In This Issue
Radioactive Waste Disposal
September 1, 2003 Volume 33 Issue 3

Will the United States Need a Second Geologic Repository?

Monday, September 1, 2003

Author: Per F. Peterson

Between 2007 and 2010, Congress must consider whether the United States needs a second repository for high-level radioactive waste.

Nuclear fission energy requires small inputs of natural resources compared to most other fossil and nonfossil energy technologies. When we consider net electricity generation (e.g., net electricity after subtracting consumption by internal plant loads and by uranium enrichment plants), the life-cycle resource inputs for nonfossil power sources are dominated by construction materials, most notably steel and concrete. The construction of existing 1970-vintage U.S. nuclear power plants required 40 metric tons (MT) of steel and 190 cubic meters (m3) of concrete per average megawatt of electricity (MW(e)) generating capacity.1 For comparison, a typical wind-energy system operating with 6.5 meters-per-second average wind speed requires construction inputs of 460 MT of steel and 870 m3 of concrete per average MW(e). Coal uses 98 MT of steel and 160 m3 of concrete per average MW(e) (Pacca and Horvath, 2002); and natural-gas combined cycle plants use 3.3 MT steel and 27 m3 concrete (Meier, 2002).

Because of this efficient use of natural resources, compared to other energy technologies, nuclear energy is an important candidate for the long-term, sustainable production of electricity and hydrogen.2 But any major role for fission will require practical approaches to spent-fuel management with environmental and public health impacts comparable to, or lower than, those of other sustainable energy technologies.

Of the resources required to produce fission energy, repository sites are arguably the only resource that has proven to be scarce. Given this scarcity, the allocation and efficient use of available repository capacity will require well informed technical and policy decisions. The 1982 Nuclear Waste Policy Act (NWPA), as amended in 1987, requires that Congress consider these questions between 2007 and 2010.3 Advanced fuel cycles (AFCs) cannot eliminate the need for repositories, but do have the potential to greatly increase repository capacity and improve performance (NRC, 1996). AFC optimization and economics will depend strongly on the licensing basis for repository sites and on the extent of AFC R&D performed prior to any large-scale deployment. In this article, I outline the major issues facing the United States in considering the role of AFC technology in nuclear waste management.

Regulatory standards to protect current and future public health, safety, and the environment are adopted through national policy-making processes. The regulatory criteria for Yucca Mountain require, among other things, that the groundwater below the Armagosa Valley near Yucca Mountain be protected for at least 10,000 years; the maximum radiation dose to an individual who drinks two liters of groundwater per day must be less than 4 mrem, in other words, less than 1.3 percent of current U.S. average natural radiation exposures. Like other regulated, engineered systems (e.g., aircraft), repository systems apply redundancy and diversity to meet these regulatory requirements with acceptable uncertainty. Uncertainty in the performance of individual repository barriers and processes are acceptable because redundancy and diversity reduce the uncertainty in total system performance. The isolation provided by deep geologic storage helps bound uncertainties by creating chemical, thermal, and hydrologic conditions that can be predicted with less uncertainty over long time scales than conditions near the surface. Indeed, the fate of geologic repositories may be one of the few important environmental and public health impacts from twentieth and twenty-first century energy production that we can predict with modest uncertainty over millennial time scales.4

In 2002, the United States selected Yucca Mountain in southern Nevada as the site for the nation’s first high-level waste repository, and Congress directed the U.S. Department of Energy (DOE) to prepare and submit a license application to the U.S. Nuclear Regulatory Commission (USNRC) by the end of 2004. This site-selection recommendation was based on the results of an integrated total-system performance assessment (TSPA-SR) model that included nine different barrier mechanisms. Repeated TSPA calculations, that varied the values of uncertain parameters in the model, generated statistical performance predictions showing that regulatory requirements for groundwater protection could be met by a factor of more than 100 with high confidence (DOE, 2002).

For the upcoming license application, improvements to the TSPA models, particularly in near-field transport and drift seepage, can be expected to reduce the need for conservative assumptions in many of the barrier models. At the same time, questions raised by the USNRC during the license review can be expected to require more conservative and wider uncertainty estimates for some barrier models. Recent "one-off" studies have shown that performance is relatively well distributed across the multiple barriers (Apted et al., 2002; NWTRB, 2002). Thus, to alter the recent TSPA-SR’s positive assessment of regulatory compliance, the USNRC would have to identify substantial deficiencies in the uncertainty and conservatism of a substantial fraction of the barrier models. This outcome appears unlikely.

