PUBLIC HEALTH CONSEQUENCES OF
SUBSTITUTING MIXED-OXIDE FOR
URANIUM FUEL IN LIGHT-WATER REACTORS
Edwin S. Lyman, PhD
Nuclear Control Institute
January 21, 1999
In January 1997, the U.S. Department of Energy (DOE) decided to pursue a "dual track" policy for the disposition of approximately 50 metric tons (MT) of plutonium produced for weapons programs that have been declared excess to military needs. The two tracks of the "dual track" refer to two different approaches for converting separated plutonium into a dilute and highly radioactive form that is more difficult to return to weapons.
Under one approach, known as "can-in-canister" immobilization (CIC), plutonium will be incorporated into chemically stable ceramic discs. These discs will in turn be embedded in canisters of "vitrified" (glassified) high-level radioactive waste (VHLW) at the Defense Waste Processing Facility (DWPF) at the Savannah River Site in South Carolina. DOE is tentatively planning to use CIC for approximately 17 MT of excess plutonium in impure forms. The CIC facility will in all likelihood be sited at SRS adjacent to the DWPF.
Under the other approach, plutonium will be used to produce "mixed plutonium-uranium oxide" (MOX) fuel assemblies, which will be loaded and irradiated in a number of U.S. commercial nuclear reactors, displacing some or all of the low-enriched uranium oxide (LEU) fuel assemblies the reactors currently use. DOE is tentatively planning to utilize this option for approximately 33 MT of weapons-grade plutonium (WG-Pu). In 1998, DOE issued a Request for Proposals, seeking vendors interested in providing MOX fuel fabrication and irradiation services. Of the three proposals submitted, two have already been eliminated for failing to meet basic requirements. It is expected that DOE will sign a contract in February 1999 with the third party, a consortium including the French fuel fabricator Cogema and the U.S. utilities Duke Power and Virginia Power. It is also expected that Cogema will build and operate a MOX fuel fabrication plant at SRS, and that the fuel will be irradiated in Virginia Power's North Anna 1&2 plants and Duke Power's McGuire 1&2 plants in North Carolina and Catawba 1&2 plants in South Carolina.
Both the immobilization and MOX tracks require large-scale handling and processing of plutonium, an extremely hazardous substance. Consequently, they will be expensive to carry out and will pose risks to human health and the environment. However, the costs and risks involved will be small compared to those experienced when the material was produced, and most arms-control advocates concur that the security benefits of plutonium disposition justify the risks.
Some analysts argue further that differences in cost and hazard associated with the two disposition approaches should not weigh heavily in policy decisions. However, this view is out of touch with both budgetary and political realities. Because Cold War-sized government coffers are not likely to be available to DOE for disarmament activities, the plutonium disposition program will be under pressure to minimize costs. Also, many environmental groups and citizens' groups near affected sites will likely oppose any disposition activities unless they clearly have low environmental and public health impacts. It is certainly sensible to reject an option with substantially greater economic and health risks, if it brings no additional benefits.
Cost and public health impact were major considerations in the process that DOE used to select MOX and immobilization from the large number of disposition options that were initially proposed. In deciding on the dual track policy, DOE argued that there are no decisive differences between the MOX and immobilization options with regard to any of its evaluation criteria, including public health impact. However, this report concludes that DOE's evaluation is inaccurate. We find that the public health risks associated with the MOX approach are significantly greater than those associated with CIC. This is due primarily to our findings that the consequences of severe accidents involving LWRs with MOX cores are likely to be greater than those involving LEU cores.
Our finding also has international implications. For instance, the U.S. and Russia are also pursuing a plan to utilize Russian excess WG-Pu in VVER-1000 light-water reactors located in Russia and Ukraine, which meet less stringent safety standards than nuclear plants in the U.S. Also, several nations, such as France, Switzerland and Japan, either use or are planning to use plutonium obtained in so-called "civil" reprocessing programs as fuel for LWRs. The "reactor-grade" plutonium (RG-Pu) used in these programs has different isotopic characteristics than WG-Pu and a different impact on reactor safety, including a greater increase in potential consequences.
In this report, the public health consequences of severe accidents at MOX-fueled pressurized water reactors (PWRs) are calculated and compared with the consequences of accidents at LEU-fueled PWRs. The acceptability of the increased risk associated with the change from LEU to MOX fuel in U.S. PWRs is then evaluated in the context of the "risk-informed" regulatory procedures now being implemented by the U.S. Nuclear Regulatory Commission (NRC).
