For years radioactive waste in the United States has been transported in regular freight trains at the behest of the nuclear industry and regulatory agencies. Since 1975, however, the railroads have advocated the use of dedicated trains, believing public safety would be enhanced. It now appears that the decision to use regular freight trains was based on incomplete cost and risk estimates and that social and institutional factors were neglected. New information reveals that the additional cost of dedicated service is lower than previously thought and a small price to pay for reduced public opposition and increased levels of public trust.
Should spent nuclear fuel and other radioactive waste be transported in regular freight trains or dedicated trains? This question has been at the heart of a fifteen-year debate and is now the subject of a study mandated by the U.S. Congress in the Hazardous Materials Transportation Uniform Safety Act of 1990. The nuclear industry and its regulatory agencies would like to consider the case closed, but this new legislation gives the railroads and other proponents of dedicated trains another chance to present their case. It is a propitious time, therefore, to ask whether earlier conclusions favoring regular trains over dedicated trains are still valid.
In the context of high-level nuclear waste transportation, a "dedicated train" is made up of a locomotive, caboose, buffer cars, and one or more flatcars carrying casks of radioactive materials. No other freight is carried. Following conventional practice, dedicated trains may not exceed 35 miles per hour (mph), and when a dedicated train passes a regular train on a parallel track, one of the trains must stop.
At one time, the railroad industry transported all radioactive materials in regular train service. In 1975, however, the Association of American Railroads (AAR) advocated that spent fuel and large-quantity radioactive waste should be shipped only in dedicated trains. Growing doubts among the railroads about issues of liability and public safety encouraged this shift in attitude. The railroads have emphasized that dedicated trains may reduce the risks of transporting high-level nuclear waste because passing is restricted, speeds are limited, time and activity in switching yards are minimized, and other hazardous freight is prohibited. The use of dedicated trains also facilitates monitoring, and may enhance public acceptance.
Under current Nuclear Regulatory Commission (NRC) regulations, rail casks for the shipment of high-level nuclear waste must be able to withstand both normal and accident conditions without the loss of shielding or any release of radioactive materials, and without "going critical"—that is, achieving a self-sustaining nuclear chain reaction. Normal conditions are defined as those that a cask might be expected to endure during routine transportation and handling. The ability of a cask to withstand accident conditions is evaluated by a sequence of tests on the same cask, including: a free drop of 30 feet onto an unyielding surface; a free drop of 40 inches onto a blunt spike; exposure to a temperature of 1,475°F for 30 minutes; and immersion in 3 feet of water for 8 hours.
The railroad industry believes the use of dedicated trains will enhance public safety, but the nuclear industry and the agencies that regulate it argue that these trains do little to reduce negligible risks.
The railroad industry has argued that these performance criteria may be inadequate. Fires lasting longer than 30 minutes at temperatures higher than 1,475°F are possible, given the variety of flammable chemicals often carried on regular freight trains. Prohibiting other freight from dedicated trains reduces the likelihood of accidents. Similarly, the railroads recommend restricting dedicated trains to 35 mph, since this is the speed attained during the drop tests. The railroads are also skeptical of the crash tests conducted by the nuclear industry in Great Britain and by national laboratories in the United States. These dramatic demonstrations have included running locomotives into casks at high speeds and crashing cask-carrying trucks into concrete barriers. The railroads have criticized these tests as spectacular exercises in public relations that fail to demonstrate the invulnerability of casks.
The railroads also have continuing concerns about liability. In 1959 they pushed for federal indemnity protection under the 1957 Price-Anderson Act, fearing that the losses arising from a major transportation accident involving the release of radioactive materials could be catastrophic. While the likelihood of such an accident was thought to be remote, the total bill for the claims could be enormous. No insurance company would offer adequate insurance, and the railroad companies could not afford self-insurance of this magnitude. By limiting the liability of operating utilities in the event of a reactor accident, the Price-Anderson Act was intended to encourage the development of commercial nuclear power. The railroads wanted similar protection, and within the year federal indemnity was extended up to $500 million.
Even in the absence of a release, however, an accident involving a shipment of spent fuel could be very costly in terms of business disruption if it were to result in the prolonged closure of rail lines. It remains unclear whether the Price-Anderson Act would cover such losses.
In contrast, the nuclear industry and the agencies that regulate it have argued that dedicated trains are not only expensive, but do little to reduce already negligible risks. They assert that dedicated trains are unnecessary and uneconomical. In support of this position, the Interstate Commerce Commission (ICC) has repeatedly concluded that the railroads have a common carrier duty to ship spent fuel and high-level nuclear waste, that dedicated trains are unnecessary and wasteful given compliance with federal regulations, and that any surcharges for such shipments are unreasonable.