The NWPA limits the capacity of the proposed Yucca Mountain repository to 63,000 MT of initial heavy metal in commercial spent fuel.5 The 103 U.S. commercial reactors currently operating will produce this quantity of spent fuel by 2014. Recently, the federal government has started to issue 20-year license renewals for U.S. nuclear plants, extending the permitted plant operating life to 60 years. As of May 2003, 16 U.S. plants had received renewals and 14 applications were under review; 22 more applications are expected in the next two years. Because of the low average production cost of nuclear electricity (1.69 cents per kilowatt-hour in 2002), it is anticipated that a substantial fraction of remaining U.S. plants will also seek renewals, thus increasing the total federal spent-fuel management obligation for current reactors to as much as 125,000 MT. Licenses for new plant construction would increase the total further.

Technical Capacity: Commercial Spent Fuel
The capacity of geologic repositories is set primarily by areal heat load limits for decay heat and by available footprint. At Yucca Mountain, spent fuel and high-level waste will be placed in corrosion-resistant canisters and emplaced horizontally in 5.5-m diameter drift tunnels. The TSPA-SR set a loading limit of 60 MT/acre, based on tunnels spaced 81 m apart and on the nominal characteristics of the first 70,000 MT of defense and commercial wastes planned to be sent to the repository. For spent fuel from pressurized-water reactors (PWRs), comprising 60 percent of commercial spent fuel, the TSPA-SR canister design permits a loading of 87 MT/acre. The boiling-water reactor (BWR) canister design permits 75 MT/acre.

With a relatively modest license amendment to increase the site capacity, an average areal loading of 75 MT/acre for commercial spent fuel is a reasonable assumption. This value has potential conservatisms that might be erased with subsequent, more aggressive license amendments. However, it is unlikely that increases in areal loading could exceed a factor of two or three for commercial spent fuel.

The maximum repository footprint at Yucca Mountain is correspondingly uncertain. An earlier viability assessment (TSPA-VA) concluded that, with a substantial new characterization, the total repository area could potentially be increased to somewhat more than 2,000 acres (8.0 km2). This suggests a minimum "technical" site capacity of approximately 75 x 2,000 = 150,000 MT of spent fuel, with a maximum site capacity greater by perhaps a factor of two or three. Thus any substantial construction of new U.S. nuclear power infrastructure in the coming decades will almost certainly create a technical requirement (perhaps as soon as 2030 to 2050) either for additional repositories or for the construction of infrastructure for recycling spent fuel.

Technical Capacity: Advanced Fuel Cycles
It is technically possible for AFCs to recycle and transmute almost all of the heavy actinide elements that contribute to decay heat, leaving only fission products and residual actinides for disposal. Only two of the fission-product isotopes- strontium(Sr)-90 and cesium(Cs)-137, both of which have 30-year half- lives - would contribute significantly to the remaining decay heat. Because these isotopes have relatively short half lives, it is technically possible to separate and manage them separately for the 200 to 300 years required for their nearly complete decay. Separation and separate management of Cs-137 and Sr-90 have already been demonstrated at large scale at the Hanford site in Washington state, where both cesium and strontium recovered from high-level waste are currently stored separately in sealed capsules.6

Without cesium and strontium, the remaining fission products and residual actinides that require geologic disposal have very small rates of decay-heat generation. Thus, it becomes relatively easy to estimate the capacity of the Yucca Mountain site. If the current canister design for defense high-level waste (capable of holding five 60-cm diameter cylinders of borosilicate waste glass) were used to hold fission products, the fission-product loading could be 500 kg/m of drift tunnel length;7 this is 7 times greater than the fission-product loading for current 21-assembly PWR canisters. A 1-GW(e) light-water reactor (LWR) (whether a BWR or PWR), which can produce energy for one million typical homes, also produces approximately 1,080 kg of fission products per year. Slightly more than two meters of Yucca Mountain drift could hold a year’s fission products from a plant this size.

