes to be made
to the HF report as a result of public comment. However, the CEPPO staff
provided no indication of the likely content of the executive summary or findings
and recommendations. Additionally, several public comments of a technical nature
were made at the SAB public hearings; these were considered by the
Subcommittee in formulating its report.
-------
3. GENERAL COMMENTS
The CEPPO staff is commended for collection of a large amount of
background information on the properties, hazards, industrial uses, regulations,
and accidental releases of HF in the short time dictated by the Clean Air Act
Amendments. Subject to important reservations noted in remarks that follow, the
HFRS found the draft report to be a well organized, well written, concise and
generally representative presentation of background information on the special
properties of HF and current practices associated with the production and uses of
HF. However, it is not clear that the information gathered to date is sufficient to
allow the EPA to make a reasonable and comprehensive assessment of the
potential risks to surrounding communities associated with the production and
uses of HF. In its presentation, the report lacks the scientific detail requisite for
a thorough technical review. Also, some of the data appear to be out of date. For
example, the future market assessment that was conducted by the Agency for use
of HF, should be updated in light of the acceleration of the adoption of the
Montreal protocol for phase out of chlorofluorocarbons (CFC's) (See Appendix A-
references #5 and #13). Additionally, the draft report should be reviewed for
small, but distracting logical errors and that primary sources should be contacted
or examined to verify information wherever possible.
As stated earlier, the draft Report to Congress did not include the
executive summary or findings and recommendations, nor any definition of the
approach to be used to synthesize the findings and recommendations. Therefore,
an assessment of the adequacy of the data base could not be determined, since the
use of the data in relation to the yet to be inserted executive summary and
findings and recommendations section is not clear. At a minimum, the HFRS
recommends that the final draft Report to Congress, including the findings,
recommendations, and public comment docket, be peer reviewed with the same
rigorous process that the preliminary draft has undergone, including stakeholder
reviews and, if requested, SAB review. The HFRS recognizes that such a review
could not be conducted quickly enough to allow the CEPPO staff time to complete
the Report to Congress on the mandated schedule. Nonetheless, the HFRS
recommends that the review of the full report be conducted, even if after the
Report to Congress has been submitted, since the approach used in evaluating HF
is important to the consideration of other hazardous chemicals.
The HFRS recognizes that there is a very real need to foster a thorough
and objective analysis and to ensure a more complete presentation of changes in
industrial practices and the ability of communities to deal with chemical
emergencies. To partially address this need, the HFRS notes that "chemical use
8
-------
trees" could be a useful visual tool to inform the reader and foster greater
awareness of the families of chemicals which are derived from HF. In particular,
the importance of this chemical to the evolution, strength, and ultimate
competitiveness of affected segments of U.S. industry should be evaluated by the
Agency in order to place its role in perspective for the Congress. Likewise, the
consequent positive and negative impacts on communities and the concern of the
public, as represented by communities and environmental organizations, needs
equal attention. If the CEPPO staff cannot place such material in the Report to
Congress, then they should recommend to Congress that such activity be
conducted by the Agency in the future.
The Agency's response to Congress should not be so narrow as to respond
only directly to Congress's request. Rather, the Agency should attempt to
anticipate and address questions which their report will generate. The
Subcommittee believes, in fact, that this is already explicit in the Congressional
charge. The most obvious of these questions is: Is HF so dangerous that HF use
should singularly be severely restricted? One aspect of this question should
include consideration of the comparative risks presented by alternatives to HF. In
one locality, the California South Coast Air Quality Management District
(SCAQMD) apparently has determined that the use of HF is unsafe in their
specific location and has regulated the phase out of HF use (See Appendix A -
reference #7). This phase-out decision could result in significant costs for
replacement or retrofit of existing alkylation units to an alternative sulfuric acid
(HgSO^ alkylation process. It should be noted, however, that the SCAQMD Rule
has been suspended by court action in California. It is conceivable that
alternatives may be as risky or perhaps more risky than HF.
The actual hazards associated with the use of substitutes do not appear to
have been adequately considered, especially concerning total process implications
such as materials transport and component regeneration. A comprehensive study
using life cycle analysis concepts of health, environment and safety should be
completed for both the use of HF and its alternatives. The evaluation of
alternatives should be more completely considered in the report and compared on
the basis of all hazards associated with the total process (i.e., life cycle
implementation). The Office of Research and Development (ORD), Risk Reduction
Engineering Laboratory (RREL) in Cincinnati, Ohio has the capability to conduct
such an analysis. If the recommended comparative life cycle analysis of
alternatives cannot be conducted in a timely manner consistent with Congressional
mandates, then this useful activity should be undertaken by the Agency as future
work. Congress must be made aware that the alternatives to HF may have
substantial risk. Another necessary consideration is the comparison of risks
associated with HF and the risks associated with somewhat similar commonly-used
-------
hazardous substances such as chlorine, fuming sulfuric acid, phosgene and
ammonia.