One particular case was especially influential in setting the terms of the debate. In 1976, a group of eastern railroads proposed to offer commercial rates for the shipment of spent fuel and large-quantity radioactive waste only if these shipments were restricted to dedicated trains. The Energy Research and Development Agency (ERDA), the NRC, the Tennessee Valley Authority, and various utilities objected. Presuming such transportation would comply with NRC and Department of Transportation regulations, the presiding administrative law judge concluded that dedicated trains would provide little additional safety. He found that the risks were already so small that dedicated trains were unnecessary and that additional costs would pose an uneconomical burden on shippers and the public. The ICC concurred.
Conrail argued that using dedicated trains to transport radioactive materials from the damaged reactor at Three Mile Island would dramatically reduce transit time
The Three Mile Island shipping campaign
In spite of repeated setbacks with the ICC, the railroads were notably successful in negotiating the use of dedicated trains for the shipment of radioactive materials from the damaged reactor at Three Mile Island (TMI). One hundred and fifty tons of highly radioactive fuel and debris were moved in 30 rail shipments from the reactor site near Harrisburg, Pennsylvania, to the Idaho National Engineering Laboratory (INEL) between July 1986 and April 1990. The route selected by the Department of Energy (DOE) involved two railroads: Conrail from Harrisburg to East St. Louis and the Union Pacific from East St. Louis to Idaho.
While the major controversy revolved around the choice of route, there was also considerable debate about the use of regular versus dedicated trains. DOE wanted to use regular trains but Conrail disagreed, arguing that dedicated trains offer several distinct advantages. Routes could be planned to use the best-quality track and to avoid population centers, and the movement of other trains along the selected routes could be controlled. Such planning and control would facilitate emergency response in the event of an accident. The speed of the train could be restricted and the train could be scheduled to avoid other rail traffic to minimize the likelihood and severity of accidents. Escorts could be provided and trains could be monitored more easily en route, thus minimizing the potential for terrorist intervention. Finally, Conrail believed the use of dedicated trains would enhance public confidence in the safety of rail transport of radioactive waste.
DOE did not accept Conrail's argument, but agreed to the use of dedicated trains because local freight service was the only alternative. (This would have taken eight days rather than two days to reach Union Pacific at East St. Louis, and the rail cars would have spent most of the time idle in sidings under reduced surveillance.) On learning of Conrail's position, Union Pacific, which had originally agreed to ship by regular trains, negotiated to move the first three shipments by dedicated train, with one cask per train. In a subsequent review of DOE's performance, Union Pacific concluded that the agency had failed to educate the public adequately on the safety of the shipments and that public insistence on dedicated trains had increased. General Public Utilities, the owner/operator of the TMI facility, then agreed to cover the additional cost of using dedicated trains for the remaining shipments in order to accelerate the de-fueling process.
Even though dedicated trains were used throughout the campaign, a series of mishaps occurred that led to a temporary suspension of shipments in March 1988. On March 24, 1987, a TMI train hit an automobile in St. Louis, and on December 22, 1987, a TMI train arrived in St. Louis during rush hour. On January 26, 1988, a regular Union Pacific train derailed at the Meramec River on the Missouri portion of the TMI shipment route. Finally, on February 9, 1988, a buffer car erroneously marked with a flammable materials placard was placed in a TMI train that later passed through St. Louis in the morning rush hour, reportedly at excessive speed. The ensuing media attention and public furor prompted Senator John Danforth of Missouri to call for the suspension of further shipments pending a review by the Federal Railroad Administration (FRA). He also called for additional safety measures to be put in place on resumption of shipping. These included the avoidance of rush hours by TMI trains, the requirement that an FRA official and a health physicist accompany all TMI trains, and the placement of three casks on each train to reduce the number of shipments.
Continuing his commitment to this issue, Senator Danforth later pressed for an amendment to the Hazardous Materials Transportation Act that would require the use of dedicated trains in shipping spent fuel and other high-level nuclear waste. In the final Hazardous Materials Transportation Uniform Safety Act of 1990, a compromise provision calls for a study on the advisability of dedicated trains to be completed by the fall of 1991. However, this deadline is unlikely to be met.
