2
Transportation Package Safety

Package performance—the ability of a transportation package1 to maintain a high level of containment effectiveness in long-term normal use and under extreme mechanical forces and thermal loading conditions (i.e., thermomechanical conditions) generated during severe accidents—is a crucial issue for transportation safety and key to understanding and quantifying transportation risks. The regulatory requirements for the design, fabrication, certification, and maintenance of these packages are substantially more rigorous than those for transporting most other types of hazardous materials (e.g., hazardous chemicals). Accordingly, the packages used to transport spent fuel and high-level waste are designed to withstand extreme thermomechanical conditions without a significant loss of containment. The slogan “safety is built into the package” is commonly used by package manufacturers and vendors to describe the ability of these packages to maintain their containment effectiveness under most conceivable accident conditions.

This chapter provides a summary of investigations carried out in the United States and several other countries to examine the performance of transportation packages. The chapter provides a brief overview of the re-

1  

Transportation containers loaded with spent fuel or high-level waste are referred to as packages in international standards, whereas the containers themselves without their contents are referred to as packagings. For simplicity, the term packages is used throughout this report. The terms casks and flasks (the latter term is commonly used in the United Kingdom) are sometimes used synonymously to refer to such packages.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States 2 Transportation Package Safety Package performance—the ability of a transportation package1 to maintain a high level of containment effectiveness in long-term normal use and under extreme mechanical forces and thermal loading conditions (i.e., thermomechanical conditions) generated during severe accidents—is a crucial issue for transportation safety and key to understanding and quantifying transportation risks. The regulatory requirements for the design, fabrication, certification, and maintenance of these packages are substantially more rigorous than those for transporting most other types of hazardous materials (e.g., hazardous chemicals). Accordingly, the packages used to transport spent fuel and high-level waste are designed to withstand extreme thermomechanical conditions without a significant loss of containment. The slogan “safety is built into the package” is commonly used by package manufacturers and vendors to describe the ability of these packages to maintain their containment effectiveness under most conceivable accident conditions. This chapter provides a summary of investigations carried out in the United States and several other countries to examine the performance of transportation packages. The chapter provides a brief overview of the re- 1   Transportation containers loaded with spent fuel or high-level waste are referred to as packages in international standards, whereas the containers themselves without their contents are referred to as packagings. For simplicity, the term packages is used throughout this report. The terms casks and flasks (the latter term is commonly used in the United Kingdom) are sometimes used synonymously to refer to such packages.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States quirements for spent fuel and high-level waste transportation packages; describes some key investigations of package loading conditions that have furthered technical understanding of package performance; and offers findings and recommendations about current standards and regulations and about improvements in package performance. There is extensive literature on package performance in response to extreme thermomechanical conditions. Some of this work appears in the “gray” literature,2 which may or may not be peer reviewed. There is also extensive classified literature on package performance in response to potential terrorist acts.3 As discussed in Chapter 1, the committee did not review this classified literature or perform an in-depth examination of package performance in response to terrorist acts. The committee does, however, comment on this issue in Chapter 5. 2.1 TRANSPORTATION PACKAGE DESIGNS AND REGULATIONS Packages for the transport of spent fuel and high-level waste are designed to meet three basic requirements both during normal conditions of transport4 and during a range of hypothetical accident conditions established in 10 CFR Part 71:5 Prevent an unsafe configuration (i.e., accidental criticality6) of spent fuel. Prevent or limit the release of radioactive contents. Limit dose rates on external package surfaces to acceptable levels. A wide range of package designs have been developed to meet these general performance requirements. 2   The U.S. Interagency Gray Literature Working Group defines gray literature as “foreign or domestic open source material that usually is available through specialized channels and may not enter normal channels or systems of publication, distribution, bibliographic control, or acquisition by booksellers or subscription agents.” (Gray Information Functional Plan, January 18, 1995, accessed at http://www.osti.gov/graylit/whatsnew.html). This literature includes technical reports, conference proceedings, and business documents. 3   In addition, some unclassified literature on this topic was removed from public circulation after the September 11, 2001, terrorist attacks. 4   Normal conditions of transport subject packages to minor mishaps due to rough handling or exposure to weather. Such conditions would not be expected to compromise the vital containment functions of the package. 5   Title 10, Part 71 of the Code of Federal Regulations, Packaging and Transport of Radioactive Material. The hypothetical accident conditions are described in subpart 73 (i.e., 10 CFR 71.73). 6   That is, a configuration that allows the establishment of a self-sustaining nuclear chain reaction as occurs in an operating nuclear reactor.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States Most packages consist of a hollow cylindrical body that is open at one end (Figure 2.1). The body is typically constructed of multiple layers of the following materials: Steel is used to provide structural strength and durability; steel, lead, depleted uranium, or concrete is used to provide shielding against gamma radiation; and water, borated polymers, or concrete is used to provide shielding against neutrons. The package closure system consists of one or two steel lids that are attached to the open end of the package body with steel bolts. Elastomer or FIGURE 2.1 Generic truck and rail spent fuel packages. In these particular designs, lead is used as the shielding material. Other materials such as steel and concrete are used in other package designs. SOURCE: Sprung et al. (2000).