Risks of MOX Use
The MOX approach consists of several stages, each of which can have a significant environmental and public health impact. A plant for fabrication of the fuel must be built and operated, the fuel must be shipped to reactor sites, and the fuel must be irradiated in reactors. By comparison, the environmental impacts of CIC immobilization are associated primarily with the operation of the ceramic immobilization plant. Because this plant will be very similar to the MOX fabrication plant in design and size, it will have similar impacts. Therefore, any risks associated with MOX transportation and irradiation increase the cumulative risk of the MOX approach to a level greater than that of immobilization.
In order to quantify and compare the public health impacts of the two options, it is necessary to understand how the risks of nuclear power plant operation change when WG-MOX is substituted for LEU. Risk is defined as the product of the probability and the consequences of a particular event, summed over all events. Nuclear power plants pose risks both as a result of routine operation (high probability and relatively low consequence events) and through the possibility of accidents (low probability and high consequence events). This report focuses on accident risk.
Carrying out a complete and accurate comparison of the accident risks of MOX and LEU cores is a difficult undertaking, for a number of reasons. Nuclear power plant accident safety analyses, or probabilistic risk assessments (PRAs), are extremely complex. In general, the substitution of WG-MOX for LEU fuel will affect both the probability of occurrence and the consequence of each accident sequence which can occur during reactor operation, so that existing PRAs will have to be extensively modified. The difficulty of doing so is compounded by the relative lack of experience with the use of WG-MOX fuel, as well as insufficient data on many technical aspects of MOX use.
Another complication results from the fact that almost every nuclear plant in the U.S. has unique features which are relevant to safety, so that the impacts of MOX use are highly reactor-specific. Also, the safety analysis will depend on details of the specific MOX irradiation plan, such as the amount of plutonium in the MOX fuel, the maximum burnup (amount of heat extracted) from each fuel assembly and the fraction of the core (from 33%-100%) that will be loaded with MOX fuel. These details have not been publicly released yet and may for the most part remain proprietary and unavailable to the public.
However, there are some safety-related problems with the use of MOX fuel which will apply to any LWR. For example, the total inventory of highly radiotoxic actinides, including plutonium-239 (Pu-239), americium-241 (Am-241) and curium-242 (Cm-242), is significantly greater in MOX cores than in LEU cores throughout the operating cycle. Our analysis shows that the public health consequences of some severe accidents will be greater for reactors fueled with MOX.
The exact quantities of plutonium and other actinides in MOX cores depend on parameters such as the concentration and isotopic content of the plutonium in the fresh fuel. For the case considered in this study we find that, compared to an LEU core, a full WG-MOX core will contain about three times the amount of Pu-239, seven times as much Am-241 and seven times as much Cm-242 at the end of an operating cycle (i.e. just before the reactor is shut down for reloading). For MOX fabricated with reactor-grade plutonium (RG-Pu), Am-241 and Cm-242 inventories are greater by additional factors of 4 and 3, respectively.
Since most of these radionuclides emit alpha particles, which are much more hazardous per decay than beta or gamma particles if inhaled or ingested, they will contribute significantly to public radiation exposures following severe reactor accidents, even if only a small fraction of the core inventory is released.
The initial draft of DOE's Storage and Disposition of Weapons-Usable Fissile Materials Draft Environmental Impact Statement (DPEIS) did not analyze the environmental impacts of accidents involving MOX-fueled LWRs. Instead, it only included an analysis of an LEU-fueled LWR. DOE justified this by claiming that
"separate studies ... indicate that the use of MOX fuel in a ... LWR does not increase the risk and consequences of accidents. This results from the fact that the other radioisotopes that are released in an accident have more serious impacts on human health than the Pu used in the MOX fuel." 
Another DOE study makes the stronger claim that the greater actinide inventories in a MOX core will not affect the consequences of an LWR accident because "plutonium and other insoluble fuel isotopes are not included in the releases to the environment." 
These statements are misleading, however. Certain severe accidents can result in the expulsion of significant quantities of actinides into the environment. Although such "beyond design-basis" accidents are expected to occur very infrequently, there are both historical precedents and regulatory requirements for considering them in safety analyses.
