Hazardous Materials: Improving the Safety of Railroad Tank Car Transportation of Hazardous Materials

Federal RegisterApr 1, 2008

Ask Donna

What actually matters in this document.

Text

DEPARTMENT OF TRANSPORTATION

Pipeline and Hazardous Materials Safety Administration

49 CFR Parts 171, 173, 174 and 179

[Docket No. FRA-2006-25169]

RIN 2130-AB69

Hazardous Materials: Improving the Safety of Railroad Tank Car Transportation of Hazardous Materials

AGENCY:

Pipeline and Hazardous Materials Safety Administration (PHMSA), Department of Transportation (DOT).

ACTION:

Notice of proposed rulemaking (NPRM).

SUMMARY:

The Pipeline and Hazardous Materials Safety Administration and the Federal Railroad Administration are proposing revisions to the Federal Hazardous Materials Regulations to improve the crashworthiness protection of railroad tank cars designed to transport poison inhalation hazard materials. Specifically, we are proposing enhanced tank car performance standards for head and shell impacts; operational restrictions for trains hauling tank cars containing PIH materials; interim operational restrictions for trains hauling tank cars not meeting the enhanced performance standards; and an allowance to increase the gross weight of tank cars that meet the enhanced tank-head and shell puncture-resistance systems.

DATES:

Submit comments by June 2, 2008. To the extent possible, late-filed comments will be considered as we develop a final rule.

ADDRESSES:

You may submit comments identified by the docket number FRA-2006-25169 by any of the following methods:

•

Federal eRulemaking Portal:

http://www.regulations.gov.

Follow the instructions for submitting comments.

•

Fax:

1-202-493-2251.

•

Mail:

U.S. Department of Transportation, Docket Operations, M-30, West Building Ground Floor, Room W12-140, 1200 New Jersey Avenue, SE., Washington, DC 20590.

•

Hand Delivery:

U.S. Department of Transportation, Docket Operations, M-30, West Building Ground Floor, Room W12-140, 1200 New Jersey Avenue, SE., Washington, DC 20590.

Instructions:

All submissions must include the agency name and docket number (FRA-2006-25169) for this rulemaking. Note that all comments received will be posted without change to

http://www.regulations.gov

including any personal information. Please see the Privacy Act heading in the “Regulatory Analyses and Notices” section of this document for Privacy Act information related to any submitted comments or materials. Internet users may access comments received by DOT at

http://www.regulations.gov.

FOR FURTHER INFORMATION CONTACT:

William Schoonover, (202) 493-6229, Office of Safety Assurance and Compliance, Federal Railroad Administration; Lucinda Henriksen, (202) 493-1345, Office of Chief Counsel, Federal Railroad Administration; or Michael Stevens, (202) 366-8553, Office of Hazardous Materials Standards, Pipeline and Hazardous Materials Safety Administration.

SUPPLEMENTARY INFORMATION:

Abbreviations and Terms Used in This Document

AAR—Association of American Railroads

ABS—Automatic Block Signal

Action Plan—National Rail Safety Action Plan

ADAMS—Automated Dynamic Analysis of Mechanical Systems

ARI—American Railway Car Institute

ATIP—Automated Track Geometry Program

BNSF—BNSF Railway Company

BTS—Bureau of Transportation Statistics

RS—Confidential Close Call Reporting System

CEQ—Council on Environmental Quality

CPC—Casualty Prevention Circular

CI—Chlorine Institute

CP—Canadian Pacific

CPR—Conditional Probability of Release

CSXT—CSX Transportation

Department—U.S. Department of Transportation

DOW—Dow Chemical Company

DOT—U.S. Department of Transportation

ECP—Electronically Controlled Pneumatic Brake Systems

ETMS—Electronic Train Management System

Federal hazmat law—Federal hazardous materials transportation law (40 U.S.C. 5101

et seq.

)

FRA—Federal Railroad Administration

HMR—Hazardous Materials Regulations

NGRTCP—Next Generation Rail Tank Car Project

NPRM—Notice of Proposed Rulemaking

NTSB—National Transportation Safety Board

OMB—Office of Management and Budget

PHMSA—Pipeline and Hazardous Materials Safety Administration

PIH—Poison Inhalation Hazard

PTC—Positive Train Control

PV—Present Value

QA—Quality Assurance

R&D—Research and Development

RSAC—Railroad Safety Advisory Committee

RSI—Railway Supply Institute

SAFETEA-LU—Safe, Accountable, Flexible, Efficient, Transportation Equity Act: A Legacy for Users, Pub. L. 109-59

SBA—Small Business Administration

SOMC—Association of American Railroads Safety and Operations Management Committee

SRT—Structural Reliability Technologies

Tank Car Manual—Association of American Railroads Tank Car Committee Tank Car Manual

TCC—Association of American Railroads Tank Car Committee

TFI—The Fertilizer Institute

TIH—Toxic Inhalation Hazard

TRANSCAER®—Transportation Community Awareness and Emergency Response

TSA—Department of Homeland Security, Transportation Security Administration

Trinity—Trinity Industries, Inc.

Union Tank—Union Tank Car Company

UP—Union Pacific Railroad Company

Volpe—Volpe National Transportation Systems Center

Table of Contents for Supplementary Information

I. Background

II. Summary of Proposals in this NPRM

III. Statutory Authority, Congressional Mandate, and NTSB Recommendations

IV. Brief Overview of FRA Programs to Continuously Improve Rail Safety Outside of Tank Car-Specific Efforts

V. Relevant Regulatory Framework

VI. Railroad Accidents Involving Hazardous Materials Releases and Accompanying NTSB Recommendations

A. Minot

B. FRA's Responses to the NTSB Tank Car Recommendations for Minot

C. Macdona

D. Graniteville

E. FRA's Responses to the NTSB Tank Car Recommendations for Graniteville

VII. Evaluating the Risk Related to Potential Catastrophic Releases from PIH Tank Cars in the Future

A. Graniteville

B. Minot

VIII. The Railroad Industry's Liability and the Impact of Accidents Involving the Shipment of PIH Materials on Insurance Costs and Shipping Rates

IX. Industry Efforts to Improve Railroad Hazardous Materials Transportation Safety

A. General Industry Efforts

B. Trinity Industries, Inc.'s Special Permit Chlorine Car

C. AAR Proposals for Enhanced Chlorine and Anhydrous Ammonia Tank Cars

D. Dow/UP Safety Initiative and the Next Generation Rail Tank Car Project

E. The Chlorine Institute Study

X. Discussion of Relevant Tank Car Research

XI. Discussion of Public Comments

A. May 31-June 1, 2006 Public Meeting

B. December 14, 2006 Public Meeting

C. March 30, 2007 Public Meeting

XII. Proposed Rule and Alternatives

XIII. Section-by-Section Analysis

XIV. Regulatory Analyses and Notices

A. Statutory/Legal Authority for This Rulemaking

B. Executive Order 12866 and DOT Regulatory Policies and Procedures

C. Executive Order 13132

D. Executive Order 13175

E. Regulatory Flexibility Act and Executive Order 13272

F. Paperwork Reduction Act

G. Regulation Identifier Number (RIN)

H. Unfunded Mandates Reform Act

I. Environmental Assessment

J. Privacy Act

I. Background

Hazardous materials are essential to the economy of the United States and to the well being of its people. These materials are used in water purification, farming, manufacturing, and other industrial applications. Railroads carry over 1.7 million shipments of hazardous materials annually, including millions of tons of explosive, poisonous, corrosive, flammable, and radioactive materials. The need for hazardous materials to support essential services means that the transportation of highly hazardous materials is unavoidable.

Rail transportation of hazardous materials is a safe method for moving large quantities of hazardous materials over long distances. The vast majority of hazardous materials shipped by railroad tank car each year arrive at their destinations safely and without incident. In the year 2004 (most recent data available), for example, out of the approximately 1.7 million shipments of hazardous materials transported by rail, there were 29 accidents in which a hazardous material was released. In these accidents, a total of 47 hazardous material cars released some amount of product; thus, the risk of a release was a tiny fraction of a percent (0.0028 percent or 47/1,700,000). The DOT Hazardous Materials Information System's ten-year incident data for 1997 through 2006 identifies a total of 17 fatalities resulting from rail hazardous materials incidents. While even one death is too many, these statistics show that train accidents involving a release of hazardous materials that causes death are rare. We recognize, however, that rail shipments of hazardous materials frequently move through densely populated or environmentally-sensitive areas where the consequences of an incident could be loss of life, serious injury, or significant environmental damage.

Historically, the Pipeline and Hazardous Materials Safety Administration (PHMSA), working closely with the Federal Railroad Administration (FRA), has issued a number of regulations to improve the survivability of rail tank cars in accidents.

Among other things, these regulations require hazardous material tank cars to be equipped with tank-head puncture resistance systems (head protection), coupler vertical restraint systems (shelf couplers), insulation, and for certain high-hazard materials, thermal protection systems. The historical safety record of railroad tank car hazardous material transportation demonstrates that these systems, working in combination, have been successful in greatly reducing the potential harm to human health and the environment when tank cars are involved in accidents.

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,

60 FR 49048 (Sept. 21, 1995);

Performance-Oriented Packaging Standards; Miscellaneous Amendments

, 58 FR 50224 (Sept. 24, 1993);

Performance Oriented Packaging: Changes to Classification, Hazard Communication, Packaging and Handling Requirements Based on UN Standards and Agency Initiative,

55 FR 52402 (Dec. 21, 1990);

Transportation of Hazardous Materials, Miscellaneous Amendments,

54 FR 38790 (Sept. 20, 1989);

Specifications for Railroad Tank Cars Used to Transport Hazardous Materials,

49 FR 3468 (Jan. 27, 1984);

Shippers, Specifications for Tank Cars,

49 FR 3473 (Jan. 27, 1984); Interlocking

Couplers and Restrictions of Capacity of Tank Cars,

35 FR 14215 (Sept. 9, 1970);

Shippers; Specifications for Pressure Tank Cars,

42 FR 46306 (Sept. 15, 1977);

Tank Car Tank-head Protection,

41 FR 21475 (May 26, 1976).

In the last several years, however, there have been a number of rail tank car accidents in which the car was breached and product lost on the ground or into the atmosphere. Of particular concern have been accidents involving materials that are poisonous, or toxic, by inhalation (referred to as PIH or TIH materials). For example, on January 18, 2002, a Canadian Pacific Railway Company (CP) train derailed in Minot, North Dakota, resulting in one death and 11 serious injuries due to the release of anhydrous ammonia when five tank cars carrying the product catastrophically ruptured, and a vapor plume covered the derailment site and surrounding area. On June 28, 2004, a Union Pacific Railroad Company (UP) train collided with a Burlington Northern and Santa Fe Railway Company (now known as BNSF Railway Company) (BNSF) train in Macdona, Texas, breaching a loaded tank car containing chlorine and causing the deaths of three people and seriously injuring 30 others. On January 6, 2005, a Norfolk Southern Railway Company train collided with a standing train on a siding in Graniteville, South Carolina. The accident resulted in the breach of a tank car containing chlorine, and nine people died from the inhalation of chlorine vapors. Although none of these accidents was caused by hazardous material tank cars, the failure of the tank cars involved led to fatalities, injuries, evacuations, property and environmental damage.

On August 10, 2005, Congress passed the Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users, Pub. L. 109-59 (SAFETEA-LU). SAFETEA-LU added section 20155 to the Federal hazmat law. 49 U.S.C. § 20155. As discussed below, section 20155, in part, required FRA to (1) validate a predictive model quantifying the relevant dynamic forces acting on railroad tank cars under accident conditions, and (2) initiate a rulemaking to develop and implement appropriate design standards for pressurized tank cars.

In response to these recent accidents and in light of Congress's mandate in SAFETEA-LU to develop and implement appropriate design standards for pressurized tank cars, PHMSA and FRA, the two operating administrations within DOT responsible for overseeing the safe transportation of hazardous materials by rail, initiated a comprehensive review of design and operational factors that affect rail tank car safety. DOT's approach to enhancing the safety of rail tank cars and transportation of hazardous materials by rail tank cars is on-going and multi-faceted. For example, DOT is utilizing a risk management approach to identify ways to enhance the safe transportation of hazardous materials in tank cars, including: (1) Tank car design, manufacture, and requalification; (2) railroad operational issues such as human factors, track conditions and maintenance, wayside hazard detectors, signals and train control systems; and (3) improved planning and training for emergency response.

Recognizing the need for public input into this review of hazardous material tank car safety, on May 31 and June 1, 2006, PHMSA and FRA hosted a public meeting to discuss the initiation of this comprehensive review and to invite interested parties to participate in the agencies' efforts to surface and prioritize issues relating to the safe transportation of hazardous materials by railroad tank car. Subsequent to the meeting, FRA established a public docket (Docket No. FRA-2006-25169) to provide interested parties with a central location to both send and review relevant information concerning the safety of railroad tank car transportation of hazardous materials and a venue to gather and disseminate information and views on the issues.

See

71 FR 37974 (July 3, 2006).

Building on the initial public meeting, FRA and PHMSA held a second public meeting on December 14, 2006. At this second meeting, FRA announced DOT's commitment to develop an enhanced tank car standard by 2008. In addition,

at this meeting, the agencies solicited input and comments in response to nine specific questions pertaining to potential methods and goals of tank car improvements. On March 30, 2007, PHMSA and FRA held a third public meeting at which FRA shared the preliminary results of its research related to tank car survivability and provided an update on DOT's progress towards developing enhanced tank car safety standards.

As discussed in Section XI below, meeting participants from both the railroad and shipping industries expressed agreement on the need for continuous improvement in the safe transportation of hazardous materials by railroad tank car, particularly in light of the Minot, Macdona, and Graniteville accidents. Accordingly, after careful review and consideration of all of the relevant research and data, oral comments at the public meetings, and comments submitted to the docket, PHMSA and FRA are proposing enhanced tank car performance standards and operating limitations designed to minimize the loss of lading from tank cars transporting PIH materials in the event of an accident.

Issuance of this NPRM does not mean that FRA and PHMSA's efforts to improve tank car safety will end. Improving the safety and security of hazardous materials transportation via railroad tank car is an on-going process. Going forward, FRA's hazardous materials research and development (R&D) program will continue to focus on reducing the rate and severity of hazardous materials releases by optimizing the manufacture, operation, inspection, and maintenance procedures for the hazardous materials tank car fleet. FRA's overall R&D program will also continue to examine railroad operating practices and the use of technologies designed to increase overall railroad safety.

II. Summary of Proposals in this NPRM

As discussed in detail in Section X below, DOT's tank car research has shown that the rupture of tank cars and loss of lading are principally associated with the car-to-car impacts that occur as a result of derailments and train-to-train collisions. Conditions during an accident can be such that a coupler of one car impacts the head or the shell of a tank car. With sufficient speed, such impacts can lead to rupture and loss of lading. When a tank car is transporting PIH materials, the consequences of that loss of lading can be significant. Based on the information currently available, DOT believes that a significant opportunity exists to enhance the safe transportation of PIH materials by railroad tank car. Accordingly, in order to enhance the safety of hazardous materials transportation, and in direct response to the Congressional directive of 49 U.S.C. 20155, DOT is proposing revisions to the Hazardous Materials Regulations (HMR; 49 CFR Parts 171-180) that would improve the accident survivability of railroad tank cars used to transport PIH materials. Specifically, in this NPRM, we are proposing to require:

• A maximum speed limit of 50 mph for all railroad tank cars used to transport PIH materials;

• A maximum speed limit of 30 mph in non-signaled (i.e., dark) territory for all railroad tank cars transporting PIH materials, unless the material is transported in a tank car meeting the enhanced tank-head and shell puncture-resistance systems performance standards of this proposal;

• As an alternative to the maximum speed limit of 30 mph in dark territory, submission for FRA approval of a complete risk assessment and risk mitigation strategy establishing that operating conditions over the subject track provide at least an equivalent level of safety as that provided by signaled track;

• Railroad tank cars used to transport PIH materials to be manufactured to meet enhanced performance standards for tank-head and shell puncture-resistance systems;

• The expedited replacement of tank cars used for the transportation of PIH materials manufactured before 1989 with non-normalized steel

head or shell construction; and

Non-normalized steel is steel that has not been subjected to a specific heat treatment procedure that improves the steel's ability to resist fracture.