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States metal seals are used between the body and lids to obtain airtight seals. The package body and lids may contain sealable openings, such as pipes and tubes, to allow for the removal of water, the addition of inert gases, leak testing of lid seals, and monitoring of internal pressures after the package is closed. The lid, lid bolts, pipes, and valves are usually recessed to protect them from damage. The package is also designed with impact limiters to absorb mechanical forces generated in the event of transport accidents and to provide thermal protection for the lid seals in case of fires. Impact limiters, which are usually attached to the ends of packages, are typically constructed of wood, rigid foam, or honeycombed metal. Metal fins may also be machined into (or welded onto) the exterior surfaces of the package body and closures for additional impact protection and heat dissipation. The interior of the package contains a basket to hold the spent fuel assemblies in a fixed configuration to ensure criticality control and minimize damage to the fuel during transport. In packages designed to hold multiple spent fuel assemblies, the baskets are typically constructed of materials containing neutron absorbers (e.g., borated metals) to provide a further margin of criticality safety. In some package designs, these baskets are placed into a stainless steel canister with a welded lid, which in turn is placed into a transportation overpack. Packages containing such canisters are sometimes referred to as canister-based packages. Packages without such canisters are sometimes referred to as bare-fuel packages because the fuel basket is placed directly into the package body. Packages designed for transport by legal-weight trucks7 typically carry between about 0.5 and 2 metric tons (0.55 and 2.2 short tons) of spent fuel. These packages are about 3 feet (0.9 meter) in diameter and weigh up to about 25 metric tons (28 short tons) when loaded. Packages designed to be transported by train can hold about 10 to 18 metric tons (11 to 20 short tons) of spent fuel. The packages typically have diameters of about 8 feet (2.4 meters) and can weigh 150 metric tons (165 short tons) or more when loaded. The number and types of spent fuel assemblies that can be carried in a package are, of course, determined by its size. Package size, in turn, depends on transportation mode (i.e., rail versus truck). Legal-weight truck package size is limited primarily by highway weight regulations, whereas rail package size is usually not weight limited but instead limited by railcar clearance requirements. Regulatory limits on radiation levels on the exterior surfaces of packages (10 CFR 71.47) also restrict the age and burn-up of spent fuel that can be carried in the package. 7   A truck having a total gross weight (i.e., including cargo) of 80,000 pounds (36,300 kilograms) or less.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States The International Atomic Energy Agency (IAEA) has established standards for the safe transportation of radioactive materials. These standards were first issued in 19618 and have undergone several revisions and amendments; the latest edition was issued in 2005 (IAEA, 2005a). The IAEA standards establish recommended requirements for a number of different package types, each designed to transport specific quantities and types of radioactive materials: Excepted packages are designed for transport of very small quantities of radioactive material such as radiopharmaceuticals. Industrial packages are designed for transport of low-specific-activity materials such as uranium ore and low-level radioactive wastes. Type A packages are designed for transport of materials of limited radioactivity—for example, uranium hexafluoride and fresh nuclear fuel. Type B packages are designed for transport of larger quantities of radioactive material including spent fuel, high-level waste, and mixed oxide fuel.9 Type C packages are designed for air transport of quantities of radioactive material exceeding a defined (large) threshold including, for example, plutonium and mixed oxide fuel. This report is concerned with Type B packages. IAEA safety standards are recommendations and are not legally binding on member states. However, the United States and many other member states adopt these standards, either in whole or in part, in their own national regulations. U.S. regulations for the packaging and transport of radioactive materials are provided in 10 CFR Part 71 (see footnote 5). The Nuclear Regulatory Commission (USNRC) and Department of Transportation (DOT) recently modified these regulations (USNRC, 2004a) to reflect the 1996 amended version of the IAEA standards (IAEA, 2000).10,11 8   The first comprehensive set of regulations for transporting radioactive materials by common carrier were established by the Interstate Commerce Commission in 1947–1948. These regulations were drafted by the National Research Council’s Subcommittee on Shipment of Radioactive Substances (see NRC, 1951). These and subsequent U.S. regulations served as an important basis for the establishment of IAEA standards (see Pope, 2004). 9   Mixed oxide fuel, or MOX, contains uranium and plutonium. 10   The updated regulations did not incorporate the IAEA standards for Type C packages, because U.S. federal law mandates more stringent requirements for air shipments of plutonium. 11   The IAEA issued updated safety standards in 2003 and 2005 (see http://www-ns.iaea.org/standards/documentpages/transport-of-radioactive-material.htm). USNRC and DOT staff were discussing whether to update U.S. regulations to bring them into conformance with these updated standards when this report was being finalized for publication in December 2005.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States Type B packages that are to be used for transporting spent fuel and high-level waste in the United States are required to be certified by the USNRC. To receive a certification, the applicant (i.e., the package manufacturer or vendor) must demonstrate to the Commission’s satisfaction that the package design will meet the testing requirements in 10 CFR Part 71, the key elements of which are described in Sidebar 2.1. Under these regulations, packages are required to survive a free-drop test onto an essentially unyielding surface (Sidebar 2.2) as well as a puncture test, an immersion test,12 and a thermal test with less than the specified loss of containment effectiveness (see Sidebar 2.1). The USNRC permits quantitative analysis (e.g., computer simulations using finite element models), scale-model (typically one-quarter or one-half scale; see Sidebar 2.3), and full-scale testing of packages or package components, and comparisons with existing approved package designs to be used to demonstrate compliance with the regulations. Testing of full-scale packages13 is not a requirement of the regulations. Testing of full-scale spent fuel packages is not carried out routinely because of the cost. A full-scale package can cost more than a million dollars, and it is generally not reusable after undergoing full-scale testing in accordance with USNRC regulations. The paucity and costs of suitable testing facilities are also impediments. There are no package testing facilities in the United States capable of handling large truck or rail packages. A new facility was recently opened in the Horstwalde region in Germany, which is located near Berlin.14 Two full-scale 9-meter (30-foot) free-drop tests on rail packages were carried out at this facility in late September 2004 in conjunction with the 14th International Symposium on Packaging and Transportation of Radioactive Materials (PATRAM 2004). A subgroup of the committee’s members visited this facility and witnessed a full-scale test of a 180 metric ton (198 short ton) rail package (Figure 2.2). USNRC regulations also require that transportation packages be de- 12   Package immersion is not discussed at much length in this chapter because the committee judges it to be of a lower concern than the thermomechanical conditions generated during truck and train accidents. 13   There is sometimes confusion about what is meant by “full scale” in regard to package testing. The regulatory tests described in Sidebar 2.1 are full-scale tests because they are carried out using actual Type B packages. These tests are referred to as certification tests. The term full-scale can also apply to tests made on actual Type B packages under simulated accident conditions such as those described in Section 2.3. These tests are referred to as demonstration tests. 14   This facility is operated by the German Federal Institute for Materials Research and Testing (BAM).

