Versm
INC.
FINAL REPORT
Pressurized Storage Tanks/
A Preliminary Assessment
6850 VERSAR CENTER • P.O.BOX 1549 • SPRINGFIELD, VIRGINIA 22151
TELEPHONE: (703) 750-3000
-------
FINAL REPORT
Pressurized Storage Tanks/
A Preliminary Assessment
Submitted by:
Versar, Inc.
P.O. Box 1549
Springfield, VA 22151
Submitted to:
Angela WHkes
Office of Solid Waste
Waste Treatment Branch
U.S. Environmental Protection Agency
401 H Street, S.W.
Washington, O.C. 20460
In Response to:
EPA Contract No. 68-01-7053
Work Assignment No. 1
June 27, 1985
-------
ABSTRACT
This preliminary EPA study suggests that the U.S. is unlikely to
suffer an industrial chemical accident on the scale of the one that
killed more than 2,000 people in Bhopal, India, last year. Safety
systems and operating procedures used 1n the U.S. for the type of
pressurized industrial tanks that caused the disaster, however, appear to
be inadequate to prevent much smaller scale releases, which occur more
frequently.
At Bhopal, according to Union Carbide's official report, up to 2,000
gallons of water entered a chemical storage tank, setting off a reaction
that four separate safety-related systems failed to bring under control.
All these systems were mechanically workable, but were not in service at
the time because of management decisions, operator errors, and
substandard maintenance. In contrast, all such fail-safe devices are
continuously on-line at the company's plant in Institute, West Virginia,
which until recently was the only plant in this country producing methyl
isocyanate (MIC), the highly toxic chemical released from the company's
Bhopal plant. Nevertheless, public concern over the Bhopal disaster has
prompted extensive design improvements at Union Carbide's Institute
plant. It has also led EPA to perform this study which investigates the
general safety of pressurized tanks used in industry to store or process
toxic materials.
No comprehensive records are kept on the number, type, and contents
of pressurized tanks in this country. The Resource Conservation and
-------
Recovery Act (RCRA) does not specifically cover pressurized^ tanks. It
does, however, control all types of tanks that store hazardous wastes,
Liquid petroleum products, and substances defined as "hazardous" under
the Comprehensive Environmental Response, Compensation and Liability Act
("Superfund")
EPA's two-month study estimates that there are at least 150,000
pressurized tanks 1n the U.S., located at up to 15,000 plants across the
country. The study estimates that they contain at least 38 regulated
toxic chemicals produced at more than 10,000 tons per year. The most
dangerous chemicals are probably methyl Isocyanate, acroleln, phosgene,
ally! chloride, chloroprene, and vinyl chloride.
A rapid search of the general press and chemical Industry literature
uncovered 54 accidents over the past 20 years that Involved significant
damage to human health or the environment These took place 1n 15 U.S.
States and 12 foreign countries. Fourteen (other than Bhopal) resulted
1n a total of 755 human fatalities. There 1s no clear trend; the
frequency of major accidents reported was steady over the period reviewed.
Minor releases of chemicals, not resulting 1n reported human health
or environmental damage, appear to be far more frequent, however. In
testimony before the House of Representatives, for Instance, Union
Carbide Identified 61 minor accidental releases of MIC at their Institute
-------
plant since January 1980. An analysis revealed that these were caused
operator error and mechanical defects, 1n both the storage and process
areas of the plant. None were caused by structural defects 1n the tanks
themselves.
Pressurized tanks tend to be Individually designed for specific
applications, making them Inherently difficult to regulate. The main
reasons for the good U.S. safety record to date have been Industry's
self-policing and the requirements of Insurance companies. Codes written
by professional societies provide design standards defining structural
and antlcorroslon requirements for tanks. Approval of design drawings by
registered Professional Engineers 1s another basic quality control-.
While most formal regulation of pressurized tanks occurs at the State
level, many States with large numbers of pressurized tanks have no
regulatlonSat all. It Is significant that no formal standards exist to
cover the design, maintenance, and operation of fall-safe systems and
other external devices, and yet'lt 1s the failure or lack of these
systems that has contributed to most of the significant accidents over
the past 20 years, Including that at Bhopal.
-------
TABLE OF CONTENTS
Page No.
EXECUTIVE SUMMARY 1
1. INTRODUCTION 1-1
1.1 Scope 1-1
1.2 Approach 1-1
1.3 Pressure Vessel Designs and Applications 1-2
1.3.1 Storage Applications 1-3
1.3.2 Processing or Treating Applications 1-5
1.3.3 Pressure Vessel Configurations 1-5
1.3.4 Pressure Vessel Design Considerations 1-13
2. EXIST ING INDUSTRY STANDARDS 2-1
2.1 ASME Code 2-1
2.2 Other Industry Standards Pertaining to Pressurized
Storage Tanks 2-2
2.3 Controls Not Addressed by the ASME Code 2-3
2.4 Emission Detection 2-5
2.4.1 Aboveground Tank Emission Detection 2-5
2.4.2 Underground Tank Emission Detection 2-8
2.5 Emission Containment 2-10
2.5.1 Double-Walled Tanks 2-10
2.5.2 Aboveground Tanks 2-10
2.5.3 Underground Tanks 2-12
2.6 Corrosion Control 2-14
3. EXISTING LEGAL REQUIREMENTS 3-1
3.1 Design Requirements 3-1
3.2 Inspection Requirements 3-2
3.3 Exemptions to Pressure Vessel Laws 3-11
4. INVENTORY INFORMATION 4-1
4.1 Available Statistics 4-1
4.2 Indirect Analysis 4-1
5. MAJOR RELEASE INCIDENTS FROM PRESSURIZED TANKS 5-1
-------
TABLE OF CONTENTS (continued)
Page No.
6. RELEASE OF METHYL ISOCYANATE AT INSTITUTE, WEST VIRGINIA,
AND AT BHOPAL, INDIA 6-1
6.1 MIC Production at UCC-Institute Facility 6-1
6.1.1 MIC Production Process 6-2
6.1.2 MIC Preliminary Storage (Unit) 6-5
6.1.3 MIC Secondary Storage (Underground) 6-6
6.2 Description of MIC Release Incidents at UCC-Inst1tute 6-8
6.2.1 Analysis of Causes 6-15
6.3 Safety Improvement at Institute Since Bhopal Event ... 6-17
6.4 Description of the Bhopal Release Incident 6-18
6.5 Comparison of Bhopal to Institute 6-19
APPENDIX A Summary of ASME Code Design Procedures A-l
APPENDIX B Selected Compatibility Information for
Tanks and Liner Materials B-l
APPENDIX C List of State Officials Contacted C-l
APPENDIX D Industrial Uses and Recommended Storage
Practices for Materials Commonly Stored
1n Pressurized Tanks 0-1
APPENDIX E Summary of Health Effects Information for
Materials Commonly Stored 1n Pressurized
Tanks E-l
APPENDIX F Glossary F-l
-------
LIST OF TABLES
Page No.
Table I. Summary of Health Effects Data for Materials
Commonly Stored 1n Pressurized Tanks • 7
Table II. Number of Facilities Likely to Employ Pressurized
Tanks to Store Materials Tested 1n Table I 8
Table III. Identified Pressurized Tank Accidents Resulting
1n Human Death 10
Table 1-1. Representative Pressurized Tank Applications 1-4
Table 2-1. Additional Industry Design Codes for
Pressurized Tanks 2-4
Table 3-1. Types of Pressure Vessels Covered by State
Design Requirements 3-3
Table 3-2. States with Cities That Have Pressure Vessel
Laws 3-6
Table 3-3. State Inspection Frequency Requirements for
Pressure Vessels 3-8
Table 3-4. Categories of Exemptions to State Pressure
Vessel Laws 3-12
Table 4-1. Telephone Survey Estimates of Pressurized Tank
Inventories by State 4-2
Table 4-2. Selected Toxic Substances Potentially Stored 1n
Pressurized Tanks, and Produced at More Than
10,000 Tons/Year 4-4
Table 4-3. Number of Facilities Likely to Employ Pressurized
Tanks to Store Materials Listed 1n Table 4-2 4-6
Table 4-4. Geographic Distribution of SIC Codes Associated
with Chemicals Possibly Stored 1n Pressurized
Underground Tanks 4-7
Table 4-5. Summary of Health Effects Data for Materials
Commonly Stored 1n Pressurized Tanks 4-9
-------
LIST OF TABLES (continued)
Page No.
Table 5-1. Summary of Readily Available Published
Information on Major Release Incidents from
Pressurized Storage Vessels 5-3
Table 5-2. Additional Industrial Gas Accidents That May
Involve Pressurized Tanks 5-8
Table 6-1. Institute Release Incident Summary 6-16
Table B-l. Chemical Resistance Guide for HYTREL B-2
Table B-2. NACE Chemical Compatibility Data Sample B-4
Table C-l. State Officials Interviewed C-2
Table D-l. Standard Industrial Classification Codes Associated
with the Manufacture and Industrial Use of
Poisonous Chemicals Potentially Stored 1n
Pressurized Underground Tanks D-2
Table E-l. Exposure Limits, Toxldty Data, and Adverse
Health Effects Associated with Chemicals
Potentially Stored 1n Pressurized Underground
Tanks E-2
-------
LIST OF FIGURES
Figure I.
Figure IIA.
Figure IIB.
Figure 1-1 .
Figure 1-2.
Figure 1-3.
Figure 1-4.
Figure 1-5.
Figure 2-1.
Figure 2-2.
Figure 2-3.
Figure 2-4.
Figure 3-1.
Figure 3-2.
Figure 6-1.
Figure 6-2.
Figure 6-3.
States with Pressurized Tank Design Laws
U.S. Accidents Involving Pressurized Tanks
Foreign Accidents Involving Pressurized Tanks ...
Pressure Vessel Configurations and Head Types ...
Horizontal Pressure Vessel with Elliptical
Heads
Horizontal Pressure Vessel with Hemispherical
Heads
Vertical Pressure Vessel with Cone Bottom
and 01 shed Top Heads
Spherical Vessel
Potential Pressurized Tank Leak Locations
Aboveground Pressurized Tank Instrumentation
Underground Pressurized Tank Instrumentation
Underground Pressurized Tank Containment
System
States with Pressurized Tank Design Laws
States with Pressurized Tank Inspection
Requl rements
UCC MIC Production Schematic
Unit Storage Schematic
MIC Storage Schematl c
Paqe No.
4
12
12
1-6
1-7
1-8
1-9
1-10
2-b
2-9
2-11
2-13
3-5
3-10
6-4
6-7
6-9
-------
EXECUTIVE SUMMARY
In the fall of 1984, Congress amended the Resource Conservation and
Recovery Act (RCRA) by adding Subtitle I, "Regulation of Underground
Storage Tanks." Shortly thereafter, the world was shocked by th'e
disaster at Bhopal, India, where leakage of toxic gas from a pressurized
tank at a Union Carbide Corporation (UCC) pesticide plant killed 2,000
people and seriously Injured thousands more. This triggered a wave of
fear 1n the United States over the possibility of a similar accident's
happening 1n this country, particularly since Union Carbide produces the
same chemical (methyl Isocyanate) 1n West Virginia. ^g-admlnUtrator rrf-
WrtfA"^ the Environmental Protection Agency 1s concerned about the adequacy
of existing controls over the tanks 1n question^
This report presents data on pertinent Issues. It Includes:
• A discussion of Industry, State, and local design standards for
pressurized tanks and their enforcement;
• A rough estimate of the number of pressurized tanks 1n the U.S.
and other countries;
• A preliminary compilation of release Incidents 1n the U.S. and
foreign countries; and
• A comparison of practices affecting methyl Isocyanate (MIC)
production and pressurized tank storage 1n the Union Carbide
plants at Bhopal and at Institute, West Virginia.
This Executive Summary contains highlights of the findings 1n each
of these areas. For clarity, 1t begins with a discussion of general
Issues pertaining to pressurized tanks, RCRA, and the Bhopal disaster.
-1-
-------
General Issues
Pressurized tanks can be used to store both liquid and gaseous
products. Pressures commonly range from less than 15 to more than 2,000
pounds per square Inch. These tanks are also used as part of production
processes; pressures 1n such applications can approach 5,000 ps1.
Since the major safety concern has, historically, been risk of
explosion, design standards are written with this 1n mind. They focus on
physical standards for the tanks themselves and on related matters like
avoiding Internal corrosion. Industry has taken the lead 1n policing
pressurized tanks, although many States have regulations as well.
The Bhopal Incident focuses attention on the environmental hazards
of pressurized tanks. The damage there was caused by leakage of toxic
gas resulting from an uncontrolled chemical reaction. This gas leakage
and the resulting air pollution problem Illustrate the multi-media nature
of underground pressurized tank Issues.
Industry and State Standards
While many States have Incorporated portions of pertinent
professional society design standards Into their laws, Industry Itself 1s
the major force 1n regulating the safety of pressurized tanks, largely
through the requirements of Insurance companies. (For Instance, even
where standards are nominally written Into State law, Inspections are
generally carried out by Insurance companies.) The other significant
private force encouraging adequate design 1s the requirement that design
drawings be approved by certified Professional Engineers.
-2-
-------
Four States believed to have a large number of pressurized tanks
(New York, Michigan, Texas, and West Virginia) are among those with no
State laws governing pressurized tanks.
Design standards, formal and Informal, fall Into four general
categories: structural design, Internal corrosion protection, external
corrosion protection, and fall-safe systems. These vary significantly 1n
scope and applicability, largely because most systems are Individually
designed for unique applications.
Structural design standards: The most comprehensive structural
design and fabrication standards for pressurized tanks are contained In
the American Society of Mechanical Engineers (ASME) code; however, these
standards must be tailored to the specific requirements at hand. Their
main purpose 1s to prevent explosions from Improper design, fabrication,
or use. Portions of the ASME code have been Incorporated by reference
Into the laws of 32 States, as shown 1n Figure I.
Internal corrosion standards: There 1s no universal source document
of standards addressing the compatibility of tank contents and tank
materials. The Corrosion Data Survey, published by the National
Association of Corrosion Engineers (NACE), 1s the most comprehensive
reference available, since 1t contains extensive charts showing the
relative compatibility of various tank materials and chemicals. Similar
compatibility charts are also available from tank and chemical
manufacturers.
-3-
-------
VTA
ALASKA
KEY isicnoNi or ASME coot IN STATE LA*I
ASME SECTIONS VIM III 121 AND K (METAL AND FIBERGLASS TANKS!
ASME IICTIONS VIII III 111 IMtTAL IANKS ONI V|
| | NO r«ESSUR»ZEO TANK LAWS
Figure I. States with Pressurized Tank Design Laws
-------
External corrosion standards: Standards for external corrosion
control are applicable to underground tanks 1n general and are not
written specifically for pressurized tanks. The most widely used
Industry/State standard 1s the National F1re Protection Association's
(NFPA) Code 30. This requires that some form of external corrosion
control be used whenever soil resistivity 1s less than 10,000 ohm-cm.
Requirements are less specific than for structural design or Internal
corrosion prevention; the designer 1s thereby given wide latitude.
Fall-safe controls: These measures Include physical design elements
to prevent leaks (e.g., double-walled tanks) or to avert damage 1f leaks
occur (collection/treatment devices like flares or scrubbers). They also
Include various types of Instruments and automated controls to give early
warning of problems or to shut down systems when problems occur. There
are no widely accepted Industry standards 1n this area similar to the
ASME, NACE, or NFPA codes.
Estimates of Numbers of Pressurized Tanks and Their Contents
There 1s no central Information base on the number of pressurized
tanks 1n use 1n the U.S. or the number of these located underground.
Industries using pressurized tanks are concentrated 1n the Industrial
States of the Northeast, Great Lakes Region, Gulf Coast, and California.
Table I shows the number of plants, by Industry type (SIC code), that are
likely (but not certain) to employ pressurized tanks. The total 1s
15,000 facilities. If each facility used 10 pressurized tanks, the
national total would be about 150,000.
-5-
-------
0370s
Table I Number of Facilities Likely to Employ Pressurized Tanks
to Store Materials
SIC Code
1321
2041
2043
2812
2813
2819
2821
2822
2823
2824
2834
2841
2869
2879
2892
2899
2911
2999
3111
3255
TOTAL
Industry
Natural Gas Liquids
Flour and Grain Hill Products
Cereal Preparations
Alkalies, Chlorine
Industrial Gases
Industrial Inorganic Chemicals
Plastic Material
Synthetic Rubber
Cellulosic Man-made Fibers
Organic Fibers
Pharmaceutical Preparations
Soap/Other Detergent
Industrial Inorganic Chemicals
Agricultural Chemicals
Explosives
Chemical Preparations
Petroleum Refining
Petroleum and Coal Products
Leather, Tanning and Finishing
Clay Refractories
No. of facilities3
187
618
118
171
599
2183
1529
311
79
165
1512
996
1160
861
172
2563
1281
108
594
263
15,469
a Source: Dun's Marketing Service (1983). Standard Industrial
Classification Characteristics.
-6-
-------
Officials 1n all 50 States were contacted directly for estimates of
the number of pressurized tanks within their boundaries. Of this group,
12 were able to provide approximations; their estimates ranged between
10,000 and 30,000 each. (It was further Indicated that most of'the tanks
are above ground.) Thus, 1n combination, these two sources of
Information suggest that the total number of tanks nationwide Is well
over 100,000, and probably several times that amount.
Table II lists those 38 regulated gases and liquids that (1) are
produced nationwide at rates believed to exceed 10,000 tons per year and
(2) are likely to be stored 1n pressurized tanks. Also shown are the
recommended maximum exposure concentrations for each substance. Methyl
Isocyanate 1s on this 11st of relatively high-volume regulated
substances; 1t also has the lowest permissible exposure concentration.
Summary of Release Incidents over Past 25 Years
This study Involved a keyword search on eight automated data bases
to Identify Incidental chemical releases. It Included general news
services (UPI, AP, and others) as well as chemical Industry data
literature. Access was also obtained to an unpublished 11st, prepared 1n
Germany, of 31 accidents that occurred between 1964 and 1979. The types
of sources used biased the search toward major releases or actual
disasters. The results, and therefore the conclusions of this study, are
not definitive.
-7-
-------
0370s
Table II Summary of Health Effects Data for Regulated Materials Conmonly
Stored in Pressurized Tanks
Chemical
Methyl isocyanate
Acrolein
Phosgene
Allyl chloride
Chloroprene
Vinyl chloride
Allyl ami ne
Sulphur dioxide
Carbon disulfide
Dimethylamine
Hydrogen cyanide
Hydrogen sulfide
Aimcnia
Ethyl ene oxide
Methyl chloride
Methylene chloride
Propane
Recommended' maximum Potential carcinogen
exposure concentration (ppm) or mutagen
0.02
0.1 *
0.1
1
1
1
5
5
10 *
10
10 *
15 *
35
50 *
100 *
500 *
1.000
1. Lowest value of OSHA, NIOSH, or ACGIH recommended limit.
Note: See Appendix E for detailed health effects data.
-8-
-------
In all, the search uncovered 54 accidents resulting 1n significant
damage to human health and/or the environment. About half Involved the
storage of liquified natural gas (LNG) or liquified petroleum gas (LPG).
The remainder Involved a variety of chemical products Including
pesticides, fertilizers, aerosols, chlorine, and ammonia. Two of the
accidents Involved methyl Isocyanate. The Identified accidents took
place 1n 15 States and 12 foreign countries. In the U.S., New York and
Texas had the most accidents.
Fourteen of the 54 accidents resulted 1n human fatalities; total
fatalities were 755 (the 11st excludes Bhopal). Seven of the 14 fatal
accidents, which accounted for 76 deaths, occurred 1n the U.S. Three of
these Involved LNG storage, and the remaining 4 stemmed from chemical and
petroleum production. Table III lists these 14 fatal Incidents.
On the basis of this sample, which 1s not statistically
representative and mainly Includes Incidents that would be noted by the
general press, no clear trends emerge. Figures Ha and lib depict trends
1n the dated Incidents, U.S. and foreign. The number of major reported
Incidents 1s unchanged or has declined over the past two decades.