HF has unique properties and hazards which should be carefully addressed.
However, the section on properties and hazards (Chapter 2) is almost entirely
qualitative, apart from the table of basic physical properties and the graphical
comparison of guideline exposure levels. The numerous chemical and physical
reactions unique to HF, some of which may create dangerous situations, appear to
be only summarily described. The draft Report to Congress did not adequately
quantify the importance of these properties as they contributed to accidental
releases. The corrosive nature of the chemical and the heat release associated
with reactions of HF with various caustic materials have clearly contributed to
past releases. Nonetheless, management of HF is not altogether significantly
different from a wide range of commonly used industrial chemicals, such as fuming
sulfuric acid, chlorine, phosgene and ammonia. For this reason, the approach used
in this study could clearly have broader implications for the manner in which
hazards associated with other chemicals are evaluated in the future. The
Subcommittee recommends that the Report to Congress should clearly state
implications of the findings of this study on the evaluation of hazards associated
with other industrial chemicals.
The Report to Congress should be careful in defining routine industrial
practices, since this was not comprehensively characterized in this study for all
U.S. facilities. Some of the general statements concerning industrial practices are
not substantiated by detailed surveys. Rather, they are anecdotal accounts of an
incomplete set of site visits and discussions with industry representatives. Follow-
up visits to sites where chemical safety audits were previously undertaken was
suggested as one way to more completely determine how current practice has
changed at individual facilities since the initial audits. Also, the Subcommittee
recommends that a survey of the state of mitigation (industrial and community
response and prevention) readiness to such releases be conducted.
4. MODELING
The Subcommittee had substantial concerns regarding the dispersion
modeling used in the assessment of worst-case accident scenarios. The detailed
nature of the model in relation to the role of dispersion modeling was not clearly
stated in the draft report. Therefore, at the time of the review, the HFRS was
not able to comment on the adequacy of the modeling effort in relation to the
evaluation that was performed. Nonetheless, the HFRS can make several
recommendations concerning the use of these models in the context of regulatory
10
-------
decision-making. The HFRS refers the CEPPO investigators to the SAB
Resolution on Use of Mathematical Models bv EPA for Regulatory Assessment and
(EPA-SAB-EEC-89-012) (See Appendix A -reference #15). This
resolution addressed a number of concerns with the use of models that should be
considered in regard to this study. These include such concerns as input scenarios
and model validation. In addition, the Subcommittee refers the reader to an
analysis of the ALOHA model completed by one of our invited technical experts,
Eh-. Jerry Havens, Distinguished Professor in the Department of Chemical
Engineering at the University of Arkansas (See Appendix A - reference #4, as well
as Appendix B Letter to Dr. Jack Kooyoomjian from Dr. Jerry Havens, dated
November 6, 1992). The EEC notes that this is an individually authored
submittal, and not necessarily a consensus position.
In this draft Report to Congress, two computer models were employed to
explore the potential consequences of postulated "worst-case" scenarios. Results of
simulations are presented in the text for three scenarios. However, based upon
the information provided in the draft Report to Congress and its Appendices, it is
extremely difficult to assess the appropriateness and limitations of the models in
these applications. The mathematical and thermodynamic underpinnings of the
models are not described. No citations of refereed journal papers, which would
support their application, are provided. The phenomena crucial to understanding
HF post-release event behavior are discussed at a basic level that is likely to be
informative to a lay audience, but does not provide the underlying material upon
which a critical peer review can be conducted. Stated another way, since the
underlying principles are not clearly elucidated, this results, at best, in a marginal
technical review which lacks significant detail. For instance, it is unclear which
phenomena are actually included in the conceptual and mathematical models
employed to characterize the propagation of the HF cloud, how critical parameters
are estimated, and how cloud movement depends on environmental conditions.
Little or no justification or discussion is given pertaining to selected model inputs.
Furthermore, the presented simulations are clearly inadequate to fully explore
consequences of postulated release scenarios. A sensitivity analysis of simulations
for a reasonable range of inputs was apparently not considered. These are serious
deficiencies which greatly diminish prospects for successful use of the simulation
results for formulation of conclusions regarding the consequences of HF accidents.
The Subcommittee suggested that the dispersion modeling framework used
(in particular the HGSystems model) can incorporate the current state-of-the-art
understanding of dispersion of HF after release. According to one expert, the
industry has conducted significant research and development on dense gas
dispersion and adapted models to include special chemical/physical and
thermodynamic properties of HF. Near field processes such as aerosol formation,
11
-------
polymerization (hydrogen bonding), and exothermic hydrolysis are unique features
of HF asserted by the Agency staff to be adequately portrayed in the HGSystems
model used.