• An allowance to increase the gross weight on rail for tank cars designed to meet the proposed enhanced tank-head and shell puncture-resistance systems performance standards.

In drafting this proposed rule, DOT has carefully considered the results of all of its research regarding tank car accident survivability, all comments received through the series of public meetings held in the course of DOT's comprehensive review of tank car safety, as well as all written comments submitted to the docket of this proceeding. DOT believes that its two-pronged approach to enhancing the accident survivability of tank cars—that is, limiting the operating conditions of the tank cars transporting PIH materials and enhancing the tank-head and shell puncture-resistance performance—represents the most efficient and cost-effective method of improving the accident survivability of these cars. DOT invites comments on all aspects of this proposed rule.

First, with regard to the proposed speed and operating restrictions, we have reviewed the results of research on the current tank car fleet used for the transportation of PIH materials. We have also reviewed recent accidents and subsequent recommendations of the National Transportation Safety Board (NTSB). As discussed in Section X below, FRA's research demonstrates that the speed at which a train is traveling has the greatest effect on the closing velocity between cars involved in a derailment or other accident situation. Specifically, the research indicates that, in general, the secondary car-to-car impact speed is approximately one-half that of the initial train speed—the speed of the train at the time of the collision or derailment. Limiting the operating speed of tank cars transporting PIH materials is one method to impose a control on the forces experienced by these tank cars.

The rail industry, through the Association of American Railroads (AAR), has developed a detailed protocol on recommended operating practices for the transportation of hazardous materials. These recommended practices were originally implemented in 1990 by all of the Class 1 rail carriers operating in the United States. In 2006, AAR issued a revised version of this protocol, known as Circular OT-55-I, with short-line railroads also participating in the implementation. Among other requirements, OT-55-I restricts the operating speeds to a maximum of 50 mph for key trains, which are defined to include trains containing five or more tank car loads of PIH materials. Pursuant to OT-55-I, most trains with tank cars containing PIH materials are transported under this speed restriction. The period in which these tank cars are picked up or delivered is the most likely time when a train might not contain a sufficient quantity of hazardous materials to meet the definition of a key train and thus not operate under the 50 mph speed restriction. However, it is likely that the class of track into the facility may already limit the speed below 50 mph. Under FRA's Track Safety Standards,

there are minimum safety requirements that a track must meet, and the condition of the track is directly tied to the maximum allowable operating speed for the track. Only the two highest categories of track typically used for freight service, Classes 4 and 5,

have a maximum allowable operating speed above 50 mph. In addition, 50% of track in the United States is non-signaled and restricted by the Track Safety Standards to a speed limit of 49 mph. We therefore believe that the proposed restrictions in this NPRM represent an effective way to control the forces experienced by the tank car during most derailment or accident conditions without imposing an undue burden on the industry. We invite commenters to address whether our assumption that most tank cars transporting PIH materials are transported in accordance with the speed restrictions in OT-55-I is accurate, particularly for smaller and short-line carriers. In addition, we invite commenters to address whether there are alternative approaches to reduce the consequences of a train derailment or accident involving PIH materials, including data and information in support of suggested alternative approaches or strategies.

See

49 CFR part 213.

FRA analyzed data from chlorine incidents between 1965 and 2005, and anhydrous ammonia incidents between 1981 and 2005, to study those incidents resulting in loss of product from head and shell punctures, cracks, and tears.

This analysis suggests that a disproportionate number of those incidents occurred in non-signaled (dark) territory, as compared to the percentage of total train miles in dark territory. Additionally, this analysis showed that at the time of these accidents, the median train speed was 40 mph and the average speed was 38 mph. This analysis also demonstrates that approximately 80% of the losses occurred at speeds greater than 30 mph. Notably, no catastrophic losses of chlorine occurred at speeds below 30 mph. Based on this data, we are proposing an interim measure to limit the speed of the existing fleet of tank cars used to transport PIH materials when traversing non-signaled territory. Specifically, we propose to limit the maximum allowable operating speed to 30 mph for tank cars transporting PIH materials over non-signaled territory unless the tank cars meet the enhanced tank-head and shell puncture-resistance systems performance standards of this proposal. We are also proposing alternate provisions that a railroad may choose to follow in lieu of the speed restriction.

See

document no. 30 in docket no. FRA-2006-25169, “Loss of TIH Product in Head and Shell Punctures, Cracks & Tears.”

Second, we are proposing enhanced tank-head and shell puncture-resistance performance standards that are designed to enhance the accident survivability of tank cars. One critical aspect of this enhancement is improved tank-head and shell puncture-resistance standards. The enhanced standards would require tank cars that transport PIH materials in the United States to be designed and manufactured with a shell puncture-resistance system capable of withstanding impact at 25 mph and with a tank-head puncture-resistance system capable of withstanding impact at 30 mph. As noted above, we are proposing these enhanced performance standards in tandem with an operational speed restriction of 50 mph. Because the secondary car-to-car impact speed in a derailment or collision scenario is approximately one-half of the initial train speed, designing and constructing tank cars to withstand shell impacts of at least 25 mph and limiting the speed of those tank cars to 50 mph will ensure that in most instances, the car will not be breached if it is involved in a derailment or other type of accident. Designing and constructing tank cars to withstand tank-head impacts of at least 30 mph would take advantage of the greater available space for impact-attenuating structures in front of the tank-head and would help mitigate possible differences between the generalized tank-head impact scenarios and the actual tank-head impacts that occur in collisions or derailments.

Empirical evidence from recent accidents and the derailment dynamics research prepared by the Volpe National Transportation Systems Center (Volpe) show that impacts happen to both tank car heads and shells. Tank car heads have historically been provided more protection than tank shells because the majority of tank car punctures occurred in rail yards to the heads of tank cars as a result of overspeed impacts. However, given the recent PIH releases in train accidents, we believe that it is time to enhance the accident survivability of the tank car, increasing the level of protection to both the tank-head and the shell.

To support the enhanced tank-head and shell puncture-resistance standards, we are proposing performance criteria, including impact test requirements. The proposed tests reflect generalized impact scenarios as a means to evaluate the performance of alternative designs. In the shell impact scenario, a rigid ram car with a punch impacts the shell of the tank car. Similarly, in the head impact scenario, a rigid ram car with a punch impacts the head of the tank car. The test procedures are based on the modeling developed by Volpe and the baseline tank car testing performed in cooperation with the Next Generation Rail Tank Car Project (NGRTCP), as discussed in Section IX below.

As proposed in this NPRM, compliance with the proposed standards can be shown by computer simulation, by simulation in conjunction with substructure testing, by full-scale impact testing, or a combination thereof. The highest level of confidence, although at the greatest cost, is provided by full-scale impact testing. The least costly and lowest level of confidence is provided by simulation alone. Substructure testing significantly increases the confidence in simulation modeling, potentially with relatively modest costs, depending on the details of the substructure test. Economic analysis indicates that freight rail industry economics should allow the development of several new tank car designs, through compliance shown with simulations and substructure testing. The performance criteria proposed in this NPRM provide for full-scale testing, scale model or component testing, simulation, or comparative analysis to an approved design. We are proposing to require designs for which no full-scale testing is performed to be submitted to FRA for review. FRA's review is necessary to ensure that modeling parameters and scale or substructure testing are sufficient to ensure that the necessary level of safety has been achieved. In evaluating a design, FRA will consider appropriate data and analysis showing how the proposed design meets the enhanced performance standards for head and shell impacts. FRA will consider proper documentation of competent engineering analysis or practical demonstrations, or both, which may include validated computer modeling, structural crush analysis, component testing, or any combination thereof. This approach is consistent with FRA's practice in determining compliance with equipment performance standards promulgated in other areas of railroad safety.

See, e.g.

, 49 CFR 229.211 (Locomotive Crashworthiness). We request comments on this proposal.

Third, to ensure timely replacement of the PIH tank car fleet, we are proposing an implementation schedule that allows for design development and manufacturing ramp-up in the first two years after the final rule becomes effective. We are also proposing that in the next three years, one-half of the existing fleet will be replaced, with the remaining fleet replacement taking place in the following three years. This schedule will allow for replacement of the current PIH tank car fleet within eight years from the effective date of the final rule.

One of the factors we have taken into consideration in developing this proposal is the NTSB's recommendations related to pre-1989 tank cars manufactured with non-normalized steel. The NTSB, in its report on the Minot, North Dakota accident,

concluded that low fracture toughness of non-normalized steels used for tank shells contributed to the complete fracture and separation of the derailed cars. While we believe that low fracture toughness of non-normalized steels is only one of many material and design characteristics that can contribute to tank car releases, the pre-1989 tank cars are reaching the upper limits of their useful life. Therefore, we believe that these pre-1989 cars, which were manufactured with non-normalized steel, should be replaced in an expedited fashion. To accomplish this safety goal, we propose to prohibit the use of tank cars manufactured with non-normalized steel heads or shells beginning five years after the effective date of the final rule. We want to emphasize that this requirement is focused on the expedited removal of the pre-1989 tank cars that were manufactured using non-normalized steel. We recognize the efforts of the AAR to incorporate requirements for normalized steel for cars manufactured after 1988. We also recognize that some tank car manufacturers began using normalized steel prior to 1988; those tank cars would not be affected by this proposal.

See infra

Section VI for a detailed discussion of the Minot, North Dakota accident.

Finally, we are proposing to allow an increase in the gross weight of tank cars allowed on rail. Improvements in tank car performance have historically relied in large part on thicker and/or stronger steel, which brings with it a corresponding addition to the empty weight of the tank car. Therefore, a potential consequence of the proposed enhanced tank-head and shell puncture-resistance performance standards in this NPRM could be a measurable increase in the total number of PIH rail shipments to convey the same quantity of product to the customer since a heavier tank car means must contain less lading to keep within the gross weight limit. As noted above, however, there is a long history of safe shipment of hazardous materials via railroad tank car, and the enhancements proposed in this NPRM will further increase the accident survivability of the tank cars used to transport PIH materials. Accordingly, we are proposing to allow an increase in the gross weight allowed on rail (up to 286,000 pounds) for tank cars that transport PIH materials to offset the potentially increased weight of the enhanced tank car.

This measure should enable shippers to continue meeting customer demands without significantly increasing the total number of PIH shipments. In proposing to allow tank cars meeting the enhanced tank-head and shell puncture-resistance system requirements to weigh up to 286,000 pounds gross weight on rail, we recognize that there are mechanical and structural concerns that must be addressed to ensure the safety of these cars during transportation. To ensure that tank cars exceeding the existing 263,000 pound limitation and weighing up to 286,000 pounds gross weight on rail are mechanically and structurally sound, we propose to require that such cars conform to AAR Standard S-286-2002, SPECIFICATION FOR 286,000 LBS. GROSS RAIL LOAD CARS FOR FREE/UNRESTRICTED INTERCHANGE SERVICE (adopted November 2002 and revised September 1, 2005), which we propose to incorporate by reference into the HMR. AAR Standard S-286-2002 is the existing industry standard for designing, building, and operating rail cars at gross weights between 263,000 pounds and 286,000 pounds. A copy of AAR Standard S-286-2002 has been placed in the docket.

We recognize that some facilities and railroads do not currently have infrastructure sufficient to support the use of a 286,000 pound tank car. We anticipate tank car designers, working with the end users, will develop tank cars that will meet the enhanced tank-head and shell performance standards while minimizing the addition of weight to the empty car. The existing tank car specifications provide flexibility that will allow some use of new technologies and materials to provide the improved accident survivability required by this proposal. DOT encourages the development of innovative engineering design changes to meet the proposed enhanced accident survivability standard while minimizing added weight to the empty tank car. We also anticipate that the growing use of rail cars with gross weight on rail exceeding 263,000 lbs. for non-hazardous commodities, such as coal and grain, will minimize the track infrastructure barriers to the use of the heavier cars over time. For these reasons, we believe that the number of PIH shipments will not be significantly increased by the proposed enhanced accident survivability standards. As in all aspects of this proposed rule, we request comments on this proposal. We are particularly interested in data and information concerning the extent to which track infrastructure has already been modified to accommodate heavier rail cars, including how those modifications were accomplished and at what cost. We also invite comments concerning additional infrastructure modifications that may be required to accommodate the heavier cars that would be permitted in accordance with the proposals in this NPRM and the extent to which PIH shipments along certain rail lines may increase because existing infrastructure may not accommodate heavier cars.

The specific proposals in this rule are explained in more detail in Section XIII, the Section-by-Section Analysis, which is set forth below.

III. Statutory Authority, Congressional Mandate, and NTSB Recommendations

The Federal hazardous material transportation law (Federal hazmat law, 49 U.S.C. 5101

et seq.

) authorizes the Secretary of DOT (Secretary) to “prescribe regulations for the safe transportation, including security, of hazardous material in intrastate, interstate, and foreign commerce.” The Secretary has delegated this authority to PHMSA. 49 CFR 1.53(b). The HMR, promulgated by PHMSA, are designed to achieve three goals: (1) To ensure that hazardous materials are packaged and handled safely and securely during transportation; (2) to provide effective communication to transportation workers and emergency responders of the hazards of the materials being transported; and (3) to minimize the consequences of an incident should one occur. The hazardous material regulatory system is a risk management system that is prevention-oriented and focused on identifying a safety or security hazard and reducing the probability and quantity of a hazardous material release.

Under the HMR, hazardous materials are categorized by analysis and experience into hazard classes and packing groups based upon the risks that they present during transportation. The HMR specify appropriate packaging and handling requirements for hazardous materials, and require a shipper to communicate the material's hazards through the use of shipping papers, package marking and labeling, and vehicle placarding. The HMR also require shippers to provide emergency response information applicable to the specific hazard or hazards of the material being transported. Finally, the HMR mandate training requirements for persons who prepare hazardous materials for shipment or who transport hazardous materials in commerce. The HMR also include operational

requirements applicable to each mode of transportation.

The Secretary also has authority over all areas of railroad transportation safety (Federal railroad safety laws, 49 U.S.C. 20101

et seq.

), and has delegated this authority to FRA. 49 CFR 1.49. Pursuant to its statutory authority, FRA promulgates and enforces a comprehensive regulatory program (49 CFR parts 200-244) to address railroad track, signal systems, railroad communications, rolling stock, rear-end marking devices, safety glazing, railroad accident/incident reporting, locational requirements for the dispatch of U.S. rail operations, safety integration plans governing railroad consolidations, merger and acquisitions of control, operating practices, passenger train emergency preparedness, alcohol and drug testing, locomotive engineer certification, and workplace safety. FRA inspects railroads and shippers for compliance with both FRA and PHMSA regulations. FRA also conducts research and development to enhance railroad safety. In addition, both PHMSA and FRA are working with the emergency response community to enhance its ability to respond quickly and effectively to rail transportation accidents involving hazardous materials.