Comparison of Practices at Bhopal and Institute. West Virginia
According to Union Carbide's official report, the Bhopal disaster
was Initiated by the reaction caused by the Introduction of between 1,000
and 2,000 gallons of water, possibly combined with high levels of
-9-
-------
0370s
Table III Identified Pressurized Tank Accidents Resulting in
Human Death
Country/State
United States
Illinois
Mississippi
New York
New Jersey
Louisiana
Texas
Massachusetts
Mexico
France
Brazil
South Africa
Argentina
Canada
West Germany
Year
1984
1982
1973
1973
1967
1964
1964
1984
1973
1972
1968
1967
1966
1966
Material
Refined Petroleum
Chlorinated Paraffin
LNG
Unidentified Chemical
Isobutane
LNG
Vinyl Chloride
Subtotal
LPG
LPG
LPG
Ammonia
LNG
Polystyrene
LNG
Subtotal
TOTAL
No. of Deaths
17
3
40
7
7
2
7
83
490
4
38
26
100
n
_3
672
755
-10-
-------
Number of
incidents
1 '-I
I
y
'6 '6 '6 '6 '6 '6 '6 7 7 7 7 7 7 7 7 7 7 -8 * 'S « *8 "8
543678901 23436789012343
Year
O U.S. Accidents without
fatalities
U.S. Accidents Involving
Fatalities
Figure MA U.S. Accidents Involving Pressurized Tanks
Number of
incidents
Will fil
'6 '6 '6 '6 '6 '6 '6 7 7 7 7 7 7 7 7 7 7 "8 "8 '8 3 "8 "8
345678901 2345678901 2345
Year
LJ Foreign Accidents without
Fatalities
• Foreign Accidents Involving
Fatal itie<
Figure MB Foreign Accidents Involving Pressurized Tanks
Not all incidents identified in the literature search were dated. The
dated ones indicated, if anything, a decline in major events reported in
the U.S., and no change in events reported in the foreign press. These
data are, however, not statistically representative.
-11-
-------
chloroform, Into a pressurized tank containing MIC. The actual cause of
the release was the failure of four system elements, Including three
fail-safe devices. The refrigeration system designed to cool the MIC
storage tanks would have slowed the reaction and allowed more time for
control, but 1t had been taken out of service earlier 1n the year. None
of the fail-safe devices — high temperature alarm, a vent gas scrubber,
and a flare — could be operated for various reasons, even though all
were mechanically 1n working order.
The Union Carbide plant at Institute has safety equipment similar to
that at Bhopal. These devices Include double-walled storage tanks with
pressure sensors between the tanks, as well as temperature, pressure, and
volume Indicators and alarms 1n the main tanks. Tanks are kept no more
than 80 percent full and are cooled by a redrculatlng coolant. Pressure
controls permit tanks to vent to a scrubber/flare tower. In addition,
Union Carbide has taken several steps since the accident to upgrade and
Improve the facility at Institute, thereby further reducing the
possibilities of an accident. These steps Include:
• Substitution of a chloroform cooling system for the original
brine cooling system. This removes a source of water from the
area that could cause a runaway reaction 1f a major breach
occurred.
• Addition of an automatic air sampling system to detect MIC at 15
locations. Any leak would cause an automatic shutdown.
• Addition of a new emergency scrubber oversized by a safety
factor of 5. (The vent scrubber has always been continuously
on-Hne.)
-12-
-------
• Installation of a surge tank on the emergency vent scrubber.
• Addition of a continuous flare to back up the emergency
scrubber, should that have to be bypassed.
• Increase 1n pressure relief valve settings by a factor of 2, to
50 ps1.
These Improvements are specifically aimed at avoiding an Incident
similar to Bhopal. The most significant undertaking 1s probably the new
chloroform cooling system, which removes the proximate cause of the
Bhopal accident.
On December 14, 1984, Union Carbide testified before the House
Subcommittee on Health and the Environment that there had been 28
releases of MIC at the Institute plant between January 1980 and that
date. UCC later amended this number to 61. This study had access to
sufficient data to analyze 26 of the original 28 Incidents. Sixteen were
attributable to operator error, 10 to defective equipment. The Incidents
also appeared to be as likely to occur 1n process areas as 1n storage
areas. It 1s therefore unclear whether the facility Improvements made 1n
the storage areas are apt to reduce the number of lesser release
Incidents, such as those addressed 1n the House hearings.
Conclusions
Based on the findings of this 2-month preliminary Investigation Into
pressurized tank-related Issues, we conclude the following:
• An accident of the type and magnitude of the Bhopal disaster has
always been less likely to occur at the Institute plant, largely
because of better and more consistent operations and maintenance
procedures 1n the U.S. plant.
-13-
-------
The likelihood of a major accident 1s less today than 1t was
prior to the Bhopal Incident, since Union Carbide made major
facility changes 1n response to the publicity generated by
Bhopal.
The Improvements at Institute do not necessarily address the
problem of recurring minor releases that appear to transpire
frequently (61 Incidents 1n a 4-year period). (Improved
operator training and operations and maintenance (0/H)
procedures may have been Implemented since 1984, but have not,
to our knowledge, been documented In UCC's official statements.)
A very preliminary review of the evidence suggests no obvious
trends, upward or downward, 1n the number of releases of toxic
chemicals from pressurized tanks.
The large number of pressurized tanks nationwide (150,000 plus)
may frequently release small amounts of toxic chemicals through
minor equipment failures or operator mistakes.
Structural standards for pressurized tanks themselves do not
appear to need Improvement. 0/M procedures and enforcement,
operator training, corrosion control, and fall-safe standards
are more appropriate areas for concern, even though the
variation 1n design requirements among pressurized tanks does
not lend Itself to setting simple prescriptive standards.
-14-
-------
1. INTRODUCTION
1.1 Scope
Pressurized tanks are often used for the production and temporary
storage of chemicals. These tanks are frequently above ground although
associated piping may be underground. The RCRA amendments specifically
exclude from regulation "flow-through" process tanks and tanks with less
than 10 percent of their volume (Including piping) below ground. We have
not attempted to exclude Information on pressurized vessels that may fall
under one or both of the above exemptions at this time. Rather, we have
tried to assemble readily available Information on all pressurized tanks
containing regulated substances.
1.2 Approach
Section 1 presents the scope and approach of this report and
explains why pressurized tanks are used 1n various applications.
Section 2 discusses a summary of current Industry standards. The
requirements of Industry codes pertaining to the design and fabrication
of pressurized tanks are Identified, and key areas not addressed by these
codes are highlighted. This 1s followed by a discussion of alternative
control technologies used to prevent, detect, and contain releases from
pressurized tanks.
Section 3 cites the results of a telephone survey to officials 1n
all 50 States regarding pressurized tank design and Inspection laws. The
results are provided 1n tabular and graphic form.
1-1
-------
Section 4 provides Inventory Information for pressurized tanks
obtained from officials 1n each State. Results are also presented of an
Indirect Inventory 1n which Versar Identified regulated materials likely
to be stored 1n pressurized tanks, based on temperature, vapor pressure,
and production characteristics. The Industries using these materials are
cited, along with the number and geographic distribution of facilities 1n
those Industries, 1n an attempt to provide an Independent Inventory
estimate. Finally, health-related Information for each of these
materials 1s compiled and evaluated.
Section 5 contains the results of a literature search to Identify
and characterize major accidents that Involved (or possibly Involved)
pressurized tanks. A more detailed assessment of release Incidents at
Union Carbide facilities 1n Bhopal, India, and 1n Institute, West
Virginia, 1s discussed 1n Section 6. This comparison 1s an attempt to
determine the potential of a major accident's occurring at Institute's
Union Carbide facility.
1.3 Pressure Vessel Designs and Applications
By definition, a pressure vessel 1s a container designed to process,
treat, or store a chemical substance at nonamblent pressures; that 1s, 1f
a substance 1s to be processed, treated, or stored 1n a vessel at a
pressure greater or less than the atmospheric pressure surrounding the
vessel, a pressure vessel 1s required. In comparison, containers that
are used to process, treat, or store substances at atmospheric pressure
1-2
-------
or slightly above are called storage tanks, flat bottom tanks, or API
tanks. Pressurized tanks may be categorized as low-pressure storage
tanks, high-pressure storage tanks, and chemical production process
tanks. Representative examples of each are presented 1n Table 1-1.
This section addresses the conditions that would necessitate the use of
pressure vessels, the types of pressure vessels available, and other
design considerations for pressure vessels.
1.3.1 Storage Applications
Each chemical substance Imposes a pressure on the walls of the
vessel 1n which 1t 1s contained. This pressure 1s called the substance's
"vapor pressure" and 1s a function of the temperature at which the
substance 1s stored. Vapor pressure always Increases with Increasing
temperature.
If the vapor pressure of a particular substance at a given storage
temperature 1s less than the atmospheric pressure, three storage
alternatives are available:
• Store the substance at Its vapor pressure. The storage
container would then be operated at a subatmospherlc pressure
(vacuum), and a pressure vessel would be required since the
storage pressure would be less than the atmospheric pressure.
• Store the substance 1n a container open to the atmosphere (I.e.,
not a pressure vessel).
• Store the substance at a pressure greater than atmospheric
pressure by "blanketing" the vessel with a gas that has no
effect on the substance stored. "Blanketing" 1s the process of
Injecting a gas Into a vessel to maintain a desired pressure.
Since the vessel 1s to be operated at higher than atmospheric
pressure, a pressure vessel 1s required.
1-3
-------
Table 1-1 Representative Pressurised Tank Applications
Application
Typical Pressure
Range
Examples
Lou-pressure storage of
volatile materials
<15 psi Methyl isocyanate
Storage of compressed gases
20-200 psi
Liquified natural
gas, chlorine
Chemical production processes 100-5,000 psi
Synthesis of low-
density poly-
ethylene
1-4
-------
If the vapor pressure of a particular substance at a given storage
pressure 1s greater than atmospheric pressure, the substance must be
stored 1n a pressure vessel.
1.3.2 Processing or Treating Applications
If a chemical substance must be processed or treated 1n a vessel at
a pressure different from the atmospheric pressure, the processing/
treating vessel must be a pressure vessel. The pressure at which the
processing/treating operation 1s performed 1s determined by the process
requirements and other economic considerations. For most applications,
an optimal processing/treating pressure exists, establishing both the
need and some of the specifications for the pressure vessel.
1.3.3 Pressure Vessel Configurations
As depicted 1n Figure 1-1, there are two general configurations for
pressure vessels. The first configuration consists of a cylinder with
two "heads" (ends) and 1s shown 1n Figure l-l(A), l-l(B), or l-l(C). The
other general configuration 1s a sphere, as Illustrated 1n Figure l-l(D).
Cylinders can be made Into any diameter or length. Heads are
typically produced as a formed entity and their sizes are standardized.
Standard head sizes are available 1n Increments of 2 Inches from a
diameter of H Inches to 42 Inches. From 42 Inches upward, diameter
Increments of 6 Inches are used. Because of this large but limited
number of head sizes, a continuum of vessel diameters 1s not available,
and the cylinders must be formed to match the diameter of the
appropriately sized head.
1-5
-------
(A)
HORIZONTAL PRESSURE VESSEL
WITH ELLIPTICAL HEADS
(See Figure 1-2)
(C)
VERTICAL PRESSURE VESSEL
WITH CONE BOTTOM AND
DISHED TOP HEADS
(See Figure 1-4)
(B)
HORIZONTAL PRESSURE VESSEL
WITH HEMISPHERICAL HEADS
(See Figure 1-3)
(D)
SPHERICAL VESSEL
(See Figure 1-5)
Figure 1-1 Pressure Vessel Configurations and Head Types
1-6
-------
Figure 1-2 Pressure Vessels with Elliptical Heads
-------
ffSraJJTY
Figure 1-3 Pressure Vessel with Hemispherical Heads
1-8
-------
Figure 14 Pressurized Vessel with Dished Heads
-------
•
Figure 1-5 Spherical Pressure Vessel
1-10
-------
Four general types of heads are available. Elliptical heads, which
are depicted 1n Figure l-l(A), are formed 1n the shape of an ellipsoid.
The most commonly used head 1s a 2:1 elliptical head where the depth of
the head 1s one-fourth the diameter. Elliptical heads are used'for
moderate-pressure services. Hemispherical heads (Figure l-l(B)) are
deeper than elliptical heads, and this depth provides greater structural
strength. Hence, hemispherical heads are used for high-pressure
services. The top head of the vessel shown 1n Figure l-l(C) 1s termed a
"flanged and dished" head. This type of head 1s shallower than an
elliptical head, diminishing Us structural strength. Therefore, flanged
and dished heads are used 1n low-pressure services. These pressure
classifications are very general, and for any specific pressure vessel,
the type of heads selected will be purely a function of economics.
Also presented 1n Figure 1-4(C) 1s a conical head that 1s used for
handling solids or partlculates present 1n the pressure vessel. The
structural Integrity of this shape, however, limits Its use to
low-pressure applications.
The shape of cylindrical pressure vessels can be categorized by the
ratio of the cylindrical length (L) to the diameter (D). Pressure
vessels used for storage generally have a Iength-to-d1ameter ratio 1n the
range of 1.5 to 6.0. As the diameter Increases and the L/D ratio
decreases, the vessel wall thickness requirement Increases and, hence, so
does the vessel cost. For any specific storage application, given a
1-11
-------
required storage volume, pressure, and temperature, the L/D ratio 1s
selected on the basis of the economic optimum cost. For processing or
treating applications, the shape of the pressure vessel 1s usually
governed by the process requirements.
Cylindrical pressure vessels can be positioned either horizontally
(Figures l-l(A) and l-l(B)) or vertically (Figure l-l(C)). Vertical
vessels are commonly used for storage of small volumes of material. This
1s because vertical vessels are generally less costly and take up less
real estate than Identically-sized horizontal vessels. As vessel volume
requirements Increase, however, horizontal vessels become more practical
and economic. This 1s due to the Increased heights associated with large
vertical vessels, which pose an accessibility problem and a need for
ladders and platforms. Additionally, as vessel heights Increase, the
effects of wind and other climatic conditions result 1n costlier
foundations and thicker vessel wall requirements. Generally, the
cylindrical length of vertical storage vessels 1s no longer than 10
feet. For processing or treating applications, the position of the
pressure vessel 1s usually governed by the process requirements.
A spherical pressure vessel 1s depicted 1n Figure l-l(D). These
types of vessels, which are generally fabricated at the plant or job
site, are limited to low- and moderate-pressure services. The selection
of a spherical pressure vessel versus a cylindrical type 1s governed
solely by economic factors.
1-12
-------
1.3.4 Pressure Vessel Design Considerations
Four parameters generally determine the design and selection of a
pressure vessel: operating temperature, operating pressure, material of
construction, and volume (shape) requirement. Once these four parameters
are established, an engineer or vessel designer can design a pressure
vessel that 1s economical and satisfies the process requirements.
The operating temperature 1s the temperature at which a substance 1s
to be processed or stored. As temperature Increases, the structural
Integrity of the vessel material of construction decreases, resulting 1n
an Increased wall thickness requirement. In addition, higher
temperatures may necessitate the use of Insulation to limit heat losses
and/or protect personnel who come 1n contact with the vessel. Hence, for
storage vessels, operating temperatures are limited by cooling the
substance to be stored, where possible.
The operating pressure 1s the pressure at which a substance 1s
processed or stored. As pressure Increases, the structural Integrity of
the vessel decreases, again resulting 1n an Increased wall thickness
requirement. For storage vessels, the operating pressure 1s a function
of the vapor pressure. Limiting the operating temperature will generally
lower the operating pressure, which results 1n a more economical design.
The material of construction selection for a pressure vessel 1s a
function of the material's compatibility to the chemical substance
stored. A material must be chosen that will resist corrosion and not
1-13
-------
affect the chemical substance contained. Some materials of construction
can catalyze undeslred reactions 1n certain chemical substances.
The pressure vessel volume requirement 1s usually a function of
process requirements and/or plant operating philosophy.
1-14
-------
2. EXISTING INDUSTRY STANDARDS
2.1 ASHE Code
The American Society of Mechanical Engineers (ASME) publishes design,
fabrication, and Inspection standards for boilers and pressurized tanks.
These standards are presented 1n the ASME Boiler and Pressure Vessel
Code, and are revised periodically. Approximately 1,000 manufacturers
have been approved to place the ASME seal on their products. Industry
members usually adhere to the ASME requirements, 1n part to fulfill
Insurance requirements and to demonstrate that they act according to
accepted Industry standards and practices should an accident occur.
Section VIII of the ASME Code addresses nonfIberglass pressurized
tanks. There are two methods of tank design 1n Section VIII, referred to
as Division 1 and Division 2. Division 1 employs a nominal safety factor
of about 4 at ambient temperature, but generally Ignores many of the
secondary stresses that act on a vessel. Division 2 allows higher
stresses (safety factor of about 3). but requires a thorough stress
analysis, Including fatigue, plus a closer control of quality In
materials and fabrication. Division 1 procedures may be used to design
tanks from 15 to 3,000 ps1, while Division 2 procedures may be used for
any tanks of 15 ps1 or greater.
Section X of the ASME Code establishes general specifications for the
glass and resin used to fabricate fiberglass pressurized tanks. It sets
limits on permissible service conditions and sets rules under which
fabricating procedures are required.
2-1
-------
While the ASHE Code provides detailed specifications 1n some areas,
other areas are covered only generally or not at all. For example, the
ASME Code contains detailed design, fabrication, and Inspection
procedures for pressurized tanks, Including pressure relief valves. The
Issue of Internal corrosion Is dealt with, however, only by saying that a
corrosion allowance should be provided if the stored material Is known to
cause corrosion of the vessel metal. No specific guidelines are given in
the Code for either setting the allowance for Internal corrosion or for
determining the compatibility of the stored material.
The issue of external corrosion control is not specifically addressed
by the ASME Code, although external corrosion is a major concern for
underground tanks. Similarly, controls such as leak detection devices,
secondary containment, or air scrubbers to collect/treat materials
released are not covered by the code. Therefore, while the ASME Code
Includes detailed design, fabrication, and inspection procedures to
prevent leaks from aboveground pressurized tanks, it appears to be
lacking 1n the areas of internal/external corrosion control and 1n
detection/containment of material should a release occur. These
deficient areas reflect the ASME Code's emphasis on preventing explosions
or catastrophic releases from the tank through structural design/
fabrication procedures. A summary of specific design practices required
by the ASME Code is presented in Appendix A.
2.2 Other Industry Standards Pertaining to Pressurized Storage Tanks
Although the ASME Code is the most comprehensive and widely accepted
industry standard pertaining to pressurized tanks, other related
2-2
-------
voluntary Industry standards are available. These codes and their areas
of applicability are shown 1n Table 2-1.
These other codes often draw heavily from the ASME Code and provide
additional requirements to meet the needs of the particular
applications. For example, National F1re Protection Association (NFPA)
58 Code for the storage of Liquified Petroleum Gases (LPG) requires that
containers be manufactured according to the ASME Code, Division VIII,
Section 1; however, other requirements such as maximum container size
(120,000 gallons) and wind loading requirements are added. The American
Petroleum Institute (API) 620 Code applies to low-pressure tanks only.
The National Association of Corrosion Engineers (NACE) publishes
extensive charts showing the relative compatibility of various
materials. These charts, which are updated periodically, may be used to
select tank construction materials to minimize Internal corrosion for the
material to be stored. While not exclusively for pressurized tanks, the
NACE charts may be used 1n selecting tank materials, which may then be
employed using ASME procedures to design pressurized tanks. An example
of a NACE corrosion chart 1s found 1n Appendix B.
2.3 Controls Not Specifically Addressed by Industry Codes
Industry standards for external corrosion control are much less
clearly defined than those for structural Integrity and Internal
corrosion. For example, NFPA 30 states that for underground tanks 1n
soils with resistivity 1n excess of 10,000 ohm-cm, or 1f other corrosive
conditions exist, some form of corrosion control should be employed.
2-3
-------
0069s
Table 2-1 Additional Industry Design Codes for Pressurized Tanks
Source
Design code
Description
American Petroleum Institute 620
(API)
American Petroleum Institute 520 and 521
(API)
National Fire Protection 58
Association (NFPA)
National Fire Protection 59
Association (NFPA)
National Association of Corrosion
of Corrosion Engineers Data Survey
(NACE)
Design procedures for
low-pressure tanks (2.5
to 15 psi).
Detailed procedures for
the design of pressure
relief systems.
Design of petroleum gas
tanks.
Design of utility gas
plant tanks.
Material compatibility
charts for internal
corrosion.
2-4
-------
No specific control practices are recommended, however. Such practices
may Include tank coating and/or electrical protection to retard external
corrosion.
Similarly, automated controls to monitor pressurized tank operating
conditions are employed by Industry. These controls may be quite
sophisticated and, 1n some cases, can shut down an entire production
process automatically. There 1s, however, no published Industry standard
comparable to the ASME or NACE standards as to the design and use of such
controls. Individual engineering design firms may have their own
requirements, and the liability that Is associated with P.E. approval of
the design offers an Incentive to Implement automated Instrumentation and
controls.
Specific Instrumentation, containment, and corrosion-prevention
technologies applicable to pressurized tanks are discussed 1n Sections
2.4, 2.5, and 2.6.