The best possible laboratory for loss-of-containment consequences, the Chernobyl accident, has demonstrated that significant and wide-ranging dispersal of actinides is possible. Recent reviews of the Chernobyl source term have concluded that approximately 3.5% of the core actinide inventory was released. Moreover, dispersal of these relatively heavy aerosols was not limited to the immediate vicinity of the plant; fuel fragments were discovered as far away as Greece and Germany, over one thousand kilometers away. 
It has often been claimed that a Chernobyl-type accident cannot happen in the West because Western reactors have robust containment structures. However, while the presence of containment domes reduces the risk of such accidents, it does not eliminate it entirely. Many accident sequences for U.S. LWRs have been identified which can lead either to massive failure or bypass of the containment, thereby allowing significant releases of core particles. In fact, the U.S. Nuclear Regulatory Commission (NRC) has estimated that actinide releases as high as several percent of the core inventory are possible in such accidents. 
In comments on the DPEIS in 1996, the Nuclear Control Institute (NCI) challenged DOE's assumption that there was no difference between LEU and MOX with regard to reactor safety.  In particular, NCI cited the possibility of accidents resulting in a relatively large release of actinides.
DOE responded to NCI's comments in the final PEIS on storage and disposition of weapons-usable fissile materials (FPEIS) by presenting the results of a calculation that took into account the different radionuclide inventories of WG-MOX and LEU cores. The FPEIS claimed that the change in accident consequences (defined as the resulting number of latent cancer fatalities) associated with the substitution of WG-MOX for LEU ranged from +8% to -7%: in other words, the number of cancer fatalities caused by some accidents could actually decrease as a result of switching to MOX fuel. 
A complete review of the FPEIS calculation is not possible because few details are provided. However, an analysis of the information that is provided reveals several obvious inconsistencies. For instance, the FPEIS calculation used a value of 0.65 for the ratio of the quantity of cesium-134 (Cs-134) in the WG-MOX core to that in the LEU core. When this ratio was "arbitrarily set to 1.0" in the FPEIS analysis, however, the observed reduction in cancer fatalities associated with switching to MOX fuel changed to an increase. The FPEIS fails to mention a fact that appears in one of its own reference documents --- namely, that various studies have calculated Cs-134 MOX/LEU ratios ranging up to 1.08, and that the value used by the FPEIS was based on a Westinghouse "advanced" PWR and not on an existing reactor type.  Our study, which was based on existing PWRs, found a value of 0.96 for the Cs-134 MOX/LEU ratio at the end-of-cycle.
Another factor that the FPEIS did not take into account is the sensitivity of the consequences of MOX accidents to the fraction of the actinides in the core that is assumed to be released. There are large uncertainties in predictions of the fraction of core actinides that can be released in severe accidents. The FPEIS assumed only very low values of the actinide release fractions.
Because of the flaws in the FPEIS risk calculation, NCI undertook its own study to evaluate the consequences of loss-of-containment accidents at PWRs with MOX cores and compare them to those at PWRs with LEU cores. The specific example of a four-loop PWR with an ice-condenser containment was chosen for analysis. Four of the six plants included in the sole bid now being evaluated by DOE --- Duke Power's Catawba 1&2 and McGuire 1&2 --- are of this type.
First, radionuclide inventories were computed for LEU and WG-MOX cores, using the Oak Ridge National Laboratory (ORNL) SCALE 4.3 code to simulate changes in the fuel composition during irradiation. Full WG-MOX cores were considered as the bounding case. Fuel management schemes were based on those in a 1996 Westinghouse report on plutonium disposition in which full-MOX cycles were developed that resembled LEU cycles as closely as possible.
Second, the accident consequences (acute fatalities, early commitment of latent cancer fatalities, and other indicators of risk) for LEU and MOX cores were evaluated for several different accidents, using NRC methodology and the NRC consequence calculation software MACCS2,  and ICRP 72 dose coefficients. Generic parameters were used for population and atmospheric data. While the absolute values of consequence measures depend strongly on these parameters, the relative consequences of MOX and LEU accidents are much less sensitive to them.
Finally, the calculated increases in risk associated with substituting MOX for LEU were compared to the acceptance guidelines contained in the recently issued NRC Regulatory Guide (RG) 1.174,  "An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis." RG 1.174 describes a methodology for grading the intensity of NRC review of requested changes to the licensing basis (LB) of nuclear power plants according to their risk significance. RG 1.174 therefore provides a framework for evaluating the regulatory significance of the increased risk associated with use of MOX fuel.