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States signed and manufactured using an approved quality assurance program.15 Packages must be designed to standards that include conservative assumptions and design margins for material properties such as yield stress and ductility. This requirement provides for a built-in “safety margin” (see Sidebar 2.4) and offers increased confidence that packages will survive thermomechanical conditions somewhat more severe than regulatory requirements. The USNRC has certified several spent fuel storage and transportation package designs for use in the United States. These include storage-only packages as well as packages that are designed for both transportation and storage (see JAI, 2005). IAEA standards and USNRC regulations for Type B packages (see Sidebar 2.1) were not derived from a comprehensive bounding analysis of all possible extreme thermomechanical conditions resulting from package mishandling and accidents. Rather, their development, which dates from the early 1960s as noted previously, was based on then-available data on typical impacts (e.g., drops from cranes, other mechanical accidents, vehicular collisions) and thermal environments (e.g., fires from spilled fuel) to which a package might be exposed in the course of transport (see Pope, 2004). As such, they do not necessarily reflect the most extreme conditions that might be encountered during spent fuel or high-level waste shipments. Nevertheless, during the committee’s information-gathering meetings, several industry presenters asserted that it is very unlikely that a certified Type B package for spent fuel or high-level waste would fail under any credible loading conditions that might be encountered during transport. This assertion is based on the confidence that these presenters place in the combination of rigorous regulatory requirements for package certification, the built-in margins of safety in current package designs (see Sidebar 2.4), and the worldwide decades-long record of spent fuel transport without a significant package failure (see Chapter 3). However, other presenters pointed out that the spent fuel transport experience in the United States is limited, and that the planned large future shipping campaigns to a federal repository could expose transportation packages to a wider range of loading conditions and longer-term use than have been experienced to date. 15   10 CFR 71.101 defines the term quality assurance as those “planned and systematic actions necessary to provide adequate confidence that a system or component will perform satisfactorily in service. Quality assurance includes quality control, which comprises those quality assurance actions related to control of the physical characteristics and quality of the material or component to predetermined requirements.” Quality assurance requirements apply to the design, fabrication, assembly, testing, maintenance, repair, modification, and use of the proposed package.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States SIDEBAR 2.1 U.S. Regulations for Type B Transport Packages Type B packages for transporting spent fuel and high-level waste are designed to withstand severe accident conditions without a loss of containment or an increase in external radiation to levels that would endanger emergency responders or the general public. USNRC regulations for Type B packages contain requirements for both “normal conditions of transport” and “hypothetical accident conditions.” Under normal transport conditions, the regulations require that Type B packages maintain their containment effectiveness as given by the following three conditions: No loss or dispersal of radioactive contents to a sensitivity of 10−6 A2 per hour.a No substantial reduction in packaging effectiveness. No substantial increase in external surface radiation levels. For hypothetical accident conditions, the regulations allow for some degradation of the packages’ containment effectiveness as given by the following four conditions: No escape of krypton-85 exceeding 10 A2 in 1 week. No escape of other radioactive material exceeding a total amount A2 in 1 week. No external radiation dose rate exceeding 10 millisieverts per hour (1 rem per hour) at 1 meter (about 40 inches) from the external surface of the package.b Compliance with these requirements may not depend on filters or mechanical cooling systems. These release and radiation limits are designed primarily to protect emergency responders and members of the public in case of an accident. The values for A2 are tabulated in 10 CFR Part 71 and are radionuclide specific. For krypton-85, the A2 value from Table A-1 in 10 CFR Part 71 is 10 terabecquerels (TBq; approximately 270 curies).c The regulations do not place any limits on the physical form of the radioactive materials (e.g., particulate size) that could be released in a hypothetical accident. The USNRC is responsible for certifying the design of Type B packages. The requirements for certification (10 CFR 71.73) are derived directly from IAEA standards: Evaluation for hypothetical accident conditions is to be based on the sequential application of the tests specified in this section, in the order indicated, to determine their cumulative effect on a package or array of packages…. [E]xcept for the water immersion tests, the ambient air temperatures before and after the tests must remain constant at that value between −29°C and +38°C which is the most unfavorable for the feature under consideration. The initial internal pressure within the containment system must be the maximum normal operating pressure, unless a lower internal pressure, consistent with the ambient temperature presumed to precede and follow the tests, is more unfavorable. a   The regulations also provide guidance on determining A2 activity values for mixtures of radionuclides. See the glossary (Appendix D). b   These increased doses would result from structural damage to the package shielding. c   The fission product krypton-85 was selected for use in this regulation because it is the only noble gas that exists in significant quantities in irradiated fuel that has been cooled for several years. While it is not a significant contributor to dose, it is likely to be among the most mobile of radionuclides in a spent fuel package.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States The following tests are specified in the regulations (10 CFR 71.73; see figure below): A free-drop test in which the specimen is dropped through a distance of 9 meters (about 30 feet) onto a flat, essentially unyielding horizontal surface (see Sidebar 2.2), with the package striking the surface in the position expected to produce maximum damage. A package dropped from this height strikes the ground at a speed of about 13 meters per second (48 kilometers per hour, or about 30 miles per hour). A puncture test in which the specimen used in the free-drop test is dropped through a distance of 1 meter (about 40 inches) onto the upper end of a 6-inch (15.2-centimeter) diameter solid, vertical, cylindrical mild steel bar mounted on an essentially unyielding horizontal surface. The package is dropped onto the bar in a position that is expected to produce maximum damage. A thermal test in which the same specimen is fully engulfed in a hydrocarbon-fuel fire with an average flame temperature of at least 800°C (about 1475°F) for a period of 30 minutes. An immersion test in which a separate, undamaged specimen is subjected to a pressure head equivalent to immersion in 15 meters (about 50 feet) of water. Additionally, 10 CFR 71.61 also specifies that for packages designed for transport of spent fuel with activity exceeding 37 petabecquerels (PBq; 106 curies), the undamaged containment system must be able to withstand an external water pressure of 2 megapascals (290 pounds per square inch) for one hour without collapse, buckling, or in-leakage of water. This pressure corresponds to a water depth of about 200 meters (650 feet), typical of maximum water depths on the U.S. continental shelf. The velocity of package impact in the free-drop test described above is lower than many real world crashes (e.g., the Central Electricity Generating Board’s [CEGB’s] full-scale crash described in Section 2.3.3). Nevertheless, the impact forces on the package are much higher in the free-drop test because the impact surface is “essentially unyielding.” Illustration of the hypothetical accident conditions in 10 CFR Part 71. SOURCE: Modified from a USNRC circular.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States SIDEBAR 2.2 What Is an “Essentially Unyielding” Surface? An essentially unyielding surface is used for impact tests on Type B packages under both International Atomic Energy Agency standards (IAEA, 2000) and U.S. regulations (10 CFR Part 71). The standards and regulations require that the impacting surface be essentially unyielding, that is, sufficiently massive and stiff to produce maximum damage to the specimen being tested. Such surfaces are usually constructed of a thick reinforced concrete slab with a steel plate floated onto its surface (i.e., slid onto the concrete surface while still wet). The concrete provides a large reaction mass and the steel plate provides stiffness. For tests of Type B packages, the slab and plate should have a combined mass at least 10 times that of the specimen being tested (IAEA, 2002). An article such as a spent fuel transportation package that is dropped onto such an unyielding surface will be subjected to higher impact forces and will consequently experience greater deformation than if the same article is dropped onto a surface that is itself deformed by the impact (i.e., a yielding surface). An article dropped onto an essentially unyielding surface in the 9-meter (30-foot) regulatory drop test (Sidebar 2.1), for example, will be traveling at about 13 meters per second (30 miles per hour, or 48 kilometers per hour) when it impacts the surface. The deformation of the article caused by this impact will be essentially identical to the deformation resulting from the head-on collision between two identical copies of that article if each is traveling at 13 meters per second. This is comparable to the difference between a collision of a moving vehicle and a parked car versus a head-on collision between two moving vehicles. The head-on collision between moving vehicles results in much greater damage. Gonzales et al. (1986) compared the hardness of an essentially unyielding target to several other target types in a series of impact tests. These tests involved impacts of a 2500 kilogram (5500 pound) cylindrical steel test unit resembling a They also questioned the adequacy of regulations that do not require full-scale testing to demonstrate package performance. There is, in fact, a good deal of quantitative information available on the performance of transportation packages under extreme loading conditions. This information comes from modeling and full-scale testing studies carried out in the United States and Europe over the past three decades. A summary of selected studies is provided in the following sections. 2.2 PACKAGE PERFORMANCE MODELING STUDIES Computer simulation models are routinely used to estimate the thermomechanical behaviors of truck and rail packages under a range of extreme loading conditions that would not be practicable to obtain from actual