The Subcommittee indicated that the model used in the EPA study has been
extensively validated against wind tunnel and idealized field data, which would
appear to be typical for other well-validated models. However, the report relies
heavily on comparisons of modeling with field tests conducted at a DOE facility in
Nevada (test series called "Goldfish") as evidence that HF dispersion behavior is
adequately understood with available modeling procedures. The effects of
humidity, slope, rough terrain and channel flows have apparently not been
experimentally determined in these tests. For instance, the relatively low
humidity of the Goldfish tests could affect HF dispersion physics. Therefore, while
the physical processes are mathematically portrayed in the models (particularly the
HGSystems), the model's ability to accurately simulate these important phenomena
have not yet been adequately verified. This limitation should be clearly stated in
the Report to Congress. Additionally, the reader is referred to Dr. Jerry Havens'
November 6, 1992 analysis of the ALOHA model in Appendix B. (Also listed in
Appendix A - reference #4).
The sensitivity of the modeling results to the uncertainty in their treatment
of humidity and boundary effects should be examined in order to estimate the
uncertainty that this lack of verification brings to conclusions drawn from the
modeling. Future efforts should be undertaken to verify results predicted for more
realistic conditions. To accomplish this task, more field data will clearly be
required. The HFRS was advised that the U.S. Department of Energy (DOE) spill
test facilities, which are apparently uniquely suited for this type of testing, are
scheduled for shutdown due to lack of operating and maintenance funds. The
HFRS expresses concern for the possible future loss of this valuable scientific tool,
and encourages the EPA to discuss this situation with DOE.
The HGSystems dense gas dispersion model accounts for many of the
critical physical/chemical processes that are considered to be important for HF
dispersion. It is a complex tool that requires expertise and familiarity with regard
to simulating chemical processes and modeling dispersion and meteorological
phenomena for assessment of parameters and applications. It is not clear to the
Subcommittee whether the users of the models in this study are experts in the use
of this model. If this modeling effort is to be an important and meaningful aspect
of the characterization of hazards associated with HF and other hazardous
chemicals, then experts should be sought who can appropriately apply the models
and interpret the results. In addition, it is not clear that this model is or should
be suited for use by the Local Emergency Planning Committees (LEPC's) who
12
-------
must address facility-specific issues associated with potential HF releases, and who
may not be modeling experts. This issue needs to be discussed and clarified in the
Report to Congress. The more simplified model, ALOHA, on the other hand, is
specifically designed for planning emergency response, and as such, utilizes several
simplifications in theory (See Appendix A-reference #1). The impact of these
simplifications should be systematically evaluated if this model is to be used for a
more comprehensive assessment of HF dispersion.
5. WORST-CASE SCENARIOS
The characterization and study of potential hazards, consequences, and
worst-case scenarios are central to the request from Congress for this study. The
HFRS noted that "hazards" are considered intrinsic properties of the potential to
do harm from the use of HF, whereas "risks" resulting from exposure to hazards
are what relate extrinsically to the community. However, no clear underlying
definitions of these concepts (potential hazards, consequences and worst-case
scenarios) were provided in the draft report. Definitions discussed during the
briefing were rather arbitrary and difficult to rigorously employ in the absence of
clear definition. It was not made clear to the Subcommittee whether
Congressional intent was to evaluate a truly worst-case scenario (even if the
scenario had a very low probability) or a credible worst-case scenario based upon a
scientifically defensible and/or significant probability of occurrence. The HFRS
recommends that the Agency CEPPO staff should develop a more rigorous
definition of these concepts in order to provide Congress with a useful assessment.
Additionally, the HFRS refers the CEPPO staff to the Deputy Administrator's
February 26, 1992 memorandum on Risk Characterization, and the appended Risk
Assessment Council (RAG) Guidance for Risk Assessment, which provides explicit
guidance to Agency personnel on how to develop exposure scenarios, both "worst-
case" and "best estimate." (See Appendix A - reference #11).
The selection and definition of the "worst-case" accidental releases scenarios
of HF is one of the most important aspects of the hazards assessment. It is the
Subcommittee's opinion that this issue has great uncertainty, is unresolved, and
may strongly dictate the outcome of the hazards, risk, and consequences analyses
yet to be performed. The Subcommittee has determined and recommends that the
identification of credible worst-case scenarios is more important than additional
background data gathering, although the currently available background
information may not be adequate to apply such models. Apparently, the plausible
worst-case accidental release scenarios investigated in the draft Report to Congress
were defined by Agency CEPPO staff after discussions with industrial
representatives, as well as many other parties. Rationale for the selection of the
worst-case scenarios should be further established, refined and presented in the
13
-------
revised Report to Congress.