As noted above, on August 10, 2005, Congress passed SAFETEA-LU, which added section 20155 to the Federal hazmat law. 49 U.S.C. 20155. In part, section 20155 required FRA to (1) validate a predictive model quantifying the relevant dynamic forces acting on railroad tank cars under accident conditions, and (2) initiate a rulemaking to develop and implement appropriate design standards for pressurized tank cars.

Prior to the Minot accident and the enactment of SAFETEA-LU, FRA had initiated tank car structural integrity research. In response to the Minot accident, the NTSB made four safety recommendations to FRA specific to the structural integrity of hazardous material tank cars. The NTSB recommended that FRA analyze the impact resistance of steels in the shells of pressure tank cars constructed before 1989 and establish a program to rank those cars according to their risk of catastrophic failure and implement measures to eliminate or mitigate this risk. The NTSB also recommended that FRA validate the predictive model being developed to quantify the maximum dynamic forces acting on railroad tank cars under accident conditions and develop and implement tank car design-specific fracture toughness standards for tank cars used for the transportation of materials designated as Class 2 hazardous materials under the HMR. In response to the Graniteville accident, the NTSB recommended, in part, that FRA “require railroads to implement operating measures such as * * * reducing speeds through populated areas to minimize impact forces from accidents and reduce the vulnerability of tank cars transporting” certain highly-hazardous materials. Each of these NTSB recommendations is discussed in more detail in Section VI below.

The Department considers this NPRM responsive to section 20155's mandate, as well as to the NTSB recommendations.

IV. Brief Overview of FRA Programs To Continuously Improve Rail Safety Outside of Tank Car-Specific Efforts

FRA implements a broad and extensive safety program directed at reducing accidents, casualties, loss of property and threats to the human environment. Through the Railroad Accident/Incident Reporting System, FRA gathers data that are employed in crafting responsive measures.

See

49 CFR part 225. FRA safety standards address track, equipment, signal and train control systems, motive power and equipment, and operating practices. These regulations set out detailed requirements for design or system performance, inspection and testing, and training. With respect to rail equipment accident/incidents (“train accidents”), the regulations seek to reduce the risk of derailments, collisions, and other losses such as fires involving on-track equipment. FRA employs the Railroad Safety Advisory Committee (RSAC), a group comprised of all of FRA's stakeholders, to help identify safety needs and to fashion responsive regulations.

FRA also conducts R&D, both independently and in concert with the railroad industry, to identify new ways to enhance safety. R&D products are as diverse as the Track Quality Index, which can help guide investments in program maintenance before safety limits are encountered, and a human-machine interface evaluation tool that can help evaluate control systems and display designs.

On May 16, 2005, DOT and FRA launched the National Rail Safety Action Plan (Action Plan) to address further the safety issues that face the nation's rail industry. The Action Plan targeted the most frequent, highest risk causes of accidents; focused federal oversight and inspection resources; and accelerated research into new technologies that can improve safety.

The Action Plan elements focused heavily on preventing train accidents caused by human factors and track—the two major categories of train accident causes. In the area of human factors, FRA has issued a proposed rule that seeks to ensure better management of railroad operational tests and inspections. The proposed rule is also intended to establish greater accountability for compliance with operating rules, particularly those that are involved in human factors train accidents, such as the handling of switches. FRA is now completing consultations within the RSAC regarding resolution of public comments on the proposed rule, and a final rule will be issued this year.

In November 2006, FRA fulfilled an Action Plan objective by releasing a study report entitled

Validation and Calibration of a Fatigue Assessment Tool for Railroad Work Schedules.

That report, and an accompanying White Paper, confirmed the impact of fatigue on human factor train accidents and announced the availability of an analytical model that can be used to evaluate crew scheduling. On February 13, 2007, DOT delivered proposed railroad safety reauthorization legislation to the Congress (introduced by request as H.R. 1516 and S. 918) that would replace the 100-year-old Hours of Service Law with science-based regulations addressing fatigue.

Because the genesis of human factors accidents is often unclear, FRA joined with a national coalition of employee organizations and railroads to launch the Confidential Close Call Reporting System (C

RS). The Bureau of Transportation Statistics (BTS) supports this effort by collecting the data and ensuring the anonymity of the persons providing reports. Local labor/management/FRA teams use the data to identify safety needs before a serious accident occurs. An initial C

RS project is presently underway at a major UP facility, and additional pilots are being planned. Other human factors initiatives include projects on “behavior-based safety” that seek peer involvement in workplace safety, initiatives to promote crew resource management, and extensive research to support further program development. In FY 2008, FRA will be seeking to integrate many of these efforts into a larger Risk Reduction Program intended to advance safety beyond what can be accomplished with traditional command and control approaches.

Recognizing that the best answer to human factor risks is sometimes technology that can “backstop” the person in cases when errors have high

consequences, FRA continues to work actively to promote Positive Train Control (PTC) systems and similar technology. For instance, FRA R&D provided funding and technical support for the BNSF's deployment of a new Switch Position Monitoring System on the railroad's Avard Subdivision. This system can detect a misaligned main track switch in non-signal territory and provide notification to the dispatcher for appropriate action. BNSF is also demonstrating track integrity circuit technology that can help identify broken rails without the full expense of a signal system. These technologies, which are forward compatible with the railroad's PTC system, known as the Electronic Train Management System (ETMS), are already being installed on additional rail lines. FRA approved the Product Safety Plan for ETMS Configuration I in December 2006, under a performance-based regulation issued with RSAC input in March of 2005. The Product Safety Plan was submitted under subpart H of 49 CFR part 236 and described in detail the train control technology, concept of operations, and results of safety analysis for the system (which in this configuration is designed for single track territory either with a traffic control system or without any signal system).

In the field of track safety, FRA is taking concrete steps in both research and enforcement. FRA research has provided a new tool to detect cracks in joint bars. This optical recognition technology can capture and analyze images for very small cracks while mounted on a hi-rail truck or other on-track vehicle. The system is already in initial use by two major railroads.

In order to ensure compliance with track geometry limits under load, FRA acquired two additional Automated Track Geometry Program (ATIP) cars instrumented for measurement of geometry at track speed, supplementing an existing Office of Safety car (and use of FRA's research cars for geometry surveys when available). This expanded ATIP capability will permit FRA to survey the core of the national rail system on an annual basis, returning to problem areas, as appropriate, without sacrificing coverage. These two additional cars were in service as of April 30, 2007.

One of the most vexing areas of track safety work is rail integrity. The concentration of rail traffic on a smaller, post-merger system together with growth in traffic, increasing gross weight of cars, and a slow pace of rail replacement has led to heavy reliance on internal rail inspections to detect rail flaws before they become service failures and pose the imminent risk of an accident. The President's Budget for the current fiscal year requested nine positions for rail integrity specialists to build a better organized and aggressive approach to oversight of railroad rail integrity programs. The Congress authorized funding sufficient to support this staffing in February, and FRA is recruiting for these positions.

Over time, strengthened oversight of compliance with railroad safety regulations, introduction of new technology such as PTC, better management of fatigue affecting safety critical employees, and other steps should yield a reduction in the risk of train accidents that could affect the transportation of hazardous materials. FRA is encouraged that, after over a decade of gradual increases in train accidents associated with the growth of rail traffic and other factors, both the train accident rate and total train accidents declined in 2006. This decline likely reflects improved compliance with regulatory requirements, reduced stress from fatigue associated with service disruptions, and other factors. However, history suggests that the underlying factors that create safety challenges, such as growing rail service demands that strain capacity, aging infrastructure, and factors beyond the effective control of the railroads (e.g., natural disasters, impacts with heavy vehicles at highway-rail crossings) will continue to introduce substantial risk even as train accident rates decline. Accordingly, it is necessary for PHMSA and FRA to take the additional actions proposed in this NPRM to reduce the probability that future train accidents will involve catastrophic releases of PIH materials. Thus, the Action Plan provided for acceleration of the research underlying this proposed rule, which is intended to make tank cars used for PIH service more resistant to product loss when a train accident occurs.

The Action Plan also noted with approval the action of major railroads to make available to emergency responders information concerning the top 25 commodities transported through their jurisdictions and called on the railroads to make additional efforts to provide emergency responders with hazardous materials information, including the location of cars hauling hazardous materials on specific trains. CSX Transportation and CHEMTREC—the 24-hour emergency assistance hotline provided as a service by chemical manufacturers—have partnered to provide a demonstration of technology that can readily provide consistent information to emergency responders. PHMSA and FRA encourage other railroads to join in this effort.

V. Relevant Regulatory Framework

Today railroad tank cars in the United States are designed, built, maintained, and operated under four primary sets of regulations and guidelines: (1) Regulations and orders issued under the Federal railroad safety laws; (2) regulations and orders issued under the Federal hazmat law; (3) the AAR's Interchange Rules;

and (4) the AAR Tank Car Committee's Tank Car Manual (Tank Car Manual).

AAR,

Interchange Rules

, Washington, DC, published annually in a “Field Manual” and an “Office Manual.”

AAR, Operations and Maintenance Dep't, Mechanical Div.,

Manual of Standards and Recommended Practices; Section C-Part III

, “Specifications for Tank Cars, Specification M-1002” (revised annually).

FRA's freight car, safety appliance, and power brake regulations in 49 CFR parts 215, 231, and 232 apply to tank cars as they do every other type of railroad freight car. Parts 215 and 232 establish minimum safety standards; railroads are free to supplement these standards with additional or more stringent safety standards that are not inconsistent with the Federal standards. 49 CFR 215.1 and 232.1.

The HMR treat the tank car as a packaging and mandate safety features, permissible materials and methods of construction, as well as inspection and maintenance standards. A material identified as a hazardous material by the HMR may not be shipped by railroad tank car unless the tank car meets the requirements of the HMR. 49 CFR 173.31(a).

A separate set of standards—the AAR Interchange Rules, issued by AAR's standing Tank Car Committee (TCC)

—govern the tender and acceptance of rail cars among carriers within the general system of railroad transportation. The AAR Interchange Rules address a range of design and operational requirements intended to promote uniformity and reciprocity in car handling, including the obligation of rail carriers to perform running repairs on equipment received in interchange. Historically, the AAR Interchange Rules also have addressed certain subjects, such as rail tank car standards, now covered comprehensively by the HMR. Most recently, as discussed below, the TCC has issued an interchange requirement (Casualty Prevention Circular 1175, as

amended by Casualty Prevention Circular 1178) that would require tank cars transporting anhydrous ammonia and chlorine to meet tank car design standards that are more stringent than those specified in the HMR.

The Mechanical Division of AAR's Operations and Maintenance Department is responsible for industry freight car standards and for administering the Interchange Rules, a body of private law that governs the acceptance and use by railroads of equipment which they do not own.

See

fn. 8,

supra.

Railroads, as common carriers, are generally required to provide transportation services in a reasonable manner, and they may not impose unreasonable requirements as a condition precedent to providing rail transportation services. Accordingly, interchange requirements, such as Casualty Prevention Circular 1178, that restrict the movement of railroad tank cars that meet DOT standards must be reasonable, and, if challenged, the burden is on the railroad to establish the reasonableness of the restriction.

See Akron, Canton & Youngstown R.R.

ICC,

611 F.2d 1162, 1169 (6th Cir. 1979);

see also Consolidated Rail Corp.

ICC,

646 F.2d 642, 650 (D.C. Cir. 1981),

cert denied,

454 U.S. 1047 (1981). Two of the factors that the Surface Transportation Board and the courts consider in determining the reasonableness of interchange requirements are whether there are Federal safety standards on point and whether a railroad has the ability to seek changes to these standards to meet the safety concerns of the railroad.

See Consolidated Rail,

646 F.2d at 651. In fact, DOT has established safety standards for tank cars carrying PIH commodities and, pursuant to this rulemaking, is proposing enhanced standards for tank-head and shell puncture resistance systems for these cars. Through participation in this rulemaking, railroads and other interested parties have the ability to influence the enhanced safety standards ultimately adopted by DOT. As discussed below, DOT has concluded that it is inappropriate at this time to establish new standards for top fittings protection, but DOT will continue to work with interested parties on research and ongoing discussions aimed at establishing enhanced consensus standards. There is, therefore, no reasonable basis for the railroads to implement Casualty Prevention Circular 1178 at this time. Railroads are free at any time to seek stricter tank car safety standards through a DOT rulemaking (49 CFR 106.95); to date, no rail carrier has petitioned PHMSA to adopt the tank car standards embodied in Casualty Prevention Circular 1178. FRA has notified the AAR that before the TCC can implement the proposed requirements in Circular 1178, the proposal must be submitted to DOT for approval.

The AAR TCC is a standing committee of the Mechanical Division of AAR's Operations and Maintenance Department. Voting members of the TCC include representatives of AAR member railroads, as well as tank car shipper and owner organizations, tank car builders, and chemical and industry associations. In addition, the Bureau of Explosives and the Railway Supply Institute have non-voting membership on the TCC. FRA and PHMSA, as the Federal agencies responsible for oversight of the safety of hazardous materials transportation by railroad, also participate in the TCC as nonvoting members.

Under the HMR, certain functions related to hazardous material tank cars are delegated to the TCC, including: (1) Approvals for construction of tank cars meeting DOT specifications; (2) procedures for repairs or alterations; and (3) recommending changes in tank car specifications.

First, the HMR require tank car manufacturers to obtain TCC approval for specific tank car designs and construction methods and materials and procedures for repairs and alterations to tank cars. The HMR authorize the TCC to make the determination that the proposed design, construction, or repair procedures conform to the applicable DOT specification requirements and to issue the approval. 49 CFR 179.3. This authority is primarily a ministerial function, designed to ensure that plans to construct, alter, or convert tank car tanks conform to DOT regulations. In accordance with 49 CFR 179.3(b), the TCC must approve construction of a tank car that meets all Federal requirements.

Federal regulations also require tank car facilities to have quality assurance programs that are approved by AAR. These programs relate to construction, life-cycle maintenance, and continuing qualification for service.

When a party seeks to construct a railroad tank car to be used in hazardous materials service that does not meet a current DOT specification (

see

49 CFR 179.10-179.500-18), the HMR authorize the TCC to review the proposed specification and report its recommendations on the proposal to DOT. 49 CFR 179.4. In this capacity, DOT benefits greatly from the technical expertise of the TCC members. However, final policy judgment lies with DOT, and only DOT is authorized to approve a new tank car specification, or, through issuance of a special permit in accordance with 49 CFR 107.101-.127, the construction and use of a tank car not meeting an existing DOT specification. DOT does not construe the procedures established in 49 CFR 179.4 as limitations on its rulemaking authority.

In addition to the approval authority noted above, in several subsections of Part 179 of the HMR, the TCC is authorized to approve fittings, attachments, materials, designs, methods, and procedures relevant to tank car design, construction, maintenance, repair, and inspection. For example, 49 CFR 179.103-2(a) provides that manway covers “shall be of approved design.” Similarly, 49 CFR 179.201-9 states that “a gauging device of an approved design must be applied to permit determining the liquid level of the lading.” In addition, 49 CFR 179.10 states that “[t]he manner in which tanks are attached to the car structure shall be approved.” In each instance, the term “approved” refers to approval by the TCC.

