[Federal Register: September 21, 1995 (Volume 60, Number 183)]
[Rules and Regulations ]
[Page 49047-49083]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
[DOCID:fr21se95-22]
[[Page 49047]]
_______________________________________________________________________
Part II
Department of Transportation
_______________________________________________________________________
Research and Special Programs Administration
_______________________________________________________________________
49 CFR Part 171, et al.
Crashworthiness Protection Requirements for Tank Cars; Detection and
Repair of Cracks, Pits, Corrosion, Lining Flaws, Thermal Protection
Flaws and Other Defects of Tank Car Tanks; Final Rule
[[Page 49048]]
DEPARTMENT OF TRANSPORTATION
Research and Special Programs Administration
49 CFR Parts 171, 172, 173, 179, and 180
[Docket Nos. HM-175A and HM-201; Amdt Nos. 171-137, 172-144, 173-245,
179-50, and 180-8]
RIN 2137-AB89 and 2137-AB40
Crashworthiness Protection Requirements for Tank Cars; Detection
and Repair of Cracks, Pits, Corrosion, Lining Flaws, Thermal Protection
Flaws and Other Defects of Tank Car Tanks
AGENCY: Research and Special Programs Administration (PHMSA), DOT.
ACTION: Final rule.
-----------------------------------------------------------------------
SUMMARY: PHMSA is amending the Hazardous Materials Regulations (HMR) to:
Require facilities that build, repair, and ensure the structural
integrity of tank cars, to develop and implement a quality assurance
program (QAP); allow the use of non-destructive testing (NDT)
techniques, in lieu of currently prescribed periodic hydrostatic
pressure tests, for fusion welded tank cars; require thickness
measurements of tank cars; allow the continued use of tank cars, with
limited reduced shell thicknesses, for certain hazardous materials;
increase the frequency for inspection and testing of tank cars for
added safety; clarify tank car pretrip inspection requirements; expand
the use of thermal protection systems and head protection on tank cars
to include certain other high hazard materials; add new requirements
for bottom-discontinuity protection; require the use of protective
coatings on insulated tank cars; prohibit the use of self-energized
manways located below the liquid level of the tank; remove
``grandfather'' provisions allowing certain uses of tank cars; and
improve the puncture resistance of tank cars used for certain high
hazard materials, including those that are poisonous-by-inhalation
(PIH) and those determined by the Environmental Protection Agency (EPA)
to pose health and environmental risks.
These actions are being taken to enhance the safe transportation of
hazardous materials in tank cars. The intended effects of these actions
are to improve the crashworthiness of tank cars and to increase the
probability of detecting critical tank car defects.
DATES: Effective date. The effective date of these amendments is July
1, 1996.
Compliance date. Voluntary compliance with the regulations, as
amended herein, is authorized November 1, 1995.
Incorporation by reference date. The incorporation by reference of
certain publications listed in these amendments is approved by the
Director of the Federal Register as of July 1, 1996.
FOR FURTHER INFORMATION CONTACT: Ed Pritchard (telephone 202-366-0509)
and James H. Rader (telephone 202-366-0510), Hazardous Materials
Division; or Thomas A. Phemister (telephone 202-366-0635), Office of
Chief Counsel, Federal Railroad Administration, 400 Seventh Street,
S.W., Washington, D.C. 20590-0001.
SUPPLEMENTARY INFORMATION:
I. Introduction
This final rule consolidates two related notices of proposed
rulemaking published under Docket HM-175A [58 FR 52574, October 8,
1993] and Docket No. HM-201 [58 FR 48485 September 16, 1993], that
address the safe performance of tank cars used to transport hazardous
materials. PHMSA believes that, by consolidating these two rulemakings,
changes to sections that are affected by both rules will be more easily
understood by readers. This preamble discusses separately, for each
rulemaking, the notices of rulemaking and comments received in response
to these notices. A consolidated ``Review by Section Summary''
summarizes the changes made under this final rule.
The Federal Railroad Administration (FRA) has enforcement authority
for tank cars and rail transportation. FRA developed these rulemakings
jointly with PHMSA.
II. Docket HM-175A--Crashworthiness Protection Requirements for
Tank Cars
A. Background
Based on research and on the FRA's continuing review of serious
accidents, involving the transportation of hazardous materials in tank
cars in the United States and Canada, PHMSA issued a number of
regulations to improve the survivability of tank cars in
accidents.<SUP>1 In these rulemakings, PHMSA required the installation
of a tank-head puncture-resistance system (head protection), a coupler
vertical restraint system (shelf couplers), insulation, and a thermal
protection system for certain high-risk hazardous material ladings. The
difference between a ``thermal protection system'' and ``insulation''
is that a ``thermal protection system'' protects a tank from a pool or
torch-fire environment. In contrast, ``insulation'' protects the lading
inside the tank from ambient, temperature differentials, much like home
insulation. The record shows that these systems, working in
combination, have greatly reduced the potential harm to human health
and the environment when tank cars are involved in accidents.
\1\The discussions in the following rulemakings provide greater
detail about each of these safety system requirements: Interlocking
Couplers and Restrictions of Capacity of Tank Cars, Docket HM-38, 35
FR 14215 (September 9, 1970); Tank Car Tank Head Protection, Docket
HM-109, 41 FR 21475 (May 26, 1976); Shippers; Specifications for
Pressure Tank Cars, Docket HM-144, 42 FR 46306 (September 15, 1977);
Shippers, Specifications for Tank Cars, Docket HM-174, 49 FR 3473,
(January 27, 1984); Specifications for Railroad Tank Cars Used to
Transport Hazardous Materials, Docket HM-175, 49 FR 3468 (January
27, 1984); Transportation of Hazardous Materials, Miscellaneous
Amendments, Docket HM-166W, 54 FR 38790 (September 20, 1989); and
Performance-Oriented Packaging; Changes to Classification, Hazard
Communication, Packaging and Handling Requirements Based on UN
Standards and Agency Initiative, Docket HM-181, 55 FR 52402
(December 21, 1990).
---------------------------------------------------------------------------
On October 8, 1993, PHMSA published a notice of proposed rulemaking
(NPRM) under Docket HM-175A (58 FR 52574) based, in part, on
recommendations issued by the National Transportation Safety Board
(NTSB) and comments received in response to an advance notice of
proposal rulemaking published on May 15, 1990 [55 FR 20242], and a
supplemental advance notice of proposed rulemaking published on August
29, 1990 [55 FR 35327]. The NPRM solicited comments on the costs and
safety benefits that would be derived should the HMR be amended in the
following areas: (1) Tank-head protection; (2) thermal protection; (3)
self-energized manways below the tank liquid level; (4) non-pressure
tank cars for PIH materials; (5) grandfather provisions allowing use of
certain tank cars conforming to former standards; (6) bottom
discontinuity protection on tank cars; (7) protective coatings on
insulated tanks; and (8) tank cars of limited and designated
specifications, with greater protection in accidents for transporting
materials determined by EPA to pose health and environmental risks.
On January 6, 1994, FRA and PHMSA held a public hearing to solicit
information to assist in deciding what actions, if any, should be taken
to improve the survivability of tank cars involved in hazardous
materials accidents. Twelve persons made presentations at the public
hearing. In addition, PHMSA received 37 written comments in response to
the NPRM from representatives of trade associations and the various
industries that own, lease, transport, or use tank
[[Page 49049]]
cars. All written and oral comments were given full consideration.
B. Tank Cars Transporting ``Thermally Reactive Materials'' (Materials
That May Violently Decompose or Polymerize When Exposed to Fire)
In the NPRM, PHMSA proposed to require the use of full-head
protection and thermal protection on tank cars used for certain
materials termed, ``thermally reactive.'' These materials, listed by
name, are thought by many to be capable of a violent decomposition or
polymerization reaction when exposed to fire. For these materials, the
critical temperature for the tank car, and its thermally reactive
lading, may be the heat at which the material undergoes decomposition
or polymerization--as opposed to the temperature at which the steel of
the tank becomes so plastic, it begins to lose tensile strength.
The proposal was based on several accidents involving thermally
reactive materials. For example, on August 2, 1988, at 9:00 p.m., in
Brazoria, Texas, 13 cars of a Union Pacific freight train
derailed.<SUP>2 Seven of the derailed tank cars contained acetaldehyde,
and none of these tank cars had a thermal protection system, which was
not required. Two acetaldehyde tank cars sustained coupler punctures
and released their contents, which ignited. The resulting fire engulfed
four other acetaldehyde tank cars, and each of them had a total failure
or rupture of the tank shell within 5 to 10 minutes after the
derailment. Witnesses reported 3-4 explosions between 9:05 p.m. and
9:10 p.m.
\2\Union Pacific Derailment at Brazoria, Texas, FRA Accident
Investigation No. 137-88, Railroad Report No. 0888H0200, August 2,
1988.
---------------------------------------------------------------------------
In another accident, NTSB found that the puncture of a tank car
containing hydrogen peroxide resulted in a release of lading and, when
the hydrogen peroxide combined with contaminants on the ground, a
chemical reaction occurred causing a fire.<SUP>3 The fire heated and
ignited nearby polyethylene pellets, causing an explosion of the
hydrogen peroxide tank car and releasing a force equivalent to an
explosion of 10 tons of TNT (trinitrotoluene).
\3\Collision and Derailment of Montana Rail Link Freight Train
with Locomotive Units and Hazardous Materials Release, Helena,
Montana, February 2, 1989, National Transportation Safety Board
Report NTSB/RAR-89/05, National Transportation Safety Board,
Washington, D.C.
---------------------------------------------------------------------------
Most commenters opposed the requirement for full-head protection or
thermal protection on tank cars used for thermally reactive materials.
In clarifying its comments on the NPRM, the Association of American
Railroads (AAR) stated that full-head protection is not necessary for
tank cars used for these materials, unless the materials pose another
hazard that warrants such protection. Other commenters, such as
American Petroleum Institute (API), Chemical Manufacturers Association
(CMA), and the Compressed Gas Association, Inc. (CGA), suggested that
PHMSA open a new ANPRM to address these materials. A commenter stated--
the creation of this category has ramifications that reach far
beyond this particular rulemaking, which deals with one mode of
transportation (rail) and one type of packaging (tank cars). We are
concerned with the likelihood that, in the future, the Department
will expand the regulation of TRMs to affect other modes of
transportation and types of packaging.
Other commenters objected to the proposal to identify by list,
rather than by definition, certain existing hazardous materials that
would be designated ``thermally reactive.'' CMA challenged the
placement of several chemicals on the list, such as ``styrene, monomer
inhibited,'' ``vinyl toluene,'' ``vinylidene chloride,'' ``sulfur
trioxide,'' and ``hydrogen peroxide.'' CMA further stated that--
[s]tyrene, for example, is flammable and can polymerize in an
accident but solidifies causing little or no harm to the
environment. For hydrogen peroxide tank cars, the proposed rule
would create a safety hazard by requiring thermal protection.