One approach suggested by the Subcommittee was the use of dispersion
modeling and probabilistic risk assessment procedures to identify and define the
release parameters and accident scenarios that most influence dose response to the
off-site public. Conducting sensitivity analysis on dense gas dispersion models may
provide further understanding of the behavior of the most important parameters.
The Subcommittee believes that some important parameters are the HF release
rate, event duration, and mitigation systems. The gathering and documenting of
accidental release data could be used to identify and statistically define the range
of these parameters that have been encountered in actual previous HF releases.
By analogy, it may be useful to examine the past releases of other compounds with
similar properties in order to "broaden" the accidental release data base
framework. If the data base is not sufficient to develop a statistically significant
selection of initial model data input, alternative methods of selecting scenarios and
data should be explored. Another approach suggested by the Subcommittee was to
examine the philosophical approach used in the Agency's Risk Assessment
Guidance for Superfund. (See Appendix A - reference #12).
For each scenario, a sensitivity analysis should be conducted in order to
clearly understand uncertainties in prediction of impacts associated with assumed
release parameters. The water spray mitigation release scenario should be
investigated as one of the plausible release scenarios, in order to define the
performance of water spray mitigation for hazards control.
The Agency should examine the final selection of worst-case accident
scenario(s) with respect to other potential scenarios. For example, the Agency
should provide an explanation as to why the screening guidance provided to
LEPCs (the "Green Book") is not a credible worst-case for this analysis. Is this
due to its oversimplification and unrealistic release scenarios? The Agency should
also examine the worst-case scenarios selected for analysis with respect to HF
release accidents that have already occurred. Finally, the Agency should examine
the assumptions made in the California South Coast Air Quality Management
District (SCAQMD) hazards analysis and discuss the differences in selected worst-
case release scenarios. Moreover, the Agency should analyze the SCAQMD
accident scenarios and dispersion modeling approach and discuss why the methods
used may or may not be adequate.
14
-------
6. DOSE RESPONSE
In the draft Report to Congress, the dose response analysis associated with
the worst-case release scenario appears to be particularly weak. The guideline
exposure levels, such as Immediately Dangerous to Life and Health (IDLH) and
Emergency Response Planning Guideline (ERPG), are based upon times of
exposure which are typically of 30-and 60-minute duration. It is inappropriate to
utilize and compare peak concentrations determined from the dispersion analysis
without consideration of actual exposure-time behavior. The Occupational Safety
and Health Administration (OSHA) Permissible Exposure Limit (PEL) standards
were appropriately not used by the Agency CEPPO staff in this analysis, since
they are designed for worker exposure assessment and use an eight-hour time
weighted average exposure (See Appendix A-reference #14). The HFRS did not
evaluate further the adequacy of the various guideline exposure levels.
7. ACCIDENT PREVENTION
Regarding significant releases of HF, no information was available
suggesting that such accidents are inevitable (i.e., lie beyond the prospect of
mitigation by human intervention). This suggests that such accidental releases
may be made to have a low probability of occurrence. Prudence dictates a need
for increased attention and diligence to workforce training, equipment inspection
and improved monitoring, maintenance and mitigation activities. The Agency
CEPPO staff have made no major effort to evaluate or rank the effectiveness of
various prevention and mitigation measures. At a minimum, the Report to
Congress should provide all available information regarding opportunities to
prevent or mitigate accidental releases, consistent with Congressional intent to
prevent pollution. The HFRS would like to stress that the CEPPO staff should
continue to examine the newly-passed (February 24, 1992) Occupational Safety and
Health Administration (OSHA) 1910 "Process Safety Management (PSM) Rule"
relative to accident prevention for highly hazardous chemicals (See Appendix A-
reference #6). The HFRS recommends that the CEPPO staff expand its
discussion of the PSM standard, and more importantly, address its role in
controlling accidental releases of HF. This newly enacted rule establishes
performance standards for safety management of hazardous materials, and HF is
one of many chemicals covered by this rule.
The use of the recently promulgated OSHA requirements for HF could be
used as a framework to define industrial practices that may be adopted in the
future. The impact of the adoption of this rule on the prevention of accidents
should be evaluated. In addition, the standard operating procedures guidance
recently proposed by the American Petroleum Institute (API) for HF alkylation
15
-------
units should be evaluated as to its adequacy for accident prevention (See Appendix
A-reference #2). The mitigation methods defined by the SCAQMD in its rule-
making decisions could be evaluated to identify risk-reduction impacts and
potential implications of these measures. Finally, the quantitative reductions in
exposure from mitigation approaches like water quench and remote-operated valves
should be determined.