See

49 CFR 179.2.

The primary document containing the standards governing these approvals of the TCC is the Tank Car Manual. The December 2000 version of the Tank Car Manual is incorporated by reference into the HMR at 49 CFR 171.7; thus, compliance with the Tank Car Manual's standards is required under the HMR. Chapter 2 of the Tank Car Manual contains the AAR requirements for DOT tank cars. As noted above, the TCC, subject to certain limitations, may establish standards for tank cars that go beyond the standards set by DOT. For example, the Tank Car Manual requires that the heads and shells of pressure tank cars constructed of certain types of steel must be normalized; although DOT participated in the discussions leading to these standards and approves of them, the tank car specifications contained in the HMR do not contain comparable requirements.

However, as indicated above, because the December 2000 version of the Tank Car Manual is incorporated by reference into the HMR, compliance with the tank car standards specified in that version of the Tank Car Manual is required under the HMR. Under the Administrative Procedure Act, compliance with any other version of the Tank Car Manual would be required under the HMR only upon the incorporation of that version into the HMR by reference through rulemaking.

Chapter 2 of the Tank Car Manual also includes additional commodity specific tank car requirements relevant to certain PIH materials which are not included in the HMR. See §§ 2.1.2 (hydrogen sulfide tank cars) and 2.1.4 (hydrogen fluoride tank cars).

VI. Railroad Accidents Involving Hazardous Materials Releases and Accompanying NTSB Recommendations

The NTSB investigated three recent accidents involving tank cars transporting PIH materials, which occurred between 2002 and 2005 in Minot, North Dakota; Macdona, Texas; and Graniteville, South Carolina. In all three accidents, the NTSB recommended that FRA study improving the safety and structural integrity of tank cars and develop necessary operational measures to minimize the vulnerability of tank cars involved in accidents.

A. Minot

The accident occurred at approximately 1:30 a.m. on January 18, 2002, near Minot, North Dakota, and resulted in the derailment of 31 cars of a 112-car train. Eleven of the 31 derailed cars were pressurized tank cars transporting anhydrous ammonia, a toxic liquefied compressed gas. Five of those tank cars (DOT 105J300W cars) received sidewall impacts to their shells, causing the cars to catastrophically rupture and instantaneously release their contents. Approximately 146,700 gallons of anhydrous ammonia were released from those five cars. As a result, a toxic vapor plume covered the derailment site and the surrounding area. The plume rose approximately 300 feet and gradually expanded five miles downwind of the accident site. The remaining six pressurized tank cars transporting anhydrous ammonia that derailed also suffered from shell impacts. Those cars, DOT 105J300W, 112J340W, and 105S300W cars, gradually released 74,000 gallons of anhydrous ammonia due to damage to the cars' fittings or small punctures and/or tears to the shells. One resident was fatally injured, and 333 people suffered other injuries (11 serious). According to the NTSB, early in the emergency response effort, the Chief of the Minot Rural Fire Department ordered residents in the affected area to shelter-in-place (

i.e.

, remain inside their homes with the windows shut). NTSB concluded that sheltering-in-place was an effective emergency response and credited this action with the relatively low number of injuries, as compared to the number of persons affected by the vapor plume (333 injuries in 11,600 persons affected).

The NTSB determined that the probable cause of the accident was an undetected defective rail. Damages to rolling stock and track, as well as monetary loss from the damaged or destroyed lading, exceeded $2.6 million. As of March 15, 2004, over $8 million has been spent on environmental remediation. Other significant costs include: evacuation costs, truck delay, rerouting and associated out of service expenses, expenses for disruption to non-railroad businesses, and expenses incurred in settling claims arising from the accident.

On October 9, 2007, a Federal judge approved a $7 million settlement in a class-action lawsuit between Canadian Pacific Railroad and individuals affected by the accident.

On March 15, 2004, the NTSB released Safety Recommendations R-04-01 through R-04-07 as a result of the Minot accident. The first three recommendations (R-04-01, R-04-02, and R-04-03) pertain to FRA's oversight of continuous welded rail maintenance programs and are not relevant to this rulemaking. The four remaining recommendations (R-04-04, R-04-05, R-04-06, and R-04-07) concern tank car structural integrity and are relevant to this rulemaking. In fact, these four recommendations served as the basis for the reformulation of FRA's tank car research program.

Recommendations R-04-04 through R-04-07 read as follows:

See

Section X,

infra,

for a more detailed discussion of FRA's tank car research program.

(

R-04-04

). Conduct a comprehensive analysis to determine the impact resistance of the steels in the shells of pressure tank cars constructed before 1989. At a minimum, the safety analysis should include the results of dynamic fracture toughness tests and/or the results of nondestructive testing techniques that provide information on material ductility and fracture toughness. The data should come from samples of steel from the tank shells from original manufacturing or from a statistically representative sampling of the shells of the pre-1989 pressure tank car fleet.

(

R-04-05

). Based on the results of the Federal Railroad Administration's comprehensive analysis to determine the impact resistance of the steels in the shells of pressure tank cars constructed before 1989, as addressed in Safety Recommendation R-04-04, establish a program to rank those cars according to their risk of catastrophic fracture and separation and implement measures to eliminate or mitigate this risk. This ranking should take into consideration operating temperatures, pressures, and maximum train speeds.

(

R-04-06

). Validate the predictive model the Federal Railroad Administration is developing to quantify the maximum dynamic forces acting on railroad tank cars under accident conditions.

(

R-04-07

). Develop and implement tank car design-specific fracture toughness standards, such as a minimum average Charpy value, for steels and other materials of construction for pressure tank cars used for the transportation of U.S. Department of Transportation class 2 hazardous materials, including those in “low-temperature” service. The performance criteria must apply to the material orientation with the minimum impact resistance and take into account the entire range of operating temperatures of the tank car.

B. FRA's Responses to the NTSB Tank Car Recommendations for Minot

In August 2004, the FRA responded to NTSB Safety Recommendations R-04-04 through R-04-07, which arose from the Minot accident. As for NTSB Recommendation R-04-04 and R-04-05, which recommended that FRA analyze the impact resistance of steels in the shells of pressure tank cars constructed before 1989 and establish a program to rank the cars according to their risk of fracture, FRA advised the NTSB that the TCC had developed a plan to sample steels from pre-1989 pressure tank cars and that a program to rank those cars would be established. Because of FRA's commitment to ranking the pre-1989 fleet, the NTSB classified Safety Recommendation R-04-05 as “Open—Acceptable Response.” The NTSB, however, classified Safety Recommendation R-04-04 as “Open—Unacceptable Response” because the Board did not believe that the necessary analysis would be completed in a timely manner. After FRA provided additional information to the NTSB about the sampling, including preliminary fracture toughness data relating to the samples from the pre-1989 tank cars, the NTSB reclassified Safety Recommendation R-04-04 as “Open—Acceptable Response.”

As for NTSB Recommendation R-04-06, which recommended that FRA validate its model to quantify the dynamic forces acting on tank cars in accident conditions, the FRA advised the NTSB that it had initiated modeling programs at Volpe and the University of Illinois at Chicago to determine in-train forces on tank cars involved in train derailments. Based on FRA's response to Safety Recommendation R-04-06, the NTSB classified the Recommendation as “Open—Acceptable Response.”

Finally, as for NTSB Recommendation R-04-07, which recommended that FRA develop tank car design-specific fracture toughness standards for steels used in pressure tank cars, the FRA responded by stating that more research was needed (approximately three years) to address tank car design-specific fracture toughness standards. Because the NTSB believed there were existing solutions and accident findings from which to gauge fracture toughness values, such as Charpy impact, in June 2005, the NTSB

classified the FRA response to Safety Recommendation R-04-07 as “Open—Unacceptable Response.” Since June 2005, AAR, in cooperation with FRA, has developed standards that ensure a minimum level of impact resistance for normalized steel and that require that Charpy tests be performed in the orientation of the sample material with the lowest impact property. In July 2006, the NTSB determined that FRA had made progress on the development of fracture toughness standards, and it reclassified Safety Recommendation R-04-07 “Open—Acceptable Response.”

C. Macdona

The accident occurred at approximately 5 a.m. on June 28, 2004, in Macdona, Texas, and resulted in the derailment of four locomotives and 36 cars belonging to two trains that collided while traveling on the same track in opposing directions. As the eastbound 123-car train was attempting to leave the main line to enter a parallel siding, it was struck midpoint by a westbound train traveling on the same main line track. The 16th car of the westbound train was a pressurized tank car transporting chlorine, a toxic liquefied compressed gas. This tank car, a DOT 105A500W car, was punctured in the lower quadrant of the tank car head and the puncture terminated one inch beyond the seam joining the tank-head to the tank shell. The tank car instantaneously released approximately 9,400 gallons of chlorine, and a toxic vapor plume engulfed the accident area to a radius of at least 700 feet before drifting away from the site. The NTSB noted that the vapor cloud drifted with the wind from the accident site and traveled in a northwesterly direction toward several residential areas within the city of San Antonio. NTSB further noted that Sea-World, a large commercial entertainment venue, was about 10 miles northwest of Macdona in the path of the chlorine vapor cloud.

The NTSB determined that the probable cause of the accident was UP train crew fatigue that resulted in the failure of the engineer and conductor to appropriately respond to wayside signals governing the movement of their train. Thirty-three persons were injured, three fatally (including the UP train conductor and two occupants of a residence located near the accident site).

Damages to rolling stock, track and signal equipment were estimated at $6.3 million. As of July 20, 2006, $150,000 was spent to clean-up environmental consequences. Other significant costs include: Evacuation costs, truck delay, rerouting and associated out of service expenses, expenses for disruption to non-railroad businesses, and expenses incurred in settling claims arising from the accident.

The crew of the striking train survived the collision and exited the locomotive unassisted, but could not escape the chlorine gas. The conductor and engineer were able to walk some distance from the collision where they were transported to hospitals. The engineer was treated and released, the conductor died several hours later from inhalation of the toxic gas. Given that both crew members survived the collision, no fatalities or serious injuries would have resulted from the accident had a tank car of chlorine not been punctured.

On July 20, 2006, the NTSB released Safety Recommendations R-06-14 and R-06-15 as a result of the Macdona accident. Although neither recommendation specifically addressed the vulnerability of tank cars involved in an accident, the NTSB stated that the successful and timely implementation of Safety Recommendations R-04-04 through R-04-07 (recommendations from the Minot accident) and R-05-16 through R-05-17 (recommendations from the Graniteville accident discussed below) may have prevented/mitigated the Macdona accident and any future catastrophic releases of hazardous materials from pressurized tank cars involved in an accident.

D. Graniteville

The accident occurred at approximately 2:30 a.m. on January 6, 2005, in Graniteville, South Carolina, when a freight train was improperly switched from a main line track onto an industry track and struck an unoccupied, parked train head-on, on a rail spur leading to a textile manufacturing facility. The collision resulted in the derailment of three locomotives and 17 cars belonging to the two trains. Three of the 17 derailed cars were pressurized tank cars transporting chlorine. One tank car, a DOT 105J500W car, was punctured in the shell by the coupler of another car, and instantaneously released approximately 9,220 gallons of chlorine, creating a toxic vapor plume that engulfed the surrounding area.

The NTSB concluded that the probable cause of the accident was the failure of a train crew to return a main line switch to the normal position after the crew completed work at the Avondale Mills' industry track. As a result of the chlorine release, 5,400 people within a 1-mile radius of the derailment site were evacuated for several days. Nine persons were fatally injured and 554 sustained other injuries (75 requiring hospitalization). The nine persons fatally injured included the train engineer, six employees of the textile manufacturing facility, Avondale Mills, a truck driver at one of Avondale Mills' facilities, and an individual in a residence south of the accident site.

Noting that emergency responders were enroute to the scene within two minutes of the accident occurring and that emergency responders used a “particularly efficient and expeditious means” of evacuating affected persons, the NTSB concluded that the emergency response efforts were “timely, appropriate, and effective.”

The Board noted, however, that despite these emergency response efforts, the eight civilian fatalities were determined to have resulted from asphyxia that occurred within minutes of exposure to chlorine gas. In other words, the fatalities occurred within the minutes that passed before emergency responders arrived on the scene or were able, because of the toxic fumes, to begin a safe search and rescue effort.

As was the case in the Macdona accident, both train crew members survived the collision (the engineer died later from exposure to the gas). Given that both crew members survived the collision, no fatalities or serious injuries would have resulted from the accident had a tank car of chlorine not been punctured.

NTSB, Railroad Accident Report, NTSB/RAR-05/04,

Collision of Norfolk Southern Freight Train 192 With Standing Norfolk Southern Local Train P22 with Subsequent Hazardous Materials Release at Graniteville, South Carolina,

(Jan. 6, 2005), at p. 40, Available at

http://www.regulations.gov

in docket no. FRA-2006-25169 and at

http://www.ntsb.gov

(Graniteville Report).

Id.

The property damage, including damages to the rolling stock and track, exceeded $6.9 million. Other significant costs include: evacuation costs, truck delay, rerouting and associated out of service expenses, expenses for disruption to non-railroad businesses, costs to affected local governments and residents, as well as expenses incurred in settling claims arising from the accident. According to financial documents produced by NS, the railroad recorded $41 million of expenses related to the accident in 2005 and it is estimated that the costs of the Graniteville accident were approximately $138 million, excluding chlorine cleanup costs.

This cost estimate likely greatly underestimates the actual costs incurred by those affected by the accident. For example, according to various South Carolina State Emergency Operations Center and U.S. Environmental Protection Agency Situation Reports,

schools were closed for several days, mail service for the

evacuated areas had to be forwarded to a neighboring post office, and preliminary estimates of costs to Aiken County were in the millions due to potential damage to electrical systems and equipment within homes and businesses, the cost of the first response and recovery operations, damage to fire and EMS response vehicles, and the treatment of the victims.

Norfolk

Southern Corporation,

Quarterly Financial Review, Fourth Quarter 2006,

at p. 4. (downloaded at

http://www.nscorp.com/nscportal/nscorp/pdf/financial_q4_06.pdf

Available at

http://www.epa.osc.org.

The fate of Avondale Mills, the textile manufacturing company with four facilities within the vicinity of the accident, illustrates the significant long-term economic impacts that may result from catastrophic hazardous materials transportation accidents. In July 2006, after spending $140 million on cleaning, re-cleaning, repairs, and damage mitigation as a result of the derailment, Avondale Mills reported that it was unable to recover financially from the derailment and closed its 10 mills in South Carolina and Georgia. The company cited irrevocable damage to its core facilities, as well as market and production losses caused by the derailment. For example, the Company was unable to identify cleaning and restoration protocols that would successfully or economically halt the chlorine's corrosive effects, repair the damage caused by the chlorine exposure, and return the affected facilities and equipment to their pre-derailment condition. As a result, the Company was faced with the expensive replacement of damaged assets in addition to the lost business, higher manufacturing costs, and lower profits related to the reduction in productive capacities resulting from the derailment.

At the time of its closure, Avondale Mills employed approximately 4,000 people.

See

Avondale Incorporated,

Notes to Consolidated Financial Statements (Unaudited),

at note 1 (Aug. 25, 2006). Available at

http://www.sec.gov.

Although the costs of associated legal claims resulting from the derailment are still accumulating, in May 2006, Avondale Mills reached a $215 million settlement with its primary property and casualty insurer for all claims related to the derailment. Even with this multi-million dollar settlement, Avondale Mills' management believed that the amount was substantially less than the full value of the losses incurred as a result of the derailment.