At 2,000 acres, the Yucca Mountain site could have 100 km of drift tunnels spaced at 81 m. Without decay heat, the spacing could be reduced to 20 m, thus increasing the drift tunnels to 400 km. Using the existing defense-waste canister design, these drift tunnels could then hold 200,000,000 kg of fission products, the energy equivalent of burning one trillion tons of coal. This means that a single Yucca Mountain could replace 170 years of current, total, worldwide coal consumption.8
Separation and separate management of cesium and strontium would require management of these materials for one to three centuries, until radioactive decay reduces their heat output sufficiently to permit their disposal. However, there are alternative strategies for managing cesium and strontium decay heat, because it drops greatly over the time scales of surface storage and repository operation.

Because the Yucca Mountain repository is located above the water table, air can be circulated through the tunnels to remove decay heat. In the current design, ventilation will continue for 50 years after the final canister emplacement. As shown in Figure 1, the ventilation extracts most of the heat from cesium and strontium. For every 30 years of operation, the ventilation system regenerates roughly half of the repository’s capacity for holding cesium and strontium.

With ventilation, the heavy actinide elements, particularly americium(Am)-241, which has a 460-year half-life, drive Yucca Mountain’s postclosure thermal response. To eliminate the need for a second reposi-tory, one must therefore cap the total inventory of heat-generating actinides, particularly Am-241, within the thermal capacity of the site.

Relatively large inventories of Am-241 build up in the spent fuel of current LWRs because of successive neutron captures in U-238. Neutron capture in U-238 yields Pu-239, and fission of this Pu-239 provides a substantial fraction of the power output from LWRs (reaching 50 percent shortly before the fuel is discharged). But with the relatively low kinetic energy of neutrons in LWRs, a significant fraction of neutron reactions with Pu-239 are capture reactions that generate Pu-240, and

Options for reducing the rate of accumulation of heat-generating actinides include adding thorium(Th)-232 into fuel as a substitute for a portion of the U-238. Neutron capture into Th-232 produces U-233, a fissile element like Pu-239 that can generate a portion of the reactor power. Unlike Pu-239, however, neutron capture in U-233 creates relatively light isotopes, thus substantially reducing the buildup of heat-generating actinides.

The generation of actinides can be further reduced by increasing reactor operating temperatures and electrical conversion efficiency, as is possible with high-temperature, gas-cooled reactors. At the same time, all of these approaches also increase the volume of fuel materials in which the actinides are contained, which could require substantial changes to the repository system design to take advantage of the reduced heat output and to increase areal loading.

Recycling spent fuel - using chemical reprocessing to separate and recycle some or all of the actinides - greatly reduces volume. Conventional reprocessing and recycling of separated plutonium into LWRs, as is currently done in France and Britain, increases the total inventory of Am-241 that requires management by increasing neutron capture into Pu-240. Therefore, for recycling to help in capping the total inventory of heat-generating actinides, new reactors capable of transmuting these actinides (e.g., fast or epithermal designs) must be developed and deployed.

The need for recycling will only arise if there is substantial construction of new reactors in the United States. Thus, the deployment of recycling infrastructure would occur in an environment with an established, large-scale technical and industrial capacity for nuclear construction.

Economics: The Nuclear Waste Fund
In the United States, national policy requires that "the costs of carrying out activities relating to the disposal of [high-level] waste and spent fuel will be borne by the persons responsible for generating such waste and spent fuel" (NWPA, Section 112). Thus the costs for civilian spent-fuel disposal are internalized by charging a 0.1 cent per kilowatt-hour fee on nuclear electricity consumption.

DOE periodically issues a report assessing whether this fee is adequate to fund the life-cycle cost of spent-fuel disposition. In 2001, when only five reactors had received 20-year license renewals, DOE estimated that the total quantity of commercial spent fuel requiring disposition would be 83,800 MT of heavy metal. Assuming that all of this spent fuel would be emplaced in Yucca Mountain, the total cost of the repository was estimated to be $57.5 billion (2001 dollars). Of that total, 29 percent is assigned to defense-waste disposal, making the cost of commercial waste disposal $490/kg (DOE, 2001).

For an average burn-up9 of 40 MWd/kg and plant thermal efficiency of 0.32, the current 0.1 cent/kWh fee generates revenues of $310/kg. The Nuclear Waste Fund accrues interest at a real rate exceeding inflation by 2.6 percent (the historical average for government bonds) to 4.2 percent (the rate for 10-year treasury notes) (DOE, 2001). Like plant decommissioning costs, this accumulated interest reduces the present cost of waste management activities that can be delayed to the future. For example, after 30 years of no-cost storage at a reactor, the fund grows to between $670 and $1070/kg of spent fuel. (Most current plants have on-site storage capacity for 30 years; all new plant designs include storage capability for the 60-year licensed life of the plant.)