2.4 Emission Detection
2.4.1 Aboveground Tank Emission Detection
Emissions from aboveground pressurized tanks may result from a
pressure control device, a loose or faulty fitting (e.g., cracked flange,
defective valve), or structural failure of the tank Itself. Causes of
such releases Include any of the following: overpressurlzatlon of the
tank; poor operative control and/or maintenance; and poor construction of
the tank Hself. (Potential leak locations are Identified 1n Figure 2-1.)
A1r emission-detection techniques for aboveground tanks Include both
visual and Instrument methods. Where pressurized tanks are used, 1t 1s
2-5
-------
r\i
i
f ROM A PROCESS
LOCATION
OVER PRESSURIZATION
OR DEFECTIVE DEVICE
PRESSURE RELIEF
DEVICE
TO VENT CONTAINMENT
OH ATMOSPHERE
TO ANOTHER PROCESS
LOCATION
Figure 2-1 Potential Pressurized Tank Leak Locations
-------
standard Industry practice for facility operators to be trained to detect
and control process leaks within their areas of responsibility. This 1s
particularly true of operators responsible for the processing and
handling of compressed and liquefied gases. Process Instrumentation 1s
therefore vital to process operators. The devices are designed to
provide operators with process variable Information and, 1n some cases,
to supply complete automatic control of the process.
Pressurized tanks may be equipped with pressure, level, and/or
temperature-senslng/control Instrumentation.
(1) Pressure Controls. Pressure-measuring equipment provides
pressure Indication 1n both field (local) and control room (remote)
locations. Field-mounted units are mounted directly onto the tank, while
remote units are connected to a control room panel. The Instrumentation
can also be equipped with alarms to Indicate critical tank pressures.
The pressure Instrumentation can automatically control valves or
other process equipment designed to aid In the pressure control of a tank
and, 1n some cases, can shut the plant down. The process Instrumentation
can also provide early warning so that proper action can be undertaken by
plant operators to avoid a potentially hazardous situation.
(2) Level Controls. Level control Instrumentation provides
operators with Information regarding liquid level (and volume) 1n the
tank. As with the pressure Instrumentation, local as well as panel-
mounted Indicators with alarms are available. During static conditions
2-7
-------
(I.e., no product transfer), a decreasing liquid level 1n the tank should
Indicate to the operator(s) that the tank may be leaking and that proper
action should be taken to correct this situation.
(3) Temperature Controls. Where Internal tank temperature.1s a
critical control variable, some units can be equipped with temperature-
sensing equipment. An operator noting a rise 1n tank temperature above
normal should associate 1t with Increased tank pressure, and take
appropriate steps to ensure that overpressure and release do not occur.
(4) Leak Detection. In addition to pressure, level, and temperature
Instrumentation, sp1ll-mon1tor1ng Instrumentation 1s available.
Instrumentation can consist of automatic air sampling devices, which feed
to a gas chromatograph, or probes, which can be calibrated for specific
substances. These probes can be placed between the walls of a double-
walled tank or on the soil near the tank. The spill-detection device 1s
tied Into an alarm system that Indicates to an operator that a spill has
occurred, and can be designed to activate other controls or shut down the
plant. A schematic diagram showing aboveground tank detection devices 1s
given 1n Figure 2-2.
2.4.2 Underground Tank Emission Detection
Underground pressurized tanks, like their aboveground counterparts,
would probably also be equipped with pressure, level, and temperature-
monitoring Instrumentation. These devices take on added Importance since
the tank 1s underground and cannot be visually Inspected. In-s1tu probes
2-8
-------
ALARMS
N>
to
ALARMS
SPILL-SENSING
INSTRUMENTATION
LOCAL
INDICATION
GAUGES
PRESSURIZED TANK
PRESSURE-SENSING
INSTRUMENTATION
LEVEL-SENSING
INSTRUMENTATION
,
TEMPERATURE-SENSING
INSTRUMENTATION
CONTROL
PANEL
WITH
PROCESS
INDICATION
Figure 2-2 Aboveground Pressurized Tank Instrumentation
-------
can be placed within a well that Is slotted to allow leaking substances
to come 1n contact with the detector's sensing mechanism. Figure 2-3 1s
a schematic representation of underground tank Instrumentation.
2.5 Emission Containment
2.5.1 Double-Walled Tanks
A double-walled pressurized tank can be constructed with each wall
designed to contain Individually the maximum anticipated operating
pressure 1n the tank. This has the dual advantage of detecting leaks 1n
the space between the two walls and containing them before they are
released to the environment. The Incremental added cost for pressurized
double-walled tanks would, however, be significantly greater than for
unpressuMzed double-walled tanks, because of the Increased design,
material, and construction costs. For this reason, double walls are not
normally used for pressurized tanks.
2.5.2 Aboveground Tanks
Process spill or leak containment for aboveground tanks containing
liquids consists of a dike (or berm) surrounding the area. The diked
area, which 1s usually made of concrete, is designed and constructed to
contain the total liquid volume of a tank (or tanks) located within 1t.
Depending upon the type of liquid to be retained within the dike, it may
be lined with an antlcorrosive membrane. Often, the dike is equipped
with a sump and sump pump, which, in the event of a spill, will transfer
the spilled liquid to a safe location where 1t will be neutralized.
2-10
-------
CONTROL
PANEL
WITH
PROCESS
INDICATION
ALARMS
TEMPERATURE-SENSING
INSTRUMENTATION
PRESSURE-SENSING
INSTRUMENTATION
LEVEtrSENSING
INSTRUMENTATION
-rl
ALARM
SPILL-SENSING
INSTRUMENTATION
LOCAL
INDICATION
GAUGES
FILL DIRT
Source: Leak Corporation, U.S.A.
Englewood Cliffs, N.J.
Figure 2-3 Underground Pressurized Tank Instrumentation
-------
Depending upon the service to be supplied by the pump and the Immediacy
with which the spill must be removed, the removal process can be
activated by manual or automatic means.
A1r emissions from aboveground tanks can be contained by conventional
air pollution control equipment such as scrubbers or flares, depending on
the chemical nature of the material being released. There may be more
than one of these systems to serve as a fall-safe mechanism 1n preventing
release of the contaminant to the environment. The degree to which such
fall-safe systems are Implemented 1s largely determined by the user.
2.5.3 Underground Tanks
A leak from an underground storage tank can be contained 1n a fashion
similar to that used for aboveground tanks. In Heu of a dike around the
tank, a resistant Uner 1s placed within the trench that will hold the
tanks. The tanks are then covered with earth as 1n a normal underground
tank Installation. This type of Installation ensures that any leaking
liquids (or gases) from tanks will be retained within the barrier of the
Uner. Removal of the released substance would likely Involve the
removal of the earth within the boundary of the Uner. The Uner must
also be chemically compatible with the material being stored. While
there are no regulatory standards for Uner compatibility, manufacturers
do provide such Information (see Appendix B). Figure 2-4 1s
representative of an underground spill-containment system.
2-12
-------
TYPICAL CONCEPTUAL VIEW
LINER
TANKS
TYPICAL CROSS-SECTION VIEW
LINER
TANK FILL DIRT
-— r
1C
—
U
'•"7-
*ES
^V
^
Figure 2-4 Underground Pressurized Tank Containment System
2-13
-------
2.6 Corrosion Control
All of the external corrosion-control alternatives available for
unpressurized tanks can usually be applied to pressurized tanks. These
alternatives include (1) cathodic protection, (2) fiberglass coating,
(3) cathodic protection plus coating, and (4) fiberglass tanks. The use
of such devices 1s normally limited to underground pressurized tanks.
There 1s no centralized Information regarding the degree to which
external corrosion controls are employed on underground pressurized
tanks, and this decision 1s made by the design engineer on a case-by-case
basis.
2-14
-------
3. EXISTING LEGAL REQUIREMENTS
Existing laws, rules, and regulations associated with the
construction, installation, and inspection of pressure vessels vary by
State and local jurisdiction. The existing laws and regulations 'for
pressure vessels in each State are reviewed and summarized in this
section. The data were derived from the synopsis of Boiler and Pressure
Vessel Laws, Rules, and Regulations, published by the Uniform Pressure
Vessel Laws Society, Inc. This publication provides a summary of
pertinent laws on a State-by-State basis, including rules and regulations
for construction, stamping, and inspections.
Information in the synopsis was confirmed and augmented by telephone
interviews with the responsible officials in each of the 50 States.
Appendix C is a list of the officials contacted. During the telephone
Interviews, efforts were made to obtain information on the degree of
enforcement of these legal requirements in the applicable jurisdictions.
Questions were asked regarding the following issues:
• Whether inspection is performed by the State, by insurance
representatives, or by others; and
• Whether inventory data on pressure vessels are available.
3.1 Design Requirements
Those jurisdictions that regulate pressure vessels generally specify
design requirements by incorporating relevant sections of the ASME Boiler
and Pressure Vessel Code (Section VIII, Divisions 1 and 2 for
nonfiberglass and Section X for fiberglass pressurized tanks). The types
3-1
-------
of pressure vessels covered by State/local laws according to ASHE
construction codes are presented 1n Table 3-1; the geographic
distribution is shown in Figure 3-1.
Note that 27 States incorporate the ASME Code sections for the
design of fiberglass and nonfiberglass pressurized tanks into their State
laws. Another 5 States Incorporate only the ASME sections dealing with
the design of nonfiberglass tanks. On the other hand, 18 States do not
incorporate any pressure vessel design requirements Into their State
laws. Among the States without pressure vessel laws are industrialized
States such as New York, Michigan, and Texas, as well as West Virginia
where Union Carbide's MIC plant is located. It should be noted that some
local jurisdictions have also adopted pertinent sections of the ASME Code
(see Table 3-2).
3.2 Inspection Requirements
Inspection requirements vary from State to State. In all States
that do require pressure vessels to be inspected, inspection reports
submitted by insurance companies are accepted in lieu of inspections made
by a State department inspector. The frequency of inspections may be
annual, biennial, or triennial, depending upon the jurisdiction. In some
States, inspection frequency is left to the discretion of the inspector
at the time of installation. In other States, vessels are inspected only
at the time of installation and reinspection Is not required. Inspection
frequency and requirements are summarized in Table 3-3. Figure 3-2 is a
map showing the geographic distribution of jurisdictions requiring
pressurized tank inspections.
3-2
-------
0080s
Table 3-1 Types of Pressure Vessels Covered
by State Design Requirements3
Jurisdiction
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Nonfiberglass
tanks
N
A
N
A
A
A
N
A
A
N
N
A
A
A
A
A
N
A
N
A
A
A
N
A
A
N
Fiberglass
tanks
N
A
N
A
A
A
N
A
N
N
N
A
A
A
N
A
N
A
N
A
A
A
N
A
A
N
3-3
-------
0080s
Table 3-1 (continued)
Juri sdi ction
Montana
Nebraska
Nevada
New Hampshire
Mew Jersey
New Hex i co
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Nonfiberglass
tanks
N
A
A
A
A
N
N
A
A
A
A
A
A
A
N
N
A
N
A
A
A
A
N
A
N
Fiberglass
tanks
N
N
A
A
A
N
N
A
N
A
N
A
A
A
N
N
A
N
A
N
A
A
N
A
N
a A = Covered by law requiring ASHE Sections VIII(1)(2) and/or X.
N = No applicable pressure vessel design law.
Source: Telephone survey of State agencies.
3-4
-------
ALASKA
V
,11
KEY (SECTIONS OF AtMi COD! IN «TATi I AMI
ASME tECTIONS VIII III 121 AND X (METAL AND FIBERGLASS TANKil
ASMt SECTIONS VIM III 111 (METAL TANKS ONI Yl
j j N0 rHESSUHIZEO TANK LAWfS
Figure 3-1 States with Pressurized Tank Design Laws
-------
0080s
Table 3-2 States with Cities That Have Pressure Vessel Laws
1. Arizona'
• Phoenix
• Tucson
2. California2
• Los Angeles
• San Francisco
3. Colorado2
• Denver
4. Florida1
• Miami
• Tampa
• Dade County
5. Georgia1
• Atlanta
6. Illinois2
• Chicago
7. Louisiana
• Jefferson Parrish
• New Orleans
8. Michigan1
• Dearborn
• Detroit
9. Missouri1
• Kansas City
• St. Louis
• St. Louis (unincorporated area)
• St. Joseph
3-6
-------
0080s
Table 3-2 (continued)
10. Nebraska2
• Omaha
11. Oklahoma2
• Tulsa
12. Tennessee2
• Memphis
13. Virginia2
• Arlington County
• Fairfax County
14. Washington2
• Seattle
• Spokane
• Tacoma
15. Wisconsin2
* Milwaukee
^States do not have pressure vessel laws.
2States have pressure laws but the cities have separate laws.
Source: Synopsis of Boiler and Pressure Vessel Laws, Rules, and
Regulations.
3-7
-------
0080s
Table 3-3 State Inspection Frequency Requirements
for Pressure Vessels
Jurisdiction
Alabama
Alaska
Ari zona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Mi nnesota
Mississippi
Missouri
Montana
Nebraska0
Nevada
Agency3
N
B
N
B
N
B
N
B
B
N
N
B
B
B
B
B
N
B
N
N
B
N
N
B
B
N
N
B
B
Frequency'*
N
*
N
1
N
*
N
2
1
N
N
2
3
3
3*
1
N
*
N
N
2
N
N
2
2
N
N
1
3*
3-8
-------
0080s
Table 3-3 (continued)
Jurisdiction
Agency3
Frequency'5
New Hampshire
Mew Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
B
B
N
N
B
B
B
B
B
B
B
N
N
B
N
B
B
B
B
N
B
N
2*
0
N
N
2
0*
1
d
3
2
1
0
N
N
2*
N
3
2
2*
2
N
3
N
a B = State accepts insurance company inspection reports in lieu of
State inspection.
N = No inspection required.
b 1 = Annual
2 = Biennial
3 = Triennial
0 = One inspection only (at time of installation)
* = Frequency determined or modified case-by-case at discretion of
the State inspector or the Department of Labor.
c Nebraska currently inspects only vessels that are at places of
employment.
d The law requiring pressure vessel inspection in Oklahoma is new;
currently no pressure vessels are being inspected.
3-9
-------
ALASKA
i.,
i
HAWAII
KEY IINSf tCTION f REQUiNCYI
ANNUAL INSPECTION
BliNNIAL INSPECTIONS
| \ IRIENNIAl INSTtCIIONS
I V \ U FREQUENCY OEIEHMINED CASE BY CASE BASIS
INSPECIIONSONLY AT TIME OF INSIAllAIION
I I INSPECTION NOT REQUIRED
Fiqure 3-2 States with Pressurized Tank Inspection Requirements
-------
Of the 30 States requiring inspections, 7 require triennial
inspections, 12 biennial, 6 annual, 3 only at the time of installation,
and 2 on a case-by-case basis. As with design requirements, some local
jurisdictions have adopted Inspection requirements that differ from their
States' requirements.
3.3 Exemptions to Pressure Vessel Laws
The types of pressure vessels subject to the design and inspection
requirements summarized above vary from State to State. For example,
pressurized tanks owned or operated by the Federal Government are
specifically excluded from State/local regulation; however, many States
exclude other categories from regulation as well. Examples are
pressurized tanks in oil refineries, on farms, at mines, and at research
facilities. Also, most States exempt from regulation tanks with
operating pressures of less than 15 psi and tanks with inside diameters
of less than 6 inches. A summary of these exemptions is presented in
Table 3-4.
3-11
-------
0080s
Table 3-4 Categories of Exemptions to State Pressure Vessel Laws
State
Exemption category
Alaska
Illinois
Indiana
Kentucky
Maine
Maryland
Massachusetts
Minnesota
Nebraska
New Hampshire
North Carolina
Oklahoma
Tennessee
Utah
Vermont
Washington
Farms
Cities in excess of 500,000
population, mines, and farms
Oil refineries
Oil refineries, farms
Transformer or circuit breaker
systems, farms
Farms
Farms
Oil refineries, farms
All nonsteam pressure vessels
Less than 3,000 psi pressure
Farms, mines, refineries
Research facilities
Farms
Rotating mechanical device systems
Farms
Farms
3-12
-------
4. INVENTORY INFORMATION
4.1 Available Statistics
Versar contacted officials 1n all 50 States (see Section 3) to
determine whether Inventory data were kept on the number, types,- sizes,
or designs of pressurized tanks in their States. Registration was
required 1n some States; however, compiled statistics on pressurized
tanks were not available from the majority. Trade associations were also
contacted to Identify any compiled Inventory data, but none were found.
Several State officials contacted were able to provide an estimate of
the total number of pressurized vessels at Industrial facilities 1n their
States. This Information 1s summarized 1n Table 4-1. Note that the
total number of pressure vessels reported 1s 1n the, tens of thousands.
4.2 Indirect Analysis
The preceding discussion Illustrates that comprehensive inventory
Information regarding pressurized tanks 1s currently not available.
Therefore, an Indirect procedure was developed to provide preliminary
Information on the (1) types of regulated substances commonly stored in
pressurized vessels, (2) hazardous characteristics of these substances,
(3) industries that use these substances, and (4) the number and
geographic distribution of facilities in those Industries.
In order to form a basis for an independent estimate of tanks, Versar
selected 38 regulated substances that met the following criteria:
4-1
-------
0080s
Table 4-1 Telephone Survey Estimates of Pressurized
Tank Inventories by State
State Estimated no. of pressure vessels
Illinois 90,000
Mississippi 15,000 - 20,000
Nevada 8.000
Oklahoma 100,000
Oregon 30,000
Utah 20,000 - 25,000
Virginia 60,000
Washington 80,000 - 90,000
Source: Fifty-state telephone survey performed by Versar. Persons
contacted in each State are listed in Appendix C.
Note: Rhode Island and New Hampshire are in the process of
computerizing inventory data on pressurized tanks in their
respective States.
Note: Missouri officials estimate 40,000 pressure vessels and boilers
in their State.
4-2
-------
• Likely to be stored 1n pressurized tanks (liquids with boiling
point less than 60°C and vapor pressure greater than 200 psi and
gases); and
• Large production (more than 10,000 tons per year).
The 11st of 38 chemicals, together with their associated boi-ling
point, vapor pressure, and production characteristics, 1s presented in
Table 4-2. Table 4-3 depicts the number of facilities in each of the
selected SIC codes. Note that the total facilities shown amount to more
than 15,000. Therefore, assuming an average of 10 pressurized tanks per
facility, the total number of pressurized tanks would exceed 150,000.
This would exclude tanks for other SIC codes or substances. Based on
this very preliminary estimate and the limited Inventory data received
from State officials, the total number of pressurized tanks may be in the
hundreds of thousands.
Table 4-4 shows a breakdown by State of the identified SIC code
facilities that have NPDES discharge permits. The data indicate that
facilities likely to use pressurized tanks tend to be concentrated 1n the
Industrial States of the Northeast, the Midwest, the Gulf Coast, and
California.
Appendix D presents a summary of the specific Industrial uses of
these materials along with recommended storage and handling practices.
Table 4-5 provides a summary of maximum recommended inhalation
exposure concentrations for the 17 chemicals for which such data were
available. Methyl Isocyanate has the lowest recommended maximum exposure
concentration (0.02 ppm). Conversely, propane exhibits the highest
4-3
-------
0080s
Table 4-2 Selected Toxic Substances Potentially Stored in Pressurized
Tanks, and Produced at More Than 10,000 Tons/Year
Boiling3 Vapor pressure3
Substance point (°C) (nro-Hg) Production15
Acrolein
Allylamine
Allyl chloride
Amronia
Argon
Carbon disulfide
Chlorodi f luoroethane
Chloropicrin mixtures
Chloroprene
Cyclohexyl isocyanate
Dichlorosilene
Dif luoroethane
Dimethylamine
Ethane
Ethyl ene
Ethyl ene oxide
Ethylene oxide/propylene oxide
mixtures
N-Ethyl tolui dines
Hydrocarbon gases
Hydrogen
Hydrogen cyanide
Hydrogen sulfide
Liquified petroleum gas
Methane (natural gas)
Methyl amine
Methyl chloride
Methylene chloride
52.5 258
55-58
44-45 390
<25
<25
<25
<25
<25
59.4 235
-
<25
<25
<25
<25
<25
<25
13.5 - 34.3
-
<25
<25
25.6 730
<25
<25
<25
<25
<25
<25
High
Medium
High
High
High
High
Medium
Medium
High
Medium
Medium
Medium
High
High
High
High
High
Medium
High
High
High
High
High
High
High
High
High
4-4
-------
0080s
Table 4-2 (continued)
Substance
Boiling3 Vapor pressure3
point (°C) (rrm-Hg) Production15
Methyl isocyanate 39.1 -59.6
Nitrogen <25
Phosgene <25
Propane <25
Refrigerant dispersant (Freon) <25
Sulfur dioxide <25
Sulfur trioxide <25
Sulfuryl chloride <25
Tetrafluoroethylene <25
Vinyl chloride <25
201 - 348 Medium
High
High
High
High
High
High
Medium
Medium
High
3Boiling point and vapor pressure data sumnarized from: Merck Index
(1983), USEPA (1982). Verschueren (1983), Versar (1984), and Worthy
(1985).