Although not directly applicable to the WG-Pu disposition program, the inventory of a typical reactor-grade (RG) MOX core was also calculated for comparison. Because RG-MOX cores have larger quantities of heavy actinides, the consequences of RG-MOX accidents are even more severe than those of WG-MOX, especially at the relatively high plutonium loadings necessary to achieve adequate utilization of the fuel.
1. The number of latent cancer fatalities (LCFs) committed within one week after a severe reactor accident will be significantly greater for both full and partial WG-MOX cores than for LEU cores. For most accidents considered, the number of prompt fatalities that result will also be greater.
(a) Compared to a PWR using LEU fuel, the number of latent cancer fatalities (LCFs) that will result from exposures immediately following a severe accident with early containment failure or bypass will be significantly greater for a PWR loaded with weapons-grade (WG) MOX fuel, for both full and 1/3-cores of WG-MOX. This is primarily due to the increased concentrations of plutonium and heavier actinides in MOX cores.
For a set of typical severe accidents that result in the release of about 1% of the inventory of plutonium and other actinides (compared with a 3.5% release for the Chernobyl accident), the number of "early" LCFs that result (those due to exposures occurring within one week after the accident), averaged over an operating cycle, was found to be 81%-96% greater for a full MOX core.
For a 1/3-MOX core, the corresponding percent increase would be 27%-32%, a factor of three smaller than for a full core. However, because the increase in consequences is essentially linear with respect to the MOX core fraction, the overall excess risk (the product of probability and consequences) associated with the disposition program will be approximately the same for both full and partial MOX core loadings. Whether 33 MT of WG-Pu is processed in six plants with a 1/3-MOX core in 15 years, in six plants with a full MOX core in 5 years, or in two plants with a full MOX core in 15 years, the total average increase in risk to the U.S. public will be approximately the same in each case (although the risk to a particular individual may be different).
(b) These increases are considerably greater than the upper limit of the +8% to -7% range cited by the Department of Energy (DOE) in its environmental impact statements on surplus plutonium disposition for full-core MOX.
(c) The actual number of additional LCFs resulting from a MOX accident depends on details of the reactor site, such as population density and atmospheric conditions. For a generic site with a population density of 100 persons/square kilometer (which is very close to the actual density in the vicinity of the Catawba and McGuire plants) the number of additional LCFs within an area of 1000 miles radius, averaged over an operating cycle, was found to range from 1,430 to 6,165 for the set of accidents analyzed for a full-MOX core. For a 1/3-MOX core, the additional LCFs range from 475 to 2055.
(d) The number of prompt fatalities resulting from acute radiation exposure is greater by around 40% for WG-MOX cores following early containment failure accidents. For containment bypass accidents, a 19% reduction was observed (from 37 to 30 cases) for a full-MOX core, and a 6% reduction for a 1/3-core. However, this reduction in prompt fatalities is tiny compared to the increase in LCFs observed.
2. The additional consequences of severe accidents involving MOX cores are sensitive to the fraction of actinides (i.e. plutonium, americium and curium) in the core that are released.
The increase in accident consequences associated with switching from LEU to MOX depends on the fraction of the actinide inventory that is released, which is a highly uncertain parameter. As the actinide release fraction is varied from 0.3% to 6%, the percent increase in LCFs resulting from an full-MOX core accident with early containment failure, averaged over an operating cycle, ranges from 38% to 131%, corresponding to an additional 1,725 to 18,180 LCFs for the generic reactor site. In the worst case, the number of additional cancers associated with a MOX accident is 60% as large as the total number of cases predicted to occur worldwide from the Chernobyl accident. For a 1/3-MOX core, the percent increases range from 13% to 44%, corresponding to an additional 575 to 6,060 LCFs.
3. The average latent cancer fatality accident risk to the population within ten miles of a nuclear plant is increased by approximately a factor of two when a full core of WG-MOX is substituted for LEU. This increase in risk is significant when compared to the risk limits in NRC's Safety Goal Policy Statement.  According to NRC's Regulatory Guide 1.174, a change in the licensing basis resulting in a doubling of risk would not be allowed for typical U.S. PWRs. The increase in risk associated with loading a 1/3-core of WG-MOX would also be unacceptably high.