Another commenter stated that ``[s]ome of the materials on the list
react violently when exposed to heat differentials and may decompose
with explosive force * * * Other materials, however, decompose through
polymerization into substances of relatively little hazard.'' The
commenter further explained that the key to the polymerization of
styrene is the absence of the inhibitor. Styrene is typically shipped
with inhibitor concentrations great enough to cover fairly lengthy,
unexpected delays in transportation. If a tank car of styrene is
exposed to extreme external heat, disregarding its flammable nature,
the inhibitor will dissipate rapidly as the temperature of the material
rises above 125 deg.F., which will allow the polymerization process to
begin. As a result of the polymerization, the internal heat of the
product will increase, and, with increasing temperature, the process
will accelerate.
Several commenters opposed the requirement for a thermal protection
system on tank cars used to transport ``hydrogen peroxide.'' One of the
commenters stated that hydrogen peroxide does not polymerize or burn,
and the products of decomposition--water and oxygen--are not toxic.
Two commenters, Eka Nobel and FMC Corporation (FMC), furnished
independent analyses of the fire effects on tank cars containing
``hydrogen peroxide.'' Eka Nobel contracted with the IIT Research
Institute (IITRI), which used FRA's computer model to analyze the fire
effects on a tank car containing hydrogen peroxide.<SUP>4 The results
of IITRI's analysis indicate that a tank car constructed from stainless
steel will meet the thermal protection criterion for withstanding the
effects of a pool fire.
\4\``Temperatures, Pressures and Liquid Levels of Tank Cars
Engulfed in Fires,'' NTIS DOT/FRA/OR&D-84/08.11, (1984), Federal
Railroad Administration, Washington, DC.
---------------------------------------------------------------------------
FMC furnished a detailed, mathematical heat transfer model using a
correlation contained in a National Fire Protection Association (NFPA)
publication, ``NFPA Pamphlet No. 30.'' FMC stated that for materials
that decompose exothermically, such as hydrogen peroxide, thermal
stability requires that the heat losses to the surroundings balance the
heat generated by the decomposition. Failure to remove the heat of
reaction could lead to runaway decomposition, and if the increased
pressure exceeds the burst pressure of the tank, the tank will fail.
Furthermore, heat input causes oxygen generation from thermal
decomposition of peroxide and vapor generation, by boiling off the
water-peroxide mixture. FMC further stated that because water is more
volatile than peroxide, the hydrogen peroxide concentration in the tank
will increase (although this may be compensated by water formation and
peroxide loss from thermal decomposition). If the peroxide
concentration reaches 74 percent by weight, the vapors in equilibrium
with the liquid (40 percent by weight of peroxide) can detonate, if
ignited, causing the tank car to fail.
The results of FMC's mathematical heat transfer model show that
tank cars containing hydrogen peroxide (having no less than a 7-percent
outage) will not fail and such tank cars will meet the thermal
protection criterion in Sec. 179.18 of this final rule for withstanding
the effects of a pool-fire. Readers who are interested in a detailed
discussion of Eka Nobel or FMC's fire studies on tank cars containing
hydrogen peroxide, should refer to the comments filed in the PHMSA
Dockets Unit.
Many commenters suggested a performance-based definition as a means
to ensure the proper identification and packaging of thermally reactive
materials, because, with increasing temperature, all materials will
reach a stability limit.
[[Page 49050]]
These commenters suggested a performance-based definition that would
include the polymerization potential; the rate of the chemical reaction
(reaction kinetics); any highly exothermic reaction; the formation of
gases, vapors, or fumes in a quantity sufficient to present a danger to
human health and the environment; and any reactive by-products that
could lead to over-pressurization of the tank. Commenters stated that a
performance-based definition was the best way to ensure that the proper
packaging requirements are attached to the appropriate hazardous
materials.
As evidenced from the comments, there is no single agreement on the
best approach to identify these materials, nor to ensure the proper
packaging requirements are assigned to these materials. Because of the
multiplicity of these yet unresolved issues, the packaging requirements
proposed in the NPRM for thermally reactive materials have not been
adopted in this final rule.
C. Tank-Head Protection
In the NPRM, PHMSA proposed several changes relating to tank-head
protection. The proposal would require tank-head protection on tank
cars, used for all Class 2 materials and for tank cars constructed from
aluminum or nickel plate, when used to transport a hazardous material.
PHMSA included Division 2.2 in its proposal to reduce the violent
rupture hazard and the asphyxiation potential to railroad workers or
bystanders exposed to the product if these tank cars are punctured. The
proposal to require full-head protection for tank cars constructed from
aluminum or nickel plate is based on the vulnerability of the tank head
to a puncture. The top-half of the tank head is vulnerable to puncture
in a derailment. Existing tank cars with half-head protection were
excluded, based on PHMSA and FRA's regulatory analysis discussed later
in this preamble. Consistent with these proposed changes, PHMSA also
proposed to eliminate a grandfather provision, in place since 1984,
following publication of a final rule under Docket HM-175, that permits
certain tank cars, with a capacity of less than 70 kiloliters (kl;
18,500 gallons), to continue in service without head protection.
PHMSA first introduced tank-head protection requirements after a
series of railroad accidents in the late 1960s and early 1970s
involving head punctures of tank cars (39 FR 27572 and 41 FR 21475).
The requirements of, and criteria for, head protection were based on
tests performed by FRA, the AAR, and the Railway Progress Institute
(RPI) Tank Car Safety Research and Test Project in the early 1970s. In
summary, these tests showed that head punctures, caused by over-speed
impacts in railroad classification yards, generally occurred at speeds
above 12 mph and often happened when a loaded tank car struck a
standing empty tank car, causing the empty car to ``jump'' and ram its
coupler into the head of the oncoming tank. A recent informal staff
analysis of data on main-line accidents showed that objects, such as
broken rails and couplers, may penetrate the top half of the tank head,
indicating that head protection is essential, even though not 100
percent effective, in a train derailment.
The NPRM referenced the recent FRA research on puncture resistance,
which shows that puncture resistance is strongly influenced by impact
location, head and jacket thickness, and insulation thickness.<SUP>5
Stated differently, research demonstrates that puncture resistance is
an inter-related function of head thickness, insulation thickness, and
jacket thickness, and that the concept of ``head protection'' must
include more than just traditional ``head shields.'' Based on the
results of this research, FRA expects that certain tank cars may meet
the 29 kilometers per hour (18-mph) threshold for puncture-resistance,
prescribed in Sec. 179.16 of this final rule, without further
modification.
\5\Coltman, M., & Hazel, M., Jr., Chlorine Tank Car Puncture
Resistance Evaluation (1992), Federal Railroad Administration,
Washington, DC (NTIS DOT/FRA/ORD-92/11).
---------------------------------------------------------------------------
Tank cars currently equipped with half-head protection. Most
commenters agreed that there is no need to require full-head protection
on existing tank cars having only half-head protection.
In comments filed in this docket, NTSB stated that the NPRM
addressed many of their concerns, but noted the proposal failed to
require existing tank cars used to transport Division 2.1 (flammable
gas) materials, or other materials with extreme hazards, to be modified
with full-head protection. Thus, these materials could be transported
indefinitely in tank cars without full-head protection modifications.
While we appreciate the concerns of NTSB, we are not able to
establish a positive benefit/cost ratio by requiring modification of
the existing tank car fleet, primarily because the half-head protection
on existing cars is already about 95-percent effective. It is not
credible to argue that greater safety gains are realized by mandating
safety improvements on tank cars that currently have a 95-percent
effective protection system, than by requiring improvements on tank
cars without a head-protection system. The regulatory evaluation
considered both approaches, with emphasis being placed on choosing the
alternative offering maximum potential benefit to society, while
imposing the least net cost. Based on the regulatory evaluation, this
final rule does not require that existing half-head protection be
removed and replaced with full-head protection.
Head protection systems for existing tank cars with capacities less
than 70 kl (18,500 gallons). PHMSA received diverse comments in response
to this proposal in the NPRM. One commenter agreed that class DOT 105
tank cars having capacities less than 70 kl (18,500 gallons) and
transporting Division 2.1, 2.2, and 2.3 materials, should have full-
head protection, unless already equipped with half-head protection.
CMA supported the proposal to require full-head protection on newly
built class DOT 105A tank cars, regardless of tank capacity, when used
to transport a Division 2.1 or 2.3 material. The Reebie Associates
report, submitted as part of CMA's comments, assumed that all tank cars
would require head protection, except those that have a tank test
pressure of 41.4 Bar (600 pounds per square inch [psi]).
The Chlorine Institute agreed that head protection systems are now
warranted for the transportation of chlorine, but recognized, based on
FRA research and the accident history, that many tank cars currently
used to transport chlorine meet the performance standard by virtue of a
thick tank-head and a tank jacket.
NTSB commented that PHMSA should require tank-head protection,
within 5 years, for all class 105 tank cars having capacities of less
than 70 kl (18,500 gallons) when used to transport a Division 2.1
(flammable gas) material as proposed in Option B of the NPRM.
RPI commented that, except for the nominal 41 kl (11,000-gallon)
capacity tank cars, existing tank cars of less than 70 kl (18,500-
gallon) capacity, transporting Division 2.1 materials or anhydrous
ammonia, should have head-protection, but only half-head protection.
RPI further commented that PHMSA should exclude tank cars having a
nominal capacity of 41 kl (11,000 gallons) from any head protection
modification program, because most tank cars in this category are near
or exceed 30 years of age; consequently, the economic life of the tank
is nearing an end.
PHMSA and FRA believe that there is no longer a justification for
excluding tank cars having a capacity less than 70
[[Page 49051]]
kl (18,500 gallons) from the modification requirements. While CMA's
report is not so optimistic on the use of DOT 105A500W specification
tank cars, PHMSA and FRA believe that most of these tank cars will meet
the performance standard by virtue of their increased head thickness,
insulation, and metal jacket. Because of the small number of tank cars
in this category, and the small incremental cost to make such head
protection modifications for those tank cars that do not otherwise meet
the performance standard mandated by this rule, in this final rule PHMSA
is removing the 70 kl (18,500-gallon) exception for existing tank cars
in current Secs. 173.314(c) and 173.323(c)(1).
Further, while most commenters supported the 10-year modification
program for existing tank cars, we agree with NTSB, that when these
tank cars are used to transport Division 2.1 materials, a 5-year
modification program (as proposed in Option B of the NPRM) will ensure
that those cars presenting the greatest risk are modified first.