8. OTHER ISSUES RAISED BY THE HFRS
The HF industry appears to be proactive in research associated with new uses
of HF, as well as assessment and mitigation of accidental releases. Much of the
current knowledge base concerning fate and transport modeling of HF after accidental
releases is attributable largely to existing industry research programs which are not
in the open peer-reviewed literature, as well as the DOE programs. The Report to
Congress should address how uncertainty in the regulation of HF will impact and
focus research and development relative to the chemical and to its alternatives. Case
histories and scenarios should be utilized to illustrate how the effects of focused
research and development can answer questions raised to deal with the current and
future regulatory requirements.
The issue of vulnerability analysis was not addressed in the draft Report to
Congress. This is of major concern to the LEPC's set up under SARA Title III.
These groups should be surveyed to determine if any of them have conducted
Vulnerability Analyses for HF in their specific communities. Clearly, this type of
technical information is desirable. These vulnerability analyses could yield some
useful information about the vulnerability of local communities to HF release events,
and could supply some very useful information that would normally be considered to
be confidential by industry.
16
-------
APPENDIX A - REFERENCES CITED
1) ALOHA-Areal Locations of Hazardous Atmospheres, National Oceanic and
Atmospheric Administration, CAMEO TM 3.0 Computer-Aided Management of
Emergency Operations, Hazardous Materials Response Branch, Seattle,
September, 1991
2) A.P.I., "Safe Operation of Hydrofluoric Acid Alkylation Units," API
Recommended Practice 751, First Edition, June 1992
3) Chemical Manufacturers Association Hydrogen Fluoride Panel, Oral Testimony
on the EPA Hydrogen Fluoride Study Report, Submitted to the EPA Science
Advisory Board by Ms. Carolyn S. Seringer on behalf of Hydrogen Fluoride
Panel of the Chemical Manufacturers Association, July 7, 1992
4) Havens, Jerry, Letter to Dr. K. Jack Kooyoomjian, U.S. EPA, Science Advisory
Board, entitled "Review of the Draft Hydrogen Fluoride Study - Report to
Congress: An Evaluation of the ALOHA Dispersion Model Results," 6 pages,
dated November 6, 1992 (See Appendix B for full text of this letter.)
5) Montreal Protocol on Substances that Deplete the Ozone Layer, As Amended
in 1990, and as Negotiated Under the Vienna Convention for Protection of the
Ozone Layer, Final Act (Nairobi UNEP 1985) [NOTE: There will be
subsequent amendments and adjustments to the Protocol.]
6) OSHA Requirements (29CFR1910) (Rule 1910) "Process Safety Management of
Highly Hazardous Chemicals, February 24, 1992.
7) South Coast Air Quality Management District (SCAQMD) Hazards Analysis
[Los Angeles County, Department of Health Services, Toxics Epidemiology
Program, Health Effects Due to Hydrogen Fluoride Inhalation: A Literature
Review. 1989. Prepared for the Hydrogen Fluoride Task Force of the
SCAQMD]
8) U.S. Congress, Clean Air Act Amendments of 1990, Section 301(n)(6) (Public
Law No. 101-549, 104 STAT. 2399), 1990
9) U.S. EPA, Chemical Emergency Preparedness and Prevention Office, "Briefing
for the Science Advisory Board's Environmental Engineering Committee on the
Hydrogen Fluoride Report to Congress," July 7, 1992
A-l
-------
APPENDIX A - REFERENCES CITED: CONTINUED
10) U.S. EPA, Hydrogen Fluoride Study: Draft Report to Congress, Office of Solid
Waste and Emergency Response, Washington, D.C., May 1992
11) U.S. EPA, Office of the Administrator, From Mr. F. Henry Habicht II, Deputy
Administrator to Assistant Administrators and Regional Administrators, Memo
entitled "Guidance on Risk Characterization for Risk Managers and Risk
Assessors," February 26, 1992 (6 pages). [See especially the appended 34 page
report from the US EPA's Risk Assessment Council, entitled "Guidance for
Risk Assessment, November 1991]
12) U.S. EPA, Office of Emergency and Remedial Response, Risk Assessment
Guidance for Superfund, Volume I, Human Health Evaluation Manual (Part A),
Interim Final, EPA/540/1-89/002 (PB90-1555581), December 1989
13) U.S. EPA, "Protection of Stratospheric Ozone, Proposed Rule," 40CFR Part 82,
September 4, 1991
14) U.S. EPA, Science Advisory Board, Environmental Health Committee (EHC),
"Superfund Site Health Risk Assessment Guidelines," (Refers to OSHA PEL
Standards), EHC Review Draft Number 3, September, 1992
15) U.S. EPA, Science Advisory Board, Resolution on Use of Mathematical Models
by EPA for Regulatory Assessment and Decision-Making, Environmental
Engineering Committee, EPA-SAB-EEC-89-012, January 13, 1989
A-2
-------
APPENDIX B - ALOHA MODEL RUN RESULTS
-------
UNIVERSITY^ARKANSAS
3202 Bell Engineering Center • Favetteville, Arkansas 72701-1201 • (501) 575-4951 • (501) 575-7926 (FAX)
College of Engineering
Department of Chemical Engineering November 6, 1992
Dr. Jack Kooyoomjian
U.S. EPA Science Advisory Board
401 M Street, SW A-101F
Washington, D.C. 20460
RE.: Review of the Draft Hydrogen Fluoride Study - Report to Congress:
An Evaluation of the ALOHA Dispersion Model Run Results
Dear Dr. Kooyoomjian:
This contains my comments on the OSWER/CEPPO Draft Hydrogen Fluoride Study:
Report to Congress, dated May 1992. My general observations of the report,
which I provided to you dated September 14, 1992, are unchanged. These
comments are directed to those parts of the report involving atmospheric
dispersion predictions.