In June 2006, a Federal judge approved a class-action settlement in excess of $10.5 million between Norfolk Southern and almost 500 individuals who claimed they suffered serious injuries after the derailment. In May 2005, Norfolk Southern announced that it had reached agreement on settlements for Graniteville residents and businesses that were evacuated as a result of the derailment, but did not seek medical attention. Under the terms of this settlement, Norfolk Southern offered each resident who was evacuated, but did not seek medical attention within 72 hours of the accident a flat amount of $2,000 for the evacuation plus $200 per person per day of the evacuation. These amounts are separate from any property damage claims. Norfolk Southern settled separately with the families of the nine people killed as a result of the accident.

Id.

On December 12, 2005, the NTSB released Safety Recommendations R-05-14 through R-05-17 as a result of the Graniteville accident. The first recommendation (R-05-14) pertains to railroad switching devices and is not directly relevant to this rulemaking. The three remaining Safety Recommendations (R-05-15, R-05-16, and R-05-17) relate to operating speeds in non-signaled territory, as well as the transportation of PIH materials and other hazardous materials that may pose inhalation hazards in the event of unintentional release. Recommendations R-05-15 through R-05-17 read as follows:

(R-05-15).

Require railroads, in non-signaled territory and in the absence of switch position indicator lights or other automated systems that provide train crews with advance notice of switch positions, to operate those trains at speeds that will allow them to be safely stopped in advance of misaligned switches.

(R-05-16).

Require railroads to implement operating measures, such as positioning tank cars toward the rear of trains and reducing speeds through populated areas, to minimize impact forces from accidents and reduce the vulnerability of tank cars transporting chlorine, anhydrous ammonia, and other liquefied gases designated as poisonous by inhalation.

(R-05-17).

Determine the most effective methods of providing emergency escape breathing apparatus for all crewmembers on freight trains carrying hazardous materials that would pose an inhalation hazard in the event of unintentional release, and then require railroads to provide these breathing apparatus to their crewmembers along with appropriate training.

In addition, noting that the punctured car was among the strongest tank cars in service, the NTSB concluded that even the “strongest tank cars in service can be punctured in accidents involving trains operating at moderate speeds.”

The NTSB then repeated its concern for crashworthiness integrity of railroad tank cars by restating what it said, in part, in response to the Minot accident:

Graniteville

Report at p. 51.

Improvements in the crashworthiness of pressure tank cars can be realized through the evaluation of alternative steels and tank car performance standards. The ultimate goal of this effort should be the construction of railroad tank cars that have sufficient impact resistance and that eliminate the risk of catastrophic brittle failures under all operating conditions and in all environments. Achieving such a goal does not necessarily require the construction of a tank car that is puncture-proof; it may only require construction of a car that will remain intact and slowly leak its contents if it is punctured.

Id.

E. FRA's Responses to the NTSB Tank Car Recommendations for Graniteville.

On June 30, 2006, the FRA responded to NTSB Safety Recommendations R-05-15 through R-05-17, which arose from the Graniteville accident. As for NTSB Recommendation R-05-15, which recommended that railroads be required, under certain conditions, to operate trains at lower speeds in non-signaled territory, the FRA informed the NTSB that the Recommendation was not feasible for operational and economic reasons. From an operational standpoint, depending on the terrain at the switches and the train make-up, train braking could prove difficult, generating excessive in-train forces that could cause derailments. From an economic standpoint, Recommendation R-05-15 would impede the movement of trains, especially on tracks where many switches exist, thereby causing train delays and an increase in running time. The FRA also explained that Recommendation R-05-15 was overly broad in that it would apply to all trains, regardless of lading. The NTSB classified Safety Recommendation R-05-15 as “Open—Response Received.”

As for NTSB Recommendation R-05-16, which suggested that FRA require railroads to position tank cars towards the rear of trains and reduce their speeds through populated areas, the FRA advised the NTSB that it would be imprudent to require the placement of tank cars carrying PIH materials at the rear of trains for several reasons. First, the placement of tank cars carrying PIH materials at the rear of trains could expose the cars to the consequences of rear-end collisions. Second, FRA's research demonstrates that the preferred location for loaded cars is towards the front of trains because, upon braking, heavy cars decelerate more slowly than empty cars. If loaded cars are placed towards the rear of trains, they would push the more rapidly decelerating cars

in front of them and generate higher buff forces. Finally, the switching of railroad cars to position tank cars containing PIH materials at the rear of trains involves the risk of increased yard accidents and employee injuries resulting from additional switching. In its response to NTSB Recommendation R-05-16, the FRA also noted several practical difficulties with slowing trains on a location-by-location basis (including the dangers of introducing additional train handling challenges, the impact of such a speed restriction on the efficiency and capacity of the rail network, as well as the potential negative effect that slowing operations could have on communities located along the track). Nonetheless, in its response, FRA stated that it would review the potential costs and benefits of slowing trains carrying certain toxic commodities. The NTSB classified Safety Recommendation R-05-16 as “Open—Response Received.”

As for NTSB Recommendation R-05-17, which recommended that FRA examine the most effective methods of providing emergency escape breathing apparatus for crewmembers on trains carrying PIH materials, FRA explained to the NTSB that it would initiate a study of potential breathing apparatus for use by crewmembers of freight trains carrying TIH materials. Based on FRA's response to Safety Recommendation R-05-17, the NTSB classified the Recommendation as “Open—Acceptable Response.”

The NTSB Safety Recommendations referenced in this section above and the publicly available responses to them may be found on the

http://www.regulations.gov

Web site under docket number FRA-2006-25169.

VII. Evaluating the Risk Related to Potential Catastrophic Releases From PIH Tank Cars in the Future

Although it is not possible to accurately determine the probability of future occurrences of railroad accidents that would result in the catastrophic release of hazardous materials, it is unrealistic to assume that absent the improvements proposed, consequences from future accidents involving hazardous materials tank cars would be of the same order of frequency and severity as in the past. In fact, absent the improvements proposed, one or more events could be significantly more severe than experienced thus far. All that would be required would be the necessary environmental conditions (concentrating and channeling a gas plume at ground level), an exposed population of scores or hundreds within the path of the plume, and an ineffective or delayed emergency response (either due to deficiencies in the emergency response process or because of safety risks posed to emergency responders prohibiting emergency responders from entering an accident area).

Each of the three accidents discussed in section VI above share certain similarities that effectively minimized the catastrophic results of the accidents. Each accident occurred in a relatively rural area, thereby limiting the population exposed to the hazardous materials release. Each accident occurred during the early morning hours, while most of the surrounding populations were in their homes and not in the immediate accident vicinity. The meteorological conditions at the time of each accident effectively limited the speed at which the resulting toxic plumes expanded and the distance over which the plumes expanded. Had any of the accidents occurred in a more densely populated area or later in the day, it is likely that many more people would have been exposed to the toxic plumes. Had the meteorological conditions at the time of any of the accidents been different (e.g., wind speed or direction, temperature, barometric pressure, or humidity) it is possible that the plumes could have expanded more than what actually occurred, again, exposing many more people to the toxic chemicals. To demonstrate the potential affects of different accident conditions, such as location, time of day, or the weather, the circumstances surrounding the Graniteville and Minot accidents are discussed below.

A. Graniteville

Graniteville is a mixed rural and suburban area of Aiken County, South Carolina, with a population of approximately 7,000.

Graniteville lies in a relatively shallow valley, approximately 200 feet above sea level. The terrain surrounding the accident site is approximately 225 feet above sea level, with the elevation of the industry track where the accident occurred moderately decreasing as the track extends north and west towards the Avondale Mills plant. The January 6, 2005, accident occurred at 2:30 in the morning, a time at which most individuals were asleep in their homes and very few individuals were on the premises of the Avondale Mills plant. At the time of the accident, a light wind was blowing in a south-southwest direction, the temperature was approximately 55° F, and humidity was high.

As of 2006, the approximate population of Aiken County was 152,000. U.S. Census Bureau, State & County QuickFacts (available at

http://quickfacts.census.gov

The NTSB concluded that approximately 120,000 pounds (9,218 gallons) of liquefied chlorine was released before emergency responders arrived on the scene.

The chlorine settled in low areas around the railroad tracks and the plume expanded to the west of the accident site and into the Avondale Mills plant, generally following the local topography, running downhill to the south and west,

before being blown to the north by light winds where it hovered. The NTSB concluded that based on emergency responder observations and the locations of those receiving fatal injuries, the cloud extended at least 2,500 feet to the north; 1,000 feet to the east; 900 feet to the south; and 1,000 feet to the west.

Note: The vaporization of liquefied chlorine at 32 °F at atmospheric pressure can generate a gaseous cloud with a volume 450 times greater than the volume of the liquid released.

See

Graniteville NTSB Report at 49 (citation omitted).

Because chlorine gas is heavier than air with a vapor density of 2.5 at 32 °F, it will seek the lowest point in the immediate area.

The area to the east of the accident site and extending in a southerly direction is primarily a residential area. To the west and extending in a northerly direction are several moderate- to large-sized industrial plant facilities, some of which operate continuously. A small commercial/retail district is just north of the accident site.

Given the demographics and topography surrounding the accident site, had the accident occurred at a different time of day, or had any of the meteorological variables been different (e.g., wind speed or direction, temperature, barometric pressure, or humidity), it is likely that many more people would have been exposed to the chlorine plume. For instance, if the accident had occurred while the Avondale Mills plant was fully staffed, or during an afternoon shift change, hundreds of individuals could have been exposed. In addition, a middle school is located approximately 1,000 feet north of the accident site (well within the area of the plume that did occur). Had the accident happened while school was in session, approximately 500 students and scores more school personnel could have been exposed to the toxic plume.

Similarly, had any meteorological variables been different (e.g., wind speed or direction, temperature, barometric pressure, or humidity), it is likely that the chlorine plume could have expanded more rapidly and affected a greater area than it did. For instance, at the time of the accident, a

light wind was blowing in a south-southwest direction. If the wind had been blowing at the same intensity, but in a south-southeast direction, the chlorine plume could have hovered over the southeasterly side of the accident site, rather than the northwesterly side. Southeast of the accident site is primarily a residential area and given the size of the plume that did result, the plume could have endangered approximately 185 homes. Given the average household size of 2.68 in Aiken County,

almost 500 people to the southeast of the accident site could have been exposed to vapors above the ERPG-3 level causing significantly more casualties and fatalities.

We note as well that the high humidity at the time of the accident limited the plume's rate of expansion because the chlorine reacted with the moisture in the area (effectively diluting the chlorine) to form a weak hydrochloric acid. This weak hydrochloric acid, a highly corrosive liquid, then accumulated in low lying areas and on the abundant vegetation surrounding the accident site, limiting the expansion of the plume. At the time of the accident the outside temperate was approximately 55 °F. As the NTSB noted, the liquefied chlorine rapidly vaporized and expanded when it spilled from the tank car, but the sudden release of the gas caused the product remaining in the tank car to auto-refrigerate and remain in a liquid state, slowing the release of additional gas.

Had it been warmer, the higher temperature could have provided additional energy for the chlorine to expand, and it is likely that the chlorine plume would have expanded faster.

U.S. Census Bureau, American FactFinder (available at

http://factfinder.census.gov).

“ERGP-3 level” refers to the American Industrial Hygiene Association's (AIHA) Emergency Response Planning Guideline level 3 which means “[t]he maximum airborne concentration below which it is believed that nearly all individuals could be exposed for up to one hour without experiencing or developing life-threatening health effects.”

See

AIHA, Emergency Response Planning Committee,

Procedures and Responsibilities,

at 1 (Nov. 1, 2006) (downloaded from

http://www.aiha.org).

According to AIHA the ERGP levels are intended as health based guideline concentrations for single exposures to chemicals and the levels are commonly used in the emergency response planning industry for assessing the adequacy of accident prevention and emergency response plans.

Id.

Graniteville Report at 11, 49.

B. Minot

The Minot accident occurred at approximately 1:30 in the morning, a time at which most individuals were sleeping inside their homes with their windows closed. Almost instantaneously, approximately 146,700 gallons of anhydrous ammonia were released as five tank cars catastrophically ruptured. A toxic vapor plume formed almost immediately. The plume rose approximately 300 feet and gradually expanded five miles downwind of the accident site and over a population of about 11,600 people (approximately one-third the population of the City of Minot). The outside temperature at the time of the accident was −6 °F, a light snow had fallen earlier in the day and a large amount of residual snow was on the ground.

Recognizing the smell of the chemical, the responsible fire chief immediately determined that the leaking material was anhydrous ammonia. Because of the large amount of anhydrous ammonia released, emergency responders were unable to enter the accident area for approximately three hours. Within 15 minutes of the accident, however, 911 operators were advising residents in the affected area to shelter-in-place (i.e., remain inside their homes with the windows shut) and the emergency room of a local hospital was notified of the derailment.

Upon notification of the derailment, the hospital activated its disaster plan and staff secured the facility against the hazardous vapors by shutting down air handlers, setting up a portable air-handling unit in the emergency room, and establishing an alternate emergency room entrance away from the vapor cloud. Within three hours of the accident, the ammonia cloud had drifted to and encompassed the hospital. Nevertheless, throughout the incident, the hospital treated approximately 300 people.

Ultimately, one resident of the neighborhood nearest the derailment site was fatally injured, two residents were seriously injured, and 60-65 residents were rescued hours after the derailment. All three residents that were seriously injured left the protective confines of their homes and were directly exposed to the anhydrous ammonia cloud for a prolonged period of time (given the time of day and widespread power outages as a result of the accident, it is unknown whether these individuals had heard or seen any of the emergency directives to shelter-in-place). As a result of the accident, nine other people sustained serious injuries, and 322 people, including the two train crew members, sustained minor injuries.

The NTSB concluded that sheltering-in-place was an effective emergency response and credited this action with the relatively low number of injuries, as compared to the number of persons affected by the vapor plume (approximately 330 injuries in 11,600 persons affected). However, had this accident happened at another time of day, possibly during the morning commuting hours when people are generally not at home, or if emergency responders did not promptly direct residents to shelter-in-place, or if the local hospital had not taken appropriate measures to protect itself from the plume, the consequences of the release could have been much worse than what occurred on January 18, 2002.

Similar to the meteorological circumstances surrounding the Graniteville accident, had the atmospheric variables been different (particularly, the temperature at the time of the accident), it is likely that many more people could have been at risk of exposure to the toxic plume. The low atmospheric temperature at the time of the accident helped to keep the ammonia plume close to ground level as it traveled downwind and also minimized the chemical's vaporization, accordingly limiting the spreading of the plume. Had this accident happened in the spring or summer, or any other time of warmer temperatures, windows in the homes may have been open and it is likely that the ammonia plume would have expanded more rapidly, thus exposing a greater population to the chemical.

Although the Minot, Macdona, and Graniteville accidents each occurred during the early morning hours, while most of the surrounding populations were in their homes and not in the immediate accident vicinity, because hazardous material transportation is not limited to early morning transportation, any of the accidents could have occurred later in the day, when neighboring factories were fully staffed, schools were in session, and unsuspecting individuals were otherwise outside of the protective confines of their homes and workplaces going about their daily routines. As an example, at approximately 11 a.m. on October 10, 2007, a CSX train transporting mixed freight of grain, lumber, and tank cars of various hazardous materials, derailed in Painesville, Ohio,

resulting in an explosion and subsequent fire as hazardous materials were released to the environment. Although the train was reportedly not carrying any toxic inhalation hazard materials, and no injuries were reported, 600 people

(including over 300 children from a nearby elementary school) within a half mile radius of the train derailment were evacuated.

Painesville is located approximately 30 miles from Cleveland and has an estimated population of 20,000.