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States spent fuel transportation package onto four targets: desert soil, a concrete runway, a concrete highway, and an unyielding target. The unyielding target was constructed of a 56 metric ton (62 short ton), 3.6-meter (11.8-feet) thick reinforced concrete slab with a steel face plate. The table below shows the results of these tests for an impact speed of 13 meters per second (30 miles per hour), which corresponds to the impact speed for the 9-meter (30-foot) regulatory free-drop test. In the table, the second column shows how far the test unit penetrated into each of the targets, and the third column shows the maximum strain in the test unit. Penetrations ranged from 48 centimeters (19 inches) for the soil target to 0 centimeters for the unyielding surface. The maximum strains experienced during the impact of the test unit on the unyielding surface were several times greater than for the other target types. The reason for this is simple: The other targets absorbed some of the impact forces, whereas the test unit absorbed essentially all of the impact forces when impacted against an unyielding surface. In fact, in only one case—impact onto the unyielding target—was the strain great enough to produce permanent deformation of the test unit. Experimental Results Obtained by Gonzalez et al. (1986) Target Test Unit Penetration, centimeters (inches) Maximum Strain, microstrainsa Soil 48 (19) 90 Concrete highway 10 (4) 400 Concrete runway 0.6 (0.25) 500 Unyielding target 0 3500 aMicrostrains are measured in parts per million, for example, centimeters per 106 centimeters. testing of full-scale articles. “Generic” truck and rail packages, which incorporate the salient design features of certified packages, are typically investigated with such models. The loading conditions used in these models are usually derived from historical accident reconstructions, but hypothetical accident scenarios may also be used to investigate conditions that exceed those of any known historical accident. The committee selected a number of modeling studies for discussion in the following subsections. These are summarized in Table 2.1. 2.2.1 Modal Study The modal study (Fischer et al., 1987) examined the expected responses of spent fuel transportation packages to thermomechanical conditions de-