^Production estimates based on information from SRI (1984) and U.S.
Department of Congress (1982).
High = >100,000 TPY
Medium = >10,000 TPY
Note: Chlorine is not on this list because, according to the Chlorine
Institute, chlorine is not stored underground (conmunication with
Bob Fredrickson, Technical Services, the Chlorine Institute,
February 4, 1985).
4-5
-------
0080s
Table 4.3 Number of Facilities Likely to Employ Pressurized Tanks
to Store Materials Listed in Table 4-2
SIC Code
1321
2041
2043
2812
2813
2819
2821
2822
2823
2824
2834
2841
2869
2879
2892
2899
2911
2999
3111
3255
Industry No.
Natural Gas Liquids
Flour and Grain Mill Products
Cereal Preparations
Alkalies, Chlorine
Industrial Gases
Industrial Inorganic Chemicals
Plastic Material
Synthetic Rubber
Cellulosic Man-made Fibers
Organic Fibers
Pharmaceutical Preparations
Soap/Other Detergent
Industrial Inorganic Chemicals
Agricultural Chemicals
Explosives
Chemical Preparations
Petroleum Refining
Petroleum and Coal Products
Leather, Tanning and Finishing
Clay Refractories
of facilities3
187
618
118
171
599
2183
1529
311
79
165
1512
996
1160
861
172
2563
1281
108
594
263
TOTAL 15,469
a Source: Dun's Marketing Service (1983). Standard Industrial
Classification Characteristics.
4-6
-------
Table 4.4 Geographic Distribution of SIC Codes Associated with Chemicals Possibly
Stored In Pressurized Underground Tanks
'****"*"**••*'
"*"*•"•*****
«***••••*•*
••"*•""••*•••
*••"•"*""**
Ntabtr of ficllUIn
locition ilCCodu Ittl
fttfion 1
no
V
M ft
1C ft
IU 0
rmt V
II 0
lit A
"• W
fcibtolili 0
fc|im II
NJ ft
n t
n o
feMoUli ft
l*|io« III
K ft
n »
M ft
WA
w
HV ft
faUotih ft
*t|ion IV
n. ft
« ft
R| 0
5 1
tuMollli t
h|lonV
a t
IN 0
HI t
* ft
Oil «
HI ft
fcMolih t
M4I t041
«A
V
ft 0
1 0
ft 0
ft 0
• A
V
1 ft
1 0
t e
3 I
1 ft
1 t
ft ft
e •
t t
• •
0 ft
? 1
1 1
7 1
4 4
1 1
1 4
1 1
1 1
1 1
II 14
tilt till
4 1
1 ft
14 3
'! I
II 13
: i
i u
1 4
1 t
3 tl
0 3
1 t
i !
14 tt
1 t
t 7
1 1
1 4
It II
J 1
II U
an tin
ft 4
tltl tU
t MM tlt4
to t ft ft
• i i • • »
« i
« n
1 IS
i ii
i ii
t i
t s
t SI
ft 1
t I
ft ft
ft IS
ft tl
ft It
t t
1 II
ft 7
1 t 1 ft
n 4 i t
U 3 4 4
14 ft ft ft
ft ft • ft
34 i 4 4
J
II
7
I ft 1
J 1
1 t
i » i •
45 S 4 S
* j
$ !
'! i
!* *
i 1
ft 1
ft ft
1 3
1 II
ft 1
U II 3 tl
II 1
7
It
t
17
1
t
1
0
1
|
1 ii U 10 1 4
till 1114
A A
1 ft
9 9
ft 4
4 7
t4 X
3 Ift
i tS
ft ft
a u
I I
4 11
1 1
ft 1
7 II
1
ft
ft ft
t IS
1
1
1
ft
t
1 t4
till tltl U71
?f A A
1 ft
«A |
f 1
1 1 1
S B t
to u i
1 li 1
1 S ft
t4 *4 t
$ 2 t
1 11 t
1 1 ft
ft 1 1
It M 4
Oil £
II •
i i i
It A
1 f
ft 1 t
* 'i ft
1 t 1
5 57 M
II 14 1
1 t
ft 1
1 10
t 1
tO 41 t
a« tin
A *
ft ft
*|
1
t i
ft 3
1 IS
ft ft
4 44
ft ft
1 3
ft 1
t 1
1 Ift
• A
! i
*A
V
ft 0
$ i
ft 1
1 T
1 3
1 ft
ft 1
1 ft
t 11
t (
7 K
tin rm mi
• A A
9 V
t ft X
ft ft 1
»• i
f f
\ 1 I
V v 1
4 ft SS
7 ft II
i • a
4 ft ft
14 ft X
i : ;
i» i i
t t i
i i t
11 4 U
*•
C
0 ft
1
11 II
u
17
7
1
11
1 10
M i 41
J2S5 MM
1
ft
1
I
V
t
II
t
0
IS
t
t
1 1
0
0
IS 1
4 1
0 0
1 t
09
C
ft 1
I ?
ft ft
7 S
t 1
ft 0
ft t
0 0
Ift t
0 4
It 7
Infill rt
1 4/1*1 f »ff
lolill li
totih N
iouii 11
Mill M
lolill III
Tot ill £0
lolill El
Mill 11
lolill 1
lottli 901
lolill t]
lolill 44
lolill IX
lotill 40
lotill 40
lolill HI
Tat it I 41
IQIIII «W
Mill 11
Mill 42
loiiii IT
lot ill t*
Mill »
lolill It
lolill M
loiiii m
Milt in
Mill U
Mill U
loiiii M
Mill 157
lolill 71
Totih 50*
-------
Table 4.4 (continued)
I
00
UMlirn IICCMfl
*m
M
U
Ml
OK
tl
fcbUUli
Kt|lM VII
II
US
NO
NE
StMolili
Itf ICM VIII
n
MB
a
UI
w
u
w
HI
HV
StMolili
IffiMl
M
MA
S«Molili
Stint lalili
Source: IFD
ISI
ft
I
ft
ft
ft
0
ft
ft
t
1
ft
t
0
2
19P
2041
ft
ft
2
14
1
ft
ft
ft
ft
2
1
4
47
2041
ft
ft
1
ft
ft
ft
ft
1
1
ft
ft
21
2112 211
1 ';
i i'
• r
S
i i
« i
• i
t
2 I
«•
) ]
1 (
4 II
ft 1
1 3
9 1
14 ISI
I 2111
I
1 ft
1 ft
1 ft
I ft
ft
i t
ft
> t
ft
ft
1 t
IA
1 t
ft
ft
ft
ft
ft
ft
t
2
2119
'8
24
31
II
20
2
27
10
14
112
2121
'1
H
43
|
14
ft
ft
1
ft
17
1
10
IX
2*22
1
It
1?
ft
1
ft
ft
ft
ft
1
ft
1
a
2121
ft
ft
ft
ft
ft
1
1
ft
ft
ft
0
ft
0
li
2124
I
1
1
ft
ft
ft
I
t
ft
0
17
Nutter
2111
1
ft
1
i
1
ft
ft
ft
1
ft
ft
41
•f ficl
2114
ft
1
2
1
ft
t
!
i
i
i
142
lillM
2141
ft
4
C
1
10
ft
:
g
i
i
ft
0
2
2
11
2141
1
31
17
1
1
ft
1
ft
7
4
J3I
an
«
j
7
t
3
i
ft
4
ft
4
0
1
1
44
a*
I
i
A
ft
ft
ft
J
i
0
0
o
0
1
a
2tn
t
4
11
2
1
ft
I
ft
0
Q
4
1
1
111
2111
22
U
M
1
II
12
1
2
11
2
10
11
2»
2m
•
4
31
ft
0
ft
I
ft
0
1
0
i
0
0
(7
llll
ft
i
1
1
1
0
I
•i
0
12
0
i
4
7
202
1239
ft
ft
ft
ft
7
ft
ft
0
2
ft
4
0
0
U
im
o
ft
i
0
0
0
0
0
I
0
0
o
1
1
o
0
0
0
»
Mih
Mih
Mih
Mill
Mih
Mih
Mill
Mill
Milt
Milt
Mih
Mill
Mill
Mill
Mill
Mill
lOllll
Mih
Mill
Mih
Mill
Mill
Mih
Mill
Mill
Mill
loUh
Mill
1
201
411
14
10
121
1
24
1
1)
141
47
73
23%
-------
0080s
Table 4.5 Summary of Health Effects Data for Materials Conmonly
Stored in Pressurized Tanks
Chemical
Methyl isocyanate
Acrolein
Phosgene
Allyl chloride
Chloroprene
Vinyl chloride
Allyl ami ne
Sulphur dioxide
Carbon disulfide
Dimethyl ami ne
Hydrogen cyanide
Hydrogen sulfide
Armenia
Ethyl ene oxide
Methyl chloride
Methylene chloride
Propane
Recommended3 maximum Potential carcinogen
exposure concentration (ppm) or mutagen
0.02
0.1 *
0.1
1
1
1
5
5
10 *
10
10 *
15 *
35
50 *
100 *
500 *
1,000
a. Lowest value of OSHA, NIOSH, or ACGIH recommended limit.
Note: See Appendix E for detailed health effects data.
4-9
-------
recommended exposure concentration (1,000 ppm). Also, the potential
carcinogens or mutagens tend to be at the lower end of the scale with
respect to recommended exposure limits. A detailed summary of health
effects data 1s found 1n Appendix E. EPA may wish to use this a'nd
similar health Information to prioritize chemicals for regulatory
purposes.
4-10
-------
SECTION 4 REFERENCES
CFR. 1985. Code of Federal Regulations. Department of Transportation
Hazardous Materials Tables (49 CFR 172.101-102). Published by the
Bureau of National Affairs, Inc. Washington, D.C. 20037. Revised as
of January 7, 1985. 221:0101-0167, 0351-0395.
IFD. 1985. Industrial Facilities Discharge File. Computerized
Database. Washington, D.C.: U.S. Environmental Protection Agency,
Office of Water Regulations and Standards, Water Quality Assessment
Branch. Retrieval March 4-5, 1985.
Merck. 1976. The Merck Index, An encyclopedia of chemicals and
drugs. Martha Wlndholz, ed. Ninth edition. Rahway, N.J.: Merck &
Co., Inc.
SRI. 1984. Directory of chemical producers - 1978-1984. Menlo Park,
Ca.: SRI International.
U.S. Department of Commerce. 1977-1982. Bureau of the Census. Census
of Manufacturers. Current Industry Reports/Industry Series: Organic
Chemicals; Inorganic Chemicals; Agricultural Chemicals. Washington,
D.C.: U.S. Government Printing Office.
USEPA. 1982. Aquatic fate process data for organic priority pollutants.
Final report. Office of Water Regulations and Standards. Washington,
D.C. 20460. EPA Report No. 440/4-81-014.
Versar. 1984. Physical - chemical properties and categorization of
RCRA wastes according to volatility. Final draft report. EPA
Contract No. 68-03-3041. U.S. Environmental Protection Agency,
Emission Standards and Engineering Division, Office of Air Quality
Planning and Standards. Research Triangle Park, N.C.
Vershueren, K. 1983. Handbook of environmental data on organic
chemicals. New York: Van Nostrand Relnhold Co.
Weast, Robert, ed. 1975. Handbook of chemistry and physics.
Cleveland, Ohio: CRC Press.
Worthy, W. 1985. Methyl Isocyanate, the chemistry of a hazard, chem.
Eng. tfevs. February 11, 1985. pp. 27-33.
4-11
-------
5. MAJOR RELEASE INCIDENTS FROM PRESSURIZED TANKS
Versar performed a computerized literature search to Identify major
release Incidents from pressurized tanks. The eight data bases'selected
for the search are briefly described below.
UPI NEWS (Files 260, 261) contains full text of news stories carried
on the United Press International Wire.
AP NEWS (Files 258-259) provides full text of national, International,
and business news from the AP data stream service, available 48
hours after data were Initially transmitted.
FACTS ON FILE (File 264) 1s a weekly record of contemporary history
compiled from worldwide news sources.
NATIONAL NEWSPAPER INDEX (File 111) covers news items printed
since 1979, updated monthly.
NEWSEARCH (File 211) 1s a dally Index of more than 2,300 news items
from over 1,700 national newspapers, magazines, and periodicals.
OCCUPATIONAL SAFETY AND HEALTH (NIOSH) (File 161) includes citations
for more than 400 journals as well as over 70,000 monographs and
technical reports. All NIOSH documents are Indexed in this file.
CHEMICAL INDUSTRY NOTES (File 19) extracts articles from over 78
worldwide business-oriented periodicals that cover the chemical
processing Industries.
NTIS (File 6) data base consists of government-sponsored research,
development, and engineering reports publicly available (1964 to
present).
Each data base was searched with a program requesting reports on
release Incidents or Industrial accidents that involved pressurized
storage tanks. The search programs limited the responses to stationary
Incidents. The keywords that were common to all the search programs used
are as follows: chemical/industry/plant; accident/explosion/spill/release/
leak; storage (process)Xtank (vessel); pipes/pipelines; pressurized.
5-1
-------
Reports that were Identified in the DIALOG data bases were furnished
with dates, authors, news carrier (1f appropriate), and other
bibliographic Information. Full copies of selected news articles were
then obtained for review.
There are several limitations to the search, Including:
• The news files searched cover the period of the mid 1970s to the
present;
• Many of the technical/engineering details that would be of
Interest are often lacking 1n the general news reports; and
• Since only major newspapers and wire services were searched, only
major disasters would be reported.
Despite these limitations, the search did reveal 16 significant
accidents as summarized in Table 5.1. Also shown are 11 LNG/LP6
accidents from a referenced document. Unfortunately, the level of detail
available does not permit an extensive analysis of the circumstances
surrounding each incident.
Versar also obtained an unpublished list prepared by Professor
E. Hampe, Weimar University, Democratic Republic of Germany (see
Table 5.2). Dr. Hampe heads the International Association of Shell
Structures. This 11st identifies 31 accidents from 1964 to 1979,
worldwide; however, the details provided usually do not allow analysis of
the circumstances surrounding the accident. Also, some of these
accidents may have Involved tank trucks and rail cars, which were
excluded from our computerized search. Nevertheless, this limited
Information indicates that major health ( i.e., over 700 deaths) and
environmental Impacts can and do result from pressurized tank accidents
1n addition to the Bhopal event.
5-2
-------
Table B-»l summary of Readily Available pub) t shed In format ton
on Major Release Incidents from Pressurized Storage Vessels
Reported
incident
date
Location
Vessel
type
Incident
details
Discharge
material
(quantity and/or date)
Consequences
References
1977 American Cyanamid Co. —
Herbicide Plant
Hannibal, Ho.
June 2, Plastifax, Inc.
1982 Gulfport, Miss.
Explosion
Plant explosion; toxic fumes (Plant produces
cloud, spilled mixture of chlorinated paraffin
hazardous chemicals and wax)
$67.2 million
insurance claim
3 dead, 3 critically injured,
61 injured
News clipping (#2)
News clipping (#4)
en
CO
Oct. 22, Aerosol Research
1981 Laboratory
Holbrook, Mass.
Nov. 15, FMC Corp.
1984 Middleport, N.Y.
1976 Givavdan
(Hoffman-LaRoche)
Seveso, Italy
1977 Bayer Corp.
Beziers, France
1949 Monsanto Co.
Nitro, W. Va.
Jan. 26, Ultrafertil Co.
1985 Vila Parisi, Brazil
(NR Cubatao)
Blast flattens and burns
building; cause not
reported
Malfunctioning pump;
9:45 a.m. chemical spill
Plant explosion; cleanup
not completed until
July 1984
Gas escaped
Industrial accident
Methyl isocyanate;
30 gallons
Dioxin (produced
incidentally in
chemical manufact.)
Methyl isocyanate
Dioxin
Plant destroyed; at least 25
persons injured
Eye irritation in 30
students/1 teacher at
school 500 yards from plant
Noted as most severe chemical
disaster in Europe; killed
hundreds of animals; caused
skin disease in nearly 200
people
Several workers poisoned
121 workers exposed and
developed typical chloracne;
subsequent mortality analysis
did not disclose any deaths
News clipping (#5)
News clipping (#6)
News wire (#16)
News clipping (#7)
News wire
News clipping (#7)
(secondary)
News clipping (#8)
Gas escaped from fertilizer Anmonia gas
plant
Town evacuation; 30 injuries News clipping (#10)
-------
Table 5-1 (continued)
Reported
incident
date
Location
Vessel
type
Incident
details
Discharge
material
(quantity and/or date)
Consequences
References
April 8,
1983
STP
Knoxville, Tenn.
Chlorine gas leak for
1 hour
Eye irritation, 4 people,
no serious injuries
News wire (#12)
Chemical (8:15 a.m.) Workers pumped
storage chemicals into wrong tank,
tank reacted with contents to
produce chlorine; subsequent
leak from tank for 1 hour
Oct. 6, American Cyanamid Co. Outdoor Tank overheated (10:00 a.m.), Malathion (pesticide) Sickened more than 100 people News wire (#13)
1984 Linden, N.J. storage steel cover blew off; fumes 97% solution in 2-mile radius; 9 people (#18)
tank discharged (65,000 gal) hospitalized (#19)
(12,000
gal)
July 23,
1984
Union Oil Co.
Romeoville, 111.
en
Jan. 11, Nobel, Co.
1985 Karlskoga, Sweden
Tank
Leaking gas caused by
mechanical failure; two
explosions; one in an
alkylation unit
Leak from burst pipe (from Gaseous sulphuric acid
frost); entire pipe contents (Oleum) 1 full tank
leaked
17 workers killed, 23 injured;
Union's largest refinery shut
down
News wire (#14)
(#20)
Nov. 19, Pemex Large • Facility: Liquid gas plant, Liquefied petrol, gas;
1984 (Petroleos Mexicanos) outdoor butane + propane dist./ 80,000 brls
(State-owned) LNG refinery (50 yrs old)
San Juan storage • Incident: Flame from waste
Ixhuatepec, Mexico tanks gas ignited gas leaking from
4 storage tanks
General evacuation; 18 people News wire (#15)
experienced minor irritations
Reports vary, 31,000 people News clipping (#21)
displaced, total 4,248 injured, News wire (#'s 22,
324-490 killed, 900 missing, 23, 24, 25, 26)
300 homeless families
Cleveland, Ohio
Storage Tank failure
tank 1.1 million gallons
Storage Shut-off valve failed;
tank spill ignited
Storage Tank overpressurization
tank
LNG
LNG; moderate size
pool
Ethylene; 150,000 gal
@ 6 gpm
Major disaster
Minor damage
None reported
Report (#27)
Report (#27)
Report (#27)
-------
Table 5-1 (continued)
Reported
incident Location
date
__
__
—
....
—
Vessel
type
Storage
tank
Storage
tank
Storage
tank
Storage
tank
Storage
tank
Incident
details
Overpressuri zation
Plate weldtnent failure
Outer tank bottom plate
weldment failures
Cryogenic shock caused by
sudden discharge valve
rupture
ING flashed + vented. Due
to introduction of weathered
Discharge
material
(quantity and/or date)
Propane; 3 million gal
Propane; moderate
quantity
LNG; 450,00 brl
capaci ty
LNG
LNG; N 400,000 Ibs
Consequences
Internal tank fire destroyed
tank and contents
Tank drained and repaired
Tank drained and repaired
(4 months)
1 employee killed; large LNG
spill on water; vapor plume
No structural damage; vapor
plume traveled far downwind
References
Report (#27)
Report (#27)
Report (#27)
Report (#27)
on
en
LNG into tank partially
filled with less dense LNG
Storage Gas leak contained to diked
tank area; outer tank floor
failure due to high
mechanical stress
LNG
Tank emptied and removed
from service 3 yrs
Report (#27)
Storage
tank
Storage
tank
Tank failure; instantaneous LPG; 70,000
spill of gas; ignition
Vapor lock in LNG line
caused spill onto carbon
steel roof, 2.5 m; brittle
fracture resulted
LNG; very small
quantity
Tank and plant destroyed;
all employees killed
(over 100 people)
Tank repaired
Report (#27)
Report (#27)
-------
Table 5.1 References
1. wall street Journal. 1980. Business Briefs. Monday, September
29, 1980. p. 36.
