Mr. Chairman and members of the subcommittee, I appreciate the opportunity to appear before you to discuss the state of railroad safety on our nation's railroads. On behalf of the Federal Railroad Administration (FRA), the agency charged with administering the nation's railroad safety laws, I extend my deepest sympathy to the families of the people who died in recent accidents and to those who were injured. My testimony will explain how FRA's railroad safety program is working daily to reduce the likelihood and severity of accidents such as these and will demonstrate that the state of railroad safety is generally very positive.
FRA's safety mission can be simply stated: help prevent fatalities, injuries, and property damage related to railroad operations and releases of hazardous materials from rail cars, and enhance the security of railroad operations. Under the Federal Railroad Safety Act of 1970, FRA's jurisdiction extends to all areas of railroad safety. We have issued rules on a wide range of subjects including track, signal and train control, locomotives and other equipment, grade crossing signal devices, and operating practices, and we enforce those rules as well as rules related to hazardous materials transportation by rail. We conduct inspections of railroad operations to determine the level of compliance with the laws and regulations, and use a variety of enforcement tools when necessary to encourage compliance. We help educate the public about safety at highway-rail grade crossings and the dangers of trespassing on railroad property. FRA has its own accident investigation authority, and works closely with the NationalTransportation Safety Board (NTSB) on those accidents that NTSB investigates. FRA investigates a broader range of railroad accidents than NTSB, including those involving three or more deaths at a highway-rail grade crossing, an employee fatality, damages that exceed $1,000,000, or serious injuries to passengers.
FRA tracks the railroad industry's safety performance very closely by requiring reports of accidents and injuries, investigating major accidents, and inspecting railroads and hazardous materials shippers extensively. FRA's safety data base is available on its Web site (see www.fra.dot.gov). FRA uses this information to guide its accident prevention efforts and continually strives to make better use of the wealth of available data to achieve its mission.
The Current State of Railroad Safety Across the Nation
As judged by most indicators, the long-term safety trends on the nation's railroads are very favorable. While not even a single death or injury is acceptable, progress is being made in the effort to improve railroad safety. Based on preliminary figures, last year marked all-time safety records in several important categories. Overall, the total number of rail-related accidents and incidents and the total accident/incident rate were the lowest on record. Also, 2001 saw the lowest number of railroad employee fatalities (22) and injuries (7,575) on record and the lowest overall employee casualty rate (3.19 per 200,000 employee hours). In the period between 1978 and 2001, the number of reported train accidents dropped from 10,991 to 2,962, and the train accident rate fell from 14.62 accidents per million train-miles to 4.17 accidents. Also during this period, the number of train accidents involving a release of hazardous material declined from 140 to 31 despite a significant increase in the number of hazardous materials tank car shipments to more than two million per year. Since 1990, a period in which railroads have transported morethan 20 million hazardous materials shipments, three persons have died as a result of the release of hazardous material lading in a train accident.
In other words, over the last two decades the number and rate of train accidents, total deaths arising from rail operations, employee fatalities and injuries, and hazardous materials releases and deaths related to those releases all fell dramatically. In most categories, these improvements were most rapid in the 1980s, and tapered off in the 1990s. (See the attached graph of train accidents and their rate since 1978.) Causes of the improvements included a much more profitable economic climate for freight railroads following deregulation in 1980 under the Staggers Act (which led to substantially greater investment in plant and equipment), enhanced safety awareness and safety program implementation on the part of railroads and their employees, and FRA's safety monitoring and standard setting.
Similarly, the grade crossing safety picture has shown great progress. In 1990, a total of 698 persons died in highway-rail grade crossing collisions. In 2001, the number was down to 419 despite an increase in exposure due to increased highway and rail traffic. Here, too, improvement has resulted from a variety of sources, including public investment in crossing warning devices and greater awareness of the risks present at crossings on the part of highway users, which was brought about by joint efforts of railroads, employees, FRA, the states, our Department of Transportation partners (Federal Highway Administration, Federal Transit Administration, Federal Motor Carrier Safety Administration, and the National Highway Traffic Safety Administration), and Operation Lifesaver®.