Although the Minot, Macdona, and Graniteville accidents each occurred in a relatively rural area, the accidents could have occurred anywhere, including in the midst of major metropolitan areas. The Minot accident was caused by an undetected defective rail. A crew's failure to appropriately respond to wayside signals governing movement of their train led to the Macdona accident. The Graniteville accident was caused by a train crew's failure to correctly align a switch. Each of these “causes” could have occurred in close proximity to a metropolitan area, thus potentially impacting a much larger population of people. The Painesville, Ohio, incident, although not an accident with catastrophic results, illustrates this point. As a Cleveland City Councilman noted, had the derailment occurred closer to Cleveland, more than 8,000 people could have been affected.

David Summers, WKYZ-TV (Cleveland, Oh), Hazardous Cargo Legislation Stalled on the Tracks (Oct. 14, 2007).

VIII. The Railroad Industry's Liability and the Impact of Accidents Involving the Shipment of PIH Materials on Insurance Costs and Shipping Rates

In 2005, railroads moved just over 100,000 carloads of PIH materials and nearly 37 million total carloads.

The 100,000 carloads of PIH materials equate to approximately 0.3 percent of all rail carloads. Despite the small fraction of the railroad industry's business constituted by PIH materials (and the limited revenue it generates), railroad industry representatives, citing the Minot, Macdona, and Graniteville accidents, have noted that transporting PIH materials has led to the imposition of “hundreds of millions of dollars of liability.”

Further, noting that “railroads can suffer multi-billion dollar judgments” from accidents involving highly-hazardous materials, in 2007 the President and CEO of AAR testified before a Congressional committee that “every time a railroad moves [a highly-hazardous shipment] it faces potentially ruinous liability” and that the “insurance industry is unwilling to insure railroads against the multi-billion dollar risks associated with highly-hazardous shipments.”

In support of this assertion, a representative of the railroad industry noted that as a result of the Minot, Macdona, and Graniteville accidents, insurance costs for the entire railroad industry have gone up by 100 percent.

Written Statement of Edward R. Hamberger, President & CEO, AAR , before the U.S. House of Representatives Committee on Transportation and Infrastructure, Subcommittee on Railroads, Pipelines, and Hazardous Materials (Jan. 31, 2007) at 7 (Hamberger Statement).

Statement of Bob Fronczak, Assistant Vice President, Environment and Hazardous Materials, AAR, at the Dec. 14, 2006 public meeting (Fronczak Statement).

See

document no. 19 in the docket.

Hamberger Statement at 7-8. An example of such a judgment is

In re New Orleans Train Car Leakage Fire Litigation,

795 So. 2d 364 (La. Ct. App. 2001). In that case, the Louisiana Court of Appeals upheld a class-action judgment of $850,000,000 in punitive damages and $2,100,000 in compensatory damages against CSX Transportation, Inc. Railroads, as common carriers, are generally required to provide transportation services in a reasonable manner and may not refuse to transport a material that the government has deemed safe for transportation.

Fronczak Statement.

This increase in railroad insurance rates, coupled with the actual costs of the accidents, has resulted in increased shipping rates for the shippers of hazardous materials. Minimally, shipping rates for PIH materials have doubled; however, many shippers report larger increases (including at least one shipper which has had its rates increased over 4.8 times in a two-year period).

IX. Industry Efforts To Improve Railroad Hazardous Materials Transportation Safety

A. General Industry Efforts

The rail industry, through the AAR, has developed a detailed protocol on recommended railroad operating practices for the transportation of hazardous materials. Although in early 1990 this protocol was implemented by only the Class 1 rail carriers operating in the United States, on July 17, 2006, AAR issued a revised version of this protocol, known as Circular OT-55-I, with short-line railroads also participating in the implementation. The Circular details recommended railroad operating practices for, among other things: (1) Designating certain trains hauling hazardous materials as “key trains,” defined as trains containing five or more tank car loads of PIH materials; (2) designating operating speed and equipment restrictions for key trains; (3) designating “key routes”

for key trains and setting standards for track inspection and wayside detectors on these “key routes”; (4) yard operating practices for handling placarded tank cars; (5) storage, loading, unloading and handling of loaded tank cars; (6) assisting communities with emergency response training and information; (7) shipper notification procedures; and (8) the handling of time-sensitive materials. The Circular also (1) Restricts key trains to a maximum speed of 50 mph; (2) requires, as practicable, that unless a siding or auxiliary track meets FRA Class 2 standards, a key train will hold main track at meeting or passing points; (3) requires all cars in key trains to be equipped with roller bearings; and (4) imposes a further speed restriction of 30 mph in the event a defect in a key train bearing is reported by a wayside detector, but is not able to be confirmed visually. A copy of the most recent version of Circular OT-55-I has been placed in the docket.

Circular OT-55-I defines the term “key routes” as “[a]ny track with a combination of 10,000 car loads or intermodal portable tank loads of hazardous materials, or a combination of 4,000 car loadings of PIH or TIH (Hazard zone A, B,C, or D), anhydrous ammonia, flammable gas, Class 1.1 or 1.2 explosives, environmentally-sensitive chemicals, Spent Nuclear Fuel (SNF), and High Level Radioactive Waste (HLRW) over a period of one year.”

In addition, FRA is aware that some carriers have individually taken voluntary steps to reduce the occurrence of accidents that can lead to hazardous material releases. For example, BNSF has implemented a derailment prevention program that includes, among other efforts, implementing advanced train control technology; utilizing various freight car condition monitoring technologies; and installing and maintaining switch point position indicators and broken rail protection in non-signaled territory. Specific to the transportation of hazardous materials through non-signaled territory, BNSF has also revised its operating practices at certain locations in its system through which a significant amount of PIH materials are transported in an effort to decrease the probability of an accident or incident involving a train hauling PIH material. A more detailed discussion of BNSF's efforts in this regard is found in the “Discussion of Public Comments” section below.

B. Trinity Industries, Inc.'s Special Permit Chlorine Car

In accordance with 49 CFR 107.105, in early 2005, Trinity Industries, Inc. (Trinity) applied for a Special Permit to manufacture, mark, and sell DOT 105J600W specification tank cars, for use in chlorine service, with a variation in design and construction of the protective housing (the “Trinity car”).

Specifically, as noted in Trinity's

application, the Trinity car varies from Federal standards because it has a protective housing welded, rather than bolted, to the tank nozzle and its maximum gross weight on rail is 286,000 pounds (due in part to a thicker head and shell than current chlorine cars).

In response to Trinity's application, several members of the hazardous materials shipping industry expressed concern with certain aspects of the proposed Trinity car. For example, commenters expressed concern regarding the proposed manway arrangement, noting that the modified pressure plate and protective housing may present difficulties for emergency responders because it was unclear whether the standard Emergency Kit C, which is used to contain leaks in and around the pressure relief device and angle valves, was compatible with the arrangement. Further, commenters expressed concern regarding the increased car pressure and corresponding pressure rating of the valves and fittings. Commenters also questioned the efficacy of increasing the thickness of the car's steel, but utilizing steel with a lower tensile strength than current chlorine cars. Furthermore, commenters expressed concern that given the increased weight of the car, some shipping and receiving facilities may not be able to handle the heavier car.

See

70 FR 12782, 12783 (Mar. 15, 2005) (Research and Special Programs Administration, List of Applications for Exemption). 49 U.S.C. § 5117 authorizes the DOT to issue special permits (previously referred to as “exemptions”) authorizing a variance from the HMR if the proposed variance is equivalent to the level of safety required by the HMR.

The HMR require bolted top fittings and provide for a tank car maximum gross weight on rail of 263,000 pounds.

See

49 CFR 179.100-12 and 179.13.

After careful review of Trinity's application, the comments received, and DOT's own analysis of the Trinity car, PHMSA issued the requested Special Permit on April 20, 2006, authorizing Trinity to manufacture, mark, and sell the car for use in chlorine service, subject to certain operational restrictions and inspection requirements.

Specifically, the terms of the Special Permit prohibit the Trinity car from being used in free interchange and require the manway nozzle welds to be requalified annually. The Special Permit was issued based on the finding that the Trinity car used under the specified conditions would provide an equivalent level of safety to current DOT specification cars and additionally would provide a way to gather data about an alternative to a regulatory standard over a relatively short time-span.

See

71 FR 47288, 47301 (Aug. 16, 2006) (PHMSA Special Permit number DOT-SP 14167). Subsequently, the Special Permit was revised on August 10, 2006 to clarify the outage and filling density requirements and specify requirements for filing agreements between carriers and filing non-destructive testing procedures. More recently, Trinity requested that the Special Permit be revised to amend the manway protective housing design.

C. AAR Proposals for Enhanced Chlorine and Anhydrous Ammonia Tank Cars

In early 2006, the Safety and Operations Management Committee (SOMC) of the AAR directed the AAR's TCC to consider improved packaging for the shipment of chlorine and anhydrous ammonia. Specifically, SOMC directed the TCC to present a plan for developing performance standards for chlorine and anhydrous ammonia tank cars that would reduce the conditional probability of a release, given an accident, by a target of 65% from the current values, as well as a plan to phase in the new improved cars within a target time frame of five to seven years. The goal of a 65% reduction was based on the findings of researchers at the University of Illinois at Urbana-Champaign's Railroad Engineering Program, which concluded that utilizing existing technology, the probability of a release of anhydrous ammonia and chlorine from a tank car involved in an accident could be reduced by 65% or more by substituting enhanced tank cars for the cars currently used to transport these materials.

The enhanced tank car contemplated in the University of Illinois research is the thicker, heavier Trinity car designed for chlorine service and subject to PHMSA Special Permit 14167. As noted in the AAR Risk Analysis, the finding of a potential 65% improvement is premised on replacing the current 263,000 pound cars for anhydrous ammonia and chlorine with 286,000 pound cars equipped with additional head protection, thicker shells, and modified top fittings protection.

Christopher P.L. Barkan, Ph.D., M. Rapik Saat, M.S., Railroad Engineering Program, Department of Civil and Environmental Engineering, University of Illinois at Urbana-Champaign,

Risk Analysis of Rail Transport of Chlorine and Ammonia on U.S. Railroad Mainlines

(Feb. 27, 2006) (AAR Risk Analysis).

In response to this directive, the TCC established a task force to develop the requested plan. The task force consisted of a wide spectrum of interested parties, including hazardous material shippers, railroads, the Railway Supply Institute (RSI), and railroad industry consultants. The task force, however, was unable to reach consensus on a recommendation to the TCC.

In July 2006, the AAR TCC considered proposals for improved tank cars in light of its mandate from SOMC to make the cars transporting chlorine and anhydrous ammonia 65% safer. At the July TCC meeting, all member railroads, supported by Trinity, proposed that anhydrous ammonia be transported in DOT 112J500W tank cars, equipped with full-height half-inch thick or equivalent head shields and top fittings protection designed to withstand a rollover with a minimum linear velocity of nine miles per hour. Similarly, the same parties proposed that chlorine be transported in tank cars built to the 105J600W specification, equipped with full-height half-inch thick or equivalent head shields and top fittings protection designed to withstand a rollover with a minimum linear velocity of nine mph. Alternatively, cars for each commodity could be designed in accordance with a formula derived from the statistical analysis in the RSI-AAR Tank Car Safety Project Report RA 05-02.

For anhydrous ammonia, this statistical formula required shell and head protection to reduce the conditional probability of release (CPR) by 32% given that the car is derailed in an accident; for chlorine, the statistical formula required shell and head protection to reduce the CPR by at least 45%.

This railroad/Trinity proposal contemplated that 50% of a car owner's fleet of anhydrous ammonia and chlorine cars would be replaced with these “enhanced cars” within approximately six years, with their entire fleets being replaced within approximately eleven years.

RSI-AAR Railroad Tank Car Safety Research and Test Project,

Safety Performance of Tank Cars in Accidents: Probabilities of Lading Loss,

RA-05-02 (Jan. 2006).

While this statistical analysis sought to advance the safety of tank cars, it does not foster new technology because the CPR was derived from empirical data.

At the same TCC meeting, all shipper members of the TCC, as well as every rail tank car builder other than Trinity, supported a proposal submitted jointly by The Fertilizer Institute (TFI) and the Chlorine Institute (CI). The TFI/CI proposal for cars constructed after a proposed effective date incorporated the Federal standard for head protection (49 CFR 179.16), with the ram car adjusted to reflect the increasing presence of cars with a gross rail load of 286,000 pounds. The TFI/CI proposal contemplated grandfathering existing cars in anhydrous ammonia and chlorine service prior to the effective date as compliant.

The initial result of this deliberation was the TCC's issuance of Casualty Prevention Circular 1175 (CPC-1175) on July 28, 2006. CPC-1175 proposed to implement the railroad/Trinity proposal introduced at the July TCC meeting. In response to CPC-1175, several members of the hazardous materials shipping

industry submitted comments to the AAR expressing concern with certain aspects of the proposal. For example, commenters expressed concern with the proposed implementation schedule, the proposed top fittings arrangement, and the scientific basis utilized for development of the standard. Commenters also questioned the efficacy of moving forward with the proposal without the benefit of the results of the FRA's Volpe research designed to quantify tank car survival conditions.

FRA also corresponded with the AAR in response to CPC-1175. In its letters, FRA first noted that the Circular contained two proposed, amended tank car specifications and two proposed, new specifications. Accordingly, FRA noted that before the TCC could implement the proposed requirements in CPC-1175, in accordance with 49 CFR 179.4, the proposals would have to be submitted to the Department. The FRA also expressed concern regarding the engineering analysis underlying the proposal, specifically related to the analysis of the top fittings, tank-head and shell, as well as the tank car's capacity.

In response to comments received from FRA and the industry, on October 18, 2006, the TCC issued Casualty Prevention Circular 1176 (CPC-1176), which adopted as a final TCC action the proposals set forth in CPC-1175 with minor modifications to the implementation period initially proposed. Specifically, the intermediate implementation goal of CPC-1175 (50% of the fleet by December 31, 2012) was eliminated and replaced by a requirement that the tank car owners' plans for implementation be submitted to AAR by December 31, 2007. Subsequently, on December 18, 2006, AAR issued Casualty Prevention Circular 1178 (CPC-1178) in response to appeals to CPC-1176. Although various aspects of CPC-1176 were appealed (e.g., the proposed implementation schedule, top fittings arrangement, and the scientific basis of the proposed design), CPC-1178 is substantially the same as CPC-1176, except the target implementation dates were delayed by one year (i.e., tank car owners' plans for implementation were required to be submitted by December 31, 2008 and tank cars were required to be 100% fleet compliant by December 31, 2018).

On August 28, 2007, the TCC issued Casualty Prevention Circular 1180 (CPC-1180) for public comment. CPC-1180 addresses certain high-hazard materials (including chlorine and anhydrous ammonia). CPC-1180 proposes an implementation period for a top fittings requirement consistent with that of CPC-1178, but also includes requirements for commodity specific tank improvement factors. The tank improvement factor requirements are new requirements for chlorine and anhydrous ammonia.

D. Dow/UP Safety Initiative and the Next Generation Rail Tank Car Project

In October 2005, the Dow Chemical Company (Dow) and UP, Dow's largest rail service provider, formed a partnership to address rail safety and security improvements for the transportation of hazardous materials. Specific goals of the agreement between UP and Dow include: (1) Reducing idle times for hazmat shipments by 50 percent in high-threat urban areas; (2) redesigning Dow's customer supply chains to cut in half the amount of “highly hazardous chemicals” shipped by 2015; (3) eliminating all nonaccidental leaks of certain hazardous materials in three years; and (4) having hazardous material shipments monitored by satellite tracking tags and other sensors.