Tank cars transporting materials in Division 2.2. A commenter
stated that the proposal to require full-head protection for Division
2.2 gases is sound and should be finalized. Several other commenters
disagreed with the proposal to require full-head protection for
Division 2.2 materials. The Reebie Associates report, submitted by CMA,
identified 467 Class 2 materials affected by the proposed rule, 11 of
which are Division 2.2 materials. The report shows that shippers used
1,448 tank cars in 1992 to transport these Division 2.2 materials, as
follows:
------------------------------------------------------------------------
Commodity Population
------------------------------------------------------------------------
Argon, refrigerated liquid................................ 2
Ammonia solutions......................................... 28
Bromotrifluoromethane..................................... 1
Carbon dioxide, refrigerated liquid....................... 1,016
Chlorodifluoromethane..................................... 145
Chlorotetrafluoroethane................................... 26
Chloropentafluoroethane................................... 37
Dichlorotetrafluoroethane................................. 164
Fertilizer, ammoniating solutions......................... 4
Trifluoromethane.......................................... 1
Xenon, refrigerated liquid................................ 24
-------------
Total................................................. 1,448
------------------------------------------------------------------------
CGA opposed the full-head protection requirement for tank cars
transporting carbon dioxide. CGA referenced the testimony presented by
RPI at the January 6, 1994 public hearing concerning recent head impact
tests that verified the adequacy of the current head protection system
on DOT 105A500W specification tank cars.
With regard to CMA's and CGA's comments, PHMSA and FRA believe that
most tank cars used for ``carbon dioxide, refrigerated liquid,'' meet
the performance standard for head protection by virtue of their tank
head thickness and metal jacket. Tank cars used for ``argon,
refrigerated liquid,'' and ``xenon, refrigerated liquid,'' also meet
the head performance standard by virtue of the authorized class DOT 113
tank car specification. These tank cars must have a minimum outer
jacket tank head of not less than \1/2\-inch thick steel. See
Sec. 179.400-8(d). A total of 1,042 tank cars, or 72 percent of the
total Division 2.2 tank car population, are used to transport these
three commodities.
A commenter opposed tank-head protection for Division 2.2 materials
stating, ``heavy walled tank and protective housing for the fittings is
adequate for the transportation environment.'' The commenter also
provided an in-house report using a computer model that claims the
asphyxiation potential from a punctured Division 2.2 refrigerant gas
tank car to be very low.'' Another commenter opposed applying head
protection to tank cars transporting Division 2.2 refrigerant gases.
This commenter stated that, in the past, DOT had judged a material
based on its hazards under normal conditions of transport, and that in
this rulemaking, DOT was over-assessing the potential for harm in a
low-probability event. RPI supported full-head protection on new,
insulated tank cars transporting Class 2 materials, but it opposed
full-head protection for new non-insulated tank cars or for existing
tank cars transporting these materials.
We believe that even though the probability of an event occurring
with these materials is low, safety concerns still need to be
addressed, because the event may lead to high consequences, such as a
large scale evacuation or an oxygen deficient atmosphere in a
concentrated populated area. Taking the safety steps adopted in this
final rule will mitigate these hazards.
We also believe that the transportation risks associated with
Division 2.2 gases are sufficient to require full-head protection for
new tank cars, and for existing tank cars without head protection, when
used to transport Division 2.2 materials. As noted above, this rule
does not require existing tank cars equipped with half-head protection
to be modified with full-head protection. PHMSA and FRA are aware of
industry concerns that the attachment of full-head protection to non-
jacketed cars is a feature not yet proven by long service. Similar
arguments were raised when head protection was first required almost
two decades ago [HM-144; 42 FR 46306, September 15, 1977]. FRA is aware
of companies with plans to attach full-head protection to their non-
jacketed tank cars. As discussed later in this preamble, a phased-in
10-year modification program is provided for existing tank cars.
Existing tank cars without head protection. Most commenters to the
NPRM supported the need to modify existing tank cars to meet the
current safety requirements. One commenter supported the need to modify
existing tank cars constructed from aluminum plate with half-head
protection, but believed full-head protection should be required when a
proven full-head shield design is available. Another commenter
suggested that DOT should specifically recognize that tank cars used in
``chlorine'' service meet the performance requirements for head
protection and that DOT should not require any additional head
protection for these tank cars.
As stated in the NPRM, the benefits of head protection are real,
predictable, and quantifiable. PHMSA disagrees with commenters who state
that full-head protection is not warranted. Where earlier rules
required head protection on tank cars, it was a matter of recognizing
the highest priority needs first. The question is not one of demanding
low-priority, safety benefits, but the need to expand the safety base
of hazardous materials transportation in tank cars. Further, the small
additional cost of installing full-head protection on cars that now
have no head protection system, as compared with adding only half-head
protection, is justified on the basis of increased safety (see Chapter
V of the Economic Impact Assessment and Regulatory Flexibility
Analysis). In this final rule, PHMSA requires existing tank cars that
currently have no head protection, to have full-head protection
installed when used to transport a Class 2 material. As explained
below, PHMSA is also requiring full-head protection for tank cars
constructed from aluminum or nickel plate when used to transport
hazardous material.
Tank cars constructed from aluminum and from nickel plate.
Commenters supported the need for head protection on tank cars
constructed from aluminum or nickel plate, but not the full-head
protection requirement proposed in the NPRM. Most commenters stated
that there is no design available for the securement of full-head
protection on tank cars without metal jackets.
[[Page 49052]]
One commenter stated that his company's new aluminum tank cars,
constructed with greater tank shell and head dimensions than standard
tank cars, offer greater protection without head protection. The
commenter stated that further testing should be done and suggested that
PHMSA and FRA submit more evidence to support the need for this
requirement.
CMA supported requiring half-head protection for new tank cars
constructed from aluminum or nickel plate, and requiring half-head
protection for existing tank cars for certain hazardous materials.
Several commenters requested that PHMSA consider the characteristics of
an individual Division 2.2 material, and that materials not subject to
the HMR, and low hazard materials should be excluded.
We realize that the use of good engineering practice and design
specifications are needed to secure full-head protection to tank cars
without metal jackets. Although there is no service experience for a
full-head protection design on non-insulated tank cars, such designs
are certainly not unreachable within the years ahead. In rulemaking
proceedings under another docket [HM-144; 42 FR 46306, September 15,
1977] introducing half-head protection, commenters offered similar
arguments regarding head protection, for which solutions were later
found as a result of technological innovation. Currently, FRA is aware
of several companies that are nearing completion on their full-head
protection designs for aluminum and nickel tank cars. We, therefore,
believe that the introduction of this requirement will not adversely
affect industry. In this final rule, the use of full-head protection
for all tank cars constructed from aluminum or nickel plate is required
when used to transport a hazardous material. As discussed later in this
preamble, PHMSA has provided for a phased-in 10-year modification
program.
D. Thermal Protection Systems
In the NPRM, PHMSA proposed to require a thermal protection system
for a Class 2 material when a thermal analysis of the tank car and
lading shows that a release will occur other than through the safety
relief valve when the tank car is subjected to either a 100-minute pool
fire or a 30-minute torch fire. The current HMR require thermal
protection for Division 2.1 (flammable gas) materials (with limited car
capacity restrictions) and certain Division 2.3 (poison gas) materials.
PHMSA proposed to expand the thermal protection requirements to include
Division 2.2 materials because, as stated by AAR, ``[a]t a chemical
accident, there are generally two reasons for an evacuation, one is to
protect the public from any toxic, poisonous, or noxious vapors or
fumes generated by the product itself . . ., the second is to protect
the public from thermal ruptures and the container debris that may be
hurled from an incident site'' [Emergency Action Guides, p. VII]. PHMSA
also proposed to expand the thermal protection requirement to include
all Division 2.3 materials.
PHMSA began to require the application of a thermal protection
system on tank cars transporting Division 2.1 materials (flammable
gases) or ``ethylene oxide'' (Division 2.3) after a series of major
railroad accidents involving fires and ruptures of non-insulated
pressure tank cars. The design of and criteria for thermal protection
systems were based on tests performed by FRA at the U.S. Army
Ballistics Research Laboratory in White Sands, New Mexico, and at the
Transportation Test Center in Pueblo, Colorado. These tests revealed
that a 127.2 kl (33,600 gallon) non-protected tank car filled with
propane (Division 2.1) will rupture, with 40 percent of the lading
remaining in the tank car, within 24 minutes after exposure to a pool-
fire. Rupture occurs when the residual strength of the tank shell falls
below the force generated by the vapor pressure of the lading exerted
on the inside surface of the tank shell. Further testing by FRA
demonstrated that a tank car filled with propane and equipped with a
thermal protection system delayed the thermal rupture of the tank car
for 94.5 minutes, by maintaining the shell temperature low enough to
vent 98 percent of the lading through the safety relief valve. The
current performance standard, requiring exposure to a 100-minute pool
fire and a 30-minute torch fire, was chosen because it provides
emergency response personnel time to assess the accident and to
initiate remedial actions, such as evacuating an area.
Division 2.1 (flammable gas) and 2.3 (poisonous gas) materials:
Several commenters supported the need for a thermal protection system
on tank cars transporting Division 2.1 or 2.3 materials, regardless of
tank car capacity. The AAR and another commenter supported a thermal
protection system for all Class 2 materials, unless a shipper could
show that a release will not occur, other than through the safety
relief valve, when the tank and lading are subject to a fire. RPI also
concurred on the need for thermal protection for all Class 2 materials,
but, except for Division 2.1, but did not support the high-temperature
performance standard proposed in Sec. 179.18. RPI stated that most
insulation materials (e.g., 4 inches of glass-fiber insulation) are
adequate.
In this regard, PHMSA stated in the NPRM that many insulation
materials also provide good thermal protection. These insulation
materials, when analyzed with the tank and the lading, may show that
nothing further needs to be installed on the tank car to achieve
passage of the pool- and torch-fire performance tests. Research
sponsored by FRA on urethane-foam and glass-fiber insulation systems
show that urethane-foam insulation will pass the pool- and torch-fire
requirements and that glass-fiber insulation will also pass both tests,
provided the insulation is held in place with a plastic or wire scrim.
Owners of tank cars with either of these systems, or another comparable
system, may find that their thermal analysis of the tank car shows the
presence of sufficient thermal protection to meet the performance
standard. In this case, the tank car owner would have to verify only
that the insulation material installed on the tank car is capable of
passing the pool- and torch-fire verification or ``proof'' tests in
Appendix B to Part 179 of this final rule. Owners may find that a tank
car will pass the performance standard with only minor modifications,
such as applying a thermal protection system to the manway nozzle.
Also in the NPRM, PHMSA stated that, in 1981, a joint effort between
the Chlorine Institute and RPI-AAR Tank Car Safety Research and Test
Project resulted in the development of an insulation system to protect
a chlorine tank car involved in a fire. The insulation system developed
maintains back plate (inside surface of the tank car shell)
temperatures below 250.56 deg.C (483 deg.F). After reviewing the
thermal resistance capabilities of the insulation system used on
chlorine tank cars, PHMSA incorporated it into the HMR in 1987. Readers
should refer for more information to Docket HM-166U, entitled
``Transportation of Hazardous Materials; Miscellaneous Amendments'', 52
FR 13034, (April 20, 1987).