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States FIGURE 2.10 Full-scale crash test of a Magnox spent fuel package carried out in Leicester, England by the CEGB in July 1984. This photograph was taken just after the locomotive, which was traveling at 100 miles per hour (45 meters per second), collided with the package and flatrol. The collision has lifted the package and flatrol off the ground. SOURCE: Magnox Electric Ltd. 2.3.4 Package Performance Study In 1999, the USNRC initiated a five-year project, referred to as the Package Performance Study, which had the following three objectives (USNRC, 2003c, p.4): Assess whether finite element analysis is a valuable tool for characterizing package and fuel response in extreme thermomechanical environments. Demonstrate the inherent safety of spent fuel package design using public outreach as a significant element. Refine dose and risk estimates to the public and workers through the collection of additional empirical data and improved transportation statistics. USNRC staff embarked on an effort to obtain public input to inform this study. Initially, the staff held public meetings and solicited comments through its Web site to identify concerns that could be addressed in this

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States study. An analysis of these comments by Sandia National Laboratories (Sprung et al., 2001) identified several common concerns: Some commenters asserted that the regulatory testing requirements in 10 CFR Part 71 were unconvincing as a demonstration of a transport package’s performance under severe accident conditions. They wanted more realistic full-scale testing to demonstrate package performance. Some members of the public also thought that the accident statistics used in the analytical studies of package performance in severe accidents (e.g., the 1987 modal study) needed to be reanalyzed in light of increased truck traffic and vehicular speed limits on interstate highways. USNRC staff then held additional public meetings to help them assess the accuracy of the Sprung et al. (2001) analyses of public comments and to obtain additional comments on a set of proposed package tests. These comments were used to develop a set of draft protocols for extraregulatory tests of rail and truck packages (USNRC, 2003c). Two draft protocols were proposed: High-speed (75 miles per hour [120 kilometers per hour]) impact tests involving drops of rail and truck packages from a tall tower onto an essentially unyielding surface. The rail package drop would be a package-corner-over-center-of-gravity test with the impact limiter in place. The truck package drop would be onto the package body in a so-called back-breaker orientation that bypassed the impact limiters. A fire test for both the rail and the truck packages that would involve a fully engulfing, optically dense hydrocarbon fire of a greater-than-30-minute duration. Both of these tests are extraregulatory in the sense that they would exceed the test requirements in 10 CFR Part 71 (Sidebar 2.1). The 9-meter free-drop test, for example, produces, neglecting windage, a terminal speed of 13.3 meters per second, which is approximately 30 miles per hour (or approximately 48 kilometers per hour). This impact speed is much lower than the 75 mile per hour (120 kilometers per hour) impact speed envisioned in the staff’s draft test protocol. USNRC staff proposed to perform full-scale testing on two packages that are currently certified in the United States. These packages are similar in construction to the generic packages that were used in Sandia’s spent fuel shipping risk reexamination (Sprung et al., 2000). These draft testing protocols were put out for additional public comment. USNRC staff identified several groups of suggestions from the public comments on the draft testing protocols (USNRC, 2004c): full-scale testing should be conducted to regulatory limits; tests should be based on realistic accident scenarios; they should be designed to test packages to failure; and