The modest $490/kg cost of direct disposal in Yucca Mountain contrasts sharply with current cost estimates for recycling. In a recent study, the OECD Nuclear Energy Agency (NEA) estimated costs of $1,000 to $2,500/kg just for spent-fuel reprocessing and noted that studies in the 1990s of sodium-cooled fast reactors for transmuting separated actinides estimated capital costs some 30 percent higher than for LWRs (NEA/OECD, 2002). Assuming real interest rates of 7 to 10 percent, NEA predicted that closed-cycle nuclear electricity prices are 0.2 to 1.0 cent/kWh higher than for LWR electricity with direct disposal of spent fuel; this is two to ten times the current U.S. Nuclear Waste Fund fee.

But if transmutation is performed primarily to cap the total inventory of heat-generating actinides within the thermal capacity of a single repository site, then transmutation infrastructure can be financed from the fees and interest accumulated in the Nuclear Waste Fund. This is equivalent to financing construction at an effective real interest rate of 3 percent, rather than at a commercial rate of 7 to 10 percent. This lower rate would reduce the capital charges for transmutation by more than 50 percent.10 The availability of low-interest-rate capital would more than offset the higher capital costs estimated for sodium fast reactors, compared to LWRs.11

Future Options
The 1982 NWPA adopted a 70,000 MT limit for commercial spent fuel and defense wastes in an attempt to ensure an equitable distribution of geologic repository sites between the eastern and western United States. Subsequent experience showed that the characterization and siting of a single repository was far more arduous - in cost, time, and acrimony - than the NWPA had envisioned. Since 1982, our understanding has changed in other areas as well. The carbon emissions of nuclear energy’s primary competitor, fossil fuel, are now understood to have potentially global environmental effects. Therefore, coal consumption in the western United States affects not only the eastern United States, but also Europe and Asia. In addition, since then the major technical elements for actinide management have been demonstrated at laboratory scale, and engineering designs for AFC demonstration facilities have been developed and remain available for further refinement.

Proponents of once-through fuel cycles commonly cite the costs of reprocessing and transmutation as arguments for direct disposal as the lowest-cost option for the foreseeable future (MIT, 2003; von Hippel, 2001). They present no compelling arguments, however, that the protracted and arduous technical and political process required to select Yucca Mountain could be repeated successfully for a second, third, fourth, and subsequent repositories. Conversely, arguments for reprocessing often do not consider that a large amount of spent fuel can be managed with a single repository before a technical need or economic motivation emerges for recycling a fraction of spent fuel.

Upcoming U.S. policy decisions for civilian spent-fuel management (beyond the current limit of 63,000 MT for Yucca Mountain) cannot be based on large taxpayer subsidies; future policy must be based on a credible Nuclear Waste Fund fee schedule. The Generation IV International Forum currently envisions a deployment goal of 2030 for advanced nuclear energy systems for actinide management (Generation IV International Forum, 2003). With our current understanding of the technical limits of Yucca Mountain’s capacity, this timing for the deployment of recycling strikes a balance between two pragmatic realities: (1) finite limits to repository capacity; and (2) the need for R&D to make recycling technology economically attractive.