Division 2.2 (nonflammable gas) materials. As noted earlier in the
preamble discussion on tank-head protection for Division 2.2 materials,
CMA commented that there were 1,448 tank cars allocated to Division 2.2
materials that had not already been captured in another service, such
as PIH. Of those, ``argon, refrigerated liquid,'' ``carbon dioxide,
refrigerated liquid,'' and ``xenon, refrigerated liquid,'' represent
1,042 tank cars, or 72 percent. CMA further commented that
[[Page 49053]]
almost 100 percent of the total would need retrofitting and that the
overall economic impact of the new regulations on this group of tank
cars amounts to $26.0 million for retrofitting and $2.59 million for
higher lease rates and additional cars in the tenth year of the
implementation period.
With regard to the issues raised by CMA, this final rule does not
contain any new thermal protection requirements for ``argon,
refrigerated liquid,'' ``carbon dioxide, refrigerated liquid,'' or
``xenon, refrigerated liquid.'' Carbon dioxide is transported in DOT
105A500W tank cars equipped with two regulator valves, a reclosing
pressure-relief device, a frangible disc, and an insulation system with
good thermal performance (a thermal conductance of 0.03 British Thermal
Units [B.t.u.] per square foot per degree Fahrenheit differential).
Consequently, existing and new tank cars in carbon dioxide service have
sufficient thermal resistance when exposed to fire. Likewise, because
with argon and xenon, refrigerated liquids are packaged under the
exceptions for atmospheric gases in Sec. 173.320, this final rule does
not impose any new thermal protection requirements. This section
exempts cryogenic atmospheric gases from the packaging requirements
when the packagings are designed to maintain pressures below 1.74 Bar
(25.3 psi) under ambient temperature conditions.
Another commenter opposed the use of thermal protection for
Division 2.2 materials on the basis that the hazards they pose do not
equate to those of Division 2.1 and 2.3 materials. The commenter
further stated that the thermal protection requirements proposed for
Division 2.2 materials do not appear to be justified by the hazards
posed, because, in many cases, these materials dissipate naturally with
little risk to the surroundings.
A commenter, primarily addressing refrigerant gases, noted that an
analysis of each Division 2.2 material, to predict the behavior of a
tank car in a 100-minute pool-fire, seemed an unnecessary precaution
because the calculations, required by the current regulations, for
sizing safety relief valves accomplish the same purpose and meet this
same standard. PHMSA and FRA disagree with this commenter's position
that the current regulations for sizing safety relief valves accomplish
the same purpose as the proposed Division 2.2 thermal protection
performance standard. The current safety relief valve-sizing
requirements make several assumptions. First, the valve sizing formula
assumes the exposure factor, that portion of the tank car exposed to
fire (represented as A<SUP>0.82), is about one-fourth of the tank. The
pool-fire computer model in this final rule assumes total engulfment.
Second, the safety relief valve sizing formula assumes that flame
temperatures will reach approximately 650 deg.C (1,200 deg.F.). The
pool-fire standard assumes flame temperatures will reach 871 deg.C
(1,600 deg.F) for a pool-fire and 1,204 deg.C (2,200 deg.F) for a
torch fire at 40 miles per hour.<SUP>6 Third, the safety relief valve-
sizing formula does not take into consideration either an overturned
tank car venting liquid or a liquid-gas mixture (two phase flow) or the
diminished burst strength of the heated tank shell in the non-wetted
area, after prolonged fire exposure.
\6\The pool-fire computer model assumes an average heat flux
over the entire tank surface, equivalent to complete engulfment in a
fire, where the flame temperature is 815.5 deg.C (1,500 deg.F). If
a higher or lower flame temperature were assumed, the parametric
analyses in the computer model would not match the actual field test
data.
---------------------------------------------------------------------------
The Fertilizer Institute did not support the requirement for
thermal protection on tank cars transporting ``anhydrous ammonia''. It
stated that the likelihood of a fire-induced rupture of a tank car
carrying anhydrous ammonia has significantly decreased since 1980
because of added safety devices, safer placement in trains, and
improved emergency response procedures. Thus, there is little, if any,
increase to public safety by imposition of the proposed thermal
protection requirements on these tank cars.
While PHMSA and FRA agree with The Fertilizer Institute that the
safety record for tank cars transporting ``anhydrous ammonia'' is good,
these cars have a potential for violent rupture similar to compressed
gas tank cars, which received thermal protection many years ago. As The
Fertilizer Institute notes, the threat of a fire-induced violent
rupture of an anhydrous ammonia tank car is more than just a
theoretical potential. Since 1990, according to figures from the AAR,
``anhydrous ammonia'' has been the sixth highest volume hazardous
material transported by railroad.
AAR and two other commenters supported the need for thermal
protection for Class 2 materials, including Division 2.2. One of these
commenters stated: ``thermal protection systems are a good, simple idea
whose time has come. The purpose of the system is to prevent rupture of
the tank car in a fire with the release of its hazardous materials
contents to the environment. Uncontrolled release of almost any
hazardous material to the environment is objectionable whether due to
toxicity, flammability, or simply clean-up costs.'' This commenter
further stated that there can be little basis for exempting anhydrous
ammonia from the thermal protection requirements simply because it is
not likely to catch fire once released. Its PIH characteristic remains,
and the potential for rupturing in a non-insulated tank car is high.
Although not all commenters agree on the need for thermal
protection for Division 2.2 materials, in this final rule PHMSA requires
such a system if, after an analysis of the effects of a 100-minute pool
fire and a 30-minute torch fire, there will be a release of the tank
car lading other than through the safety relief valve. Because tank
cars may transport different ladings, and because changing ladings may
affect the whole system, owners or shippers may choose to perform a
``worst case'' analysis based on all the commodities the car is likely
to carry.<SUP>7
\7\Owners are reminded that 49 CFR 173.31(a)(4) limits the use
of tank cars to those commodities for which they are authorized.
Authorized (or approved) commodities are those listed on the
certificate of construction or an AAR R-1 form. (See the AAR
Specifications for Tank Cars Section 1.4.3.1 and Appendix R, Section
R4.04.)
---------------------------------------------------------------------------
Based on these comments and FRA's research, this final rule
requires the owner or the shipper of a Class 2 material, with the
exception of ``carbon dioxide, refrigerated liquid,'' ``chlorine,'' and
``nitrous oxide, refrigerated liquid'' as explained above, to perform
an analysis of the characteristics of the material and of the thermal
resistance capabilities of the tank car, taking into consideration the
safety relief valve start-to-discharge pressure setting and relief
capacity and all areas of the tank car that are not afforded protection
from fire (such as stub sills, bolsters, and protective housings).
Tank cars constructed from aluminum and nickel plate. Most
commenters said that the lading within a tank car constructed from
aluminum or nickel plate should determine the need for a thermal
protection system.
We agree. The NPRM proposed to require a thermal protection
analysis for aluminum and nickel plate cars carrying Class 2 materials.
Based on the comments received, we believe that all such tank cars will
need protection and that such protection is essential.
This final rule requires the owner of an aluminum or nickel plate
tank car used to transport a Class 2 material to perform an analysis of
the tank car in a 100-minute pool fire and in a 30-minute torch fire
using FRA's Tank Car Fire model. If the analysis shows that a release
of the lading from the tank car,
[[Page 49054]]
will occur, other than through the safety relief valve, a thermal
protection system will be required. This final rule adopts a 10-year
phase-in period for those existing tank cars required to have thermal
protection.
E. Shell Protection
For tank cars transporting of a material poisonous by inhalation
(PIH), PHMSA proposed that they have ``shell protection conforming to
Sec. 179.100-4.'' That is, the optional use of an insulated DOT 105S
tank car or a non-insulated, but thermally protected, DOT 112J or 114J
tank car having a metal jacket. Although PHMSA used the term ``shell
protection'' to identify these systems, the intent of the NPRM was to
require tank cars transporting a PIH gas (Division 2.3) to conform to
the same requirements as tank cars transporting a PIH liquid. For a
complete discussion, see Performance-Oriented Packaging Standards;
Miscellaneous Amendments, Docket HM-181F, 58 FR 50224 (September 24,
1993). In the final rule issued under that docket, PHMSA authorized the
optional use of an insulated DOT 105S tank car or a non-insulated, but
thermally protected, DOT 112J or 114J tank car for poisonous liquids
having a PIH hazard.
In its comments to the NPRM, one commenter supported the need for
shell protection for PIH materials. Another commenter suggested that,
in lieu of a metal jacket, PHMSA should establish a performance
standard, as with thermal and head protection. Until a performance
standard is established, shell-protection resistance should be
equivalent to a tank car having a tank test pressure of 20.7 Bar (300
psi) constructed from carbon steel and with a 1/8-inch carbon steel
jacket. The commenter stated that the shell-puncture resistance should
be based on either a total metal thickness, or an approved calculation.
We agree with this commenter that a performance-based standard for
shell-puncture resistance may have merit over specification-based
standard adopted in this final rule. However, such performance based
standards have not been proposed.
Another commenter opposed the use of a metal jacket on pressure
tank cars transporting a PIH material on the basis that the FRA's
proposal did not support the conclusion that jacketing improves
puncture resistance. The commenter further questioned the use of a tank
jacket over thicker tank shells, since ``jackets provide thermal not
puncture protection.''
In response to similar remarks, PHMSA discussed in the NPRM a 1987
RPI report on the vulnerability of pressure tank car shells to
puncture.<SUP>8 RPI found that shelf couplers, hardboard insulation
(cork), increased shell thickness, thermal protection, small tank car
size and increased jacket thickness proved effective towards reducing
the frequency of shell punctures. The RPI report summarizes a 20\1/2\-
year history of accident data on shell punctures of pressure tank cars
and concludes that the 11-gauge steel jacket provides a measure of
shell protection. In addition to RPI's report, FRA also found, in a
research contract awarded to the AAR, that puncture resistance is
strongly influenced by impact location, by head and jacket thickness
and by insulation thickness.<SUP>9
\8\Phillips, E.A., Review of Pressure Car Shell Puncture
Vulnerability, RA-09-6-52, (1987), AAR-RPI Railway Tank Car Safety
Research and Test Project, AAR Technical Center, Chicago, Illinois.
\9\[Coltman, M., & Hazel, M., Jr., Chlorine Tank Car Puncture
Resistance Evaluation, (1992) Federal Railroad Administration,
Washington, D.C. (NTIS DOT/FRA/ORD-92/11).