As Dow noted at the May 31-June 1, 2006, PHMSA/FRA public meeting, the companies' joint effort focuses on six areas for improvement: (1) Supply chain redesign; (2) next generation rail tank car design; (3) improved shipment visibility; (4) a strengthened commitment to TRANSCAER®;

(5) improved rail operations safety; and (6) hazardous material shipment routing.

John D. Boyd,

UP, Dow Sign Safety Pact,

Traffic World (Mar. 19, 2007).

TRANSCAER® (Transportation Community Awareness and Emergency Response) is a voluntary national outreach effort that focuses on assisting communities to prepare for and respond to a possible hazardous materials transportation incident. TRANSCAER® members consist of volunteer representatives from the chemical manufacturing, transportation, distributor, and emergency response industries, as well as the government. For more information on TRANSCAER® see

http://www.transcaer.com/public/about.cfm.

With regard to supply chain redesign, Dow is evaluating potential ways to reduce the number and distance of shipments involving high-hazard materials. In this connection, Dow is evaluating the potential for co-location of production and consuming facilities; the use of pipelines instead of rail in some instances; and the conversion of highly hazardous products to less hazardous derivatives before shipping.

At the same public meeting, Dow noted that since 1999, the company has reduced the amount of chlorine it ships in the United States by 80%. Dow also noted that the company's current commitment is to have further reduced by 50 percent the number of shipments of highly hazardous materials (i.e., PIH materials and flammable gases) and container miles traveled by those shipments by 2015. Recognizing that the temperature, pressure, and other characteristics of the material being shipped affects the consequences of any hazardous materials release, Dow is also focusing its efforts on improving shipment visibility and tracking. Specifically, by the end of 2007, Dow's stated goal is to have implemented shipment tracking via GPS technology to know, in real time, exactly where its tank cars containing PIH materials are located and what condition they are in. Through TRANSCAER(r), Dow has also publicly committed to “touch every community” through which its highly hazardous materials travel within the next five years. Through this initiative, Dow's stated intent is to provide community awareness and emergency responder training to help ensure that the communities through which their highly hazardous materials travel are better prepared for potential chemical transportation emergencies.

See

Transcript of May 31-June 1, 2006, public meeting in docket no. FRA-2006-25169.

We invite commenters to provide data and information concerning the extent to which other companies are voluntarily implementing measures to reduce the transportation safety risks associated with the transportation of PIH materials in tank cars. We are particularly interested in efforts planned or underway to modify or redesign supply chains, reduce the number of shipments and the time-in-transit of shipments, or enhance shipment visibility and tracking. We ask commenters to consider whether implementation of these and similar risk-reduction measures industry-wide would militate against the need to improve the accident survivability of the current PIH tank car fleet, as proposed in this NPRM.

With regard to improving rail tank car design, Dow, UP, and the Union Tank Car Company (Union Tank), which had joined the Dow/UP Partnership specifically to participate in the NGRTCP, initiated the NGRTCP for the stated purpose of collaborating on the design of a next generation railcar for the transportation of certain hazardous materials. The project is multi-generational with the first generation focusing on designing a breakthrough next generation tank car for the transport of PIH materials that will meet or exceed the AAR TCC performance requirements and provide a five- to ten-fold improvement in the safety and security performance of existing rail tank cars in PIH service. Subsequent generations of the project would build on the first generation to leverage the process, methodology, and criteria used in designing the next generation PIH tank car to design a tank car appropriate

for other hazardous materials, such as flammable gases or chemicals that pose a significant risk to the environment if released. Dow's stated goal is full implementation within the company of a next generation PIH tank car by the end of 2014, and full implementation of further generations of tank cars for flammable gases and environmentally-sensitive chemicals by the end of 2029.

The NGRTCP team includes industry leaders and representatives from Dow, UP, Union Tank, as well as an external advisory panel of academic, industry, and former regulatory leaders to help guide the development of the next generation rail tank car design. Recognizing the significant opportunities to leverage government and industry resources in designing this next generation rail tank car, in January 2007, FRA signed a Memorandum of Cooperation (MOC) with the companies involved in the NGRTCP. This MOC provides for extensive information sharing and cooperation between ongoing FRA and industry research programs to improve the safety of rail shipments of hazardous commodities such as PIH materials. FRA hazardous materials safety and R&D personnel are actively involved in the project.

The MOC was amended in early 2007 when Transport Canada joined the project.

The NGRTCP is following a six sigma approach (i.e., a data driven approach and methodology for eliminating defects) to tank car design, evaluating such issues as: (1) Coupler penetration to tank sides and heads; (2) hydrostatic failure; (3) ability of tanks to withstand ballistic impacts; (4) fittings protection; (5) operational efficiency (including payload, infrastructure, maintenance and re-qualification); as well as (6) fire and thermal protection. Recognizing that the traditional method of enhancing tank car survivability (i.e., utilizing thicker, stronger steel) is limited, the project is evaluating the use of alternative technologies and design concepts from other industry sectors (e.g., automotive and aerospace). The general framework for the modeling and testing contemplated by the NGRTCP consists of the use of quantitative analysis (computer simulation using finite element analysis), component testing, quarter- to half-scale model testing, and limited full-scale testing. The project also involves a comparison of any potential new design with existing designs (e.g., the DOT 105A500W base car, the DOT 105J600W tank car with full head shields and top fittings protection).

Additional discussion of the NGRTCP may be found in the “Discussion of Public Comments” section below and in the transcript to the December 14, 2006, public meeting (document no. 22 in FRA docket no. FRA-2006-25169).

E. The Chlorine Institute (CI) Study

In late 2005, CI established a research program to investigate tank car puncture resistance and the potential development of alternative materials tests (e.g., un-notched Charpy test) to develop and validate alternative fracture criteria. The CI study recognizes that considerable advances have been made in the design of tank car steels to improve and increase the ductile-to-brittle transition temperature and that these improvements have resulted in recent tank car failures occurring in a ductile fashion due to an overload of the tank. The CI research is looking at several alternative strategies to increase the ductile performance of tank car design, including the development of novel material tests to better establish a relationship between overloading and material failure from specimens that do not include a pre-existing crack. This information will be used to refine how modeling of tank car failures occurs and to help with the evaluation of the alternative strategies being reviewed.

X. Discussion of Relevant Tank Car Research

The process of improving the safety of railroad tank cars has been ongoing for decades. It involves railroads, tank car builders, chemical companies, and government regulators. Historically, FRA has conducted, and continues to fund and co-fund, a substantial amount of tank car safety research and development projects with Transport Canada, as well as with RSI and AAR, through their cooperatively funded RSI-AAR Railroad Tank Car Safety Research and Test Project. The RSI-AAR Railroad Tank Car Safety Research and Test Project conducts tank car safety research in two principal ways: (1) By maintaining a comprehensive database on the details of the damage suffered by tank cars in accidents, to enable better understanding of tank car design strengths and weaknesses, and (2) by conducting engineering analyses of specific problems. The FRA further collaborates with industry through the TCC to develop standards for designing, constructing, maintaining, and safely operating railroad tank cars in North America.

Historically, the Department's research has focused on developing information on damage tolerance for tailoring inspection intervals for specific tank car designs; developing non-destructive evaluation and testing techniques and methodologies; improving fittings protection and gaskets; reviewing tank car operating environments; and developing new linings, coatings, and tank car steels. Since the 1970s, based on the combined research efforts of the Department and industry, DOT has issued a number of regulations to improve the survivability of tank cars in accidents. For example, DOT has promulgated regulations requiring the installation of tank-head puncture-resistance systems (head protection), coupler vertical restraint systems (shelf couplers), insulation, and thermal protection systems on tank cars used to transport certain hazardous materials.

Despite these safety improvements, as noted above, in the last several years there have been a number of rail tank car accidents in which the tank car was breached and product was lost on the ground or into the atmosphere. FRA's research focus changed after the tragic occurrence of these accidents. Specifically, as discussed in Section VI above, the NTSB issued seven safety recommendations to FRA as a result of the Minot derailment. Four of these recommendations concern tank car structural integrity (R-04-04, R-04-05, R-04-06, and R-04-07), and these four recommendations served as the basis for the reformulation of FRA's tank car research and development program. The current FRA tank car research program objective is the development of effective strategies to maintain tank integrity during train derailments or accidents. The key metric identified for this research is the maximum speed for which tank integrity is maintained. This metric has been identified because of the comparable ability for other researchers to perform large deformation analysis. Ascertaining the specifics of material failure through analysis is still extremely challenging. The ability to model tank car integrity with confidence will be critical to the ability of tank car manufacturers to develop new designs that conform to the performance standards proposed in this NPRM.

Specifically, in response to NTSB's Minot recommendation R-04-07, work was conducted on the testing of tank car steels to examine the dynamic fracture toughness of such steels as a function of service temperature. This work included standardized fracture mechanics tests and the comparison of results from these tests with Charpy V-notch impact energies at different temperatures. Due to inherent material variability, the results from the fracture toughness tests are scattered by a factor of four, which would require a safety factor of at least 2 in a quality assurance (QA)

specification. This means, for example, the samples taken from a production heat of steel would have to average at least twice the toughness needed for service.

Tightening the QA on steel products can result in inordinately expensive steel costs and most likely would be cost prohibitive. Alternatively, an unacceptable gain in structure weight may be required to sufficiently decrease the applied stresses to meet the safety factor with achievable material performance. Additionally, a specification will not provide an absolute guarantee of safety because, despite the implementation of any QA specification, some materials released from production may not meet the minimum fracture toughness standard. Accordingly, although FRA is in the process of completing the dynamic fracture toughness testing, it does not appear that a workable steel specification could be developed based on the results. Instead, in this NPRM, the Department has chosen to explore advances in tank car safety through engineering redesign of tank car structures to increase the amount of energy absorption a tank car experiences prior to a breach. The Department will continue to examine the dynamic fracture toughness of steels used in the construction of pressure tank cars in hazardous materials service and will incorporate any workable tank car design-specific fracture toughness standards into the HMR as appropriate in future rulemakings.

Also in response to NTSB's Minot recommendations, a risk model framework was developed to provide the technical basis to rank the factors affecting catastrophic failure of tank cars in derailments or collisions. The risk model framework focuses on determining whether the risk of lading loss in an accident situation could be minimized by specifying a particular material, e.g., normalized versus non-normalized steel. A hierarchal approach (i.e., Level 1, Level 2, and Level 3) was applied and as research results become available they will be incorporated.

In Level 1, a qualitative ranking is conducted by identifying the factors that are perceived to affect risk. These factors are then grossly sorted in terms of their expected impact on risk (e.g., high, medium, or low impacts). A simple Level 1 risk ranking has been completed. In Level 2, a systematic framework will be developed to provide a technical basis for ranking the risk factors. In this semi-quantitative method, a probabilistic approach will be used to account for uncertainties due to physical randomness and/or limited information. Different probability distributions (e.g., normal, Weibull, triangular, etc.) have been used to assess various uncertainties in the model. In Level 3, a quantitative risk ranking, the information obtained from other research programs will be incorporated with the goal of ranking tank cars that are perceived to be the most vulnerable to catastrophic failure. Although material properties play an important role in the performance of a tank car subjected to fatigue type loading, for overload conditions such as those experienced in collisions or derailments, the ranking developed is not expected to provide a tool for improving tank car performance. Instead, as noted above, in the NPRM, the Department has chosen to examine the potential redesign of the tank car structure to minimize the effect of the overload conditions, e.g., to absorb more energy prior to incipient rupture and spread the load over as large an area as possible.

Currently, FRA's research focusing on the accident survivability of railroad tank cars involves a three-step process to assess the effects of various types of train accidents (e.g., derailments or collisions) on tank cars. Each phase involves the development of computational models with different objectives. The first phase involved the development of a physics-based model to analyze the gross motions of rail cars in a derailment (i.e., a derailment dynamics model). This derailment dynamics model was then used to estimate the closing speeds, peak impact forces, and angles of incidence between an impactor (e.g., the coupler of another car) and the tank car head or shell. The second phase involved the development of structural finite element analysis models to simulate the structural response of the tank car head or shell to an assumed scenario (i.e., penetrator shape, initial closing velocity, and effective collision mass). The third phase is an assessment of the damage created by the impacting loads, which entails the application of fracture mechanics testing and analysis methods. The research is being conducted by Volpe and is summarized below. In addition, a more detailed discussion of the research can be found in the transcript to the March 30, 2007, public meeting (document no. 29 in docket no. FRA-2006-25169) and in FRA's “Research Results” (document no. 24 in docket no. FRA-2006-25169).

The first phase of FRA's current research program developed information about the performance of a train consist after a derailment occurs. Initially, this phase of the research was aimed at developing a derailment model effectively recreating the Minot derailment. However, due to the chaotic events and inherent complexities (e.g., track layout and condition; the three dimensional topography of the local terrain; car types in a train; and the location of each car in a train) of derailment situations, the initial and boundary conditions that lead up to specific derailment scenarios are very poorly understood. Early in its research effort, FRA realized that the exact circumstances and boundary conditions of the Minot derailment could not be accurately reproduced.

Accordingly, FRA revised its objective in this first phase of research from trying to replicate the conditions of the Minot accident, to identifying all of the salient features of derailment situations based on historical accident consequence review, as well as active accident investigations, thereby creating a generalized accident scenario with well-defined initial and boundary conditions. This information was then used to establish more easily analyzed impact scenarios. Specifically, the derailment dynamics model was used to estimate the post-derailment car-to-car interactions; that is, the gross motions of the cars as they come off the track after a derailment, the closing impact speeds, and the orientations at which the derailed cars come together in a generalized derailment scenario.

Sensitivity studies were then performed to assess the relative effect of various factors on derailment severity. The factors analyzed included: (1) The number of cars derailed; (2) the secondary car-to-car closing speed; (3) the peak forces that the couplers experience; and (4) the lateral displacement of the derailed cars from the point of derailment. Although there are several potential alternative analysis techniques that could be employed, FRA used two different types of models to calculate the gross motions of rail cars during a derailment scenario. One model was a purpose-built model using an explicit derivation of the equations of motions for a two-dimensional lumped-parameter representation. The second model involved a commercially-available, general-purpose model for rigid-body dynamics, commonly accepted within the rail industry. The inputs for the models included: (1) Operational factors such as the number of cars in the train and the masses of the cars; (2) descriptions of the initial conditions such as the longitudinal speed of the train just prior to derailment and the initial angular velocity used to perturb the train set and cause the derailment; (3) the coefficients

of friction between the tank car trucks (i.e., the swiveling frames of wheels under each end of the tank car) and the rail or the ground; (4) specific coupler characteristics such as length, dead band, stiffness, and maximum swing angles; and (5) higher-level model assumptions such as how the couplers break, the car-to-car contact forces, and lumped mass simplification.

The input parameters were varied by as much as +/−fifty percent. The models consistently demonstrated that significant sensitivities are associated with initial train speed and ground friction. The higher the initial train speed, the higher the post-derailment car-to-car impact closing speed and the greater the number of derailed cars. However, the results indicate that, in general, the secondary car-to-car impact speed is one-half that of the initial train speed across the variation in input parameters. Additionally, the resulting car-to-car impact speeds are negatively affected by increases in ground friction. That is, for higher ground friction, the resulting car-to-car closing speeds are lower and fewer cars derail. Of interest was the finding that within the parameters of the modeling, the mass of the cars was not a significant factor on post-derailment car-to-car closing speeds or on the number of cars derailed.