The draft report correctly identifies the issue of the appropriate
surface roughness for use in simulation of releases in an urban
environment. This is a general question which should be given high
priority for research, since it is a problem which is basic to the
understanding of the dispersion of dense gas clouds in the areas
where they are most likely to be formed. Since the majority of gases
which can constitute a major hazard if accidentally released are
denser-than-air, the requirement for information in this area extends
to most of the hazardous gases in commerce--hence the top-priority
requirement for some resolution of this question. I hope that the
SAB's findings in this review process will foster the required
research.
The analysis of the accident scenarios considered consists mainly of
the presentation of predictions of the maximum downwind distances to
which concentrations of 30 ppm and 50 ppm hydrogen fluoride in air
would be experienced.
Two dispersion models are used, HFSYSTEM and ALOHA-5.1.
The HFSYSTEM model is specifically designed for use by persons who are
trained in its use. I am familiar with the methods used in the HFSYSTEM
model, and there are many similarities between the HFSYSTEM and DEGADIS
models. However, without additional information no determination can be
made of the accuracy or applicability of the predictions presented.
B-l
The University of Arkansas is an equal opportunity/affirmative action institution.
-------
Dr. Jack Kooyoomjian
November 6, 1992
Page 2
The ALOHA (Areal Locations of Hazardous Atmospheres) model is a joint
development of NOAA and EPA. ALOHA is designed for rapid estimation of
downwind hazard extent for emergency response. The DEGADIS model (of which
I am a coauthor) is distributed by EPA and the Gas Research Institute, and
ALOHA (5.1) incorporates a simplified form of DEGADIS. Since ALOHA is
designed for emergency response use, the incorporation of DEGADIS required
simplification to allow rapid prediction. The ALOHA directions-for-use
distributed by NOAA specifically state that ALOHA-DEGADIS is designed for
emergency response application, and that the (parent) DEGADIS model should
be used for calculations made for risk assessment purposes where it is
frequently required to consider carefully the effects of transient releases
and other complicating factors. I have recently completed an evaluation of
the ALOHA-DEGADIS model (which has been incorporated in ALOHA) and am
therefore in a position to consider this question. I can state the
following observations which are important to the consideration of the
predictions presented by OSWER/CEPPO in the Draft HF Study-Report to
Congress:
• The incorporation of DEGADIS in ALOHA (by NOAA) is generally accurate
and acceptable. However, the incorporation of DEGADIS did involve
simplifications, and there are effects of these simplifications which
bear directly on the predictions presented in the draft report.
• The prediction of HF dispersion for the scenarios presented in the
draft report requires model treatment of the complex effects of
aerosol formation and the associated thermodynamic effects which can
occur in an HF cloud. My initial comparison of NOAA-DEGADIS and
DEGADIS indicates that such differences are not likely to be very
important in this application, although there are some differences.
• The primary simplifications of DEGADIS in ALOHA relate to the
modeling of transient releases. When transient releases are modeled
(in ALOHA) the time-varying release rate is approximated by five
piecewise "steps" of uniform rate. More importantly, for the
calculation of the maximum downwind distance to the concentration of
concern, ALOHA presents the distance calculated for the maximum (in
the case at hand, the initial) release rate, as if it occurred
continuously. The ALOHA directions-for-use specifically state that
this simplification (made to decrease run-time) can result in
significant overestimation of the downwind extent for transient
releases.
B-2
-------
Dr. Jack Kooyoomjian
November 6, 1992
Page 3
The modeling assumptions made for each of the three scenarios are severe and
somewhat arbitrary. Specifically:
Different surface roughnesses are specified for the ALOHA and
HGSYSTEM models (3 cm and 1 cm respectively). The comparisons should
be made for the same input variables.