---------------------------------------------------------------------------
PHMSA explained earlier, in Docket HM-181, that the purpose of a
metal jacket is to provide ``both accident damage and fire protection''
for certain [liquid] PIH materials.<SUP>10 This final rule expands that
philosophy to all PIH materials [including compressed gases] and
authorizes the use of an insulated class DOT 105S tank car or a non-
insulated, but thermally protected, class DOT 112J or 114J tank car.
\10\See the final rule on Performance-Oriented Packaging
Standards; Miscellaneous Amendments, Docket HM-181F, 58 FR 50224
(September 24, 1993), and the NPRM, 58 FR 37612 (July 12, 1993).
---------------------------------------------------------------------------
F. Self-Energized Manways Located Below the Liquid Level of the Lading
PHMSA proposed in the NPRM to prohibit the use on tank cars of a
self-energized manway located below the liquid level of the lading. The
proposal was based on a September 8, 1987 railroad yard incident in New
Orleans, Louisiana.<SUP>11 In this incident, a tank car equipped with a
self-energized bottom manway and loaded with butadiene developed a leak
and caught fire. At one point during the incident, the flames were
large enough that both spans of a bridge on Interstate 10 were
engulfed. After the investigation, NTSB concluded that ``it is unlikely
that a hazardous material leak through a bottom manway during
transportation could be stopped.'' NTSB urged FRA to prohibit the
transportation of tank cars that have a manway opening located below
the liquid level of the lading in hazardous materials service. Because
the design of bottom manways depends in part on the weight of the
product and the pressure in the tank to make the seal fully effective,
this type of closure system becomes vulnerable to releasing product
when the lading is displaced within the tank. Therefore, we agree with
NTSB's conclusion.
\11\Butadiene Release and Fire from GATX 55996 at the CSX
Terminal Junction Interchange, New Orleans, Louisiana, September 8,
1987, National Transportation Safety Board Report NTSB/HZM-88/01,
National Transportation Safety Board, Washington, D.C.
---------------------------------------------------------------------------
In its comments to the NPRM, the AAR, RPI, and several other
commenters supported the proposal to remove self-energized manways
located below the liquid level of the lading. A commenter stated that
their design incorporates an externally elliptically shaped ring clamp
which is bolted to the manway closure plate with numerous closely-
spaced studs around the circumference of the ring. This commenter holds
two DOT exemptions (DOT-E 5493 and DOT-E 6117) to operate tanks cars in
hydrogen sulphide service with this design. PHMSA and FRA believe that
this design is certainly preferable to that used on the car that leaked
and burned in New Orleans and is similar to a more conventional
external flange, however, we believe this design still remains a
potential source of leaks since it is located below the liquid level of
the lading. Based on these reasons, PHMSA will grant the exemption
holder a reasonable amount of time to phase out the use of these tank
cars.
While some commenters agreed with a 2-year phase out program of
self-energized manways, NTSB stated that PHMSA should immediately
prohibit such manways, and the AAR suggested a one-year phase-out
program.
Based on these comments, this final rule prohibits the construction
of new tank cars having an internal self-energized manway located below
the liquid level of the lading. This prohibition is added in
Sec. 179.103-5. Based on NTSB's comments, compliance with this
provision is required beginning on the effective date of this final
rule.
G. Non-Pressure Tank Cars for Materials Poisonous by Inhalation
In the NPRM, PHMSA proposed to prohibit the use of non-pressure tank
cars (e.g., class DOT 111A) for materials poisonous by inhalation.
In a recent research report, FRA found that, in a single-car
national risk profile, the transportation of ethylene oxide in a DOT
111A100W4 tank car involves significantly greater risk than
transportation of the same material in a
[[Page 49055]]
DOT 105J500W tank car.<SUP>12 Characteristics and parameters evaluated
in this assessment included the toxicity, fire hazard, and explosion
hazard. In comments to the ANPRM, RPI reported that, during the time
period of 1965 through 1986, class DOT 111A tank cars involved in
accidents and damaged were slightly more than three times as likely to
lose lading as were class DOT 105 cars in similar situations.<SUP>13
\12\Raj, P.K., and Turner, C.K., Hazardous Materials
Transportation In Tank Cars/Analysis of Risks--Part 1, NTIS DOT/FRA/
ORD-92/34, (1993), Federal Railroad Administration, Washington D.C.
\13\Phillips, E.A., Analysis of Tank Cars Damaged in Accidents
1965 through 1986, RA-02-6-55, (1989), AAR-RPI Railway Tank Car
Safety Test and Research Project, AAR Technical Center, Chicago,
Illinois.
---------------------------------------------------------------------------
The Raj/Turner report amply demonstrates (and AAR/RPI Tank Car
Safety Test and Research Project data support) that it is
``improbable'' to assume that any single tank car (e.g., DOT 111A or
DOT 105) would be involved in an accident. However, based on FRA
accident data referenced earlier regarding DOT 111A and DOT 105 tank
cars, a significant number of such cars will be involved in accidents
during their service life.
Several commenters supported disallowing the use of non-pressure
tank cars for the transportation of PIH materials. Because of the
hazards associated with PIH materials and the performance superiority
of the so-called ``pressure'' tank cars for this service, PHMSA agrees
with the commenters. This final rule removes the class DOT 111A tank
car as an authorized packaging for Division 2.3 materials on the
effective date of this final rule.
H. Phasing Out of Various ``Grandfather'' Provisions
In the NPRM, PHMSA proposed to remove from the HMR several
grandfather provisions that affect tank cars. The grandfather
provisions allow tank cars built before a certain date to remain in
service without modification. As an example, in Sec. 173.314(c), Notes
23 and 24 allow the continued use of class DOT 105A tank cars for
certain compressed and flammable gases if they were built before
September 1, 1981, while tank cars built after that date must meet a
more stringent class DOT 105S or 105J standard.
NTSB stated, in a March 1, 1988 letter to PHMSA, that tank cars
failing to meet current minimum safety requirements should no longer be
used for transportation of hazardous material under grandfather
provisions. NTSB stated that these grandfather provision could result
in a reduced level of safety. The AAR also petitioned PHMSA to amend
Sec. 173.314(c) Note 30 (P-1138), stating that it does not provide any
assurance that tank cars with head protection will be used for PIH gas
service in the foreseeable future because companies will be able to use
tank cars without head protection for PIH compressed gas service for
the next 30 years. Other commenters agreed that the grandfather
provisions proposed for removal in the NPRM are no longer compatible
with the needs of safety.
Based on these comments, PHMSA is removing certain grandfather
provisions. In Sec. 171.102, special provision ``B63'' is removed to
disallow the use of DOT 105A100W, 111A100W4, 112A200W, and 114A340W
tank cars for ``ethyl chloride'' and ``ethyl methyl ether.'' Prior to
the issuance of Docket HM-181, these two materials were classed as
flammable liquids. Because these tank cars do not have head protection
or thermal protection systems, they do not provide an equivalent level
of safety compared to other tank cars used for Division 2.1 materials.
Also, special provision ``B63'' is removed from column 7 of the
Sec. 172.101 table entries for these two hazardous materials, thereby
prohibiting the use of non-protected tank cars.
Other changes are made to disallow the use of class DOT 111A non-
pressure tank cars for Class 2 (compressed gas) materials, such as
``ammonia solutions,'' ``ethylamine,'' ``ethyl chloride,'' and ``ethyl
methyl ether.'' This final rule also removes the DOT 111A100W4 car as a
packaging for ``ethylene oxide'' in Sec. 173.323(c)(1).
I. Bottom-Discontinuity Protection for Bottom Outlets
In the NPRM, PHMSA proposed to require bottom-discontinuity
protection (e.g., for bottom outlets) on tank cars. The proposed
requirements were intended to simply adopt the requirements published
by the AAR. In July of 1979, the AAR required bottom-discontinuity
protection for new tank car construction. Over a period of years, these
requirements were extended to existing tank cars on a priority schedule
determined by the nature of the commodity transported. The AAR's
program for bottom-discontinuity protection consists of either a metal
``skid'' protecting the portion of the bottom outlet that protrudes
beyond the shell or the machining of a ``breakage groove'' in the valve
assembly.
AAR, the Chlorine Institute, CMA, and several other commenters
supported the adoption of bottom-discontinuity protection for tank
cars, provided such protection was consistent with the AAR
requirements. API asked PHMSA to clarify the requirements for bottom-
discontinuity protection in this final rule. API and several other
commenters stated that the proposed rule would require the modification
of a number of tank cars, built before July 1, 1979, because most were
modified according to Appendix Y and not paragraphs E9.00 or E10.00 of
the AAR Specifications for Tank Cars. Appendix Y permits three levels
of protection for allowing the types of discontinuity: bottom outlets
that extend 1 inch or more; blind flanges and washouts that extend 2
and \5/8\ inches or more; and sumps and internally closed washouts that
extend 5 inches or more. Paragraphs E9.00 and E10.00 generally require
the protection of each valve and fitting from mechanical damage by the
tank, an another protective device, or the underframe.
Several other commenters stated that the proposed rule would also
require the modification of all existing tank cars, including those
that do not transport hazardous materials. The Sulphur Institute and
another commenter opposed the need to add bottom-discontinuity
protection to existing tank cars that transport sulfur, molten,
claiming that such protection has little practical benefit.
In the public hearing held on January 6, 1994, in Washington, D.C.,
FRA stated that it was not the Department's intention to require the
modification of previously modified tank cars, nor to require bottom-
discontinuity protection for tank cars that transport materials not
subject to the HMR.
In this final rule, PHMSA requires bottom-outlet protection that
conforms to paragraphs E9.00 and E10.00 of the AAR Specifications for
Tank Cars, M-1002, for all new tank cars equipped with bottom unloading
devices. Existing tank cars, without bottom-discontinuity protection,
used for the transportation of hazardous materials must conform to the
above paragraphs no later than 10 years after the effective date of
this final rule. Existing tank cars that conform to the bottom-
discontinuity protection requirements of Appendix Y of the AAR
Specifications for Tank Cars, M-1002 may continue in use after the
effective date of this final rule. This final rule does not require the
modification of existing tank cars that transport materials not subject
to the HMR.
J. Protective Coatings on Insulated Tank Cars
In the NPRM, PHMSA proposed use of protective coatings on the
exterior of a
[[Page 49056]]
tank car and the interior of a tank car jacket to retard rust or
corrosion. The proposal was in response to an AAR petition (P-1050) and
FRA's findings of severe corrosion or pitting on the outer surface of
the tank shell, or the inner surface of the tank jacket, of insulated
tank cars. It is not known whether the corrosion stems from the
physical properties of the insulation itself or whether the corrosion
develops when insulation becomes impregnated or contaminated with water
or a chemical from the atmosphere in which the tank car operates.
Research within the industry has led to the development of protective
coating materials.
Most commenters supported the proposal. One commenter stated that
acid-resistant protective coatings should be applied. The commenter
further stated that several manufacturing and repair shops are using
non-acid resistant latex coatings under polyurethane-foam insulations.