Despite the impression one might get from news accounts of recent accidents, rail remains an extremely safe mode of transport for passengers. In the five-year period between1997 and 2001, just two passengers were killed in train collisions and derailments, and 13 more in grade crossing collisions, out of the 2.3 billion passengers who rode our nation's commuter and intercity passenger trains. According to the National Safety Council (see attached chart on passenger death rates), the number of deaths per 100 million railroad passenger-miles is quite comparable to the rate for airline passengers, both of which are a fraction of the rate for automobile passengers. Given the strength of rail passenger equipment and the fact that rail passengers are distributed throughout a train in such a way as to minimize the impact of a collision or derailment for many, rail passenger accidents-while always to be avoided--have a very high survival rate.
Unfortunately, not all of the major safety indicators are positive. In recent years, rail trespasser deaths have replaced grade crossing fatalities as the largest category of deaths associated with railroading. In 2001, a total of 508 persons died while on railroad property without authorization, which was an increase of nearly 10 percent over the previous year. Track safety has also emerged as a growing problem. The number and rate of "track-caused" accidents have actually increased over the last few years. For the first time in many years, in 2001, track causes actually exceeded human factors as the largest category of train accident causes. In that year, track causes were cited in about 38 percent of all reported train accidents, while human factors accounted for about 34 percent, equipment causes were responsible for about 14 percent, signal-related factors were causal in about one percent, and miscellaneous causes accounted for the remainder.
Any discussion of the railroad accident data, however, must take into account the fact that, under the current reporting threshold, any train mishap resulting in at least $6,700 indamage to railroad equipment or structures must be reported as a "train accident." This means that many "fender benders" and mechanical malfunctions that pose no danger to either the public, railroad workers, or railroad operations meet the reporting threshold and are classified by FRA as train accidents. For example, FRA recently analyzed the number of train accidents in its database that occurred on Amtrak's Northeast Corridor over the past five years. While the raw data contained 101 events that were classified as train accidents, closer examination revealed that 84 incidents involved mechanical malfunctions or damage to the overhead electrical equipment. These malfunctions cause a loss of electrical power that interrupts train service but causes no harm to the passengers. There were also three cases of vandalism to trains, five cases of trains striking debris and animals on the track, three incidents in which no passenger train was involved, and one fire caused by a cigarette in restroom debris. In fact, of the 101 total accidents reported on Amtrak's Northeast Corridor over the five-year period, there were only three train derailments, two of which occurred at very low speeds, and there were two cases where an Amtrak train struck unsecured equipment protruding from passing freight trains.
Another factor to consider when discussing train accidents is that the severity of accidents can vary greatly. More than half of all train accidents occur in yards where train speeds are low, resultant damages are minor, and casualties are rare. Consider, for example, that train accidents, as FRA uses the term, resulted in only six of the 966 deaths associated with railroading in 2001. The vast bulk of those fatalities involved grade crossing incidents (419 deaths) and trespassers (508 deaths). Given the limited usefulness of the aggregate data, FRA tries to continually mine the accident and inspection data at its disposal to find where the major pockets of risk exist and then determine how its actions can produce the biggest safety returns.
FRA is also quite concerned at the number of recent train collisions in which human performance appears to be a primary contributing factor. Since the Placentia, California collision in April of this year, there have been seven more serious collisions. In many of these cases, we believe that compliance with the railroad's own operating rules on signals and restricted speed may have prevented the accident. As explained more fully below, FRA has recently launched a nationwide, focused effort to examine how the railroads are implementing their own programs for testing their employees' compliance with these important safety rules.