Adams, M.L. 2002. Sustainable energy from nuclear fission power. The Bridge 32(4): 20-26.
Apted, M.J., D. Langmuir, D.W. Moeller, A. Joohhong, A.E. Waltar, and D. von Winterfeldt; R.C. Ewing, and A. Macfarlane. 2002. Yucca Mountain: should we delay? Science 296(5577): 2333-2335.
Bryan, R.H., and I.T. Dudley. 1974. Estimated Quantities of Materials Contained in a 1000-MW(e) PWR Power Plant. ORNL-TM-4515. Prepared for the U.S. Atomic Energy Commission. Oak Ridge, Tenn.: Oak Ridge National Laboratory.
DOE (U.S. Department of Energy). 2001. Nuclear Waste Fund Fee Adequacy: An Assessment. DOE/RW-0534. Washington, D.C.: U.S. Department of Energy.
DOE. 2002. Yucca Mountain Site Suitability Evaluation. Washington, D.C.: U.S. Department of Energy.
Generation IV International Forum. 2003. Generation IV Roadmap. Available online at:
MIT (Massachusetts Institute of Technology). 2003. The Future of Nuclear Power. Cambridge, Mass.: MIT. Also available online at:
Meier, P.J. 2002. Life-Cycle Assessment of Electricity Generation Systems and Applications for Climate Change Policy Analysis. UWFDM-1181. Madison, Wis.: Fusion Technology Institute, University of Wisconsin.
NEA/OECD (Nuclear Energy Agency/Organization for Economic Cooperation and Development). 2002. Accelerator-Driven Systems (ADS) and Fast Reactors (FR) in Advanced Nuclear Fuel Cycles: A Comparative Study. Report 3109. Paris: OECD. Also available online at:
NRC (National Research Council). 1996. Nuclear Wastes: Technologies for Separations and Transmutations. Washington, D.C.: National Academy Press.
NWTRB (Nuclear Waste Technical Review Board). 2002. Minutes of the September 10, 2002, meeting. Available online at:
Pacca, S., and A. Horvath. 2002. Greenhouse gas emissions from building and operating electric power plants in the upper Colorado River basin. Environmental Science and Technology 36(14): 3194-3200.
von Hippel, F.N. 2001. Plutonium and reprocessing of spent nuclear fuel. Science 293(5539): 2397-2398.

1 Based on the current U.S. average capacity factor of 90 percent, with data from Bryan and Dudley (1974). 2 Uranium is abundant, with average concentrations in U.S. soils of 1.8 ppm, or about 2.7 metric tons per square kilometer in the top meter of soil ( Known, economically recoverable, high-quality ores contain 3.3 million MT of uranium and 4 to 6 million MT of thorium, which if used in closed fuel cycles hold energy equal to 1,500 times current total worldwide annual energy consumption (Adams, 2002).
3 Specifically, NWPA states that "The Secretary [of the DOE] shall report to the President and to Congress on or after January 1, 2007, but not later than January 1, 2010, on the need for a second repository."
4 At its current statutory capacity limit of 63,000 MT of commercial spent fuel, Yucca Mountain displaces energy equivalent to 5 billion tons of coal, or six years of current U.S. coal consumption. Advanced fuel cycles might expand this capacity by a factor of more than 50. Coal mining mostly occurs at or near the surface, and its combustion products are widely dispersed. Thus, it is difficult even to speculate about environmental and public health consequences in 10,000 years.
5 Specifically, NPWA states that "The [Nuclear Regulatory] Commission decision approving the first such application shall prohibit the emplacement in the first repository of a quantity of spent fuel containing in excess of 70,000 metric tons of heavy metal or a quantity of solidified high-level radioactive waste resulting from the reprocessing of such a quantity of spent fuel until such time as a second repository is in operation." Of this quantity, 7,000 metric tons is commonly assumed to be allocated to defense wastes.
6 Cesium-135 is found in very small concentrations in fission products, and has a 2.6 million year half-life.
7 This is the fission-product loading for a glass density of 2,700 kg/m3, at a fission-product mass fraction of 15 percent.
8 Nuclear fission releases 1 GW-day/kg, or 86 x 106 MJ/kg of fission products. Coal releases around 32 MJ/kg. In 2001, worldwide coal consumption was 3.0 billion MT.
9 Burn-up is the amount of energy released by fission (megawatt-days) in a given initial mass of fuel (kilograms) and is directly proportional to the mass of fission products in the fuel.
10 NEA Case 3a for recycling minor actinides into fast reactors gives total electricity costs from 5 to 25 percent higher than the once-through electricity cost of 3.8 cents/kilowatt-hour. Assuming uniform capital outlays over a four-year period and operation for 60 years, capital charges drop by 52 percent if the interest rate is reduced from 7 percent to 3 percent (NEA/OECD, 2002).
11 In practice, the construction of reprocessing infrastructure could be funded by direct appropriations from the Nuclear Waste Fund. Transmutation services could be procured through long-term contracts with commercial reactor operators, which would permit commercial reactor operators to obtain favorable commercial financing.
About the Author:Per F. Peterson is professor and chair of the Department of Nuclear Engineering at the University of California, Berkeley.