Another commenter suggested that the rule should be clarified to
exclude tanks or jackets manufactured with self-protective materials
such as stainless steel. Still another commenter asked PHMSA to consider
adopting a recommended practice for applying protective coatings on
tank cars that is now under development by the National Association of
Corrosion Engineers.
With regard to these comments, this final rule simply modifies
Secs. 179.100-4 and 179.200-4 by removing the exception for
polyurethane-foam insulations. Each of the current sections, and the
proposed rule, only require a protective coating on a carbon steel tank
shell and tank jacket. Concerning the comment on acid-resistant
coatings, PHMSA agrees that applied coatings should prevent any
corrosive attack to the tank metal. PHMSA and FRA will explore, in
cooperation with the AAR, CMA, and RPI, the need for and development of
acid-resistant coating standards.
NTSB commented that the proposed rule does not sufficiently address
the potential problem of existing tank cars. NTSB further noted that a
requirement to apply a protective coating on an existing tank car, only
when the jacket is removed to repair a tank, cannot ensure that
corrosion problems will be detected before the tank corrodes through
and releases its lading. NTSB stated that, at a minimum, tank cars
currently in use without protective coatings should be inspected
periodically for corrosion damage and tank cars found with corrosion
damage should be required to have appropriate repairs.
We agree with NTSB, and in this final rule require, under Docket
HM-201, new inspection intervals for materials that are corrosive to
the tank and a thickness performance measurement to ensure that the
tank shell is not corroded below the minimum shell thickness as
prescribed by the AAR. PHMSA and FRA believe that HM-201 is responsive
to NTSB's concerns.
In this final rule, PHMSA is requiring protective coatings for all
new tank cars and for existing tank cars when a repair to the tank car
requires the complete removal of the jacket, as suggested by
commenters.
K. Halogenated Organic Compounds (HOC)
To address a 1991 NTSB safety recommendation,<SUP>14 PHMSA proposed
in the NPRM to require the use of a tank car with enhanced puncture
resistance if the tank is used to transport one or more of the 100 HOC
compounds listed in 40 CFR Part 268 Appendix III. The Appendix III list
was developed by EPA pursuant to statute (42 U.S.C. 6924) in order to
prohibit the land disposal of certain compounds having a carbon-halogen
bond, and that have the potential to harm human health and the
environment (these EPA compounds were identified as the ``California
List'' under the statute [See also 40 CFR 268.32]).
\14\Transportation of Hazardous Materials by Rail, National
Transportation Safety Board Safety Study, Report NTSB/SS-91/01,
National Transportation Safety Board, Washington, D.C. (Safety
Recommendations R-91-11 and R-91-12).
---------------------------------------------------------------------------
Many commenters opposing regulation of the EPA compounds suggested
that PHMSA should continue to only regulate the compounds identified as
hazardous substances in Appendix A to Part 172. Commenters further
suggested that DOT should not consider the HOC concentration threshold
for those compounds. Several commenters stated that the regulatory
action proposed by PHMSA is unnecessary, that PHMSA should discontinue
its efforts to regulate these EPA compounds, and that PHMSA should not
consider extending enhanced tank car standards to those carrying the
more than 1,000 chemicals prohibited from land disposal.
API, CMA, and several other commenters suggested that the threshold
quantities for the EPA compounds are too low for transportation
purposes. The EPA threshold in 40 CFR 268.32 is 1,000 milligrams per
liter (mg/l) for liquids and 1,000 milligrams per kilogram (mg/kg) for
solids.
CMA furnished a benefit/cost analysis, prepared by Reebie
Associates, that used 1992 TRAIN II data; thereby updating the previous
work performed by AAR, CMA, and RPI addressed in the NPRM. The CMA
report shows that a total of 3,893 tank cars transported an EPA
compound. CMA's list and the number of tank cars used for such
compounds follows:
------------------------------------------------------------------------
AAR/CMA/RPI
Hazardous CMA's 1992 agreement Currently in
substances population (based on pressure Remaining
1988 data) tank cars
------------------------------------------------------------------------
1,1-
Dichloroethylen
e.............. 1 ............ ............ 1
1,2-
Dichloroethane. 236 236 ............ ............
1,2-
Dichloropropane 31 ............ ............ 31
Carbon
tetrachloride.. 312 312 ............ ............
Chlordane....... 10 ............ ............ 10
Chlorobenzene... 105 105 ............ ............
Chloroethane
(ethyl
chloride)...... 106 ............ 106 ............
Chloroform...... 227 227 ............ ............
Chloropropene... 7 ............ ............ 7
CIS 1,3-
dichloropropane 42 ............ ............ 42
Dichlorodifluoro
methane........ 224 ............ 224 ............
Dichlorofinrorom
ethane......... 2 ............ ............ 2
Dichlorofluorome
thane.......... 1 ............ ............ 1
Hexachlorocyclop
entadiene...... 8 ............ 8 ............
[[Page 49057]]
Methylene
chloride....... 2 2 ............ ............
o-
Dichlorobenzene 15 15 ............ ............
p-
Dichlorobenzene 82 82 ............ ............
Pentachloropheno
l.............. 10 ............ ............ 10
Tetrachloroethan
e.............. 13 13 ............ ............
Trichlorobenzene 6 ............ ............ 6
Trichloromonoflu
oromethane..... 4 ............ 4 ............
Vinyl chloride.. 2,449 ............ 2,449 ............
-------------------------------------------------------
Totals...... 3,893 992 2,791 110
------------------------------------------------------------------------
Commenters stated that PHMSA should not include materials that are
transported as a solid because, when released, the clean up of these
materials is easily achieved. This statement assumes that accidents
will not occur near lakes, rivers or streams, or that rainfall will not
carry solid residue to such water sources. It is PHMSA's and FRA's
experience that these types of accidents can occur as evidenced by the
metam sodium spill in the Sacramento River in California.
As discussed in the NPRM, these materials were also evaluated by
the AAR in an effort to identify materials that have the potential to
harm human health and the environment. The AAR analyzed the EPA
compounds using a computer model based on EPA and standard chemical
dispersion equations. The AAR model describes a method of evaluating
the relative environmental hazard of chemicals shipped in tank
cars.<SUP>15 In addition to the computer model, the AAR surveyed the
railroad industry for the clean-up costs associated with a spill of an
EPA compound. The AAR considered in their analysis: (1) Compounds that
were permitted in non-pressure tank cars by the DOT in 1988; (2) at
least one shipment of the compound reported to TRAIN II<SUP>16 in 1988;
(3) the compounds with an EPA reportable quantity (RQ) of less than
1,000 pounds in 1988; (4) the compounds prohibited from land disposal
by the EPA; and (5) the compounds suggested by the railroads' hazardous
materials or environmental staff, or the AAR contractor on the project.
The results of the 1988 survey identified 10 compounds, transported in
class DOT 111A tank cars at that time, that pose a potential threat to
human health and the environment. These compounds were:
\15\Lowenbach, William, A., Consequence Models of Hazardous
Materials Releases on Railroads, Association of American Railroads
(1989), Washington, D.C.
\16\The Association of American Railroads (AAR) data network,
Tele-Rail Automated Information Network (TRAIN II), collects
information on approximately 90 percent of the rail traffic
originating and terminating in the United States. Users of the
network can trace individual car movements or gather information on
a particular cargo moving by rail. The AAR uses the data to develop
statistical trends in both car movement and commodity flow.
Carbon tetrachloride
Chlorobenzene
Chloroform
Dichlorobenzene
Ethylene dibromide (1,2-Dibromomethane)
Ethylene dichloride (1,2-Dichloroethane)
Methyl chloroform (1,1,1-Trichloroethane)
Methylene chloride (Dichloromethane)
Perchloroethylene (Tetrachloroethene)
Trichloroethylene (Trichloroethene)
The results of AAR's analysis show that, within the last 10 years,
the release of these compounds in railroad accidents has resulted in
environmental clean-up costs exceeding $50 million. Even though these
materials accounted for less than one percent of the total volume of
hazardous materials, their releases accounted for 60 percent of all
railroad environmental clean-up costs. Based on the results of the
analysis, the AAR, CMA, and RPI have agreed that by January 1, 2000,
these 10 compounds should be transported only in a DOT 105S200W or a
DOT 112S200W tank car manufactured from AAR TC-128 normalized steel.
One of the 10 compounds, ``ethylene dibromide,'' is a compound that is
poisonous by inhalation (Zone B).
As shown by CMA, 3,893 tank cars were used to transport these ``EPA
compounds''; of that total, ``chloroethane,''
``dichlorodifluoromethane,'' ``hexachlorocyclopentadiene,''
``trichloromonofluoromethane,'' and ``vinyl chloride'' represent 2,791
tank cars, or 72 percent of the total. Because the packaging
authorizations for these compounds currently require the use of classes
DOT 105J, 112J, 112T, 114J, 114T tank cars, these tank cars currently
meet the proposed standard.
As noted above, AAR, CMA, and RPI agreed to use only DOT 105S200W
and 112S200W (or better) tank cars: These compounds are transported in
992 dedicated tank cars. CMA identified an additional 110 tank cars
that are used to transport an EPA compound, but lie outside of the
industry agreement. Because these 110 additional tank cars represent a
potential risk to human health and the environment, PHMSA believes it is
reasonable to require the same level of protection for the additional
tank cars identified by CMA, based on the 1992 TRAIN II data, as those
identified by the AAR, CMA, and RPI, based on the 1988 TRAIN II data.
It simply cannot be argued that the shipment of an EPA compound
identified after 1988 poses less risk in transportation than if the EPA
compound would have been identified by the AAR, CMA, and RPI in 1988.
Furthermore, because the AAR, CMA, and RPI agreement does not preclude
the use of a non-protected tank car in transportation by any one member
or nonmember of the agreement, such cars may still be used.
After considering each of the comments, PHMSA agrees it should only
regulate those EPA compounds listed in the HMR. After reviewing the 100
EPA compounds (listed in 40 CFR 268 Appendix III), PHMSA found that all
but 16 of the compounds are currently identified as a hazardous
substance. The 16 compounds are:
Bis(2-chloroethoxy)ethane
Bis(2-chloroethyl)ether
Bromomethane
2-Chloro-1,3-butadiene
3-Chloropropene
1,2-Dibromomethane
Dibromomethane
Hexachlorodibenzo-p-dioxins
Hexachlorodibenzofuran
Iodomethane
Methylene chloride
Pentachlorodibenzo-p-dioxins
[[Page 49058]]
Pentachlorodibenzofuran
Tetrachlorodibenzofuran
Tribromomethane
1,2,3-Trichloropropane
More than 30 of the compounds are listed by proper shipping name in the
Sec. 172.101 Table. As a group, the EPA compounds include: volatiles
(35 compounds); semivolatiles (33 compounds); organochlorine pesticides
(20 compounds); phenoxyacetic acid herbicides (3 compounds); PCBs (all
PCBs); and dioxins and furans (7 compounds).