2. wall street Journal. 1981. American Cyanamld Sues American Home
Assurance Over Claim. Friday, August 28, 1981.
3. New York Times. 1982. Leaking Chemicals 1n California's Silicon
Valley Alarm Neighbors. Thursday, May 20, 1982. p. A22.
4. New York Times. 1982. Around the Nation. Third Victim Is Found
1n Explosion at Plant. Friday, June 4, 1982.
5. New York Times. 1981. Around the Nation. 25 Injured in
Explosion at Aerosol Can Factory. Friday, October 23, 1981.
6. New York Times. 1984. Special Precautions Used in Handling Toxic
Gas. Thursday, December 6, 1984. p. A12.
7. New York Times. 1984. Gas Leak Touches a Nerve in Europe.
Friday, December 7, 1984. p. All.
8. Wall Street Journal. 1979. No Excess in Deaths Is Found in a
Study of Dloxin Exposure. Tuesday, October 23, 1979. p. 48.
9. Washington Post. 1984. Industrial Accidents Rock Brazil.
Sunday, April 1, 1984. pp. F3 - F4.
10. New York Times. 1985. 5,000 Flee Gas Cloud from Brazilian
Plant. January 27, 1985. p. 5.
11. Wolf JH. 1983. UPI. Dioxin. Jefferson City, MO. August 25, 1983.
12. UPI. 1983. Chlorine. Knoxville, TN. April 8, 1983.
13. UPI. 1984. Leak. Linden, NJ. October 6, 1984.
14. UPI. 1984. Refinery. Romeovllle, IL. September 27, 1984.
15. AP. 1985. Sweden - Gas Leak. Karlskoga, Sweden. January 11, 1985.
16. AP. 1984. Chemical Leak. Mlddleport, NY. December 5, 1984.
5-6
-------
Table 5.1 References (continued)
17. AP. 1984. Worst Accidents. December 4, 1984.
18. AP. 1984. Pesticide Leak. Linden, NJ. October 7, 1984.
19. AP. 1984. Pesticide Leak. Linden, NJ. October 6, 1984.
20. Pollack D. 1984. AP. Refinery Explosion. Romeoville, IL. July
24, 1984.
21. New York Times. 1984. In Devastated Mexican Area, the Anger
Persists. Thursday, December 6, 1984. p. A12.
22. UPI. 1984. Mexico. Mexico City. December 27, 1984.
23. Romand Z. 1984. UPI. Inferno. Mexico City. November 30, 1984.
24. Bussey J. 1984. UPI. Environment. Mexico City. November 21,
1984.
25. Galan M. 1984. UPI. Explode. Mexico,City. November 21, 1984.
26. AP. 1984. Worst Accidents. December 4, 1984.
27. Wesson and Associates, Norman, OK. Summary of Accidents/Incidents
at Operating Refrigerated Liquified Gas Storage Facilities.
28. Melgard J. 1985. (Feb. 15) South Dakota Department of Public
Safety; Deputy Secretary. Transcribed Phone Conversation with Pat
Wood, Versar, Inc.
NOTE: The source of most of the above entries is DIALOG Information
Retrieval Services, Palo Alto, California (on-line retrieval).
5-7
-------
0188s
Table 5-2 Additional Industrial Gas Accidents
That Hay Involve Pressurized Tanks
1963 Texas
Fire and explosion in low-pressure polypropylene polymerization
unit.
Loss approximately $6,000,000.
1964 Texas
Fire and explosion resulting from an escape of ethylene being
ignited by electrical switchgear.
Loss approximately $3,000,000.
1964 Texas
Fire ruptured a high-pressure ethylene pipe and gave rise to a
vapor cloud explosion.
2 died. Loss approximately $4,000,000.
1964 Massachusetts
Leaking sightglass on reactor was being tightened while
pressurized. Escaping vinyl chloride was ignited and exploded.
7 died. 40 injured. Loss approximately $5,000,000.
1965 Louisiana
Gas escaped from 8-inch diameter pipe in ethylene plant and was
followed by explosion and fire.
12 hurt. Loss approximately $3,000,000.
1966 West Germany
3 killed, 83 injured in LNG explosion.
1966 LaSalle, Canada
Sightglass failure believed to have created styrene/air mixture
which exploded in a polystyrene plant building.
11 killed. Loss approximately $4,000,000.
1967 Buenos Aires, Argentina
100 killed and 400 houses destroyed in LNG explosion.
1967 Lake Charles, Louisiana
Leaking pit valve in alkylation unit liberated a cloud of
isobutane which exploded. Fires and secondary explosions
continued for 2 weeks.
7 killed. Loss approximately $20,000,000.
5-8
-------
0188s
Table 5-2 (continued)
1968 Alaska
2 people injured in LPG accident.
1968 Tarrytown, New York
3,500 people evacuated in LPG accident.
1968 South Africa
A horizontal (cigar-type) pressurized tank burst open in an
ammonia plant. An amnonia cloud covered a major part of the
plant.
26 people died.
1969 Libya
12 injured in LNG accident.
1970 Mitcham, United Kingdom
Houses and 2 cars destroyed in LPG explosion.
1970 Port Hudson, New York
Houses destroyed in LPG explosion.
1972 Brazil
38 killed, 75 injured in LPG explosion.
1973 Saint-Agimand Les Eaux, France
4 killed. 2 missing. 37 injured in LPG explosion.
1973 New York
40 killed in LNG explosion.
1973 Lodi, New Jersey
Released vapors from a chemical reactor descended from the
emergency relief line, above roof level and exploded on reaching
a boiler house more than 100 feet away.
7 killed. Loss approximately $2,200,000.
1973 Cologne, West Germany
Flange rupture in P.V.C. plant released 10 tons of vinyl
chloride in 30 seconds. Fire and explosion followed.
Loss approximately 50 million marks.
1973 Tokuyama, Japan
Inadvertent interruption of instrument air was followed by
exothermic reaction and leakage which resulted in fire and
explosion.
Loss approximately $25,500,000.
5-9
-------
0188s
Table 5-2 (continued)
1974 Flixborough, England
Pipe failure between cyclohexane oxidation reactors liberated a
large vapor cloud which ignited and caused widespread
destruction.
28 killed. 89 injured. Loss approximately 19,000,000 Pounds
Sterling.
1974 Wenatchee, Washington
A railcar, containing a 15% slurry of monomethylamine nitrate in
water, exploded and destroyed 30 buildings.
2 killed. 66 injured. Loss approximately $5,000,000.
1974 Los Angeles, California
A leaking semi-trailer, containing organic peroxides, exploded
and caused widespread damage.
Loss approximately $5,000,000.
1974 Beaumont, Texas
A large spill of isoprene was believed to have formed a vapor
cloud which exploded and was followed by fire in a synthetic
rubber plant.
2 killed. 10 injured. Loss approximately $16,000,000.
1975 Antwerp, Belgium
Ethylene leakage from compressors in a polyethylene plant was
followed by an explosion which destroyed a large part of the
plant.
6 killed. Loss approximately $50,000,000.
1975 South Africa
LNG explosion.
7 killed. 7 injured.
1977 Cassino, Italy
LPG explosion.
1 killed. 9 injured.
1977 Jacksonville, Florida
LPG accident.
2,000 people evacuated.
1978 Waverly
LPG accident.
12 killed. 50 injured.
Source: Professor E. Hanpe, Weimer University, Democratic Republic of
Germany.
5-10
-------
6. RELEASE OF METHYL ISOCYANATE AT INSTITUTE, WEST VIRGINIA, AND
AT BHOPAL, INDIA
The major release of methyl Isocyanate (MIC) from the Union Carbide
Plant 1n Bhopal, India, has triggered a wave of public Interest.in the
possibility of a similar disaster's occurring in the United States.
Limited details are available concerning the release incident in Bhopal;
however, Union Carbide owns and operates a sister plant 1n Institute,
West Virginia (UCC-Institute), for which considerable Information
concerning process design, controls, and release incidents is available.
In this section we will (1) compile information obtained through
government contacts and through Union Carbide concerning the Institute
plant; (2) compare this information to what is known about the Bhopal
plant and the major disaster there; and (3) evaluate what effect any
differences between the two plants may have on the possibility of a major
disaster's occurring at Institute.
6.1 MIC Production at UCC-Institute Facility
The UCC facility at Institute (UCC-Institute), West Virginia, has a
total area of 1,435 acres, of which 350 acres are developed (OSHA 1985).
MIC (which is a pesticide, Insecticide, and herbicide intermediate) is
one of 150 products produced at this plant and is used on the site in the
production of Sevin and Larvln brand insecticides. MIC is distributed to
the following four facilities in the U.S.:
• UCC - Woodbine, Georgia;
• Morton Chemical Company - Weeks Island, Louisiana;
• FMC - Middleport, New York; and
• DuPont - La Porte, Texas.
6-1
-------
All four facilities use MIC as an Intermediate In the production of
either pesticides or herbicides. These four and UCC-Instltute are the
only facilities 1n the U.S. presently using MIC 1n a chemical production
process (OSHA 1985).
Following the Bhopal incident, UCC stopped production at the
Institute plant, thereby cutting off the MIC supply to all of the above
plants. One result of this action 1s that Dupont 1s currently adding an
MIC production line to its LaPorte, Texas, plant. MIC will be produced
and used as needed, thereby eliminating the need for storage.
6.1.1 MIC Production Process
Methyl Isocyanate is produced by a gas phase reaction involving
methylamlne and phosgene. Both of these chemicals are fed simultaneously
into a process reactor where, with the addition of heat, they combine to
form MIC and hydrogen chloride (HC1). The chemical equation for the
reaction 1s:
M u
--
H' A Cl
+ ^ C=0 -r» H-C.-N=C=0 + 2 HC1
Cl' * H
Methylamlne Phosgene Methyl Isocyanate Hydrogen Chloride
The MIC and HC1, along with any unreacted methylamlne and phosgene, are
removed from the reactor and chilled. The methylamine, phosgene, and
hydrogen chloride are separated from the MIC product stream via
distillation and absorption techniques, and the resultant pure liquid MIC
is then transferred to the unit storage process.
6-2
-------
The production process is operated automatically, with manual
controls as backup (USEPA 1984). A block diagram of the production
process 1s given 1n Figure 6-1.
Methyl Isocyanate Is flammable and reactive with a number of
compounds, most notably water. MIC and water react to form methylamlne
and carbon dioxide. The methylamlne reacts further with MIC or other
reaction products to yield either 1,3-d1methylurea or 1,3,5-tr1methyl-
bluret. These reactions tend to be vigorous and exothermic, and a trace
of add or base will promote the reaction. If sufficient quantities of
MIC and water are present, and 1f the heat 1s not removed, the MIC will
begin to boll violently. MIC also reacts with Itself to form trlmethyl
Isocyanurate. This reaction 1s promoted 1n the presence of a catalyst
such as strong bases (e.g., NaOH), metallic chlorides (e.g., FeCl_),
and others. These reactions are also exothermic and can occur with just
a trace of catalyst.
All process vents on the production unit are connected to a vent gas
scrubber (VGS). This unit 1s designed to process releases of MIC, HC1,
methylamlne, or phosgene from the production process. Furthermore, the
production unit 1s designed to shut down 1n the event that any of the
aforementioned compounds 1s released from the process reactor (USEPA
1984). Also, the vent gas scrubber has been designed to shut down the
MIC production unit 1f any process gas releases 1n excess of scrubber
capacity occur (USEPA 1984).
Any emissions from the scrubber are vented to a flare tower, where
gases not neutralized by the scrubber should be destroyed (C&EN 1985).
6-3
-------
MIC PROCESS
1 '
PHOSGENE
MONOMETHYLAf
PHOSGENE
STRIPPING
t
^ REACTION
^ SYSTEM
CHLOROFORM
•TAILSs^^^ F
\ '
' \
HYDROGEN
CHLORIDE
i '
'YROLYSIS L£S££!L^.
| MIC
1
t
MIC
DESTRUCTION
t
UNIT
VENTS
MIC
REFINING
PRODUCT
MIC
STORAGE
PRODUCT
MIC
DERIVATIVES
Figure 6-1 MIC Production Schematic
SOURCE: UNION CARBIDE CORPORATION
-------
6.1.2 MIC Preliminary Storage (Unit)
The purified MIC from the production is fed into one of three 14,000-
gallon aboveground unit storage tanks. These tanks operate at up to
80 percent capacity. Should the level in any one of these tanks exceed 80
percent, the MIC production process is shut down (USEPA 1984). At this
stage of the production process, MIC is sampled for product quality and
possible contamination (OSHA 1985). The MIC unit tanks are kept"under a
slight pressure with dry nitrogen gas. The nitrogen system serves to
blanket the tanks and to keep transport lines from freezing. Also,
nitrogen is used to move the MIC from unit tanks to the main storage
tanks. Each of the unit storage tanks is equipped with temperature,
pressure, and volume monitors and alarms. The tanks are also fitted with
pressure relief devices which, in the event of an overpressure, vent to the
storage area scrubber (storage scrubber). This scrubber, larger than the
process scrubber, uses caustic soda (NaOH) to neutralize any vapors emitted
by the tanks (USEPA 1984). Any fugitive emissions from the scrubber are
vented to a flare tower for ultimate destruction (USEPA 1984). The flare
is equipped with two pilot lights, an Igniter, and boost gas to ensure that
any MIC is incinerated.
MIC in the unit storage tanks is cooled by an external heat exchanger.
The cooling medium is brine. All tanks and dedicated piping are
constructed of stainless steel. The tanks are protected with an external
ballistic shield and equipped with a water deluge system.
6-5
-------
Contaminated MIC can be transferred to a 30,000-gallon capacity or
emergency tank. This tank can also be used to provide Intermediate storage
for a tankful of MIC that may be undergoing an "uncontrolled reaction,"
using chloroform to slow down the reaction (OSHA 1985). The MIG can then
be diverted to the scrubber for neutralization. The unit tank farm is
surrounded by a dike with a 42,000-gallon capacity and equipped with a sump
and sump pump (OSHA 1985). Figure 6-2 1s a block diagram of the unit
storage facility.
6.1.3 MIC Secondary Storage (Underground)
MIC that meets specifications 1s transported to one of three 30,000-
gallon capacity underground storage tanks. As previously stated, rather
than using pumps, this type of operation uses pressurized nitrogen to
move the MIC from one tank to another. UCC believes that this type of
transfer operation eliminates leaks from pumps and their seals. These
three tanks all have a double-wall construction for environmental
protection. Also, the double wall enables operators to detect leaks in
the storage tank. A pressure gauge that measures the pressure of the
space between the walls of the tanks should show an increase in pressure
1n the event of a leak 1n the primary wall. As with the unit tank, the
storage tanks are kept no more than 80 percent full and use dry nitrogen
as a tank blanket (OSHA 1985). MIC is cooled by a redrculation cooler
using chloroform as a cooling medium. The tanks are equipped with
temperature, pressure, and volume indicators and alarms. Also, pressure
6-5
-------
CTi
_ rnriAnr \
•^ V CAUSTIC
' V
nnv NiTiinrrfiia- -to*- * *
i <
MIC FROM \^
PRODUCTION f Uf
13 T*
^
\
i
PRESSURE W
CONTROL DEVICE ^
t A
1 ' /
— -J ^ »^ ^ / VENT FROM
>V , »* - P- - -v STORAGE
f A
JIT
IAGE
NKS) \ / SCHUBBtH
PRESSURE
J CONTROL DEVICE
MIC TO \
STORAGE ./
^ / MIC FROM
wuwr innn *• ^ ~~ N^ STORAGE
r
— — •• •"•••
MICTO
PRODUCTION I J
Figure 6-2 Unit Storage Schematic
-------
control devices, which allow the tanks to be vented to the storage
scrubber/flare tower, ensure that tank overpressures do not occur.
Should MIC become contaminated 1n storage, 1t could be transferred to the
30,000-gallon dump tank or the storage scrubber. A block diagram showing
the MIC storage process 1s given 1n Figure 6-3. MIC 1n storage Is
transferred to Sevln or Larvln production, or to distribution via
centrifugal pump (C&EN 1985).
6.2 Description of MIC Release Incidents at UCC-Institute
On December 14, 1984, during a hearing by the U.S. House of
Representatives Subcommittee on Health and the Environment, Union Carbide
Corporation (UCC) stated that 28 releases of methyl Isocyanate (MIC) had
occurred at UCC-Inst1tute from January 1980 to the hearing date. Data
regarding these Incidents are contained In the U.S. Environmental
Protection Agency Inspection report of UCC-Inst1tute. Since then, after
a more thorough Investigation, UCC has stated that 61 Incidents regarding
MIC occurred between January 1980 and December 1984. Data regarding the
28 Incidents originally alluded to by UCC are analyzed on the following
pages. Data on the remaining 33 Incidents were not available at the time
of report preparation.
Each of the 28 Incidents has been reviewed to yield the following
data:
Date of occurrence;
Work shift of occurrence (A, B, C, or D);
Location of spill (production or storage);
Cause of spill; and
Quantity of MIC spilled.
Data summaries of each of the Incidents follow:
6-8
-------
DETAIL OF STORAGE TANK
O>
I
IO
UNI l muuuci
f HOMUMII
1IOMAOI
IIUN*AIIMI MIIHMI
WuMIICMf iMOHIIOal
ODOUMDIIVII \L^/ Xj"/
^'fry^ J I
S I I ^^^
Ml All IHMAOI
umi
nnuum
MIIK
OIVICI
f^
WC
UNII
IAMII
111
V*~X
IMKGIMir
VlN<
A
1
&
ICHUHIIIN
,(
*\
MMMAl VI HI
mi HUM
Ml III
Dl VlCI
^ s
IMiHGINCV
VtHl
^
X
Ml IHOMVI
UNlt TAN*
UIS1HIHIIH)M
lANNtm
Figure 6-3 MIC Storage Schematic
SOUIICt: ClllMICAl AND tNlilNH HIN(i NLWS. FtHMDAMY 11. IUU'.i "METIIVL
ISOCVANAIE HIE CMIMISIDY OF HAZAIIO"
-------
Date
2/1/80
Shift
D
2/5/80
2/7/80
3/14/80
4/24/80
Comments
A spill occurred on a pump seal of an MIC
transfer pump. The pump was used to transfer
MIC from storage to drums. MIC was released
Into the process sewer and atmosphere.
Improper maintenance performed on the pump
seal was deemed the cause of accident. The
report stated that "problems are frequent with
pumps."
Quantity of MIC lost: Not specified.
A spill occurred from an Improperly blanked
pump header at the storage area. MIC was
released Into the process sewer and
atmosphere. Improper maintenance on the blank
assembly procedure was deemed responsible for
MIC spill.
Quantity of MIC lost: Not specified.
A leak 1n chemical process supply line to #2
MIC production unit. MIC was released Into
the process sewer and atmosphere. An
equipment deficiency was deemed responsible
for accident.
Quantity of MIC lost: Not specified.
A leak in an MIC line from the MIC production
facility to unit storage resulted in MIC being
released to the atmosphere. An Improper weld
was deemed responsible for the leak.
Quantity of MIC lost: 20 gallons.
A leak in an MIC line from the MIC production
facility resulted in an MIC emission. This
apparently was the same leak that occurred on
3/14/80 and was still not repaired, or was
Improperly repaired again. The fact that the
report stated that the leak clamp was still
covering the faulty weld leads one to believe
that the hole was never repaired. Improper
maintenance was deemed responsible for product
loss.
Quantity of MIC lost: Not specified.
6-10
-------
Date
7/4/80
Shift
B
9/9/80
4/2/81
4/14/81
5/25/81
Comments
A safety valve on an MIC transfer pump to
Sevln production was activated. MIC was
released to the process sewer and atmosphere.
Responsible party was not specified.
Quantity of MIC lost: 2-3 gallons.
Shift maintenance, 1n removing a gauge on the
MIC recycle line from the storage area,
aggravated an already leaking pipe nipple so
that It cracked, releasing MIC to the
atmosphere. Defective materials were deemed
the cause of accident.
Quantity of MIC lost: Not specified.
The vent system 1n the storage area became
filled with liquid MIC. The knock-out pot at
flare tower also became filled with MIC; Us
safety valve had liquid MIC flowing out of the
discharge line connected to 1t. MIC was
released to process sewer, flare, and
atmosphere. Operator execution deficiency and
Inadequate procedures were responsible for the
accident.
Quantity of MIC lost: 5,000 pounds.
A pump seal blew out on a pump located 1n MIC
production area, resulting 1n a loss of MIC.
This release was to the process sewer and
atmosphere. Defective equipment was
responsible for the accident.