Projected Number of Fatalities from Rail Transport of Spent Nuclear Fuel to the Proposed Yucca Mountain Repository Over 25 Years
Cost and risk estimates
The comparative costs and risks of dedicated and regular trains have been estimated in a number of government studies. The costs were addressed most recently by the Battelle Memorial Institute in a report prepared for the DOE Office of Civilian Radioactive Waste Management. This report estimated the total cost of transporting high-level nuclear waste from reactors to the proposed waste repository at Yucca Mountain, Nevada, with the possibility of interim storage at a monitored retrievable storage (MRS) facility in the eastern United States. With no MRS, the Battelle report estimated this cost to be $247 million with regular trains and $373 million with dedicated trains, a relative increase of about 50 percent. These estimates changed to $539 million and $583 million, respectively, with an MRS in the system. In this case, the relative increase in cost is only about 10 percent. The costs are smaller with an MRS primarily because DOE assumes that dedicated train service would be used from the MRS to the repository even if regular trains were used between reactors in the eastern United States and the MRS and between reactors in the western United States and the repository.
The cost of dedicated train service is greater than that of regular train service, but the difference may be overestimated.
Included in these estimates are the costs of acquiring and maintaining casks, hauling, security escorts, and inspection. Owing to the surcharge on the basic freight rates and the increase in the total number of train-miles if dedicated trains are added to the existing system, the hauling costs account for most of the additional expense. Assuming a surcharge of $48 per train-mile, these costs amount to roughly half of the total cost of using dedicated trains without an MRS, and about two-thirds with an MRS. However, the surcharge may be much less than Battelle assumed. For example, in the case of the TMI shipping campaign, the shipper and the carrier negotiated a surcharge of only $24 per mile.
Battelle's approach to estimating costs is illuminating since it accounts for all of the important costs, not just the hauling costs. By contrast, in the case of Commonwealth Edison Co. et al. v. Aberdeen & Rockfish Railroad Co., the Interstate Commerce Commission limited its attention to hauling costs. In this case, in which the utility was seeking reimbursement for past surcharges on the ground that they were unwarranted, the ICC found only that the railroad's special train charges were "several times" the charges for regular train service. Thus, the ICC apparently did not recognize the savings in any of the other costs. Some advocates of regular trains have overstated the difference in hauling costs even more, estimating the rate charged for dedicated train services at more than twelve times the regular freight rate.
As regards risks, one of the most widely-referenced research efforts is the so-called modal study conducted by the Lawrence Livermore National Laboratory for the NRC. This study estimated the annual expected number of spent fuel casks involved in accidents in regular train service to be 0.82 per year (1.7 x 10^-6 accidents per car-mile times 735 miles per trip, times 652 shipments per year). Of these, the fraction involving a release of radioactive materials would be 0.6 percent due to mechanical forces and less than 0.1 percent due to the thermal effects of a fire. The probability of a release due to either cause is thus 8.8 x 10^-6 per year (0.82 divided by 652, times 0.7 percent). This risk is small but not negligible. It is about the same as the annual individual risk of death from accidental gas or vapor poisoning in the United States.
The Battelle study for DOE also compared the expected number of fatalities resulting from the shipment of 63,000 metric tons of uranium with the proposed Yucca Mountain repository over the course of 25 years (see table, p. 12). Radiological fatalities result from exposure to radiation during normal operating and accident conditions. Nonradiological fatalities result from accidents and expo-sure to nonradiological pollutants (such as diesel fumes from locomotives). With regular trains, the expected number of radiological fatalities is very small whether there is an MRS or not; the expected number of nonradiological fatalities is substantially higher in each case. With dedicated trains, the expected number of radiological fatalities is somewhat lower whether or not there is an MRS. However, this advantage is swamped by the increase in the expected number of nonradiological fatalities in each case. This increase is directly related to the greater number of trains needed for a dedicated service and is almost entirely due to the associated rise in grade-crossing accidents and accidents in which trespassers are hit.
The ICC's ruling in favor of regular trains in the Commonwealth Edison case and the cases preceding it assumes that the expected number of fatalities increases with dedicated trains since the total number of train-miles increases. This assumption may not be valid or may be exaggerated given that the operational advantages of dedicated trains—such as shorter train length, lower train speed, better track quality, and increased yard avoidance—received little or no attention in the supporting analyses. Even if valid, this assumption does not address the major public concern: catastrophic accidents involving large releases of radiation. None of the analyses to date has estimated the risk of such accidents. In other words, better estimates of the probabilities of high-fatality accidents resulting from radioactive releases are needed, rather than estimates of the average number of fatalities resulting from accidents in general.
Public confidence and support
Institutional factors are as important to resolving the debate as are the estimates of costs and risks. Public trust in the institutional arrangements for transporting nuclear waste is a paramount consideration.
The first major lesson to be learned from the experience of past shipping campaigns is that shipments of nuclear waste are impossible without some measure of public confidence and support. This is illustrated by the public opposition that led to the interruption of the TMI shipping campaign in 1988. Even relatively minor mishaps are often seen by the public as indicative of larger, more pernicious and systemic management problems. This underscores the railroad industry's contention that major losses can result even from relatively minor nonradiological events.