FRA's Safety and Security Program
FRA's safety program is the heart and soul of the agency. The program has several elements: setting safety standards, ensuring compliance with those standards, focusing attention on serious safety problems whether or not covered by current standards, educating the rail industry on the federal standards and the public on rail safety issues, focusing on emerging security issues, investigating accidents and employee fatalities, conducting research and development on safety issues, and setting the tone for safety efforts in the industry.
The program's most important element, of course, is its people. Our Office of Safety headquarters staff of 100 works on the gamut of activities including rulemaking, compliance, data analysis, and program management. Our field force of 486 (which includes safety inspectors, support staff, and managers) works on inspection and compliance activities, investigations, and outreach to communities and the public on safety issues. More than 160 certified state safety inspectors from 30 states supplement the efforts of our field forces in all of these areas. Supporting the Office of Safety is the Safety Law Division of the Office of Chief Counsel, our Office of Administration (which provides human resource, budget, informationtechnology, and procurement support), our public affairs staff, and our research and development office.
Setting Safety Standards
Congress has authorized FRA, as the delegate of the Secretary of Transportation, to issue necessary regulations and orders for every area of railroad safety. Since FRA's inception in 1967, the agency has issued a wide range of standards on subjects such as track safety, signal inspection, freight car safety, passenger car safety, locomotive safety, power brakes, alcohol and drug testing, operating rules and practices, accident reporting, hours of service recordkeeping, railroad communications, roadway worker and bridge worker protection, engineer qualifications, grade crossing signal maintenance, and passenger train emergency preparedness. FRA also assists the Department of Transportation's Research and Special Programs Administration (RSPA), which issues hazardous materials standards for all modes of transportation, in developing standards for rail transportation of those materials.
In 1996, FRA established the Railroad Safety Advisory Committee (RSAC) to develop consensus recommendations on safety issues. RSAC contains representatives from all major groups interested in railroad safety, including railroads and their associations, railroad labor organizations, the states, suppliers, and public interest groups. The NTSB and representatives from Mexico and Canada are associate members of the committee, as are a number of groups added to ensure RSAC's diversity. FRA seeks RSAC's recommendations on specific tasks; on each task, RSAC can decide whether or not to accept it and begin work. On those tasks it accepts, RSAC members appoint a working group of those most involved with the subject covered by the task. If the working group's recommendations are unanimously adopted by thatgroup and by a majority of the full RSAC, they are sent to the FRA Administrator. While FRA is free to accept or reject RSAC's recommendations, we fully engage ourselves in the working group process to ensure that the recommendations are consistent with FRA's goals for the rulemaking project. As a result, our proposed and final rules that arise from RSAC recommendations usually incorporate those recommendations substantially.
This consensus approach to rulemaking has produced notable successes: revised track safety standards that include rules for high speed operations, revised communication standards reflecting technological advances in the field, and updated certification standards for locomotive engineers. More important, RSAC has helped engender a cooperative approach to developing new safety rules in which the railroad industry's major players have the opportunity to shape FRA's, and each other's, thinking from the start and feel more invested in the final product.
FRA's recent standard-setting accomplishments include the first standards for passenger cars, issued in 1999, which were the product of a rule-specific consensus process separate from RSAC; power brake standards for freight service, which FRA issued in 2001 without the benefit of consensus recommendations after an unsuccessful attempt to achieve consensus; and, issued just this year, the first standards for locomotive cab sanitation, which are the product of the RSAC process. Late in 2001, we issued an interim final rule establishing a United States locational requirement for dispatching domestic train operations.
FRA has several important regulatory projects under development. We are developing, through the RSAC process, standards for processor-based signal and train control systems (discussed more fully below), which will lay the foundation for integrating such systems into the existing rail network. We hope to have a final rule out this year. We are also using the RSAC process to develop revised event recorder standards to facilitate movement to a new generation of recorders and standards for the crashworthiness of locomotives. One major rulemaking on which we are not using the RSAC process is our final rule on the use of train horns at grade crossings. While very broad-based, RSAC membership is not sufficiently broad to include all the interests that might be directly affected by this rule. Instead, to address this sensitive subject, we held a dozen public hearings across the country and a technical conference and have engaged in extensive outreach with local communities.