Quantity of MIC lost: Not specified.
A flexible transfer line on a circulating pump
1n production area ruptured, resulting 1n a
spillage of MIC. MIC was released Into
process sewer and atmosphere. Defective
equipment was responsible for the Incident.
Quantity of MIC lost: Not specified.
6-11
-------
Date
6/2/81
6/4/81
8/27/81
3/3/82
10/3/82
Comments
A pipe nipple leaked on an HIC transfer line
1n the production area. (Apparently this line
had been out of service for a period of time.
Comments Indicated that line should have been
checked for leaks prior to being put bacjc Into
service.) As a result, HIC was released Into
the atmosphere. Defective equipment was
responsible for the spill.
Quantity of MIC lost: Not specified.
A valve Inadvertently left open resulted 1n
liquid HIC flooding the knock-out pot and
flare tower 1n the storage area. HIC
extinguished the flare and eventually spilled
out of the top of the tower Into the
atmosphere. Lack of knowledge and procedure
errors were responsible for the spill.
Quantity of MIC spilled: Not specified.
The knock-out pot at the flare tower 1n the
storage area overflowed Into the vent and
released MIC to the atmosphere (result of
8/26/91 Incident).
Quantity of MIC spilled: Not specified.
An equipment failure caused an HIC leak 1n the
production area. MIC was released Into the
cooling tower sewer and atmosphere. Defective
equipment was responsible for spill.
Quantity of MIC lost: Not specified.
A pump plug corroded on a pump that was
providing transfer service 1n the production
area (Improper plug for MIC service). The
failed plug eventually caused an HIC leak Into
the process sewer and atmosphere. Defective
equipment and Improper maintenance were
responsible for the loss of HIC.
Quantity of HIC lost: Not specified.
6-12
-------
Date
7/5/83
Shift
B
9/22/83
11/1/83
12/13/83
12/28/83
Comments
A faulty rupture disc, which was not properly
checked prior to being replaced, caused an MIC
leak into the atmosphere. Lack of knowledge
and faulty equipment were responsible for the
failure.
Quantity of MIC lost: Less than 1 ounce.
An MIC transfer line in storage area was not
properly evacuated for maintenance. When
maintenance broke into the line, liquid MIC
escaped into the atmosphere. Improper
evacuation procedures were responsible for the
accident.
Quantity of MIC lost: Not specified.
An operator was unplugging a valve on an MIC
line in the storage area when MIC leaked from
the valve. Defective equipment was
responsible for loss.
Quantity of MIC lost: Not specified.
Maintenance working on the MIC transfer lines
in storage area caused MIC trapped in line to
escape to the atmosphere. Improper operator
execution caused the release incident.
Quantity of MIC lost: Nil.
MIC trapped between a valve and a slip blank
was released to the atmosphere during a line
purging operation in the storage area. MIC
was subsequently released into the
atmosphere. Operator error was responsible
for accident.
Quantity of MIC lost:
(UCC 1983).
A few pounds at most
6-13
-------
Date
12/31/83
Shift
2/8/84
2/9/84
3/28/84
8/1/84
9/28/84
Comments
An accumulation of ice on an overhead line
dropped on a line carrying MIC in the
production area. The MIC line snapped,
resulting in a spillage of MIC to the
atmosphere and process sewer. No blame placed
for the incident.
Quantity of MIC lost: 14,000 pounds. (This
was corrected by UCC to 840 pounds and even
further corrected to 5 pounds.)
An MIC transfer line in the storage area was
not properly evacuated for maintenance. When
maintenance broke into the line, liquid MIC
escaped into the atmosphere. Improper
evacuation procedures were responsible for the
accident.
Quantity of MIC lost: Not specified.
A faulty replacement of a gasket in an MIC
header in the production area resulted in a
spill of MIC to the atmosphere. Faulty
equipment was responsible for the loss.
Quantity of MIC lost: Not specified.
An overpressure of MIC in a storage area line
caused a spill of MIC to the atmosphere from a
leaky pipe nipple. Defective equipment was
responsible for the spill.
Quantity of MIC lost: Minimal.
The evacuation jets in the production area
became overloaded with MIC. MIC eventually
spilled into the process sewer and
atmosphere. Improper operating procedures
were responsible for the spill.
Quantity of MIC lost: Minimal.
During an attempt to remove liquid MIC from a
cooler discharge line 1n the production area,
MIC was released to the atmosphere and process
sewer. Improper maintenance practices and
faulty equipment were responsible for the
spill.
Quantity of MIC lost: 2 gallons.
6-14
-------
6.2.1 Analysis of Causes
Table 6.1 presents a summary of the causes (operator vs. equipment)
and locations (process vs. storage) of these 28 Incidents. Note that of
the 26 for which sufficient information was available, 16 can be
attributed to operator error and 10 to equipment defect. Regarding
location, exactly half of the releases occurred in the MIC production
area, while the other half occurred in the MIC storage area. Although no
trend was observed in the results from year to year, the ratio of
operator to equipment-related leaks was greater in the storage area (9:4)
than in the production area (7:6).
Based on this analysis, releases of MIC at Institute appear to be
caused more frequently by operator error than by equipment defect. If
this trend holds true for chemical releases at other facilities, then
pressurized tank standards that address only design requirements (and not
operator proficiency) might be ineffective in dramatically reducing
release incidents.
The data further show that release incidents of regulated materials
from chemical plants may be as likely to occur in process areas as in
storage areas. Therefore, regulation of storage tanks without regulation
of production processes may also fail to substantially reduce release
incidents from chemical plants.
6-15
-------
0068s
Table 6-1 Institute Release Incident Summary
Year
1980
1981
1982
1983
1984
Total
Storage
Operator Equipment
error defect
2 1
3 0
0 0
3 2
I _!
9 4
Production
Operator
error
3
1
1
0
2
7
Equipment
defect
0
3
1
0
2
6
Total
Operator
error
5
4
1
3
3
16
Equipment
defect
1
3
1
2
3
10
1. One incident in the storage area and one in the production area could not be
attributed to either equipment or operations deficiencies based on available data.
2. Operations include facility operation and maintenance.
6-16
-------
6.3 Safety Improvements at Institute Since the Bhopal Event
The Union Carbide Institute plant resumed production in May 1985.
Prior to plant startup, Union Carbide implemented major safety
improvements (USEPA 1985). These improvements are:
1. Replacement of brine coolant with chloroform for the unit storage
tanks;
2. Installation of redundant storage tank monitoring instrumentation;
3. Modification of the design of the safety valves and vent lines
for the storage tanks;
4. Increased caustic storage capacity for the emergency vent gas
scrubber;
5. An additional tank that increases residence time for liquid MIC
to be 1n contact with the caustic in the scrubber during
emergency operation;
6. Provision for steam addition to maintain temperature in the
emission vent scrubber to achieve better reaction rates;
7. Improved caustic concentration control for the scrubber;
8. Modifications to the flare ignition system to increase its
reliability;
9. Installation of an air sampling leak detection system for MIC;
10. Installation of a computerized emission warning system (SAFER®)
that would assist emergency response officials in deciding which
corrective actions should be taken if a significant release of
MIC were to occur;
11. Improved internal emergency contingency plans and early warning
notification procedures establishing specific criteria for
initiating notification of local emergency response
organizations; and
12. Installation of additional alarm systems for the storage tank
monitoring instruments, directly wired to the shift
administrator's office.
6-17
-------
All of these Improvements reduce the possibility of a major
disaster's occurring at Institute. However, it is not known how
effective they will be in reducing the lesser-release incidents of the
type previously described.
6.4 Description of the Bhopal Release Incident
After more than four months of investigation, Union Carbide
Corporation released its report of the Bhopal accident in March 1985 (UCC
March 1985). Based on the information in that report, a number of safety
controls were not operable at the time of the disaster, including the
following:
• The refrigeration system that cools the MIC storage tanks had been
nonoperational since June 1984. Had the refrigeration system been
operable, the lower temperature would have slowed the reaction and
allowed more time for corrective action.
• The high temperature alarm had not been reset; therefore, it did
not signal the increased temperature at the start of the reaction.
• The vent gas scrubber (VGS) was not in service at the time of the
incident. Had the VGS been in service, at least some of the MIC
could have been rendered harmless prior to its release.
• The flare was not lit. This is the last fail-safe device intended
to destroy MIC prior to release to the environment.
The UCC report hypothesized that the cause of the reaction in the MIC
storage tank was the introduction of 1,000 to 2,000 gallons of water,
possibly combined with high levels of chloroform. The source of the
water was not conclusively established. Regardless of the source, the
nonoperational status of the refrigeration system, pressure alarm, vent
gas scrubber, and flare contributed to the magnitude of the release,
which resulted in over 2,000 deaths.
6-18
-------
There 1s insufficient information in the Union Carbide report to
determine what degree of protection would have been afforded had these
safety devices been operational (i.e., capacity of VGS, flare, etc.).
However, any prevention, warning, and/or containment provided by -these
systems may have reduced the magnitude of the disaster.
6.5 Comparison of Bhopal to Institute
There are many similarities in the safety controls of the Bhopal and
Institute plants. These include:
• Temperature/pressure alarms;
• Refrigeration of tanks;
• Vent gas scrubber to neutralize MIC releases; and
• Backup flare to incinerate MIC not neutralized by the scrubber.
The new continuous MIC air sampling system installed at Institute is
an added protection-not present at Bhopal. Considering only those
systems common to both plants, however, the operating conditions at
Institute would make a disaster on the scale of Bhopal much less likely.
For example, the VGS is continuously on line while the Institute plant is
in operation. Also, the newly installed flare system at Institute will
automatically relight should it go out. If the flare is extinguished for
more than 30 minutes, the entire production process automatically shuts
down. The continuous MIC sampling system can also automatically stop MIC
production if a release is detected.
We do not believe that safety equipment and operating procedures at
Institute can guarantee that a major release will never occur. The
chances of such a release are, however, much less likely than under the
circumstances leading to the Bhopal event.
6-19
-------
SECTION 6 REFERENCES
Baker, O.P. Jan. 29, 1985. Bhopal gas leak said Initially treated as
wasted. The Washington Post.
Chadwick, O.H., and Cleveland, T.H. 1981. Organic Isocyanates. Ln
Kirk Othmer encyclopedia of chemical technology. New York: John
Wiley and Sons.
Feb. 1, 1985. More chemical leaks at West Virginia disclosed by Union
Carbide Corp. The Washington Post.
Laskowski, S.I. 1985. Inspection report - Union Carbide, Institute,
Vest Virginia. Philadelphia, Pa.: U.S. Environmental Protection
Agency.
Occupational Safety and Health Administration. 1985. OSHA's response
to methyl isocyanate concerns. Washington, D.C.
Poulson, T.H. 1984. institute MIC ii OS/HS. South Charleston, W. Va.:
Union Carbide Corporation.
Union Carbide Corporation. March 1985. Bhopal Methyl Isocyanate
Incident Investigation Team Report.
Worthy, W. 1985. Methyl Isocyanate. The chemistry of a hazard, chem.
Eng. News, February 11, 1985. pp. 27-33.
6-20
-------
APPENDIX A
Summary of Design Procedures and Criteria Specified by
the ASME Code Section VIII, Divisions 1 and 2
(Note: The information in this appendix has been extracted from the ASME
Code, Section VIII. Figures, appendices, etc., mentioned herein can be
found in that document.)
A-l
-------
DESIGN PHILOSOPHY OF SECTION VIII
Although there are differences 1n the factors of safety and design
criteria for Section VIII - D1v. 1 and Section VIII - D1v. 2, many
similarities also exist. In fact, some analysis methods are Identical 1n
each division and use the same allowable stress values.
These similarities and differences will be Identified and examined
In the following discussions of each division.
SECTION VIII - DIVISION 1
Design Requirements
Design rules and formulas for the ASME Code, Section VIII - D1,v. 1,
are primarily given In Subsection A on General Requirements, 1n
Subsection B - Part UW on Welded Vessels, and 1n the Mandatory and
Non-Mandatory Appendices.
The purchase order or design specification usually contains the
design requirements Including design temperature and design pressure.
The vessel shall be designed for at least the most severe condition of
coincident pressure and temperature at normal operation. No temperature
shall exceed the maximum temperature listed for each material
specification given 1n the tables 1n Subsection C or the External
Pressure charts. Other loading Information, such as job site, must also
be provided to establish wind and/or earthquake loads.
A-2
-------
Minimum Design Loadings
According to UG-22, the minimum loadings that must be considered
are: Internal and external design pressure; Impact and fluctuating
loads; weight of vessel and contents Including pressure due to s'tatlc
head of liquid; external loads from attached equipment, piping, etc.;
wind and earthquake loads; local loads from lugs, saddles, etc.; and
temperature gradients.
Allowable Tensile Stress
The maximum allowable tensile stress values permitted for various
materials are found In Subsection C 1n Table UCS-23 for Carbon and
Low-Alloy Steels, 1n UCS-27 for Welded Carbon Steel Pipes and Tubes, 1n
Table UNF-23 for Nonferrous Metals, 1n Table UHA-23 for High-Alloy
Steels, In Table UCI-23 for Cast Iron, 1n UDC-23 for Cast Ductile Iron,
and UHT-23 for Heat Treated Steels.
The basis for establishing the allowable tensile stress values 1s
given 1n Appendix P. At temperatures below the creep range, except for
bolting materials, the allowable tensile stresses are based on the lowest
value of the following:
1/4 of the specified minimum tensile strength at room temperature;
1/4 of the tensile strength at temperature;
5/8 of the specified minimum yield strength at room temperature for
ferrous materials;
5/8 of the yield strength at temperature for ferrous materials;
A-3
-------
2/3 of the specified minimum yield strength at room temperature for
nonferrous materials; and
2/3 of the yield strength at temperature for nonferrous materials.
In addition, for certain austenitic and nonferrous materials, another
higher set of allowable stresses is permitted by increasing one criterion
to 90 percent of the yield strength at temperature. These higher
allowable stresses are not recommended for flanges and other strain
sensitive uses.
For bolting material, the basis for establishing allowable stresses
1s the same as for other materials except for those bolting materials
whose strength is Increased by heat treatment or strain hardening. At
temperatures below the creep range, for these special conditions, the
following additional limits shall not be exceeded:
1/5 of the specified minimum tensile strength at room temperature; and
1/4 of the specified minimum yield strength at room temperature.
At temperatures 1n the creep range, the allowable tensile stresses for
all materials including bolting are based on the lowest value of the
following:
100 percent of the average stress for a creep rate of 0.01 percent in
1,000 hours,
80 percent of the minimum stress rupture at the end of 100,000 hours;
and
67 percent of the average stress rupture at the end of 100,000 hours.
Allowable Longitudinal Compressive Stress in Cylinders
The maximum allowable longitudinal compresslve stress used for the
design of cylindrical shells and tubes subjected to longitudinal
A-4
-------
compression 1s the lower of the maximum allowable tensile stress value or
the value of Factor B determined according to UG-23(B), using the design
charts 1n Appendix V.
Design Criteria
In D1v. 1 the basic design of specific parts, such as heads and
shells, 1s established by a set of design rules using the allowable
stresses as determined above. D1v. 1 does not require a detailed stress
analysis or stress report other than that called for 1n contract
specifications and/or by the stress analyst and the Inspector to satisfy
U-2 (g) and UG-22. This analysis may be necessary to provide details of
design and construction consistent with the rules of this division. If
the vessel 1s subject to severe cyclic operation, 1s 1n some other severe
service, or has a complex geometry not covered by the rules of Div. 1, It
will be necessary to decide whether additional stress analysis Is
necessary.
Strength Theory for Internal Pressure and Tensile Loadings
D1v. 1 uses the maximum stress theory for Internal pressure and other
tensile loadings. This theory considers stress 1n each direction
Independently.
For cylindrical shells, the stress 1n the circumferential or hoop
direction and 1n the longitudinal or axial direction is used to determine
the minimum thickness or maximum allowable working pressure according to
A-5
-------
UG-27. The stress in the radial direction is usually not included in the
design rules of Div. 1.
In order to maintain equilibrium in the hoop direction, the Internal
pressure force exerted against the inside vessel wall must be resisted by
the strength of the metal in the wall thickness for seamless shells and
by the longitudinal weld joints in welded shells. Development of the
formula in UG-27(c) (1) follows. Assuming a thin-wall vessel, for each
length of shell or weld (assume 1 inch), the internal pressure exerts a
force equal to (P) (2R) (1 inch) and the wall thickness or weld exerts a
resisting force at the cross sections of 2(t)(l inch)S. When these are
equated, 2PR=2tS or solving for t = PR/S. If a weld joint efficiency is
included to modify the allowable stress, the formula becomes:
T = PR or P * SEt
SE R
Where
E = longitudinal joint efficiency.
This thin-wall formula was used in the Code until the 1942 edition
when it was modified to more accurately calculate results for thicker
walls due to higher pressures and/or temperatures. The formula became:
T = PR or P = SEt
SE-0.6P R+0.6t
These formulas are given in UG-27(c)(l) of the 1983 edition and are
limited to a thickness not exceeding one-half of the inside radius and P
A-6
-------
not exceeding 0.385 SE. When these limits are exceeded, the rules of
Appendix 1 (l-2)(a)(l) must be followed.
The stress formula for the longitudinal or axial direction 1s
developed 1n the same way. In order to maintain equilibrium 1n the axial
direction, the Internal force exerted against the heads must be resisted
by the strength of the metal In the cross section of the vessel for
seamless shells or by the circumferential weld joints for welded shells.
2
The pressure force equals w/4 D P, while resisting force equals
wDtS. When these are equated, w/4 (D2P) = * DtS or T = PO/4S =
PR/2S and with E added, T = PR/2SE or P = 2SEt/R.
This longitudinal stress formula was also modified 1n the 1942
edition for the same reasons as the hoop equation. It became:
T = PR or P = 2SEt
2SE+0.4P R-0.4t
These formulas are 1n UG-27(c)(2) of the 1983 edition and are limited
to a thickness not exceeding one-half of the Inside radius and P not
exceeding 1.25SE. Again, when these limits are exceeded, the rules 1n
Appendix 1 (l-2)(a)(2) must be followed.
For spherical shells and hemispherical heads, the same thin-wall
formula 1s obtained as for the longitudinal stress 1n the cylinder; that
1s:
T = PR
2SE
A-7
-------
This formula was modified in the 1942 edition for the same reason as
the other formulas to give:
T = PR for spherical shells 1n UG-27(d)
2SE-0.2P
and
T = PL for hemispherical heads 1n UG-32(f).
2SE-0.2P
In a thin-wall cylindrical pressure vessel with hemispherical heads,
the average hoop stress in the shell 1s about twice the average axial
stress in the shell and twice the average stress in any direction of the
hemispherical head. The average radial stress on the cylinder and head
is compresslve and equal to one-half the internal pressure. If the
average hoop stress is Sh, the allowable stress 1s S, and the internal
pressure is P, the relationships of stresses 1n various directions are:
Average Hoop Stress 1n Cyl. = Sh < S
Average Axial Stress in Cyl. = Sh/2 < S
Average Hoop and Meridional Stress in Head = Sh/2 = S
Average Radial Stress in Cyl. and Head = -P/2 = S* (*S1nce P 1s small
for thin-walled vessels.)
This shows that for internal pressure, average hoop stress usually
controls.
For formed head with pressure on the concave side, the minimum
required thickness and maximum allowable pressure are given in UG-32 and
Appendix 1 (1-4). The formulas were based on thin-shell analysis, which
A-8
-------
was adjusted for Increased wall thickness based on a combination of
analytical stress analysis, experimental stress analysis, and
experience. Formulas for common geometries based on inside dimensions
are given in UG-32; formulas for general geometries based on both" inside
and outside dimensions are found in Appendix 1 (1-4). Specific design
requirements and limitations are also given in UG-32. Geometries
included are:
Ellipsoidal Heads - (2:1) - given in UG-32(d) and others in Appendix
1 (1-4);
Torispherical Heads (6 percent knuckle) - given in UG-32(e) and
others in UA-4(d);
Hemispherical Heads - same as for spherical shell;
Conical Heads with <30° - given in UG-32(g) and Appendix 1 (1-5);
Conical Heads with >30° - must be based on special analysis; and
Toriconical heads - follow the rules of UG-32(h) for knuckle design
and conical head rules for the conical part.
For unstayed flat heads and covers, the formulas of UG-34 are used to
calculate the thickness. The basic formula is:
t = 0 CP/SE
This must be adjusted for various details of corner design, methods of
closure (bolted), and weld details given in Fig. UG-34. Rules are also
given for spherical dished covers and quick-opening closures.