Distrust of the agencies that regulate the nuclear industry fuels public concerns about the risks of nuclear waste transportation.
Large volumes of shipments to a repository from multiple origins over an extended time will create a situation that is immensely more complicated than any past shipping campaigns. Smooth and successful shipping will require the support of multiple state and local jurisdictions and, to the extent possible, of various public interest groups. Since opposition at any point could threaten the functioning of the entire program, it is prudent to foster as much public support as possible.
Evidence on public attitudes toward dedicated trains is limited but persuasive. State and local officials generally favor dedicated trains, in part because they enhance monitoring capabilities and are more acceptable to the public. Environmental and public interest groups strongly favor the use of dedicated trains for their perceived safety and environmental benefits. Public insistence was one of the major reasons why Union Pacific argued for the continuation of dedicated services in the TMI campaign. The railroads are sensitive to the fact that public protests over spent fuel shipments can delay and disrupt their entire operations, even in the absence of an accident.
The other major lesson to be learned is that the creation and maintenance of public confidence and support is an institutional and social problem, not a technical one. Confidence and support depend on the development of trust. Unfortunately, the public is very distrustful of the agencies responsible for regulating the nuclear industry, particularly DOE. Public concerns over the risks of nuclear waste transportation are fueled by and exacerbate this lack of social trust, and provoke intense public opposition.
Once trust has been lost, recovering it is likely to be a long and difficult process. This process is as yet poorly understood. Certainly, improved risk communication efforts play a part, but redoubled programs of public education and information are not by themselves likely to bring the public's perception of the risks into line with experts' assessments.
Five key dimensions of trust can be identified: openness, commitment, caring, competence, and predictability. The use of dedicated trains would build on each of these dimensions, and would offer DOE a relatively low-cost way to demonstrate its commitment to regaining public trust.
Opting for dedicated trains necessarily demonstrates a willingness by the nuclear industry and its regulatory agencies to be open about their activities. Dedicated train service avoids the impression that the industry and the agencies are trying to sneak the waste through by night or disguised among other freight. Openness can be enhanced by engaging in extensive information and notification programs, which are easier to conduct with a dedicated service incorporating regularly scheduled shipments. The overwhelming number of requests for information from DOE during the TMI campaign demonstrates the need for extensive prenotification activities that satisfy public demands for information.
The decision to use dedicated train service would also demonstrate the commitment of the nuclear industry and the regulatory agencies to public safety, regardless of the additional cost. Public officials favor dedicated trains, and the available evidence indicates that the public perceives dedicated trains to be safer than regular freight services. This commitment to public safety is synonymous with a caring attitude.
In addition, the selection of dedicated trains is likely to enhance public perceptions of competence. Any minor mishaps will be perceived as indicators of more general incompetence, so extra efforts (such as the use of dedicated trains) should be taken to avoid them. The logistics of dedicated services are much more simple, and inspection, monitoring, and emergency preparedness are likely to be improved. Competence in the transport sector, however, will be of little use if it is not matched by similar competence in other sectors of radioactive waste management.
Finally, the use of dedicated trains is likely to enhance the public's perceptions of predictability. Schedules would be known ahead of time. Adherence to schedules would be more likely if dedicated trains were to be given priority in passing and were to avoid sidings and switching yards. The resulting services would be both faster and more routine.
The nuclear industry and the regulatory agencies would like to consider the subject of dedicated trains closed. With the recent reauthorization of the Hazardous Materials Transportation Act, however, the railroads have another chance to present their case for the use of dedicated trains. Improved cost analyses will likely demonstrate that the true costs of a dedicated service are only marginally greater than those of a regular freight service. Improved risk analyses will likely confirm that the risk of a catastrophe with either alternative is low, but not negligible, since the number of shipments will be high.
While such analyses necessarily inform the debate, they also serve to obscure the major issue—namely, that the benefit perceived by the public outweighs the additional cost of a dedicated system. This cost is a small price to pay for increased public acceptance or, at least, for reduced public opposition. Contrary to widely held beliefs, the public does not expect zero risk. It does expect industry and government to demonstrate a sincere concern for and commitment to public safety. Opting for dedicated trains would neither eliminate the risks nor completely restore the credibility of the nuclear industry and the agencies that regulate it, but it would help demonstrate that they are more deserving of public trust.
Theodore S. Glickman is a senior fellow in the Center for Risk Management at RFF. Dominic Golding is a fellow in the center. Research on which this article is based is expanded on in RFF discussion paper CRM91-04.
A version of this article appeared in print in the June 1991 issue of Resources magazine.
A version of this article appeared in print in the June 1991 issue of Resources magazine.