Whether or not we employ the consensus process of RSAC, in all of our standard-setting activities we strive to avoid unnecessary regulation, consider all reasonable options, and issue rules that embody a fitting balance between benefits and burdens, are clearly stated, and are enforceable. However, neither the consensus rulemaking process nor the more traditional process is designed for quick action. Rulemaking can take a very long time. My philosophy is to try to do fewer things better and more quickly rather than trying to write simultaneously every rule that might have found its way onto the agency's agenda. This fits with the Department's renewed emphasis on rulemaking timeliness, which entails enhanced methods of coordinating and monitoring regulatory projects and tighter control of the clearance process.
Encouraging Compliance and Safety Improvements
The railroads, of course, have the responsibility for compliance with the standards FRA sets and to perform the necessary inspections and tests to ensure that they do comply. There are more than 650 railroads in the nation operating more than 1,000,000 pieces of equipment over more than 200,000 miles of track. FRA's inspection force cannot possibly observe all railroad activity. Instead, FRA monitors railroads to determine their level of compliance with thosestandards and employs a variety of tools to encourage compliance. We start with the assumption that railroads and their employees want to do the safe thing for their own benefit, not just because a law or regulation requires it. And we also understand that the Code of Federal Regulations is not the sole source of wisdom on safe practices; there are, in fact, safety problems not covered by existing rules that require a solution nonetheless.
FRA calls its approach to compliance the Safety Assurance and Compliance Program (SACP). The basic principles of SACP are to look for root causes of safety problems, try to develop solutions to those problems cooperatively with railroad management and employees, and focus both inspection activity and the use of enforcement tools on the most serious safety risks, as revealed by our inspections and our accident data. On each of the major railroads, SACP teams include FRA inspectors and managers, railroad officials, and employee representatives. The SACP teams provide a forum for resolving both compliance issues and safety problems not within the four corners of existing rules. Issues can be resolved through informal agreements or formal action plans. At the same time, FRA continues its normal review of railroad activities through regular inspections of facilities, vehicles, operations, and records and investigation of complaints.
FRA's policy is one of focused inspection and enforcement. That is, we try to concentrate our inspection efforts on detecting conditions that are leading causes of accidents, injuries, and hazardous materials releases, and, where noncompliance is found, we try to focus our enforcement efforts on violations that may cause such events. Where routine inspections reveal minor defects that pose little risk, FRA will certainly address the noncompliance with the railroad but is not likely to take enforcement action. Where a railroad has acknowledged theexistence of a serious safety problem, developed a plan for alleviating it, and implemented that plan in a timely way, FRA will ordinarily take no enforcement action in the absence of some immediate hazard. However, FRA is very likely to use its enforcement tools where FRA discovers serious safety violations causing an immediate and unacceptable risk that the railroad should have found and corrected on its own. FRA is also likely to take enforcement action where, even though there is no immediate hazard, FRA has identified serious rail safety problems requiring concerted action by the railroad to prevent an unacceptable risk from developing, and the railroad has failed to make a good faith effort to implement a specific remedial program to fix those safety problems by a date certain, despite having agreed to do so.
Where enforcement appears necessary to encourage compliance, the tool we use will depend on the circumstances. Civil penalties are the most frequently used tool. In fiscal year 2001, for example, FRA collected over $7.6 million in penalties from railroads and hazardous materials shippers. Our Office of Chief Counsel, based on the recommendations of our field inspectors and working closely with the Office of Safety, assesses and collects these penalties. As the safety statutes encourage us to do, we settle nearly all of these cases through negotiations with railroads and shippers, and determine settlement amounts by applying the settlement criteria stated in the safety statutes. The settlement negotiations provide an excellent forum for addressing the most current and serious compliance issues that have not been resolved through more cooperative methods.