Design Theories for External Pressure and Axial Compression
The design bases for external pressure of cylindrical and spherical
vessels as applied in UG-28 and for axial compression of cylinders as
applied in UG-23(b) are elastic instability (also called buckling and
A-9
-------
collapse) and yielding from compressive stress. For cylindrical vessels,
the critical length and thickness are varied to prevent buckling or
yielding. Stiffening rings are added according to UG-29 and 30. For
spherical vessels, the wall thickness must be increased. Design 'charts
for many materials are given in Appendix 5. Minimum required thickness
or maximum allowable pressure is determined on a "trial-and-error" basis.
For formed heads, the design bases are also elastic instability and
compressive yielding. Rules for formed heads with pressure on the convex
side are given in UG-33 as follows:
Ellipsoidal Heads - Use the larger t from UG-32(d)
with P = 1.67 P0 or UG-33(d) rules along with Appendix 5.
Assumes R is K-|D0.
Torispherical Heads - Use the larger t from UG-32(e)
with P = 1.67 P0 or UG-33(e) rules where the UG-33(d) rules
are followed with R = L-j outside radius of crown.
Hemispherical Heads - Same as for spherical shell using UG-28(d).
Conical Heads - Determine t from UG-33(f) using various methods
depending on the apex angle.
A-10
-------
SECTION VIII - DIVISION 2
Design Requirements
The design requirements for Section VIII - D1v. 2 are given in Part
AD. Part AD contains specific design rules and formulas for common
configurations. When rules and formulas are not given, or when the
designer chooses, a stress analysis can be made. The thicknesses,
however, cannot be less than those obtained by using the specific rules
and design formulas.
Minimum Design Loadings
According to AD-110, the minimum loadings to be considered are:
Internal and external pressure;
Dead weight of vessel and contents;
Loads from other equipment;
Wind and earthquake loads;
Local loadings;
Impact loads; and
Temperature conditions.
Table AD-120.1 establishes the coincident pressure and temperature
relationships that must be considered by the designer. These include
conditions for design pressure and temperature, operating temperature and
pressure, test pressure, and safety valve setting.
The stress intensity factor, k, for various loading combinations is
given in AD-150 and Fig. AD-150.1. These are k = 1.0 for sustained
loadings, including pressure and dead loads, k = 1.2 for sustained
loadings plus wind load or earthquake load, k = 1.25 for hydrostatic
test, and k = 1.15 for pneumatic test.
A-ll
-------
AD-160 presents rules that establish whether a fatigue evaluation is
required. According to AD-160.1, operating experience under certain
restricted conditions may be considered along with other specified
requirements.
Design Stress Intensity Values
The design stress intensity values permitted for various materials
are given in Part AM - Material Requirements. The basis for establishing
these values, which are contained in Appendix 1, is as follows:
Except for bolting materials, the design stress intensity values are
based on the least of the following:
1/3 of specified minimum tensile strength at room temperature;
1/3 of the tensile strength at temperature;
2/3 of the specified minimum yield strength at room temperature; and
2/3 of the yield strength at temperature except for austenitic,
stainless steels and certain nonferrous materials; this factor may be
as large as 90 percent of the yield strength at temperature.
Bolting materials have their design stress Intensity values
established by similar, but more restrictive, criteria. For bolting used
with bolted flange connections of Appendix 3, the least of the following
1s used:
1/4 of the specified minimum tensile strength at room temperature for
nonheat-treated materials;
1/5 of the specified minimum tensile strength at room temperature for
heat-treated materials;
1/4 of the tensile strength at temperature;
A-12
-------
5/8 of the specified minimum yield strength at room temperature for
nonheat-treated materials;
1/4 of the specified minimum yield strength at room temperature for
heat-treated materials; and
5/8 of yield strength at temperature.
The criteria for design stress intensity values for bolting materials
for use with Appendix 4 are:
1/3 of the specified minimum yield strength at room temperature; and
1/3 of the yield strength at temperature.
No design stress intensity values have been established in the creep
range for Oiv. 2. When a vessel is to be designed at a temperature that
is above the highest temperature permitted for that material as given in
the tables of design stress intensity values, for restricted applications
where fatigue is not a consideration, stresses may be established
according to Code Case 1489, which uses allowable stresses from Div. 1.
Design Basis
The design basis of Section VIII, Div. 2, is to provide design
formulas and rules for the more common configurations of shells and
formed heads for temperatures below the creep range. Detailed
evaluations of actual stresses in complex geometries and/or with unusual
loadings are made. Those calculated stresses are assigned to categories,
along with different allowable stress values. The categories and
subcategories are:
A-13
-------
A. Primary Stress
(1) General Primary Membrane Stress
(2) Local Primary Membrane Stress
(3) Primary Bending Stress
B. Secondary Stress
C. Peak Stress
Primary stress 1s that stress caused by loadings that are necessary to
satisfy the laws of equilibrium between applied forces and moments.
Primary stress 1s not self-limiting. Secondary stress is the stress
developed by self-constraint of a structure. A basic characteristic of
secondary stress is that It is self-limiting. Finally, peak stress is
the highest stress in a structure and 1s usually due to a stress
concentration caused by geometry. It is important in consideration of a
fatigue failure.
Potential failure modes arid various stress categories are related.
Limits on primary stresses are set to prevent plastic deformation and
ductile burst. The primary-plus-secondary stress limits are established
to prevent plastic deformation leading to incremental collapse and to
validate the use of an elastic analysis to make a fatigue evaluation.
Finally, peak stress limits are set to prevent fatigue failure.
Basic Stress Intensity Limits
The basic limits for the various categories are based on limit design
theory, assuming no strain-hardening and applying an adequate safety
A-14
-------
factor. The primary stress limit permits no yielding to occur, while the
primary-plus-secondary stress limit allows only that amount that will
"shake down" to elastic after one or two cycles.
The basic stress Intensity limits for various categories are:-
Factor Based Factor Based
on Allowable on Allowable
Stress Intensity Category
General Primary Membrane (Pm)
Local Primary Membrane (PL)
Primary Membrane Plus Bending
(PL *
Value
Value
Factor Based
on Yield
Strength
1
1
.5
.5
Sm < 2/3 Sy
^m - Sy
5m ^ 5..
<
<
<
1/3
1/2
1/2
S
S
S
u
u
n
Primary Plus Secondary
(PL + Pb + Q)
3 S
'm
23,
Fatigue Analysis
In contrast to Div. 1, when Div. 2 considers fatigue as a mode of
failure, rules are given to design against fatigue failures. Curves,
which give design cycles for various values of peak stress including the
effects of the stress concentration factors, are provided for many
materials.
Rules are also given for combining a variety of design stress cycles
that a vessel may undergo during its design lifetime. These various
cycles are combined by a linear damage relationship called cumulative
fatigue damage.
When the design considerations given in AD-160.2 and AD-160 are met,
no design fatigue evaluation is required, even though the vessel may be
designed and subjected to some cyclic operation.
A-15
-------
Strength Theory
The strength theory used for Section VIII, Div. 2, 1s the maximum
shear stress theory, also known as the Tresca criterion. Both the
maximum shear stress theory and the distortion energy theory are*
considered to be better than the maximum stress theory in predicting
yielding and fatigue failure. The maximum shear stress theory is easier
to use and thus was chosen for Div. 2.
Maximum shear stress is defined as one-half of the algebraic
difference between the largest and the smallest of the three principal
stresses. This theory states that yielding occurs when the maximum shear
stress is S /2. Therefore, yielding occurs when:
1/2 ("largest - Smallest) = 1/2 S or (°L - °S) < S
Using an appropriate factor of safety, the term S is replaced by S ,
1.5S and 3S . This new term of the actual difference to compare to
m m
these limits is called "stress intensity."
A-16
-------
APPENDIX B
Selected Compatibility Information
For Tank and Liner Materials
B-l
-------
Table B-l Chemical Resistance Guide for HYTREL
The data tabulated betow summarize the effects of a broad variety of fluids on HYTREL polyester elastomers. As a general rule the resistance of HYTREL
elastomers to fluids and chemicals increases as the polymer hardness increases. Unless otherwise notid the ratine? shown in the table apply to all
hardness trades.
Rating Key: A-Fluid has little or no effect
B-Fluid has minor to moderate effect
C-Fluid has severe effect
T-No data-likely to have minor effect
X-No data-likely to have severe effect
Ratings are at 22°C [72*F] unless otherwise specified. Concentrations of aqueous solutions are saturated, eicept where noted.
We emphasize that the data contained herein should be used as a guide only. The tabulation is based primarily on laboratory tests but does not take into
account all variables than can be encountered in actual use. Therefore it is advisable to test the material under actual service conditions before
specification. If this is not practical, tests should be devised that simulate service conditions as closely as possible.
CHEMICAL
Acetic Acid. 20%
Acetic Acid. 30%
Acetic Acid, Glacial
Acetic Acid. Glacial (38°C) [ 100eF]
Acetic Anhydride
Acetone
Acetylene
Aluminum Chloride Solutions
Aluminum Sulfate Solutions
Ammonium Chloride Solutions
Ammonium Hydroxide Solutions
Ammonium Sulfate Solutions
Ammonium Sulfate Solutions
Amyl Acetate
Amyl Alcohol
Aniline
ASTMOilNo. l(149°CM300Dn
ASTM Oil No 3
-------
Table B-l (continued)
CHEMICAL
Formaldehyde. 40%
Formic Acid
FREON-111"
FREON-12
FREON-113(550C)[1300F]
FREON 114
Gasoline
Glue .
Glycerin
fl-Hexane
Hydrazine
Hydrochloric Acid. 20%
Hydrochloric Acid. 37%
Hydrocyanic Acid
Hydrofluoric Acid. 48%
Hydrofluoric Acid. 75%
Hydrofluoric Acid. Anhydrous
Hydrogen
Hydrogen Sulfide
Isooctane
lupropyl Alcohol
JP-4 Jet Fuel
Kerosene
Lacquer Solvents
Lacquer Solvents
Lactic Acid
Linseed Oil
Lubricating Oils
Magnesium Chloride Solutions
Magnesium Hydroxide Solutions
Mercuric Chloride Solutions
Mercury
Methyl Alcohol
Methyl Ethyl Ketone
Methyl Ethyl Ketone
Methylene Chloride
Mineral Oil
Naphtha
Naphthalene
Naphthalene
Nitric Acid. 10%
Nitric Acid. 30%
Nitric Acid. 60%
Nitric Acid. 70%
Nitric Acid, Red Fuming
Nitrobenzene
OleicAcid
Oleum. 20-25%
RATING
B
B
A
A
A
C
B
C
T
X
X
X
A
A
A
A
A
T
B (40. 550)
A (63. 720)
T
T
A
T
T
T
A
A
B (40. 550)
A (63. 720)
C
A
A
B (40. 550)
A (63. 720)
B
C
C
C
C
C
A
C
CHEMICAL
Palmitic Acid
Perchloroethylene
Perchloroethylene
Phenol
Pickling Solution (20% Nitric Acid. 4% HF)
Pickling Solution (17% Nitric Acid. 4% HF)
Potassium Dichromate Solutions
Potassium Hydroxide Solutions
PYDRAUL 312"a
Pyridine
SAElOOil
Sea Water
Silicone Grease
SKYDROL 500B"»
Soap Solutions
Sodium Chloride Solutions
Sodium Dichromate. 20%
Sodium Hydroxide. 20%
Sodium Hydroxide, 46Vi%
Sodium Hypochlorite. 5%
Soybean Oil
Stannous Chloride. 15%
Steam (100'C) [212°F]
Steam <110°C)[230«F1
StearicAcid
Styrent
Sulfur. Molten
Sulfur Dioxide. Liquid
Sulfur Dioxide, Gas
Sulf uric Acid, up to 50%
Sulturic Acid, above 50%
Sulfuric Acid. Fuming (20% Oleum)
Sulfurous Acid
TannicAcid. 10%
TartaricActd
Tetrahydrofuran
Tetrahydrofuran
Toluene
Toluene
TREFLAN"8
TricMoroethylene
Trichloroethytene
Triethanolamine
Trisodium Phosphate Solution
Tung Oil
Water (70°C) [158°F]
Water (lOO'C) [212*F]
Xylene
Xylene
Zinc Chloride Solutions
RATING
A
C (40. 550)
B (63. 720)
C
X
X
T
A
A
A
A
T
A
B
A
T
T
B
C
T
X
T
T
T
A
C
C
B
A
T
8 (40,550)
A (63. 720)
B (40. 550)
A (63. 720)
B
C (40. 550)
B (63. 720)
C
A
T
A
B
B (40. 550)
A (63. 720)
A
Sourre: DuPont Company, Elastomer Division.
Wilmington, Delaware. (Product of brochure HYT-504A).
B-3
-------
Table B-2 NACE Chemical Compatibility Data Sample
CORROSION R»IE
• LESS IHAN 0 00?" PER VEAR
O IESS IHAN 0 0?!!» II
& I
\ I
•
I
!
a «• iiA»«iu wnnvn UJUM ,_f^**11,,
• ji A*I iu m« M OUUMII x «»« * " j uu
G^LuMJ *
EM '
Kj ft fr
ii i * ii
MM
QUMWfl
OHMHOM
i J; i ..:. . ;:
;
w i i$« ^81 •*«• • x^
Pi ffiii I':I
1155. .
in i* n
- H t> 'OK MTU*I 1 ! -
iiliJ 8 I ii;: - .
1 Al til All
1 . tsstnt 1 1
--....-- -j- - - _ . _ ..j
-it !
wr MI
' i . "'
•
R BASE ALLOYS NICKEL BASE AILOYS
tMldvy UKXH
HAU HAU I*»UI Mf>« -*.rt l.r.M
niTTTTTf [ ! ' | 1 !J
! h-7! :*H
. - ' . »L .. •[ J. Bti. K
1*!
mil oiaii
,. . 1 5 9
!
• • i« •« » »M 'J(ir^J« • r^KJI «»• j«
| f| f ill
I
' I * HI A HE
1 il (_ ! 1 !
1
.V , i
)
1 ' : I
i
-------
APPENDIX C
Individuals Contacted 1n Telephone
Survey for Pressurized Tanks
C-l
-------
0359s
Table C-l State Officials Interviewed
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
o
i
1X1 Connecticut
Delaware
District of
Columbia
Florida
Georgia
Hawaii
Name of official
James Stephenson
R. 0. Gather
Jim Little
Peter Watson
George Horn
Sill Cimino
Robert O'Connor
J. G. Cvar
Roderick Freeman
Louis Price
Joe Tanner
Howard t. Farwell
Title
Administrator
Chief, Mechanical
Inspection
Boiler Inspection Chief
Inspector
Principal Safety Engineer
Chief Boiler Inspector
Boiler Inspection Director
Acting Director
Chief Inspector
Administrator
Comnissioner
Chief Inspector
Agency
OSHA
Oept. of labor, Pressure
Vessel Section
OSHA
Dept. of Labor, Boiler
Inspection Division
OSHA
Division of Labor
Bureau of State Fire
Marshal
Division of Boiler Safety
Building and Zoning
Regulation Division
Safety Standards Section
Oept. of Labor
Boiler and Elevator
Phone number
(205) 261-3460
(907) 264-2447
(602) 255-5559
(501) 375-8442
(415) 557-0437
(303) 289-5641
(203) 238-6034
(302) 571-3247
(202) 727-7554
(904) 488-9660
(404) 656-3011
(808) 548-5400
Inspection Bureau
-------
Table C-l (continued)
State
Idaho
Illinois
Indiana
Iowa
Kansas
o Kentucky
co
Louisiana
Maine
Maryland
Massachusetts
Michigan
Mi nnesota
Name of official
Harry Nichols
Ouane Gallup
James Moore
Harold McLamb
William E. Brown
Harmon C. Mills
Carl Thompson
Robert Sullivan
John Grail
Ray Archambault
Stanley Mierzwa
Title
Inspector
Superintendent and Chief
Acting Chief Inspector
Chief Inspector
Chief Inspector
Chief Boiler Inspector
State Fire Marshal
Chief Inspector
Chief Inspector
Supervising Inspector
Chief Boiler Inspector
Assistant Chief Inspector
Agency
Oept. of Labor and
Industrial Services
Office of the State Fire
Marshal
Indiana Boiler and Pressure
Vessel Board
Bureau of Labor
Dept. of Human Resources
Office of State Fire
Marshal
Office of State Fire
Marshal
Bureau of Labor
Division of Labor and
Industry
Oept. of Public Safety
Dept. of Labor
Dept. of Labor and
Phone number
(208) 334-2327
(217) 785-0969
(317) 232-1921
(515) 281-3647
(913) 296-5000
(502) 564-3626
(504) 925-4911
(207) 289-3331
(301) 659-4180
(617) 727-7686
(517) 373-9435
(612) 296-6107
Industry
-------
0359s
Table C-l (continued)
State
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
o
** New Jersey
New Mexico
New York
North Carolina
North Dakota
Name of official
Henry McEwen
Jana Bopp
Paul Rafferty
John Mickels
Samuel Bartholomew
Veronique C. Soucy
Mr. Hawley
Louis Garcia
Edwin Hicks
Mr. Eaton
R. Reetz
Title
Chief Inspector
Fire Marshal Assistant
Boiler Supervisor
Chief Boiler Inspector
Mechanical Inspection
Coordinator
Inspector
Chief Inspector
Chief Inspector
Assistant Director
Chief Inspector
Agency
OSHA
Dept. of Public Safety
Dept. of Labor and
Industry
Dept. of Labor
Dept. of Industrial
Relations
Dept. of Labor
Dept. of Labor and
Industry
Mechanical Board of the
Construction Industries
Commission
OSHA, Bureau of Boilers
Dept. of Labor
Workmen's Compensation
Phone number
(601) 354-6026/36
(314) 751-2930
(406) 444-6419
(402) 554-3097
(702) 885-4583
(603) 271-3176
(609) 984-3001
(505) 827-6253
(518) 457-2722
(919) 733-3034
(701) 224-2700
Ohio
Bill Brennan
Chief of Boiler Division
Bureau
Workmen's Compensation
Bureau
(614) 466-3271
-------
0359s
Table C-l (continued)
State
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
V South Dakota
en
Tennessee
Texas
Utah
Vermont
Virginia
Name of official
J. W. Greenwalt
Charles Walters
Denise Siders
Gary Wheat on
Sharon Dantzler
James Helgaard
Morris L. Snow
Steven H. Matthews
Raymond K. Blosch
M. J. Wheel
W. E. Long
Title
Chief Boiler Inspector
Chief of Boiler and Pressure
Vessel Division
Boiler Division Assistant
Chief Boiler Inspector
Attorney
Deputy Secretary
Director, Chief Inspector
Director, Chief Inspector
Chief Inspector
Director
Chief Inspector
Agency
Dept. of Labor
Dept. of Commerce
Dept. of labor and
Industry
Dept. of Labor, Division
of Occupational Safety
OSHA
Dept. of Public Safety
Dept. of Labor
Dept. of Labor and
Standards
Industrial Comnission
Fire Prevention Division
Boiler and Pressure Vessel
Phone number
(405) 521-2461
(503) 229-5755
(717) 787-5279
(40)) 277-2756
(803) 758-2851
(605) 773-3541
(615) 741-2123
(512) 475-4799
(801) 530-6872
(802) 862-8970
(804) 786-3160
Washington
George W. Folta
Chief Inspector
Safety Division
Dept. of Labor and
Industries
(206) 281-1290
-------
o
en
0359s
Table C-l (continued)
State
Name of official
Title
Agency
Phone number
West Virginia Lawrence Barker Commissioner
Wisconsin John J. Duffy Chief Inspector
Wyoming
Rick Obstar
Chief Inspector
Oept. of Labor
(304) 348-7890
Dept. of Industry, Labor, (608) 266-3131
and Human Relations
OSHA
(307) 777-7786
-------
APPENDIX D
Industrial Uses and Recommended Storage Practices for
Materials Commonly Stored 1n Pressurized Tanks
D-l
-------
0089s
Table 0-1 Standard Industrial Classification Codes Associated with the Manufacture and Industrial Use of
Poisonous Chemicals Potentially Stored in Pressurized Underground Tanks
Chemical
Synonyms
SIC Codes
CAS Reg. H Manufacture
Use
Cannon industrial
uses
Storage/handling
precautions
Acrolein
Allylamine
ro Allyl chloride
Anmonia
Argon
Carbon disulfide
Chlorodifluoro-
ethane
2-Propenal
Acrylaldehyde
Propenal
Allylaldehyde
3-Ami nopropy1ene
2-Propenylamine
2-Propen-l-amine
Chlorallylene
3-Chloroprene
3-Ch1oropropy1ene
107-02-8
2813 2818*
NA
Carbon bisulfide
Dithiocarbonic-
anhydride
NA
2821 2869
5161 2834
0721
107-11-9
107-05-1
2822 2834
2869 2879
Anhydrous ammonia 7664-41-7
3872 2819
2824 2892
2821 5161
2821 2822
2834 2869
2879 5161
2879 2834
423 46 -
5161
7440-37-1
75-15-0
75-68-3
2821 2823
2879 2911
2813
2869 5161
2822 2869
5161
Plastics; perfumes; organic
synthesis; military poison;
refrigerant; production of
glycerine, polyurethane,
polyester resin; herbicide,
biocide; pharmaceutical; etc.