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States they should address terrorist acts. USNRC staff commented in its recommendations to the Commissioners (USNRC, 2004c, p. 3) that having realistic accident scenarios and testing to failure are incompatible goals. Staff also noted that terrorism was being addressed in other studies currently under way within the USNRC that “are not suitable for the public participatory approach.” USNRC staff rejected the suggestion to test packages to failure as unrealistic and unworkable for the following reasons: There is no readily agreed-upon definition of package failure among various stakeholders; package failure is a design-specific issue that would have little generic application to risk insights; and there are no realistic accident scenarios that are sufficiently severe to lead to package failure. USNRC staff also decided to eliminate extraregulatory testing. In its place, staff recommended to the Commissioners that one or more of the following tests be carried out on full-scale truck and train packages: A test of a rail package to the regulatory limit. This would subject a rail package with impact limiters to the regulatory tests in 10 CFR Part 71. A test of a truck package with its impact limiters to the same regulatory limits, except that the 2-megapascal pressure test might be eliminated if the truck and rail packages are tested in combination. A full-scale crash demonstration of a rail package with impact limiters and its railcar with a simulated bridge abutment at about 75 miles per hour (120 kilometers per hour), followed by a fire from a ruptured tank car. A full-scale crash test of a truck package with impact limiters and its trailer by a locomotive traveling at about 75 miles per hour (120 kilometers per hour) on a grade crossing followed by a fire. The staff proposed to the Commissioners a testing protocol involving various combinations of these tests, as well as an option to perform the extraregulatory tests originally outlined in USNRC (2003c). In December 2004, the Commissioners approved a modified full-scale test that will be carried out when funds are made available by Congress (USNRC, 2004d). USNRC staff was directed to plan for a demonstration test involving a single rail spent fuel package involved in a “viable” transportation accident. The Commissioners directed that the test should consist of a simulated rail crossing with a train traveling at an appropriate speed colliding at a ninety degree angle with a transportation cask on its rail carrier car in a normal transportation configuration…. The test will consist only of the collision and the natural results of that collision. No separate fire testing or immersion testing will be conducted on the cask.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States The Commissioners directed USNRC staff to prepare an information paper that provided details and estimated costs of the test. The requested information paper was submitted to Commissioners in March 2005 (USNRC, 2005a). The staff proposed a test involving a full-scale package of the kind that is likely to be used by the Department of Energy (DOE) in its Yucca Mountain transportation program. The package would contain surrogate fuel elements and would be mounted on a rail carrier car placed at 90 degrees to a simulated rail crossing. The rail package would be subjected to a collision with a locomotive and several freight cars traveling at 60 miles per hour (96 kilometers per hour). In accordance with the Commissioners’ December 2004 directive, no fire or immersion testing was included in the proposed testing plan. The Commissioners approved this proposed test in June 2005 (USNRC, 2005b). In addition, the Commissioners reversed their earlier instructions concerning the thermal test, directing the staff to “add a fire test scenario … involving [a] fully engulfing, optically dense, hydrocarbon fire for a duration of one-half hour post-collision.” The Commissioners directed staff to inform it of the details and estimated costs for the proposed fire test. They also noted that the proposed test plan provided by staff (USNRC, 2005a) “is not the final word on this issue, as the project is subject to additional modifications and Commission direction once additional information becomes available.” The Commissioners directed staff to engage with higher-level management in DOE to request financial support for the demonstration test and to request increased funding from Congress if DOE is unable to provide support. 2.4 CRITICAL ASSESSMENT OF PACKAGE PERFORMANCE It is clear to the committee that package performance under severe accident conditions is a major concern for transportation safety among many members of the public, especially those who live and work along shipping routes. Finding a way to resolve this issue continues to be a challenge to regulators in the United States and may eventually become a challenge for DOE and the private sector in their commercial spent fuel transportation programs. The packages used to transport spent fuel and high-level waste are designed to contain their radioactive contents under normal transport conditions and to withstand accident conditions without an increase in external radiation to levels that would endanger emergency responders or the general public. The package performance standards developed by the IAEA and embodied in USNRC regulations (Sidebar 2.1) were developed to help ensure that properly manufactured, certified packages provide such containment effectiveness. The 9-meter regulatory free-drop test onto an essen-

