Corrosion of Spent Nuclear Fuel: The Long-Term Assessment

Ewing, Rodney C.

University of Michigan, Ann Arbor, Michigan (United States)

关键词: Uraninites;    Radioactive Wastes;    Oxides;    Yucca Mountain;    54 Environmental Sciences;   

DOI  :  10.2172/833759
RP-ID  :  EMSP-73751
RP-ID  :  FG07-97ER14816
RP-ID  :  833759

美国|英语

来源: UNT Digital Library

【 摘 要 】

The successful disposal of spent nuclear fuel (SNF) is one of the most serious challenges to the successful completion of the nuclear fuel cycle and the future of nuclear power generation. In the United States, 21 percent of the electricity is generated by 107 commercial nuclear power plants (NPP), each of which generates 20 metric tons of spent nuclear fuel annually. In 1996, the total accumulation of spent nuclear fuel was 33,700 metric tons of heavy metal (MTHM) stored at 70 sites around the country. The end-of-life projection for current nuclear power plants (NPP) is approximately 86,000 MTHM. In the proposed nuclear waste repository at Yucca Mountain over 95% of the radioactivity originates from spent nuclear fuel. World-wide in 1998, approximately 130,000 MTHM of SNF have accumulated, most of it located at 236 NPP in 36 countries. Annual production of SNF is approximately 10,000 MTHM, containing about 100 tons of ''reactor grade'' plutonium. Any reasonable increase in the proportion of energy production by NPP, i.e., as a substitute for hydrocarbon-based sources of energy, will significantly increase spent nuclear fuel production. Spent nuclear fuel is essentially UO{sub 2} with approximately 4-5 atomic percent actinides and fission product elements. A number of these elements have long half-lives hence, the long-term behavior of the UO{sub 2} is an essential concern in the evaluation of the safety and risk of a repository for spent nuclear fuel. One of the unique and scientifically most difficult aspects of the successful disposal of spent nuclear fuel is the extrapolation of short-term laboratory data (hours to years) to the long time periods (10{sup 3} to 10{sup 5} years) as required by the performance objectives set in regulations, i.e. 10 CFR 60. The direct verification of these extrapolations or interpolations is not possible, but methods must be developed to demonstrate compliance with government regulations and to satisfy the public that there is a reasonable basis for accepting the long-term extrapolations of spent fuel behavior. In recent years ''natural analogues'' for both the repository environment (e.g., the Oklo natural reactors) and nuclear waste form behavior (e.g., corrosion and alteration of uraninite, UO{sub 2+x}) have been cited as a fundamental means of achieving confirmation of long-term extrapolations. In particular, considerable effort has already been made to establish that uraninite, UO{sub 2+x}, with its impurities, is a good structural and chemical analogue for the analysis of the long-term behavior of the UO{sub 2} in spent nuclear fuel. This proposal is based on the study of uraninite and the naturally occurring alteration products of UO{sub 2+x} under oxidizing and reducing conditions. The UO{sub 2} in spent nuclear fuel is not stable under oxidizing conditions. In oxic solutions, uranium has a strong tendency to exist as U{sup 6+} in the uranyl molecule, UO{sub 2}{sup 2+}. Uranyl ions react with a wide variety of inorganic and organic anions to form complexes. Throughout most of the natural range of pH, U{sup 6+} forms strong complexes with oxygen-bearing ions like CO{sub 3}{sup 2-}, HCO{sup 3-}, SO{sub 4}{sup 2-}, PO{sub 4}{sup 3-}, and AsO{sub 4}{sup 3-}, which are present in most oxidized stream and subsurface waters. In arid environments, the U{sup 6+} ion can precipitate as a wide variety of uranyl oxide hydrates, uranyl silicates and uranyl phosphates. This is well demonstrated in experimental work, e.g., in long term drip tests on UO{sub 2} and is confirmed by natural occurrences of UO{sub 2} in which a wide variety of uranyl phases form as alteration products. The most striking feature of these studies is the very close parallel in the paragenetic sequences (i.e. phase formation sequence) between the very long term (10 year tests) and the young (therefore, low-Pb uraninites) of the Nopal I deposit in Mexico.