All three scenarios (apparently) assume that all of the HF goes
downwind as an aerosol, with no (or negligible) rainout. This
assumption would appear to be most questionable for Scenario #3 which
describes a leaking tank. Since the formation of aerosol upon
release depends critically on the temperature and pressure from which
the material is released, this scenario description is incomplete.
• EPA states that "F stability occurs in overcast, pre-dawn, calm
hours...". F stability can occur in pre-dawn, calm hours, but it is
normally associated with clear skies (which allow for rapid radiation
cooling of the surface).
• EPA states that "surface roughness conditions are an estimate of the
effect of surface terrain and the presence of high buildings or other
man-made structures that will impact the cloud's movement and
dispersion ...". One of the main points of contention in dense gas
dispersion prediction is the extent to which the effect on dispersion
of large "obstacles" (such as buildings) whose height is comparable
to the dispersing gas cloud can be represented as surface roughness.
This statement fails to acknowledge the importance of this unsolved
problem, and it is contradictory to EPA's identification of the
surface roughness "issue" in a later section of the report.
I have repeated the simulations with the ALOHA model, using the input data
provided in the draft report, of the three HF release scenarios and compared
the results with DEGADIS. The input data are summarized below:
B-3
-------
Dr. Jack Kooyoomj ian
November 6, 1992
Page 4
Inputs to Aloha (V5.1) Computer Model
Scenario #1 Scenario #2
Release Type
Atmospheric Conditions
Surface Roughness
Air Temperature
Relative Humidity
Concentrations of concern
3000 Ib/min
for 1 minute
5.2 mps, D
1.5 mps, F
3 cm
72 F
50%
30 ppm/30 min
50 ppm/60 min
160 Ib/min
(continuous)
5.2 mps, D
1.5 mps, F
3 cm
72 F
50%
30 ppm/30 min
50 ppm/60 min
Scenario #3
1800 Ib/min
(continuous)
5.2 mps, D
1.5 mps, F
3 cm
72 F
50%
30 ppm/30 min
50 ppm/60 min
I have verified with ALOHA the ALOHA Model results presented in Exhibits
8-10, 8-12, and 8-13 of the draft report; the results (distances) are
summarized below:
ALOHA Model Results
Scenario #
5.2 m/s,
- Transfer Line Failure
D stability
1.5 m/s, F stability
Scenario #2 - Pump Seal Failure
5.2 m/s, D stability
1.5 m/s, F stability
Scenario #3 - Vessel Leak
5.2 m/s, D stability
1.5 m/s, F stability
Distance to IDLH
(miles)
3.6
>6
0.7
1.8
2.5
5.5
Distance to ERPG
(miles)
2.8
5.5
0.6
1.3
2.0
4.0
Observations on Scenario #1 - Transfer Hose Failure
The ALOHA footprint assumes release at 3000 Ib/min for one hour. This is
clearly stated in the ALOHA directions-for-use. Furthermore, the ALOHA model
provides for calculation of the concentration and exposure (dose) at a
specified downwind distance, and the resulting concentration reflects the
transient nature of the release. EPA appears to have simulated the release
of 3000 Ib as instantaneous. The minimum duration of release for the
B-4
-------
Dr. Jack Kooyoomjian
November 6, 1992
Page 5
ALOHA model is 1 minute. Therefore the release is simulated by ALOHA as 50
Ibs/second for 60 seconds, and the footprint (from which the maximum downwind
distance to the level of concern is determined) reflects the maximum release
rate of 3000 Ibs/min. This reported distances of >6 and 5.5 miles for F
stability, 1.5 m/s and D stability, 5.2 m/s respectively are gross
overpredictions which do not reflect the transient (short-lived) nature of
the release. The ALOHA model indicates that there is no significant exposure
at either of the presented distances for the 3000 Ib release. DEGADIS
indicates the maximum downwind distance to be less than 1 mile for both
cases. (This result, and my knowledge of the similarity of the DEGADIS and
HFSYSTEM models, leads me to view the associated HFSYSTEM predictions with
suspicion.)
Observations on Scenario #2 - Pump Seal Failure
The ALOHA footprint prediction assumes 160 Ibs/min released for one hour,
whereas the actual release that was to be modeled lasted for 20 minutes. The
result would be to overpredict the distance, since the along-wind dispersion
of the cloud is neglected in the steady state prediction. EPA incorrectly
concluded that the ALOHA simulations of Scenarios #1 and #2 indicate that the
"release duration seems to play much less of a role (than the release rate)."
DEGADIS predictions appear to be in reasonable agreement with the ALOHA
predictions for this case.