Intermediate for pharma-
ceutical s and organic
synthesis.
Chemical intermediate.
Fertilizer; explosives;
synthetic fibers; manufac-
ture of nitric acid,
hydrogen cyanide, acrylo-
nitrile; refrigerant;
Pharmaceuticals.
Manufacture rayon, cellophane
CC14; soil disinfectant
solvent.
Do not store/ship uninhibited.
Vapors (uninhibited) may form
polymers in vents or flame
arresters of storage tanks,
stopping vents. Store/ship under
a blanket of oxygen-free inert
gas.
Store as standard flamnable
liquid.
General storage procedures.
Store in stainless steel.
Pressure of 175 psi required to
keep gas in liquid state. Water
spray effective in dispersing
vapors. 3-4% transported by
pipe.
Can be stored in iron, aluminum,
glass, porcelain, teflon. Ship
under blanket of inert gas.
Chloropicrin
mixtures (gas)
NA
2869 2879
0721
-------
0089s
Table D-l (continued)
Chemical Synonyms
Chloroprene Chlorobutadiene
2-Chloro-1-3-
butadiene
Cyclohexyl NA
isocyanate
Dichlorosiline NA
Difluoroethane NA
Dime thy 1 ami ne
o
i
CO
Ethane NA
Ethylene Ethene
Ethylene oxide EO, ETO, ETOX
Oxirane
Ethylene oxide/ NA
propylene oxide
SIC Codes
CAS Reg. tt Manufacture Use
126-99-8 2822 2821 2869
5161
3173-53-3 2821 2869
4109-96-0 2869
75-37-6 2821 2869
124-40-3 2818* 2841 2821 2834
3111 2869 2899
5161 2822
74-84-0 1321 2821
2869 2911
46-
74-85-1 3255 3548 2821 2869
2911 46-
2879
75-21-8 2043 2879 2821 2869
2911 46 -
9003-11-6 2821 2869
5161
Cannon industrial
uses
Manufacture synthetic rubber.
Rubber; vulcanizing; tanning;
soaps and detergents.
Welding and cutting metals;
plastics; ripening citrus
fruits; manufacture alcohol;
mustard gas organics.
Foodstuffs and textiles;
fumigant; fungicide;
agricultural; sterilize
surgical instruments; manu-
facture acrylonitrile;
organic synthesis.
Storage/handl ing
precautions
Store as standard flammable
liquid. Avoid all contact with
mercury. All equipment should
grounded.
be
Steel pressure vessels. Isolate
from oxygen, chlorine, combust-
ibles, organic and oxidizing
materials.
Steel containers. Keep below
86°F, insulated/cooling tanks.
Provide waterspray system,
diking, and drainage.
mixtures
-------
0089s
Table D-l (continued)
SIC Codes
Chemical
Synonyms
CAS Reg. It Manufacture
Use
Cormon industrial
uses
Storage/handling
precautions
N-Ethyl toluidines NA
(o-, m-, and p-)
94-68-8
102-27-2
622-57-1
2834 2869
2899
Hydrocarbon NA
gases, n.o.s.
2813 2821
2869 2899
2911 46 -
5161
Hydrogen
NA
1333-74-0
i Hydrogen cyanide Hydrocyanic acid 74-90-8
*" Prussic acid
Formonitrile
2879 2818*
2813 2911
46-
2819 2821
2869 2899
2879 0721
Agricultural chemicals; Standard combustible liquids
insecticide; rodenticide; storage. Handle in gas-tight
manufacture of acrylonitrile, equipment.
acrylates, cyanide salts,
lactic acid, dyes; metallurgy;
mining.
Hydrogen sulfide Hydrosulfuric acid 7783-06-4
Sulfur hydride
1381 2819
2869 2911
2999
Chemicals in metal; reagent;
purification of HC1 and
H2S04- precipitating
sulfides of metals; manu-
facture of elemental sulfur.
Steel pressure vessels. Storage
should be isolated/detached.
Avoid nitric acid, strong
oxidizing or corrosive materials.
Liquified petroleum (see: Propane)
gas (LPG)
74-98-6
106-97-8
68476-85-7
Methane or natural
gas
NA
74-82-8
64741-48-6
1311 2819
1381 2911
2869 46—
2999
-------
0089s
Table 0-1 (continued)
SIC Codes
Chemical
Synonyms
CAS Reg. * Manufacture
Use
Corrmon industrial
uses
Storage/handling
precautions
Methyl amine
Monomethylamine
Ami none thane
Methyl chloride Chloromethane
74-89-5
74-87-3
2818*
2813 2833
Methylene chloride Dichloromethane
75-09-2
2821
a
i
CD
Methyl isocyanate
Nitrogen
2869 3111
2821 2822
2869 2879
2899 5161
282) 2822
2869 2879
2899 5161
MIC
Methyl isocyanic
acid
Methycarbylamine
NA
624-83-9
2821 2869
2879 0721
7727-37-9
2813 2879
5161
Tanning; organic synthesis.
Refrigerant; methylating
agent, dewaxing/degreasing
agent; catalytic solvent in
synthetic rubber production;.
organic synthesis.
Solvent for cellulose
acetate; industrial solvent
in insecticides, metal
cleaners, paint and varnish
removers; aerosal propel 1ants;
blowing agent; degreasing.
Chemical intermediate for
carbamate insecticides/
herbicides.
Gas or liquid. Storage
should be outdoor/detached.
Avoid contact with mercury.
Shipped as liquid under pressure.
Cool below 40°C. Outside or
detached storage. Avoid water,
sources of ignition.
Avoid moisture in tanks. Do
not use plastic/rubber hose for
unloading.
Store in stainless steel recepta-
cles, nickel, or perfectly
vitrified materials. Dangerous
when exposed to heat, flame, or
oxidizers. Keep cool.
Nitrogen dioxide
(also trioxide
and tetroxide)
Phosgene
Nitrogen peroxide
10102-44-0
10544-72-6
10544-73-7
Carbonyl chloride 75-44—5
2041 2818 2813 2879 Nitric acid; sulfuric acid
2819 2892 5161 nitration; explosives; hemor
static; cotton bleach; floor;
rocket fuel.
2818* 2869 Organic synthesis; warfare
intermediate
Store in cool, ventilated area.
Avoid all releases of vapor.
-------
0089s
Table 0-1 (continued)
SIC Codes
Common industrial
Storage/handling
Chemical
Propane
Synonyms CAS Reg. 0 Manufacture
Liquified petroleum 74-98-6 2813
gas (IPG)
Use
1321 2911
4423 46 -
5161
uses
Fuel, gas, refrigerant,
chemical intermediate.
precautions
Liquified form common
shipping. Less than
ported by pipe.
for
IX trans-
o
01
Refrigerant/
dispersant gases
(freon)
Sulfur dioxide
Sulfur trioxide
MA
Sulfurous oxide 7446-09-5
Sulfurous anhydride
NA
3999
2869 5161
2819
2819
Preservative, breweries,
bleach, glue, pulp.
Pressurized vessels for shipping.
Store outdoors/detached.
Sulfuryl chloride NA 7791-25-5 2819 2821
2869
Tetrafluoroethylene NA 116-14-3 2821 2869
Vinyl chloride Chloroethene 75-01-4 2813 2821 2869 Plastics, refrigerant, Outside/detached storage. Free
Chlorethylene 2822 4423 organic compounds. of heat and ignition sources
5161 Approximately 81 transported
pipe.
.
by
Source: NIH/EPA 1985. National Institutes of Health/U.S. Environmental Protection Agency. Computer Printout (Chemical Information System); OHMTADS Data
Base (Oil and Hazardous Materials Technical Assistance Data System).
* 2818 has been changed to 2869 (Standard Industrial Classification Manual, OMB 1972).
NA - Information not available from NIH/EPA (1985), OHMTADS Database. SIC codes for industrial use based on knowledge of the chemical industry.
-------
APPENDIX E
Summary of Health Effects Information for
Materials Commonly Stored 1n Pressurized Tanks
E-l
-------
Table E-1 Exposure Limits, Toxicity Data, and Adverse Health Effects Associated with Chemicals Potentially Stored
in Pressurized Underground Tanks
Inhalation exposure Animal
Chemical
Acrolein
Regulations
OSHA PEL (TWA)
mg/nr ppm
0.25 0.1
Recommendations LC50
NIOSH IOLH ACGIH STEL (15 min) ppm
ppm mg/rrr ppm
5 0.8 0.1 8 (rat)
toxicity
LD50
mg/kg
30 (mouse)
26; 46 (rat)
7.1 (rabbit)
Mutageni.city/
carcinogen! city
H: potential
C: insufficient
testing
Adverse human
health effects
Vapors cause lacrimation.
Bronchitis. Death by
pulmonary edema, respiratory
paralysis. Irritation - skin,
mucosa. 153 ppm fatal to man in
10 minutes.
Allylamine
i Allyl chloride
ro '
Ammonia
Lowest toxic inhalation dose reported in man is 5 ppm for 5 min. Highly toxic via oral, inhalation, and dermal routes. Animal
experiments produced irritation to nose and mouth, congestion of the eyes. Extended exposure leads to irregular respiration, cyanosis,
excitement, convulsions, and death or permanent injury.
1
300
700 (rat)
2050s(rabbit)
35
50
500
27
35
7,338/1 hr
(rat)
4,837/1 hr
(mouse)
10,066/1 hr
(rabbit)
Vapor may cause lung and
eye injury. Absorbed through
skin. Combustion releases toxic
HC1 gas.
Exposure to high cone, can
cause temporary blindness
and eye damage. Direct
contact with liquid causes
severe eye and skin burns.
5,000-10,000 ppm rapidly fatal
for short exposure. Symptoms:
bronchitis, corneal damage,
cough, dyspnea, headache, nausea
and vomiting, pulmonary edema,
salivation.
Argon
NA
-------
Table E-l (continued)
Inhalation exposure Animal toxicity
Regulations Recorrmendations LC50 LD50
Chemical
OSHA PEL (TWA)
mg/rn^ ppm
NIOSH IDLH
ppm
ACGIH STEL (15 min) ppm mg/kg
mg/m^ ppm
Mutagenicity/ Adverse human
carcinogen! city health effects
Carbon disulfide
20
500
30(NIOSH) lO(NIOSH)
300SC(rabbit)
M: potential,
exposure
associated with
chromosome
aberrations.
T: positive
Affects central nervous
system: anesthetic,
euphoria, restlessness,
nausea, vomiting, terminal
convulsions. Thyroid hypo-
function (first symptom).
Poisoning from inhalation,
ingestion, absorbed through
skin. Appears to be etiological
factor in pathogenesis of
coronary disease.
Chlorodifluoro- MA
ethane
Chloropicrin NA
mixtures (gas)
Chloroprene
90(skin) 25(skin)
400 3.6(NIOSH) 1(NIOSH) 285g/m3(rat) 2900(LD100, rat)
605 (LC100/8
hr., rat)
Irritation of respiratory tract,
depression of respiration, and
with continued exposure
asphyxia. Human exposure has
resulted in dermatitis,
conjunctivitis, corneal necrosis,
anemia, temporary loss of hair,
nervousness, and irritability.
Cyclohexyl-
isocyanate
NA
-------
Table E-l (continued)
Chemical
Inhalation exposure
Regulations Recommendations
OSHA PEL (TWA) NIOSH IDLH ACGIH STEL (15 min)
mg/m3
Animal toxicity
LC50 LD50
ppm
mg/kg
Mutagenicity/
carcinogenicity
ppm
ppm
ppm
Adverse human
health effects
Dichlorosiline NA
Difluoroethane NA
Dimethylamine 18
Ethane
Ethylene
10
2,000
540;698 (rat)
649 (mouse)
NA
No OSHA regulations. ACGIH recommended-asphyxiant.
Extremely toxic with inhalation,
ingestion. Direct or prolonged
contact causes burns, cough,
sneeze, headache, nausea.
Anesthetic, unconsciousness with
moderate concentrations. Simple
asphyxiant.
Ethylene oxide 90
N-Ethyltolui-
dines
(o-, m-, p-)
Hydrocarbon
gases
50
800
135(NIOSH) 75(NIOSH)
4000/4 hr 330/14 day
(rat) (rat)
836/4 hr 270/14 day
(mouse) (gn pig)
973/4 hr
(dog)
H: Known mutagen,
chromosome
aberrations
C: ACGIH suspect
carcinogen
NA
NA
Moderately toxic by inhalation.
Irritating. Anesthetic. Can
cause pulmonary edema.
Hydrogen
NA
-------
Table E-l (continued)
Inhalation exposure
Chemical
Regulations
OSHA PEL (TWA)
mg/nr ppm
Reconrnendations
NIOSH IDLH
ppm
ACGIH STEL
mg/rrr
(IS min)
ppm
Animal toxicity
LC50
ppm
LD50
mg/kg
Mutagenicity/
care inogeni city
Adverse
health
human
effects
Hydrogen cyanide 11(skin) 10(skin)
50
10(ceiling) lO(ceiling) 544/5 min 4 (mouse)
(mouse)
169/30 min
(rat)
300/3 min
(dog)
potential, chromo-
some breaks
no evidence
no evidence
i
en
Hydrogen sulfide
20
300
21
15
712/1 hr
(rat)
634/1 hr
(mouse)
600/5 hr
(human)LC
Poison. Can be absorbed through
skin (100+ppm). No warning
properties. At toxic level,
110-135 ppm fatal in 30 min.
181 ppm fatal in 10 min. Causes
dizziness, confusion, vertigo,
weakness, and finally loss of
consciousness. In humans,
exposure of 4-45 ppm for several
hours caused headaches, nausea,
weakness, and nervous and
circulatory system problems.
Respiratory effects: irritant to
nose, throat, lungs. Bronchitis,
pulmonary edema. Systemic
effects: headache, dizziness,
nausea, fatigue, diarrhea,
vomiting, insomnia, weakness,
tremors, and numbness in
extremities, shock, convulsions,
unconsciousness. Irritant and
asphyxiant. Small amounts -
nervous system depressant, larger
amounts - stimulant. High
amounts - paralysis of
respiratory center. Low
level/chronic poisoning.
Methane or NA
natural gas
Methylamine
12
10
100
0.021 ppm - recognized odor in
air.
-------
Table E-l (continued)
Inhalation exposure
Chemical
Regulations
OSHA PEL JTWA1
Recomtendations
NIOSH IDLH ACGIH STEL
(15 min)
Animal toxicity
LC50
ppm
LD50
mg/kg
Mutagenicity/
carcinogenicity
Adverse human
health effects
mg/rn^
ppm
ppm
rng/m^
ppm
Methyl chloride — 100
10,000 205 100 95.100/ 1800 (rat)
LC100
(gn pig)
M:
C:
T:
positive in
Salmonella, negative
in Drosophila
potential
potential
Anesthetic and narcotic. Attacks
liver and kidneys. Causes dizzi-
ness, drowsiness, confusion,
nausea, convulsions, and coma.
Inhalation of high cone, causes
serious central nervous system
damage, sometimes death.
Methylene
chloride
500
5,000 1,740
500
2,00IP (rat) H: positive in Narcotic: light headedness,
Salmonella, negative mental confusion, nausea,
rn
en
Methyl
isocyanate
0.05
(skin)
0.02
(skin)
20
in Drosophila
C: inadequate data
T: no evidence
M: no evidence
vomiting, and headache; possibly
loss of consciousness. Strongly
corrosive to eyes. Can be
absorbed through the skin.
Severe cases: disturbance of CNS
function, depression. Deaths
reported as cardiac injury and
heart failure.
Lacrimator. Respiratory effects:
asthma, bronchitis, cough.
Dyspnea, headache, nausea,
vomiting, pulmonary edema.
Sensitizer. Chest pain. In
humans: 1 to 5 minutes at 0.4
ppm, no irritation of eyes, nose,
or throat. At 2 ppm, irritation
and lacrimation. At 4 ppm,
symptoms marked. At 21 ppm,
exposure unbearable.
Nitrogen
NA
-------
Table E-l (continued)
Inhalation exposure
Regulations Reconnendations
Chemical OSHA PEL (TWA) NIOSH IDLH ACGIH STEL (15 min)
mg/nr ppm ppm mg/rn^ ppm
Nitrogen dioxide -- — 50 10
Phosgene 0.4 0.1 2
Animal toxicity
LC50 LD50
ppm mg/kg
216mg/m3 —
(rat)
315 ppm
(rabbit)
700/30 min
LC (humans)
75 (rat) —
110 (mouse)
1,087
(monkey)
400-500/1 hr
LC (humans)
Hutagenicity/ Adverse human
carcinogenicity health effects
Irritating. Lung inflanrnation.
Pulmonary edema. 200 ppm by
inhalation lethal.
Prolonged contact may predispose
to disabling lung conditions,
pneumonia. Vapor affects eye,
nose, and throat, as irritant.
Hay cause pulmonary edema or
pneumonia. At 3-5 ppm,
irritating. At 25 ppm,
dangerous. At 50 ppm, rapidly
fatal .
Propane
1,800 1,000 20,000
Refrigerant/ NA
dispersant gases
(freon)
Asphyxiant.
-------
Table E-l (continued)
Inhalation exposure
Regulations Recommendations
Chemical
OSHA PEL (TWA)
mg/m^ ppm
NIOSH IDLH
ppm
ACGIH STEL
mg/nr
(15 min)
ppm
Animal toxicity
LC50
ppm
LD50
mg/kg
Hutagenicity/
carcinogenicity
Adverse
heal th
human
effects
Sulfur dioxide 13
100
10
Sense of suffocation (500 ppm).
Special hazard to eyes. Nose and
throat irritation. Eye
irritation, 20 ppm. 10,000 ppm
irritates moist areas of skin
immediately. Hay cause pulmonary
edema and respiratory paralysis.
400-500 ppm imnediately dangerous
to life.
i Sulfur trioxide NA
CO
Sulfuryl
chloride
NA
Tetrafluoro- NA
ethylene
Vinyl chloride
300
2.55(NIOSH) 1(NIOSH)
500 (rat)
C: studies indicate
positive from
occupational
exposure.
OSHA, ACGIH human
carcinogen
Hay be anesthetic. Handling of
uninhibited material has caused
circulatory and bone changes.
-------
Table E-l Footnotes
Source: NIH/EPA (1985) - Chemical Information System, OHHTADS Database.
NA - No information available for this chemical within the OHHTADS Database.
OSHA PEL (TWA) - OSHA permissible exposure limit, time-weighted average, for a normal 8-hour workday, 40-hour workweek (29 CFR 1910).
(Skin) - Can also be absorbed through the skin.
NIOSH IDLH - NIOSH recommended concentration, immediately dangerous to life or health.
ACGIH STEL (15 min) - The American Conference of Governmental Industrial Hygienists, short-term exposure limit, maximum concentration for an exposure period
up to 15 minutes continuously.
(NIOSH) - NIOSH ceiling (maximum concentration) for an exposure period up to 15 minutes continuously. Given when no ACGIH STEL is available.
(ceiling) - ACGIH ceiling (maximum concentration) for an exposure period up to 15 minutes continuously.
LC50 - Unless otherwise noted, lethal concentration by inhalation for 501 of the test animals in a 48-hour period.
LD50 - Unless otherwise noted, lethal dose (oral) for 501 of the test animals in a 48-hour period.
S - Skin application rather than oral exposure.
SC - Subcutaneous injection rather than oral exposure.
IP - Intraperitoneal injection rather than oral exposure.
M - mutagenicity.
C - carcinogenicity.
T - teratogenicity.
i
vo
-------
APPENDIX F
Glossary
F-l
-------
GLOSSARY
Cathodic Protection A method of protecting underground tanks from
external corrosion using electrical principles.
Exothermic
Flare
A chemical reaction that gives off heat.
A steel tower to permit burning of gases and
vapors.
Pyrolysis
A chemical reaction 1n which the reaction is
caused by the application of heat.
Reactor
A vessel or pipe in which a chemical reaction
takes place.
Vent Gas Scrubber
A device providing contact between upward flowing
gases and a downward flowing liquid for the
purpose of preventing release of the gases.
F-2
-------
VERSAR CENTER
------- |