FRA has several other enforcement tools. Our inspectors can issue special notices removing locomotives or freight cars from service until they are repaired, or lowering the speed of track to a speed at which the track segment is in compliance with the standards. Wesometimes enter into compliance agreements with railroads in which the railroad promises specific remedial actions and, should it fail to deliver on its promise, agrees to the imposition of a compliance order, emergency order, or particular fines. The FRA Administrator can address an imminent safety hazard by issuing an emergency order, with opportunity for review of the order after its issuance. Civil penalties are available against individuals who willfully violate the safety regulations, and FRA may disqualify individuals from safety-sensitive service if their violation of safety regulations demonstrates their unfitness for such service. Criminal penalties apply for certain willful violations of the hazardous materials rules and knowing and willful violations of recordkeeping or reporting requirements. We have made increased use of these criminal penalties in recent years, especially for serious violations of the rules concerning proper documentation of hazardous materials shipments.
Accident Investigations
Nearly a century ago, Congress gave FRA's predecessor, the Interstate Commerce Commission (ICC), the authority to investigate railroad accidents. FRA inherited that authority and continues to implement it. Where the NTSB decides to investigate, its investigation generally has priority over those of all other federal agencies, but does not extinguish the investigative authority of those agencies. In those cases, which usually involve the most serious accidents, our investigators work closely with NTSB and serve on NTSB's teams. As previously noted, FRA also investigates a broader category of accidents and incidents than does NTSB.
Most or all of the recent accidents that concern this Committee are still under investigation by NTSB , FRA, or both. Final determinations of probable cause will not be issuedfor some time. I refer you to NTSB's testimony for any details of its investigations that the Board may be able to share at this time.
The final, detailed reports that NTSB and FRA produce concerning accidents are a very important tool in identifying risks and determining what actions FRA may need to take to reduce those risks. While FRA pays very close attention to major accidents to determine what conditions might require immediate agency action, those accidents sometimes involve such unique combinations of causal factors and often take so long to analyze effectively that they do not offer immediate insights into actions that might prevent similar accidents. However, because FRA's role is regulatory and not just investigative, where FRA gleans any useful information from investigations while they are underway, we use it immediately to try to prevent a recurrence.
Research and Development
FRA has an extensive research and development (R&D) program. Although that program resides in our Office of Railroad Development rather than our Office of Safety, its primary mission is to serve the safety program. Our R&D efforts also serve the railroad industry, railroad employees, and suppliers of railroad equipment. FRA owns the Transportation Technology Center near Pueblo, Colorado, which is operated under contract by a subsidiary of the Association of American Railroads (AAR).
FRA's R&D program includes these elements:
The Railroad System Issues element encompasses research on technological and operational developments in the industry that may affect safety; system safety planning; and physical and cyber security in the railroad system.
The Human Factors element focuses on human performance in railroad operations (e.g., the effects of fatigue) and at grade crossings (e.g., the interface between highway users and visual and audio warnings).
The Rolling Stock and Components element focuses on improvement of equipment defect detection and control via wayside and onboard technology and the development of advanced materials.
The Track and Structures element focuses on improved methods of detecting hazardous conditions that can lead to failure of rails or structures.
The Track/Train Interaction program assesses improved methods for reducing derailments due to interactions of track structures and vehicles.
The Train Control program involves facilitation, risk analysis, testing, and evaluation of new train control systems, including positive train control.
The Grade Crossings program focuses on technical aspects of crossings such as train presence detection, crossing geometry, and warning device technology.
The Hazardous Materials element addresses the design and structural integrity of tank cars.
The Occupant Protection element looks at the structural crashworthiness of locomotives and passenger cars through simulations, laboratory tests, and full scale fire and impact tests.