Observations on Scenario #3 - Vessel Leak
The only difference between the ALOHA simulations of Scenarios #2 and #3 is
the 11.25 times greater release rate for Scenario #2. However, since the
conditions of pressure and temperature of storage (in the tank) are not
specified, it is questionable to model this release as an aerosol plume
instead of an evaporating pool. EPA also states that "A pool of HF forms
which flashes and aerosolizes into a dense, white cloud that begins to travel
downwind". It is unlikely that an aerosol cloud of HF would result from an
evaporating pool. The conditions for aerosol formation are generally agreed
to involve flashing accompanied by high shear such as occurs during high-
velocity discharge from a container. Pool formation and (complete) aerosol
cloud formation are contradictory. DEGADIS predictions appear to be in
reasonable agreement with the ALOHA predictions for this case.
In summary, important errors have been made in the presentation and analysis
of ALOHA predictions in the draft report. The errors are attributed to lack
B-5
-------
Dr. Jack Kooyoomj ian
November 6, 1992
Page 6
of understanding and familiarity with the ALOHA model and to failure to
observe ALOHA's limitations.
Sincerely,
Jerry Havens
Distinguished Professor
cc: Dr. W. R. Seeker
Energy & Environmental Research Corp.
18 Mason Street
Irvine, California 92718
JH:vh
B-6
-------
APPENDIX C - GLOSSARY OF TERMS AND ACRONYMS
ALOHA AREAL LOCATIONS OF HAZARDOUS ATMOSPHERES (A
MODEL DESIGNED SPECIFICALLY FOR PLANNING
EMERGENCY RESPONSE)
API AMERICAN PETROLEUM INSTITUTE
CAA CLEAN AIR ACT
CEPPO CHEMICAL EMERGENCY PREPAREDNESS AND
PREVENTION OFFICE (CEPPO) (U.S. EPA)
CFC's CHLOROFLUOROCARBONS
DOE U.S. DEPARTMENT OF ENERGY
EEC ENVIRONMENTAL ENGINEERING COMMITTEE
(SAB/EPA, ALSO REFERRED TO AS 'THE COMMITTEE")
EEGL EMERGENCY EXPOSURE GUIDELINE
EHC ENVIRONMENTAL HEALTH COMMITTEE (SAB/EPA)
EPA U.S. ENVIRONMENTAL PROTECTION AGENCY (U.S. EPA, or "THE
AGENCY"
ERPG EMERGENCY RESPONSE PLANNING GUIDELINE
HoSO4 SULFURIC ACID
HF HYDROFLUORIC ACID (ALSO HYDROGEN FLUORIDE)
HFRS HYDROGEN FLUORIDE REVIEW SUBCOMMITTEE (EEC/SAB/EPA,
ALSO REFERRED TO AS "THE SUBCOMMITTEE")
HG (REFERS TO HG SYSTEMS MODEL)
IDLH IMMEDIATELY DANGEROUS TO LIFE AND HEALTH
LEPC's LOCAL EMERGENCY PLANNING COMMITTEES
OMB OFFICE OF MANAGEMENT AND BUDGET
OSHA U.S. OCCUPATIONAL SAFETY AND HEALTH
ADMINISTRATION
OSWER OFFICE OF SOLID WASTE AND EMERGENCY RESPONSE (U.S.
EPA)
PEL PERMISSIBLE EXPOSURE LIMIT
PSM PROCESS SAFETY MANAGEMENT STANDARD (PROMULGATED
BY OSHA ON FEBRUARY 26, 1992)
RAG RISK ASSESSMENT COUNCIL (U.S. EPA)
RREL RISK REDUCTION ENGINEERING LABORATORY, OFFICE OF
RESEARCH AND DEVELOPMENT, CINCINNATI, OHIO (U.S. EPA)
SAB SCIENCE ADVISORY BOARD (EPA)
SARA SUPERFUND AMENDMENTS AND REAUTHORIZATION ACT
SCAQMD SOUTH COAST AIR QUALITY MANAGEMENT DISTRICT
U.S. UNITED STATES
c-i
-------
DISTRIBUTION LIST
Deputy Administrator
Assistant Administrators
EPA Regional Administrators
EPA Laboratory Directors
Deputy Assistant Administrator for Office of Pollution Prevention and Toxics (OPPT)
Deputy Assistant Administrator for Office of Research and Development (ORD)
Director, Office of Environmental Engineering and Technology Demonstration
(OEETD)
Deputy Director, OEETD
Director, Center for Environmental Research information (CERI)
Director, Office of Technology transfer and regulatory Support (OTTRS)
Deputy Director, OTTRS
Deputy Assistant Administrator for Office of Solid Waste and Emergency Response
(OSWER)
Director, Chemical Emergency Preparedness and Prevention Office (CEPPO)
EPA Headquarters Library
EPA Regional Libraries
EPA Laboratory Libraries
------- |