Based on this review, this final rule requires that, when the EPA
compounds listed in the HMR are transported in large capacity tank
cars, the tank cars must conform to a limited and designated
specification with greater protection in accidents. Also, to ensure the
proper identification and packaging of these materials, PHMSA is listing
(with the exception of Class 2 materials [compressed gases], PIH
materials, and the 16 materials not now identified as hazardous
substances) in Sec. 173.31(f), all EPA compounds listed in 40 CFR Part
268, Appendix III. As explained elsewhere in the preamble, PHMSA is no
longer authorizing Class 2 materials or PIH materials in low-pressure
tank cars, e.g., class DOT 111A.
Because PHMSA is listing the EPA halogenated-organic compounds as
hazardous substances, in this final rule, the threshold quantity is the
reportable quantity of the hazardous substance. As an example, if the
material in the tank car (including its mixtures and solutions) (1) is
listed in Appendix A to Sec. 172.101, (2) is in a quantity that equals
or exceeds the reportable quantity (RQ) of the material listed in
Appendix A, and (3) is listed in Sec. 173.31(f), it must be transported
in a tank car of limited and designated specification to offer greater
protection in the event of an accident.
In the NPRM, PHMSA proposed that any of the halogenated organic
compounds identified by EPA must be transported in a tank car meeting
DOT 105S200W, DOT 112S200W with an 11-gauge metal jacket, or DOT
112S340W without a metal jacket. PHMSA stated that the metal jacket and
head protection on these tank cars blunt the impacting forces from
couplers, wheels, track, and other objects along the carrier's right-
of-way. According to FRA research, this blunting effect is directly
proportional to the thickness of the tank jacket or head shield and is
effective in preventing tank punctures.<SUP>17 The NPRM would have
allowed the use of any class DOT 105 or DOT 112 tank car regardless of
its date of construction. Older tank cars would be allowed, including
those constructed with an older steel specification, such as ASTM A212
Grade B. Because the older steels have less puncture resistance than
the steels currently in use, the NPRM proposed the use of an external
metal jacket to help blunt any impacting force, as a result of an
accident, to the tank shell.
\17\Coltman, M., & Hazel, M., Jr., Chlorine Tank Car Puncture
Resistance Evaluation, Report DOT/FRA/ORD-92-11, Federal Railroad
Administration (1992), Washington, D.C.
---------------------------------------------------------------------------
At the January 6, 1994, public hearing, a commenter asked PHMSA to
consider the use of a non-jacketed DOT 112S200W tank car, provided that
the tank car was constructed from an AAR normalized high-strength steel
specification, AAR TC-128. This steel specification has high tensile
and yield strength. In addition to the higher tensile and yield
strengths, commenters stated that normalization of the steel adds extra
puncture resistance. A commenter further stated that a tank car
constructed from the AAR's TC-128 steel specification would provide a
level of puncture resistance comparable to that of tank cars proposed
for use in the NPRM, and would also render a indisputable benefit/cost
ratio. Upon further review, PHMSA agrees that a tank car constructed
from AAR TC-128, normalized, would provide a level of puncture
resistance equivalent to a tank car constructed from any steel
specification proposed in the NPRM. In this final rule, PHMSA has
provided for the use of a DOT 112S200W (non-jacketed tank car)
constructed from AAR TC-128 normalized steel as an authorized
packaging, as suggested by the commenter.
L. Implementation of New Requirements
In the NPRM, PHMSA proposed two implementation dates. Under ``Option
A,'' most of the compliance dates were set at 10 years from the
effective date of this final rule. This is a period that also coincides
with the duration frequently specified in typical full-term tank car
leases, whether a true lease or a financing vehicle; and with the
``thorough inspection'' interval for tank cars in Interchange Rule
88.B.2.<SUP>18 Under ``Option B,'' PHMSA proposed that certain tank car
types and car/commodity combinations be considered for shorter retrofit
periods, with 5 years given to bring existing cars into compliance. For
instance, aluminum and nickel tank cars are more vulnerable to
puncture, and tanks used for transporting PIH materials present special
hazards.
\18\Field Manual of the Interchange Rules, adopted by the
Association of American Railroads, Mechanical Division, Washington,
D.C., 1992. At intervals not to exceed 10 years, major components of
the car must be inspected, including body bolsters and center
plates, center sills, crossbearers, crossties, draft systems and
components, end sills, side sills, and trucks.
---------------------------------------------------------------------------
Option A was supported by commenters. Although urging PHMSA to adopt
the 10-year time limit, RPI stated that, because of start-up
complexities, it will not be reasonable to accomplish this on a 10-
percent per year basis. Instead, RPI suggested that its members were
willing to modify 50 percent of the fleet in the first 5 years and 50
percent in the second 5 years. This accomplishes the desired goal while
minimizing scheduling problems and maximizing efficiency.
Option B was supported by NTSB who stated that PHMSA should require
tank-head protection, within 5 years, for all class DOT 105 tank cars
having capacities of less than 70 kl (18,500 gallons) when used to
transport a Division 2.1 material (flammable gas).
Most commenters supported the 10-year modification program for
existing tank cars. PHMSA believes, however, that a 5-year modification
program is more appropriate for class DOT 105 tank cars that have a
capacity less than 70 kl (18,500 gallons) when used to transport a
Division 2.1 material. Mandating an accelerated modification program
for these particular tank cars will ensure that those cars presenting
the greatest risk are modified first. Therefore, this final rule
requires that each tank car built on or after the effective date of
this final rule conform to this final rule. For tank cars built prior
to the effective date, the phase-in period is 10 years: at least 50
percent of the fleet in the first 5 years and the balance in the second
5 years. The phase-in-period for tank cars transporting a Division 2.1
material is 5 years, with at least 50 percent within 2\1/2\ years and
the balance in the second 2\1/2\ years. For existing tank cars
constructed with an internal self-energized manway located below the
liquid level of the lading, the compliance date is the effective date
of this final rule.
III. Docket HM-201--Detection and Repair of Cracks, Pits, Corrosion,
Lining Flaws and Other Defects of Tank Car Tanks
A. Background
On September 16, 1993, PHMSA published in the Federal Register a
NPRM under Docket HM-201; Notice No. 93-15 [58 FR 48485]. The NPRM
contained proposals to: (1) require the development and implementation
of a quality assurance program (QAP) at each facility that builds,
repairs, or ensures the structural integrity of tank
[[Page 49059]]
cars; (2) require the use of non-destructive testing (NDT) techniques
in lieu of the current periodic hydrostatic pressure tests for fusion
welded tank cars to more adequately detect cracks in principal
structure elements (PSE), the failure of which could cause catastrophic
failure of the tank; (3) require thickness measurements of tank cars;
(4) allow for the continued use of tank cars with limited reduced shell
thicknesses; (5) increase the inspection and test intervals for tank
cars; and (6) clarify the tank car pretrip inspection requirements.
Readers are referred to the NPRM preamble for a complete background,
including a more extensive discussion of issues and citations to
research data summarized in the final rule.
PHMSA received 31 comments in response to the NPRM from members of
the various industries that own, lease, transport, or use tank cars.
PHMSA and FRA have given full consideration to all comments in the
development of this final rule. Following is a summary of the written
comments, a summary of the final rule, and the actions taken by PHMSA
and FRA in this final rule:
B. Damage-Tolerance Fatigue Evaluations
In 1992, the NTSB issued a report on the inspection and testing of
tank cars. The report disclosed that many tank car defects are not
routinely detected. These defects may suddenly grow to a critical size
resulting in failure of the tank car. The NTSB recommended that FRA and
PHMSA develop requirements for the periodic inspection and tests of tank
cars to help ensure the detection of cracks before the cracks propagate
to a critical length. Such requirements would establish inspection and
test intervals based on the defect size detectable by the inspection
and test method used and on the stress level and crack propagation
characteristics of the PSE based on a ``damage-tolerance'' approach.
The Federal Aviation Administration (FAA) defines a structure as damage
tolerant if the structure has been evaluated to ensure that, should
serious fatigue, corrosion, or accidental damage occur within the
operational life of the structure, the remaining structure can
withstand reasonable loads without failure or excessive structural
deformation until the damage is detected (FAA Advisory Circular AC No.
25.571-1A). Damage-tolerance assumes that flaws exist in the structure
and that the design of the structure is such that these flaws will not
grow to a critical size and cause catastrophic failure to the structure
within a specified damage detection period. The damage detection period
depends on the characteristics of each PSE, each element's
susceptibility to severe corrosive environments, the inspectability of
each element, the inspection method, and procedures used and
maintenance practices.
In the NPRM, PHMSA proposed to allow tank car owners to use an
alternative inspection and test procedure or interval based on the
completion of a damage-tolerance fatigue evaluation. The evaluation
procedures would be reviewed by the AAR and approved by the Associate
Administrator for Safety, FRA. As stated in the NPRM, FRA believes that
some tank car owners may be able to reduce inspection and test costs by
using damage-tolerance fatigue evaluation procedures that incorporate:
(1) In-service inspection and test using techniques such as ultrasonic
or acoustic emission; (2) sampling of individual designs with a 100
percent inspection and test of the design if a crack is found; (3)
inspection and test intervals unique to each tank car component; and,
(4) inspection and test intervals based on the degree of risk a
material poses (i.e., high risk materials have shorter inspection and
test intervals than those with low risks).
Most commenters stated that the damage-tolerance approach is a
significant step toward advancing the detectability of defects and well
suited to a tank car and its associated structure. They suggested that
PHMSA and FRA expand the damage-tolerance approach, for fatigue, to
include other types of damage mechanisms, such as corrosion, corrosion
fatigue, original fabrication defects, stress corrosion cracking,
impact damage, and damage caused by an accident.
PHMSA and FRA agree that the use of a damage-tolerance approach to
periodic inspection and test of tank cars would substantially increase
the likelihood of the detection of cracks and crack-like defects before
such defects propagate to a critical size. PHMSA and FRA also believe
that the inspection interval for each PSE should be based on the
inspection method used, the stress level in each PSE, and the crack
propagation characteristics of each PSE.
The agencies realize, however, that in order to fully implement a
damage-tolerance program, it will take years for each owner or
manufacturer of a tank car to analyze each element on the tank car, and
to support the results of such analysis with test evidence and service
experience. FRA is currently working with the AAR Tank Car Committee,
the RPI, tank car owners, lessors, and manufacturers to develop
acceptable non-destructive testing techniques, and to develop an
inspection and test program based on damage-tolerance principles. These
programs include finite element analysis of the stub sill and its
attachment to the tank shell to identify the PSE on the tank car that
should be examined, over-the-road tests to define the typical
environmental loading spectrum expected in service, and a damage-
tolerance evaluation of the structure.