When a full core of WG-MOX is substituted for LEU, the average increase in latent cancer fatality risk to the population near a reactor site nearly doubles. This is equivalent to the increase in risk that would occur if the probability of a severe accident with a large early release of radioactivity (the Large Early Release Frequency, or LERF) were doubled. For the PWR considered in this study, this would correspond to an increase in LERF of about seven in a million (7x10-6) per year for a full MOX core, or more than two in a million (2x10-6) per year for a 1/3-MOX core. In both cases, these exceed the threshold of one in a million (1x10-6) per year established in NRC's Regulatory Guide 1.174 for allowable increases in LERF.
4. The use of WG-MOX in U.S. PWRs is not likely to lower the probability that a severe loss-of-containment accident may occur and may in fact increase it significantly.
Some reasons why this is the case are listed below.
(a) The ability of high-burnup MOX fuels in current use to withstand severe accident conditions is inferior to that of LEU fuel.
It has been observed that MOX fuel assemblies fabricated with current techniques are inferior to LEU fuel with regard to their integrity during abnormal events that cause rapid heating of the fuel, such as reactivity insertion accidents (RIAs) and loss-of-coolant accidents (LOCAs). Based on the results of a series of RIA tests at the Cabri test reactor in France, French regulators have concluded that "MOX fuel shows a higher failure potential than UO2 at comparable burnup." In particular, a MOX fuel rod with a burnup of 55 gigawatt-days per metric ton (GWD/MT), which is typical of burnups achieved in U.S. PWRs today, experienced a violent rupture and dispersal of fuel particles, while two LEU rods of comparable and higher burnups were able to withstand similar conditions without rupture.  Based on this test, a French regulator recently concluded that this was a MOX-related phenomenon and that there is a "very high potential for rupture" of MOX fuel in RIA situations. 
(b) A MOX-fueled PWR may have a greater risk of experiencing pressurized thermal shock of the pressure vessel.
Due to a more rapid cooldown of the reactor vessel following a break in a main steam line, a MOX-fueled PWR may have a greater risk of experiencing pressurized thermal shock (PTS) than one fueled with LEU. PTS is a very severe event in which the reactor vessel becomes brittle at low temperature (below about 180 C or 350 F) and ruptures at high pressure, causing core debris to be expelled into the containment. The resulting phenomenon, known as high-pressure melt ejection (HPME), can result in a very rapid heating and pressurization of the containment atmosphere (direct containment heating, or DCH) which can cause containment failure.
(c) Ice-condenser containments may be more vulnerable to early failure in a severe accident than large dry containments.
Four of the six PWRs that have been offered for MOX irradiation services in the sole remaining proposal being evaluated by DOE, Duke Power's Catawba 1&2 and McGuire 1&2, have ice-condenser containment structures, which "do not have the same inherent capacity to withstand the credible DCH loads from all scenarios as other Westinghouse plants," according to NRC. 
Together, these facts raise the concern that if U.S. utilities plan to irradiate MOX fuel to a burnup comparable to that of LEU fuel, the risk of violent rupture and fuel dispersal that makes cooling of the core debris more difficult will be increased. Moreover, because such accidents can result in both dispersal of the core into the containment and early containment failure through the phenomenon of direct containment heating, they are also associated with release of solid core materials, such as actinides, into the environment. Therefore, both the consequences and the probability of this class of accidents may increase when MOX is substituted for LEU in PWRs.
5. A severe accident at a PWR with a reactor-grade MOX (RG-MOX) core would cause up to twice as many latent cancer fatalities (LCFs) as would an accident at a PWR with a WG-MOX core.
The number of LCFs resulting from a severe accident at a PWR fueled with a full core of RG-MOX, at the end of an operating cycle, was found to be 123%-486% greater than that resulting from an accident at a PWR fueled with LEU, depending on the actinide release fraction. This is more than twice as many cases as would result from an accident involving a WG-MOX core. This dramatic increase in risk should be taken into consideration by nations that are currently using or planning to introduce RG-MOX in their nuclear plants. Recently, some U.S. policy-makers who regret the U.S. decision not to pursue commercial spent fuel reprocessing and plutonium recycling have been seeking to take advantage of the current political difficulties of siting a geologic repository for spent fuel to revive the reprocessing option in the U.S. The results of this article provide an additional validation of the U.S. decision and another argument why reprocessing and recycle should be avoided.