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States tially unyielding surface, for example, is a more severe mechanical test of package performance than has been produced in any of the severe full-scale crash tests described in this chapter (see Section 2.3). The safety of transporting spent fuel and high-level waste depends to a great extent on the “inherent” safety of transportation packages to contain their contents even under severe accident conditions. The committee observes that while the full-scale crash tests described in this chapter have imposed mechanical conditions on packages that are less severe than the regulatory free-drop test, some of these tests have resulted in small releases from package containments.36 However, these releases would not have exceeded regulatory limits, which have as their goal the protection of emergency responders and the general public. The committee judges that the regulatory free-drop test imposes mechanical forces that are severe enough to bound conditions likely to be encountered in foreseeable real-life accidents. At the same time, the committee understands that any releases of radioactivity could have substantial social risk implications (see Chapter 3). The accident reconstruction and modeling studies described in Section 2.2.3 suggest that there may be a very small number of credible accident conditions involving very long duration, fully engulfing fires37 that are potentially capable of damaging the seals on transportation packages if such fires are allowed to burn in an uncontrolled manner for long periods of time. Such damage could potentially lead to the release of radioactive material from the package through the processes described in Sidebar 2.5. The potential for radioactive material releases from packages involved in such fires and the consequences of such releases are incompletely understood at present. USNRC staff has completed a solid technical analysis of the response of one package design to a realistically extreme thermal event resulting from the Howard Street tunnel fire. Additionally, staff are completing analyses of the performance of two additional package types for this fire exposure (see footnote 22 in this chapter). The committee judges that additional analyses of this type are needed to better understand package performance for realistic accident conditions involving very long duration fires. 36   The releases described in this chapter consisted of water that leaked from the package interiors. This water may have been slightly contaminated with radioactive material from the fuel rod cladding (crud), but the fuel itself remained intact. It is important to note that in the United States, packages now in use to transport commercial spent fuel are filled with inert gas, not water. Inert gas is a much less efficient medium than water for transporting radioactive contamination out of leaking packages. 37   The committee uses this term to describe fires that burn for periods of hours (or longer). Such fires could produce thermal loading conditions that exceed those for the regulatory thermal test specified in 10 CFR 71.73. Some of the real-world accidents described in this chapter involved fires that burned for hours to days.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States The number of additional analyses might be small in number, especially if they focused on fire scenarios that would essentially bound expected real-world accident conditions,38 and if they examined the performance of a representative set of package designs that are likely to be used in future large-quantity shipping programs. The analyses have to address both the containment effectiveness of packages in response to very long duration fires and the consequences of any radioactive releases that result from package seal failures. The purpose of these analyses would be to inform changes, if needed, to regulatory and operational practices to reduce the likelihood of occurrence and potential consequences of accidents involving very long duration fires. At least two options are available to regulators and implementers if such changes are needed: Operational steps to reduce the likelihood of occurrence of long-duration fires during spent fuel transport, and/or Design or testing requirements to improve the thermal resistance of transportation packages. The first option would clearly be preferable from a cost and implementation standpoint. The objective of taking such operational steps would be to reduce the potential for accidents between trucks or trains carrying spent fuel and other vehicles carrying large quantities of flammable materials, especially flammable liquids, in places where it might be difficult to mount an effective firefighting response.39 Operational controls are used routinely by railroads to control transportation hazards. Such controls would probably be more easily implemented and enforced when shipments of spent fuel and high-level waste are made by dedicated train. The second option would be more difficult and expensive to implement, because it very likely would require expensive changes to regulatory testing requirements and some existing package designs. However, there would be no reason to change the testing requirements or package designs if effective operational controls could be implemented. The committee received several comments at its meetings on the testing of transportation packages under extreme loading conditions, especially 38   The historical accident record provides perhaps the best available information to identify bounding accident scenarios. For example, the Livingston, Louisiana, fire (see Table 2.2) and Summit Tunnel fire (see Section 2.2.3 and Figure 2.4) are possible examples of bounding scenarios. 39   Tunnel fires are of special concern because of access restrictions that can delay or prevent an effective firefighting response. Such delays could allow the fires to burn in an uncontrolled manner for long periods of time.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States with respect to the USNRC’s Package Performance Study. Many commenters asserted that the regulatory testing requirements in 10 CFR Part 71 are unconvincing as a demonstration of a transport package’s performance under severe accident conditions. They wanted realistic full-scale testing to be conducted to demonstrate package performance. The State of Nevada suggested that comprehensive full-scale testing would improve the overall safety of the package and vehicle system and enhance confidence in risk analysis techniques. Full-scale testing was proposed as a way to potentially increase the acceptance of spent fuel shipments by state and local officials and members of the general public and to “potentially reduce adverse social and economic impacts caused by public perception of transportation risks” (Hall, 2003, p. 3). Full-scale testing can be used to determine how packages will perform under both regulatory and credible extraregulatory conditions. To be of value, however, the committee believes that extraregulatory tests must be designed to closely mimic conditions that would reasonably be expected to be encountered in actual service. The test program undertaken by the CEGB in the 1980s (Section 2.3.3) provides an excellent example of the appropriate use of full-scale testing. In that program, the full-scale test was the end point of a much larger deliberative investigation of the conditions that could be encountered during transport of spent fuel. The full-scale test was used both to validate the analytical and scale-modeling work that had been carried out beforehand, and at the same time to provide a demonstration of the containment effectiveness of the packages in a way that could be appreciated by the general public. The USNRC’s Package Performance Study has elements of an appropriate full-scale testing program. The design for a full-scale test was developed through a deliberative process that involved technical analysis and provided many opportunities for public comment. However, none of the tests that emerged from this process were selected by the USNRC for execution. In fact, the scenario that was selected (a collision between a locomotive and a package mounted on a railcar placed at 90 degrees to the rail crossing in an upright position; see Section 2.3.4) duplicates many features of the rail-crossing impact test carried out by Sandia National Laboratories more than 25 years ago. One very important difference between the Sandia test and the currently proposed test is the packages selected for testing. The Sandia test used an obsolete truck transport package. The proposed test would use a rail package that is currently certified and of the type to be used by DOE for transport of spent fuel to the federal repository. If executed properly, this full-scale test will be valuable for validating scale-model and computer simulations of package performance and thereby increasing confidence in current regulatory testing approaches. The test is also likely to be