A theme running through virtually all of the R&D program elements is the use of sensors, computers, and digital communications to collect, process, and disseminate information to improve the safety, security, and operational efficiency of railroads. Along the lines of the Intelligent Transportation Systems being developed in the highway and transit industries, FRA and the railroad industry are working on the development of Intelligent Railroad Systems that would, in an integrated way, incorporate the sensor, computer, and digital communications technologies into train control, braking systems, grade crossing protection, track and equipment defect detection, and scheduling systems as well.
The R&D program also includes the Next Generation High-Speed Rail Technology Demonstration Program, which will help develop and demonstrate the utility of positive traincontrol, a high-speed non-electric locomotive, innovative grade crossing warning systems for application on high-speed corridors, and innovative methods of constructing track and structures suitable for high-speed passenger operations and heavy axle load freight operations. Our R&D office is also implementing the Magnetic Levitation Technology Deployment Program.
FRA's Strategies for Accident Prevention
FRA combines all of the elements of its safety program to address current problems that are likely causes of accidents, injuries, and hazardous materials releases. Railroad safety contains several sub-fields, or disciplines. For each discipline, I will give some examples of how the safety program elements have been brought to bear on safety problems.
Human Factors
Human performance, especially that of railroad employees and their immediate supervisors, is critically important to railroad safety. Human factors cause about a third of train accidents and a large portion of employee injuries every year. In the 1980s, FRA identified abuse of alcohol and drugs by operating employees as a major contributor to serious railroad accidents. In 1985, the agency issued the nation's first alcohol and drug testing requirements for private sector employees. At first, railroad employee organizations opposed those rules all the way through the Supreme Court, where the rules were upheld in a landmark case in 1988. The rules have proven enormously successful and have virtually eliminated the use of alcohol and illegal drugs as a cause of train accidents. Although no one likes being tested, many employees have praised these rules as having greatly improved the safety of the industry and, in some cases, the lives of individual employees whose substance abuse has been addressed because of the rules. FRA is currently exploring the subject of legal drug use as a factor in accident causation, having been urged to do so by NTSB.
A more recent example of FRA's efforts to use the various elements of its safety program to address an area of serious safety risk is the Switching Operations Fatality Analysis (SOFA) Working Group. In the late 1990s, FRA realized that an increasing number of employee fatalities and serious injuries were occurring in the context of switching operations. FRA organized the SOFA Working Group to develop recommendations for preventing such casualties. Representatives of the AAR, the United Transportation Union, the Brotherhood of Locomotive Engineers, and The American Short Line and Regional Railroad Association analyzed 76 fatal switching incidents that occurred between 1992 and 1998. The Working Group recommended five basic practices (the "SOFA lifesavers") that, if followed invariably, would prevent such fatalities: notification to the engineer before fouling the track; extra precautions when two or more crews are working on the same track; a safety briefing before the work begins; proper radio communications; and paying special attention to crew members with less than one year of service. The recommendations were voluntarily adopted by railroads across the nation. The Working Group continues to track and report on switching incidents. Switching fatalities have dropped from thirteen in 2000, to eight in 2001, to two so far this year, while both the number and rate of yard accidents declined 8 percent and 4.6 percent, respectively, in 2001. This is an example of how consensus, non-regulatory actions can be very effective in some circumstances.
Even more recently, FRA has taken action to address a sudden spate of train collisions in which human performance appears to be a primary contributing factor. On April 23, 2002, in Placentia, California, a Burlington Northern Santa Fe freight train collided with a SouthernCalifornia Regional Rail Authority passenger train, resulting in two fatalities and 161 injuries. We believe the freight train passed a restrictive signal. In just the past two months, there have been seven additional train collisions. Including Placentia, four of these collisions involved passenger trains and resulted in two fatalities and 258 injuries, and the other four collisions involved freight trains and resulted in one fatality and 21 injuries.
While the investigations of these accidents are not yet complete, in each case the early indications are that human error appears to have been a primary causal factor. The errors included running past restrictive signals, failing to co