Results of the derailment dynamics modeling also demonstrated similar car-to-car interactions as observed in real world accident situations. For example, one type of impact occurs when two cars come together and the second car impacts the head of the first car (e.g., the Macdona accident). A second type of impact is associated with side/shell impacts (e.g., the Minot accident). Both the derailment dynamics models, as well as real world incidents documented in the RSI-AAR Tank Car Accident Damage Database, demonstrate that these head/shell impacts occur both at the centerline of the car as well as at the ends of the cars above the trucks/bogies. By combining this information, simple impact scenarios were developed that could be readily analyzed to compare the performance of different types of tank car designs (whether from the existing fleet or newer proposed designs).

The second phase of FRA's current research program utilized the information generated from the derailment dynamics modeling to assess the forces to which cars can be subjected in the event of a collision or derailment. This work required the development of large deformation finite element models capable of analyzing post buckling/plastic deformations. Both head and shell impacts were analyzed, but emphasis was placed on head impacts because there is a greater body of knowledge available on head performance.

In cooperation with FRA, extensive head puncture testing was conducted by the RSI-AAR Test Project throughout the 1970's and 1980's. This research, conducted on both empty, non-pressurized and loaded, pressurized tank cars, led to the HMR's current specification for head protection. It is important when developing such complicated models to start simply and build up in levels of complexity. Because head impacts are better understood, as is the deformation of a tank car unloaded and unpressurized, FRA initially modeled an empty, unpressurized tank car. There is greater uncertainty associated with pressurized fully-loaded cars, as well as understanding the stress states the cars experience prior to rupture. Results from the RSI-AAR head impact data, empirical puncture models, and three-dimensional laser mapping of the damage from the cars in Graniteville were used to help establish the validity and fidelity of the models. FRA intends to continue its modeling efforts to increase the level of complexity to analyze a loaded, pressurized car.

The third phase of the FRA's current research program is an extension of the model development and assessment of damage to tank cars from prescribed impact loading conditions that may lead to catastrophic failure. The results from full-scale tests will be used to validate the second and third phases of the research.

The FRA and the NGRTCP group are conducting a series of shell impact tests to provide information about the performance of conventional PIH tank cars under the collision conditions defined from the previous research program. In addition to providing baseline performance data, the test conditions developed are intended to aid in the development of a testing process that can be used to assess the relative performance of different designs, as well as to qualify a design. The full-scale testing approach involves a generalized impact condition based upon the scenarios defined previously and is designed to be simple to set-up, safe to conduct, and readily analyzed. It is also designed to provide consistent and repeatable results. The test conditions developed are not intended to replicate any specific accident conditions but are rather intended to result in similar failure and deformation modes as observed in accidents. This is a very similar approach that parallels the automotive 30 mph barrier test.

Three full-scale tests have been conducted to date, on April 11, 2007, April 26, 2007, and July 11, 2007. These tests involved a side impact between a rigid ram car with a stylized punch striking a standing pressurized DOT specification 105 tank car broadside at the centerline of the tank, both horizontally and vertically. The ram car was ballasted to a weight of 286,000 pounds. The standing tank car was pressurized to 100 psig and was loaded with clay slurry with a density equal to liquid chlorine with an outage of 10.6%. The ram car was pulled back to a predetermined position on the slightly graded tangent track and released to achieve the desired impact speed. Just prior to impact with the standing tank car, the air brakes on the ram car were activated, such that upon rebound, a second impact would not occur. In the first two tests, the punch face size was approximately 23 inches by 17 inches; in the third test, the punch face size was approximately 6 inches by 6 inches.

The first test was a limited instrumented assurance test designed to develop information about how the colliding equipment interact and to better understand the gross motions of the two cars. Because the test was designed to develop more detailed information about the interacting cars' behavior, and puncturing the standing car would have unnecessarily complicated the analysis and test set-up, the test speed was defined such that no puncture would occur. Specifically, the first test was conducted at 9.6 mph, and as predicted, no puncture occurred. The limited instrumentation on both the ram car and the standing tank car were analyzed and the force-time histories measured and predicted. The measured force-time histories from the collected data were within the standard deviation of the predicted test results.

The second test that was conducted had a fully-instrumented standing tank car. The additional instrumentation helped to define load path into the tank car, the evolution of the plastic dent growth, and recovery. It also refined the measurements of the gross motions of the colliding cars' interaction. The test was conducted at 14.0 mph. As with the first test, this test speed was chosen so that puncture would not occur. The ram car was again released from a pre-defined location and allowed to roll freely under gravity and the grade to impact the standing tank car. The analysis of the test data are on-going, but preliminary review suggests that again the force-time histories of the ram car and the struck tank car are within

the standard deviation of the predicted test results.

After the second test, a careful inspection of the ram car showed that a modest amount of damage was inflicted on the lead truck and its carbody attachment. This damage was attributed to the off-axis vertical motions resulting from the difference in the centerline of the impactor and the height of the center-of-gravity of the ram car.

In order to safely run a test to puncture the baseline car, either a smaller punch would be needed and the test speed maintained at 14 mph, or the center-of-gravity of the ram car would have to be raised to be more in line with the centerline of the punch, to minimize ram car vertical motions for impact speeds greater than 14 mph. The option selected was to reduce the punch size to 6 inches by 6 inches. There was equal confidence in simulating the influence of punch size and impact speed on tank rupture. DOT is seeking to significantly increase the impact speed at which tank cars carrying PIH materials can protect their lading. For a wide range of sizes, this goal is independent of punch size. In order to allow for safer test procedures and lower test speeds, it was decided to use the smaller punch size in the regulation.

Because of the results of the second test, in the third test, the punch face size was approximately 6 inches by 6 inches. The standing tank car that was used during the third test was fully-instrumented. The test was conducted at 15.1 mph, and this test speed was chosen so that puncture would occur. The third test was designed to confirm that material failure of the tank car and puncture would occur at 15 mph with a smaller impactor. The test also provides a comparative baseline reference for the enhanced tank car designs. As with the second test, the ram car was again released from a pre-defined location and allowed to roll freely under gravity and the grade to impact the standing tank car. The analysis of the test data are on-going, but preliminary review suggests that again the force-time histories of the ram car and the struck tank car are within the standard deviation of the predicted test results.

Additional tank car testing is planned. The further testing will provide additional insight and validation to the modeling. The additional tests include material, full-scale sub-assembly, and full-scale prototype car tests. Materials tests improve the constitutive models applicable to the specific sub-components used in alternative designs, such as behavior of composites, foams, and multi-layered metal structures. The full-scale sub-assembly tests build confidence in the fidelity of the models used as they capture both material and geometric nonlinear behavior exhibited by larger scale components. Finally, in conjunction with the NGRTC program, full-scale prototype cars will be subjected to side and head impact and over-the-road testing. Each additional test enhances the modelers' ability to predict and capture increasingly complicated behavior under extreme accident loading conditions. As noted in the discussion of the proposed rule text, the proposed head and shell performance standard is based on the model that has been developed by Volpe. As more testing is completed, any new information or refinements to the test procedure will be considered for incorporation in this proposed rule.

For the reasons outlined above, FRA's research has focused on ways to enhance the accident survivability of tank cars through implementation of an enhanced performance standard for head shields and tank shells. We recognize that there may be a number of different ways for tank car manufacturers to meet this performance standard, including different design-types, variations in materials of construction, and the like. We invite commenters to suggest specific measures that would be utilized to meet the proposed performance standard. In addition, commenters may wish to provide data and information that would support alternative strategies for achieving the goal of improved tank car accident survivability.

XI. Discussion of Public Comments

As noted above, recognizing the need for public input as part of DOT's comprehensive review of design and operational factors affecting rail tank car safety, PHMSA and FRA held three public meetings inviting interested parties to comment on relevant aspects of tank car safety. As part of the public comment process, FRA established a public docket (Docket No. FRA-2006-25169), providing interested parties with a central location to both send and review relevant information concerning the safety of railroad tank car transportation of hazardous materials. The FRA docket contains several submissions from FRA (e.g., transcripts of the three public meetings, relevant Congressional testimony, research reports), as well as comments from numerous members of the regulated community. Specifically, written comments were received from the following organizations: BASF Corporation, the Institute of Makers of Explosives, Dow, TFI, Trinity, Applied Solutions, Inc., the Brotherhood of Railroad Signalmen, Agrium U.S. Inc., CI, and PPG Industries (PPG). Many of these same organizations attended the public meetings and provided oral comments at those meetings. The following discussion provides an overview of the written and verbal comments that were received. Where appropriate, a more detailed discussion of specific comments and how DOT has chosen to address those comments in this proposed rule can be found in Section XIII below, the Section-by-Section analysis portion of this preamble.

A. May 31-June 1, 2006 Public Meeting

The primary purpose of the first public meeting, held on May 31-June 1, 2006, was to surface and prioritize issues relating to the safe transportation of hazardous materials in railroad tank cars. Attendees included representatives from the railroad industry, shipping industry, railroad tank car manufacturing and repair companies, labor organizations, the NTSB, Transport Canada, and the Transportation Security Administration (TSA). At this meeting, commenters from both the railroad industry and the hazardous materials shipping industry expressed the view that rail is the safest mode of transportation for hazardous materials over land. For example, the AAR explained that since 1980, the rate of rail accidents with a hazardous materials release per thousand rail carload has dropped by 89%. RSI noted that approximately 1.7 million carloads of hazardous materials are transported by rail throughout the United States each year and 99.98% of those shipments reach their destinations without incident. Similarly, RSI commented that statistics demonstrate that it is 16 times safer to move hazardous materials by rail, as compared to highway. Noting that it would take approximately four cargo tank trucks to deliver the amount of hazardous materials that can be carried in one rail tank car, several shippers expressed concern that if shippers were forced to transport hazardous materials via highway, the overall safety risk would increase because of the increased number of shipments on the nation's roads. Several representatives of the hazardous materials shipping industry expressed the view that rail transportation of hazardous materials is essential to the competitiveness of the U.S. chemical and agricultural industries, to the public health, safety and welfare, as well as to the economy of the United States. Dow, the largest chemical company in the world,

indicated that its North American business model is based on the belief that the rail transportation of hazardous materials is the safest, most efficient, most economical, and most socially acceptable way of shipping hazardous materials over land.

Despite these safety statistics, meeting participants from both the railroad and shipping industries expressed agreement on the need for continuous improvement in the safe transportation of hazardous materials by railroad tank car, particularly in light of the Minot, Macdona, and Graniteville accidents. However, participants expressed differing views on how to accomplish that goal. Many representatives of organizations that depend on railroads for shipping hazardous materials stated that improvements in the safe transportation of hazardous materials by railroad tank car should be made only after a “holistic” consideration of the rail transportation system. For instance, several commenters expressed the view that not only should tank car design improvements be considered, but safety improvements should also address railroad operating and maintenance practices; railroad routing practices and how to reduce ton miles PIH materials travel due to inefficient routes; shipper commodity handling practices; and emergency response procedures. Both the Brotherhood of Locomotive Engineers and Trainmen (BLET) and the United Transportation Union (UTU) echoed several of these same concerns, particularly noting human factors issues, the prevalence of non-signalized territory, the training of crews to handle hazardous materials, and crews' access to personal protective equipment in the event of an incident. One commenter specifically suggested that DOT adopt AAR Circular OT-55-I as a regulation. Several commenters noted that the tank car is only one component of the rail transportation system, and no single component of the system can provide the entire means to improving tank car safety. Accordingly, many commenters expressed a desire for DOT to take a leadership role in addressing the safe transportation of hazardous materials by railroad tank car on a system-wide basis.

FRA and PHMSA generally agree with these commenters. Although this NPRM focuses on enhancing the tank car packaging, it also proposes certain operational restrictions specific to tank cars transporting PIH materials, and DOT's comprehensive review of design and operational factors affecting rail tank car safety is not so limited. As noted above, DOT's rail safety efforts are multi-faceted, and DOT is addressing operational issues such as human factors, track conditions, and signal and train control systems designed to prevent accidents in the first place, as well as emergency response issues intended to ensure that in the event of an incident, emergency responders are able to respond appropriately. In addition, PHMSA has issued a proposed rule that would require railroads to gather traffic and commodity data on certain explosive, radioactive, and PIH materials they transport; analyze safety and security vulnerabilities of current and alternative routes used for these materials; and select the routes that pose the least safety and security risks after considering any mitigation measures that could be implemented.

See

71 FR 76834 (Dec. 21, 2006).

Other commenters noted the voluntary efforts already underway by many hazardous materials shippers to improve the safe transportation of their materials by rail. One example of an industry effort to address the safe transportation of hazardous materials in tank cars is the partnering of Dow and UP in a series of initiatives to improve rail safety and security, including the NGRTCP. These initiatives are discussed in more detail in Section IX above.

Railroad participants, including the AAR, CP, and BNSF, expressed the view that the railroad industry itself has taken many voluntary steps to reduce the occurrence of accidents that can lead to hazardous materials releases. For instance, a representative from BNSF presented information on the carrier's derailment prevention efforts aimed at track caused derailments, equipment caused derailments, as well as derailments relating to operating practices. BNSF's efforts include implementing advanced train control technology; utilizing various freight car condition monitoring technologies; installing and maintaining switch point position indicators and broken rail protection in non-signalized dark territory; as well as modifying the carrier's operating practices when transporting a significant amount of PIH materials over non-signalized territory. Specifically, noting that nearly 50% of BNSF's PIH movement is over non-signaled territory, BNSF explained changes in its operating practices aimed at ensuring the safe transport of PIH materials over this type of territory. BNSF noted the following changes in operating practices when transporting PIH materials over dark territory: (1) Inspecting the route prior to operating trains carrying PIH materials; (2) restricting the speed of trains carrying PIH materials to 35 miles per hour; (3) requiring that trains hauling PIH materials hold the main line during meets; and (4) requiring trains on sidings to stop before PIH trains pass. Additionally, a representative from CP presented information on the carrier's efforts, dating back to 1995, to address human factors issues in the railroad environment, including efforts directed at crew resource management, and fatigue risk management.

Noting member railroads' efforts to reduce the occurrence of accidents that can lead to hazardous materials releases, the AAR expressed the view that “[r]esponsible planning must consider that accidents can occur” and “in addition to the efforts to prevent accidents, industry must also do everything it can to reduce the probability of a release of TIH [materials], such as anhydrous ammonia and chlorine, should an incident occur.” Based on its research through the University of Illinois, AAR noted that there appears to be a significant opportunity to reduce the probability of a release of anhydrous ammonia and chlorine in the event of an accident.

AAR indicated that the University of Illinois research concluded that, utilizing existing technology, the probability of a release of anhydrous ammonia and chlorine from a tank car involved in an accident could be reduced by 65 percent or more by substituting enhanced tank cars for the cars currently used to transport these materials. AAR explained that this conclusion was premised on replacing the current 263,000 pound tank cars used for transporting anhydrous ammonia and chlorine with 286,000 pound tank cars equipped with additional head protection, thicker shells, and enhanced top fittings protection (i.e., the Trinity car).

Most commenters representing members of the hazardous materials shipping industry generally expressed support for the efforts of the AAR TCC to improve the transportation of hazardous materials by rail. However, those commenters expressed concerns with several aspects of the TCC's recent proposals. First, commenters stated that the implementation period proposed by AAR (i.e., replacing the entire chlorine and anhydrous tank car fleet within five to seven years) was unrealistic, particularly given tank car manufacturing capacity. One commenter, Terra Industries (Terra), a shipper of anhydrous ammonia, objected to AAR's proposal noting that the

This text is long and has been trimmed here. Open the source document for the complete record.

This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.