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States visually powerful and could provide the public with useful information about the performance of transportation packages. The committee also endorses package testing studies that integrate full-scale testing, scale-model testing, and finite element and associated structural analysis methods to demonstrate compliance with regulatory standards. These analysis methods allow a much wider variety of failure scenarios to be examined than is possible with full-scale testing alone. The analysis codes (e.g., LS DYNA, PRONTO) have been well benchmarked for these applications. Full-scale testing has two primary uses in this integrated test regime. First, it can be used to validate the computer codes for each package design, thus providing additional confidence in package performance under real-world conditions. Second, it provides a direct demonstration of a package’s ability to meet specific regulatory test conditions. Several participants at the committee’s information-gathering meetings strongly urged that testing to failure should be carried out on all package designs that will be used to transport spent fuel and high-level waste to Yucca Mountain. However, they were not clear on what would constitute a failure. When pressed for specifics, the committee understood that what was generally meant was that full-scale testing should be carried out to destruction, presumably to establish the ultimate strength of transportation packages. The State of Nevada specifically recommended that the USNRC undertake an evaluation of the costs and benefits of destructive testing of a randomly selected production model cask (Hall, 2003, p. 3). The principal argument in favor of destructive testing is that it can provide information on the magnitudes of thermomechanical loads required to eliminate containment effectiveness, thus establishing a “safety factor” for a given package design. There are significant drawbacks to this approach, however. Most importantly, each of the tests would involve only one thermal or mechanical loading condition out of the set of many possible conditions, and the selected loading condition would likely far exceed what could reasonably be expected to occur even in the most extreme real-world accidents. Moreover, a separate package might have to be acquired for each test, which would greatly increase the costs of a testing program for even a single package design. It is important to recognize that any transportation package could be destroyed if no limits are placed on the loads that act upon it. Moreover, the failure of a package, in the sense that it can no longer perform its intended containment function, will generally occur under conditions that are much less severe than needed for destruction. Consequently, even if costs were not prohibitive, testing to destruction would provide little or no insight into the conditions that would cause a loss of package containment under real service conditions.

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States Even full-scale testing to failure could be problematical. Because examining the multitude of possible accident conditions and failure scenarios through full-scale testing is clearly impractical, the committee judges that the bounding approach that the IAEA has established is entirely appropriate. The approach is one that reflects four plausible accident-like conditions: free drop, puncture, thermal exposure, and water immersion (see Sidebar 2.1). The important question to be answered by testing is not whether a package could be made to fail; as noted previously, it would certainly be possible to design tests that would accomplish this goal. Rather, the question that needs to be answered is whether there are credible accident conditions that would result in releases of radioactivity to the environment that would endanger emergency responders or the general public. It is clear from the modeling and full-scale tests described in this chapter that transportation packages are extremely rugged. The committee judges that packages designed, fabricated, used, and maintained under current regulatory standards are very unlikely to encounter loading conditions under real-world conditions, with the possible exception of very long duration fires, that would lead to releases in excess of regulatory limits. The committee recognizes, however, that even minor releases from package containment might have important social implications. These are discussed in Chapter 3. 2.5 PACKAGE PERFORMANCE FINDINGS AND RECOMMENDATIONS The committee offers the following findings and recommendations based on the analysis of package performance provided in this chapter. FINDING: Transportation packages play a crucial role in the safety of spent fuel and high-level waste shipments by providing a robust barrier to the release of radiation and radioactive material under both normal transport and accident conditions. International Atomic Energy Agency package performance standards and associated Nuclear Regulatory Commission regulations are adequate to ensure package containment effectiveness over a wide range of transport conditions, including most credible accident conditions. However, recently published work suggests that extreme accident scenarios involving very long duration, fully engulfing fires might produce thermal loading conditions sufficient to compromise containment effectiveness. The consequences of such thermal loading conditions for containment effectiveness are the subject of ongoing investigations by the Nuclear Regulatory Commission and other parties, and this work is improving the understanding of package performance. Nonetheless, additional analyses and experimentation are needed to demonstrate a bounding-level understand-

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States ing of package performance in response to very long duration, fully engulfing fires for a representative set of package designs. RECOMMENDATION: The Nuclear Regulatory Commission should build on recent progress in understanding package performance in very long duration fires. To this end, the agency should undertake additional analyses of very long duration fire scenarios that bound expected real-world accident conditions for a representative set of package designs that are likely to be used in future large-quantity shipping programs. The objectives of these analyses should be to Understand the performance of package barriers (spent fuel cladding and package seals); Estimate the potential quantities and consequences of any releases of radioactive material; and Examine the need for regulatory changes (e.g., package testing requirements) or operational changes (e.g., restrictions on trains carrying spent fuel) either to help prevent accidents that could lead to such fire conditions or to mitigate their consequences. Strong consideration should also be given to performing well-instrumented tests for improving and validating the computer models used for carrying out these analyses, perhaps as part of the full-scale test planned by the Nuclear Regulatory Commission for its package performance study. Based on the results of these investigations, the Commission should implement operational controls and restrictions on spent fuel and high-level waste shipments as necessary to reduce the chances that such fire conditions might be encountered in service. Such effective steps might include, for example, additional operational restrictions on trains carrying spent fuel and high-level waste to prevent co-location with trains carrying flammable materials in tunnels, in rail yards, and on sidings. FINDING: The committee strongly endorses the use of full-scale testing to determine how packages will perform under both regulatory and credible extraregulatory conditions. Package testing in the United States and many other countries is carried out using good engineering practices that combine state-of-the-art structural analyses and physical tests to demonstrate containment effectiveness. Full-scale testing is a very effective tool both for guiding and validating analytical engineering models of package performance and for demonstrating the compliance of package designs with performance requirements. However, deliberate full-scale testing of packages to destruction through the application of forces that substantially exceed credible accident conditions would be marginally informative and is not

OCR for page 55
Going the Distance? The Safe Transport of Spent Nuclear Fuel and High-Level Radioactive Waste in the United States justified given the considerable costs for package acquisitions that such testing would require. RECOMMENDATION: Full-scale package testing should continue to be used as part of integrated analytical, computer simulation, scale-model, and testing programs to validate package performance. Deliberate full-scale testing of packages to destruction should not be required as part of this integrated analysis or for compliance demonstrations.