In this final rule, PHMSA is revising the regulatory text for the
damage-tolerance fatigue evaluation proposed in Sec. 180.509(k). This
revised requirement provides that an acceptable damage-tolerance and
fatigue evaluation include other types of damage mechanisms and is
supported by test evidence and, if available, by service experience.
C. Inspection and Test Intervals
FRA found that cracks may reach a critical size in a PSE within
about 400,000 miles of railroad service [see ``Owners of Railroad Tank
Cars; Emergency Order Requiring Inspection and Repair of Stub Sill Tank
Cars,'' (Emergency Order Number 17) 57 FR 41799, September 11, 1992].
To ensure against premature failure, common procedures for NDT allow
for two opportunities to inspect an item before predicted failure.
Because tank cars travel an average of about 18,000 miles per year and
most cracks become critical at about 400,000 miles of railroad service,
in the NPRM, PHMSA proposed an inspection and test interval, based on a
simplified damage-tolerance evaluation, of 10 years to allow for two
opportunities to inspect an item before predicted failure.
For the sake of efficiency, and to increase safety margins for most
cars, PHMSA proposed to implement the 10-year inspection and test
interval starting at what would otherwise be the next scheduled tank
hydrostatic pressure test. For tank cars within a 20-year test cycle,
PHMSA proposed that the next inspection and test date be the publication
date of this rule plus one half of the remaining years to what would
otherwise be the next scheduled tank hydrostatic test. After that the
tank would require an inspection and test on a 10-year interval.
For materials corrosive to the tank and shipped in non-lined or
non-coated tank cars, PHMSA proposed an inspection and test interval
based on the lower of (1) the corrosion rate of the material on the
tank shell or (2) the fatigue life of the tank structure as discussed
above. PHMSA and FRA developed a test interval to ensure that the
calculated thickness of the tank at the next inspection and
[[Page 49060]]
test will not fall below the proposed allowable minimum wall thickness.
The inspection and test interval in this case is calculated by
subtracting the actual thickness (measured at the time of construction
or any subsequent inspection and test) from the allowable minimum
thickness and then dividing that difference by the corrosion rate of
the hazardous material on the tank. Consequently, as the shell
thickness corrodes throughout the service-life of the tank, the tank
must receive an inspection and test more frequently.
Commenters supported the proposed inspection and test program for
most tank cars. They suggested, however, that PHMSA consider the
availability of tank car facility space and the practicality of
implementing the new inspection and test and quality assurance programs
without immobilizing a large number of tank cars. In particular,
commenters suggested that PHMSA not reduce the inspection and test
intervals for tank cars constructed during the 1975-1979 period that
are now subject to a 20-year hydrostatic pressure test interval. As
proposed, these particular tank cars become due for inspection and test
during the years 1995 through 1997. A major oil company stated that
these particular tank cars represent at least 20 percent of its tank
car fleet.
Several commenters stated tank cars used to transport chlorine,
unlike other tank cars, are currently tested every two years. As such,
all 8,000 tank cars in chlorine service would have to be brought in
conformance with the new inspection and test requirements within two
years. One company stated that it maintains 3,000 tank cars in chlorine
service and it would have to inspect 5.7 tank cars per day, which may
not be feasible because companies must first determine efficient
inspection techniques and provide training to inspection personnel.
Commenters further argue that because tank cars that transport chlorine
have an insulation system and a metal jacket, the inspectability of
certain PSE on these tank cars is difficult; accordingly, PHMSA should
not mandate the new requirements in the short-term until the industry
and the government specify the acceptable NDT techniques for inspecting
tank cars that have metal jackets.
The RPI suggested that PHMSA phase in the new procedures slowly by
beginning with tank cars without a metal jacket and then tank cars
having a metal jacket when appropriate inspection techniques are
developed. Although RPI did not explain the basis for its comment, PHMSA
and FRA assume that the reason behind RPI's comment is the difficulty
of inspecting PSE on a tank car having an insulation system covered by
a metal jacket or a thermal protection system; consequently, tank car
facilities will need time to develop the inspection methods and to
train inspection personnel on the use of those methods. Only after
identifying the appropriate inspection method and by training
inspection personnel, will there be a high probability of defect
detection.
Several commenters requested that PHMSA not require, in proposed
Sec. 180.509(b)(3), an inspection and test [requalification] of the
tank each time it is transferred into or out of a service that is
corrosive to the tank, which one commenter stated could occur 4 times
per month. Another commenter stated that the program is redundant with
proposed Sec. 180.509(c)(3)(ii) and, therefore, the section should be
deleted. The Chemical Manufacturers Association (CMA) suggested that
PHMSA amend the proposal to allow for routine transfers, so long as the
tank car is within the established intervals for the periodic
inspection requirements. A commenter suggested that localized
modifications to a tank, such as modifying nozzles or bottom outlets,
should not subject the tank to a complete requalification.
Based on the comments received, PHMSA is not adopting proposed
paragraphs (b) (3) and (4). Paragraphs (b) (5) and (6) are renumbered
accordingly.
PHMSA and FRA also agree that local repairs or modifications should
not subject the tank to the full inspection and test program, because
the repair or modification must be done according to Appendix R of
AAR's Specifications for Tank Cars. Appendix R specifies the procedures
for repairs, alterations, and conversions of tank cars and the
appropriate non-destructive testing method to ensure that the repairs,
alterations, or conversions were performed correctly.
PHMSA and FRA agree that the new inspection and test methods,
combined with other FRA mandated inspection programs, may cause a
tremendous backlog of tank cars awaiting inspection. Therefore, to
maintain an acceptable level of safety, but also to allow for an
orderly and acceptable phased-in NDT inspection and test program, PHMSA
will delay the compliance date of this final rule for 24 months for
tank cars without metal jackets and 48 months for tank cars having a
metal jacket or a thermal protection system. Before the compliance
date, tank cars may be given an inspection and hydrostatic test in
accordance with the current requirements or the requirements contained
in this final rule. After the compliance date, each tank car must be
given an inspection and test according to the requirements contained in
this final rule on or before the next scheduled tank hydrostatic
pressure test date.
D. High-Mileage Tank Cars
FRA realizes that some tank cars can travel in excess of 18,000
miles each year and, by doing so, the tank cars may reach 200,000 miles
of railroad service before their first periodic inspection and 400,000
miles before their second.
The NTSB expressed its concerns that the proposed regulations
recommend, but do not require, more frequent inspections and tests for
tank cars with mileage rates that exceed the average. Further, because
there is no requirement to maintain cumulative mileage on individual
tank cars, the NTSB expressed concern that high-mileage tank cars would
not be identified for the more frequent inspections and tests, thereby
increasing the possibility of a non-detected fatigue crack propagating
and causing a structural failure within the 10-year inspection and test
cycle.
PHMSA and FRA agree with the NTSB that high-mileage tank cars should
receive an inspection and test prior to reaching 200,000 miles of
railroad service. However, no requirement for the maintenance or
retention of car mileage records was proposed. Because car owners keep
records of car mileage, the owners can ensure that tank cars having
high-mileage are inspected more frequently than the inspection and test
intervals adopted in this final rule. Current Sec. 173.24(b) provides
that each package used for the shipment of hazardous materials shall be
so designed, constructed, and maintained . . . so that under conditions
normally incident to transportation--the effectiveness of the package
will not be substantially reduced. Thus, an owner has an obligation to
ensure the continuing effectiveness of a tank car. This duty is not
unlike that of an owner of an automobile who replaces the tires on his
or her car when worn and not based on the warranty period. FRA will,
during its inspection activities, assess the need for a rulemaking (1)
to require owners to retain car mileage records and (2) to inspect
their tank cars before the cars accumulate more than 200,000 miles of
railroad service.
E. NDT Techniques
In the NPRM, PHMSA proposed to require that the bottom shell of
fusion welded tank cars be inspected periodically by appropriate NDT
techniques, such as optically aided visual inspections, ultrasonic,
[[Page 49061]]
radiographic, magnetic particle, and dye penetrant testing methods, in
lieu of hydrostatic pressure tests.
All commenters supported the use of NDT techniques to assess the
integrity of a tank car in lieu of a hydrostatic pressure test. Several
commenters stated that the use of qualification procedures will require
formal NDT techniques in defined areas where no previous requirements
existed and will improve the overall safety of tank cars.
Several commenters suggested that PHMSA should authorize the use of
acoustic emission testing to qualify tank cars for further use. One
commenter stated that acoustic emission testing is widely used in the
chemical process industry to assure the integrity of pressure vessels,
tanks, and piping. The commenter further stated that the overall
reliability of a series of local tests (ultrasonic, dye penetrant,
radiography, etc.) is incorrectly compared with the reliability of a
single global test (hydrostatic, acoustic emission) and that
substitution of multiple local tests for a single global test may
endanger, rather than enhance the safe transportation of hazardous
materials.
PHMSA and FRA do not agree with the commenters's conclusion about
the potential danger of multiple local tests as compared with a single
global test. PHMSA and FRA believe that multiple local tests, focusing
on known areas of tank car stress, have a safety advantage over single
global tests, at least with the current state of development of
acoustic emission testing in the tank car industry. The NDT methods
mandated by this rule are a safety improvement. As noted immediately
below, the agencies have underscored their belief in the potential
benefits acoustic emission testing offers by granting an exemption that
will permit its development and refinement in a railroad industry
context.
Outside the scope of this rulemaking, but related to it by means of
subject matter, Monsanto Chemical Company applied for a DOT exemption
to use acoustic emission technology, in lieu of the current hydrostatic
retest, for the tank cars it owns. The procedures developed by Monsanto
to support its exemption were recently evaluated under a research
contract administered by the government of Canada. (McBride, S. L.,
Acoustic Emission Tank Car Test Method Review & Evaluation, Transport
Canada Report No. TP 12140E (1994) Montreal, Quebec). The results of
that research show that Monsanto's acoustic emission testing procedures
appear to be sound. The report suggests, however, minor refinements in
the acoustic emission procedures. Taking this into account, PHMSA issued
Monsanto an exemption on September 9, 1994 (DOT-E 10589). The following
companies were granted ``party to'' status on the Monsanto exemption:
Union Tank Car Company, Testing Associates, and Physical Acoustics
Corporation.
This final rule does not include acoustic emission testing as an
authorized NDT technique. PHMSA and FRA are committed, however, to
explore new technologies for inspecting and testing tank cars and will
continue to evaluate the possibly of authorizing the acoustic emission
testing procedure in the future. In support of this commitment, FRA
issued a research contract to further explore and refine the use of
acoustic emission testing procedure and other NDT techniques in
determining the integrity of insulation and lining covered welds of
tank cars.
F. Leakage Test
In the NPRM, PHMSA proposed a leakage test that would include all
product piping with all valves and accessories in pla |