1. Licensing of U.S. reactors to use MOX will have to take place primarily on a site-specific level. In addition, an NRC finding that MOX use poses "no significant hazards" under 10 CFR 50.92 clearly would not be justified.
A key question in the procedure for licensing reactors to use MOX fuel will be whether NRC will rule, under the procedures outlined in 10 CFR 50.92, that the introduction of MOX fuel into existing reactors involves a "significant hazards consideration," which would obligate the NRC to conduct public hearings prior to issuance of a license amendment. Prospective industry participants in the MOX program have indicated that they intend to have the MOX reload core methodology licensed on a generic basis, thereby removing most MOX-related issues from consideration on a plant-specific level. In this way, they hope to facilitate an NRC finding of "no significant hazards" in individual plant license amendment proceedings and thus prevent the possibility of site-specific hearings that could lead to substantial delays in introducing MOX fuel into reactors.
However, the results of this study indicate that site-specific considerations, such as the public health impacts associated with changes in the licensing bases of existing plants to use MOX fuel, will indeed be substantial, and therefore it should not be possible for NRC to justify issuing a finding of "no significant hazards" on the plant-specific level.
2. Limitations on MOX fuel burnup to below 36 GWD/MT should be imposed unless high burnup safety issues are resolved.
Concerns with the performance of high-burnup MOX fuel in the event of an accident have led the French nuclear safety authority DSIN to restrict the burnup of MOX fuel to 36 GWD/t, whereas LEU fuel is permitted to reach 47 GWD/MT. The French national utility Electricit de France (EdF) has concluded that to achieve burnup parity with LEU, a new MOX fuel type will have to be developed. Such an effort could cause substantial additional delays to the MOX mission. The U.S. should follow France's lead and restrict MOX burnup pending resolution of these safety issues, even though this will be a costly inconvenience for U.S. nuclear plants.
3. Licensees who wish to use WG-MOX will have to demonstrate to NRC that the Large Early Release Frequencies (LERFs) of their plants are below one in a million (1x10-6) per year. Even if they can meet this requirement, the request will be subject to an intensive NRC technical and management review, and the underlying probabilistic risk assessment (PRA) calculation will have to undergo peer review and satisfy quality control requirements.
We have shown that the introduction of a full core of MOX fuel into PWRs will result on average in a doubling of the risk of a severe accident leading to a large early release of radioactivity. This increase in risk is equivalent to that which would occur if the Large Early Release Frequency (LERF) of the plant were doubled. According to NRC's RG 1.174, a change to the plant licensing basis resulting in a doubling of the LERF would only be considered for plants with a baseline LERF of one in a million (1x10-6) or below. For a 1/3-MOX core, the corresponding threshold would be three in a million (3x10-6).
The guidelines in RG 1.174 are not absolute. In particular, an applicant may argue that quantitative increases in risk are offset by "unquantified benefits" and that a less strict NRC response is warranted. Even so, plants wishing to use MOX will have to undergo intensive site-specific reviews by NRC, and may have to conduct full-scope (Level 3) probabilistic risk assessments (PRAs), which very few plants have done to date because of the time and expense involved. These will be necessary to document that the Large Early Release Frequencies of the plants are sufficiently low that the increased risk associated with a large early release from a MOX-fueled plant are "small and consistent with the intent of the Commission's Safety Goal Policy Statement." Moreover, PRA documentation will have to be done more carefully and in more detail in the future. Because of the great variability in the content and quality of PRAs that have been carried out to date, NRC is in the process of developing a quality control standard for PRAs submitted in support of risk-informed regulatory proceedings.
4. The U.S. plan to encourage Russia to use WG-MOX in Russian and Ukrainian VVER-1000 LWRs poses even greater risks than the plan for U.S. domestic use of WG-MOX.
Russian VVER-1000s do not meet Western safety standards in such critical areas as fire protection and instrumentation and control systems. Although the U.S. is encouraging Russia to commence a program for using WG-MOX in VVER-1000s, and has provided a portion of the initial financing, there will be no simultaneous effort to upgrade these plants so that they fully meet Western safety standards, which would cost on the order of $150 million per unit, according to recent estimates. In fact, Russia has to date been reluctant to accept Western assistance for plant safety upgrades. Given that the use of MOX will increase risk even in plants that do meet Western standards, encouraging Russia to use MOX in its less robust plants without ensuring maximum possible adherence to safety is nothing short of reckless.
5. Risks associated with irradiation of WG-MOX in both U.S. LWRs and Russian VVER-1000s could be averted if both nations implemented an all-immobilization policy for the entire stockpile of excess WG-Pu. The use of MOX is unnecessary and should be avoided.
The significant additional public health risks of MOX use in existing nuclear plants cannot be justified in terms of the security benefits of plutonium disposition, because a less risky alternative exists --- immobilization. The insistence of the Russian Ministry of Atomic Energy (MINATOM), along with U.S. and European nuclear interests, that immobilization is not an acceptable approach for either the U.S. or Russia, is one of the driving forces behind the heavy emphasis on MOX in both countries. However, the U.S. should not be compelled by a handful of bureaucrats and industry lobbyists to adopt an outdated, shortsighted and technically flawed approach that will unnecessarily endanger the health of its citizens. Rather than proceeding with the MOX plan, the U.S. should recognize and highlight the environmental, economic and security advantages of immobilization and explore creative ways of enhancing its acceptability both at home and in Russia.
1. U.S. Department of Energy, Storage and Disposition of Weapons-Usable Fissile Materials Draft Programmatic Environmental Impact Statement, DOE/EIS/0229-D, February 1996, Volume II, p. 4-690.
2. Oak Ridge National Laboratory, FMD LWR PEIS Data Report, Rev. 3, ORNL/MD/LTR-42, December 1995, p. B-22.
3. L. Devell et al. "The Chernobyl Reactor Accident Source Term: Development of a Consensus View," OECD/NEA, OECD/GD(96)12, November 1995.
4. U.S. Nuclear Regulatory Commission, "Severe Accident Risks: An Assessment for Five U.S. Nuclear Power Plants," NUREG-1150, 1990.
5. Edwin S. Lyman, "Comments on the Department of Energy's Storage and Disposition of Weapons-Usable Fissile Materials Draft Programmatic Environmental Impact Statement: Public and Occupational Health and Safety Issues," Nuclear Control Institute, Washington, DC, June 7, 1996 (rev. Oct. 9, 1996).
6. U.S. Department of Energy, Storage and Disposition of Weapons-Usable Fissile Materials Final Programmatic Environmental Impact Statement, DOE/EIS-0229, December 1996, p. S-37.
7. Oak Ridge National Laboratory (1995), op cit.
8. D.I. Chanin and M.L. Young, Code Manual for MACCS2: Volume 1, User's Guide, SAND97-0594, Sandia National Laboratories, March 1997. In the course of generating data for the present paper, the author discovered an error in the MACCSS2 software which resulted in the overcounting of cancer fatalities among individuals receiving committed effective doses (CEDs) greater than 10 sievert (Sv) and a consequent overestimation of population-averaged cancer risk. While this error will not be fixed until release of the next version of MACCSS, an "unofficial" corrected version of the code was provided to the author. Although the corrected version has not been officially validated, the results agree well with calculations carried out by the author by hand. All data in this report has been generated with the corrected, unofficial version of MACCSS2.
9. U.S. Nuclear Regulatory Commission, "An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis," Regulatory Guide 1.174, July 1998.
10. U.S. Nuclear Regulatory Commission, "Safety Goals for the Operations of Nuclear Power Plants: Policy Statement," Federal Register, 51 FR 20028, August 4, 1996.
11. DOE has recently claimed that the Cabri test is not relevant to the U.S. MOX program, arguing that (1) the burnup was higher than that which MOX rods will experience in U.S. reactors, and (2) the Cabri test rod was an obsolete fuel type with a high degree of heterogeneity. Both these statements are false. PWR fuel assemblies are authorized in the U.S. for burnups up to 62 GWD/t, and reactor operators expect that MOX and LEU fuel assemblies will be fully interchangeable. The Cabri rod was fabricated using the MIMAS process, which the French and Belgian industries have been using since 1984 and which is expected to be the process that a U.S. MOX fabrication plant will utilize. DOE is not encouraging the development of improved MOX fuel for the U.S. program because of the delays that would occur in its qualification.
12. A. MacLachlan, "International Meeting Fails to Resolve Questions Surrounding Cabri Future," NuclearFuel, July 27, 1998, p. 6.
13. U.S. NRC, "Status of the Integration Plan for Closure of Severe Accident Issues and the Status of Severe Accident Research," SECY-98-131, June 1998.
Plutonium Disposition Page What's New Home Page