RWMAC's Advice to Ministers on the Radioactive Waste Implications of Reprocessing
DRY STORAGE AND DISPOSAL OF MAGNOX SPENT FUEL
This annex considers the issues relating to the possibility of dry storage and disposal of spent Magnox reactor fuel. It first sets out a statement that BNFL provided to RWMAC on this matter and then adds the Committee’s own comments. The discussion surrounding this annex is contained in section 4.2 of the main report.
The statement provided to RWMAC by BNFL reads as follows.
This paper has been provided in response to RWMAC’s request to provide a statement of how unreprocessed Magnox fuel may be dealt with, noting potential difficulties and possible worker dose implications (a key issue for the Sellafield Working group was to decide whether Magnox fuel reprocessing was unavoidable).
BNFL, UKAEA, CEGB, Nuclear Electric and Magnox Electric have all studied the Magnox dry storage issue over a period of decades. The earlier references to dry storage occur with UKAEA / BNFL investigating options for delaying reprocessing of Magnox fuel by preventing corrosion using a dry atmosphere. The CEGB also investigated this problem for similar reasons in the 1970’s. A major driver for BNFL had been that problems with the decanning equipment at Sellafield in the mid – 1970’s had caused large backlogs of material to be reprocessed. The consequent corrosion of fuel had caused problems with contamination and clouding of water in storage ponds, both at the reactors and at Sellafield.
The later CEGB work begins to explore the safety and operations aspects of building a dry store for wetted Magnox fuel, including testing and selecting optimal fuel for storage.
Nuclear Electric reviewed dry storage in 1994 primarily in support of the AGR programme, but they also referred to Magnox dry storage and the evidence to the 1986 Rossi committee. They noted that Magnox dry storage was uneconomic, created a liability to build and licence dry stores, had significant safety issues and with the approaching end of the Magnox programme was less than viable.
Magnox Electric were more positive in their review of the options for long term storage and direct disposal of Magnox fuel. This was largely driven by commercial considerations – the review was prior to integration of Magnox Electric with BNFL. Their review also provides a good overview of overseas work to date on dry storage, albeit somewhat selective in the data included. It concludes however by recognising the difficulty of proceeding with dry storage in the absence of a reprocessing alternative.
Magnox fuel comprises a uranium metal bar clad in a magnesium (Magnox) alloy sheath. The cladding is normally ‘finned’ to assist gas flow and heat transfer. Both the Magnox cladding and the uranium fuel itself are chemically reactive. Both are susceptible to corrosion under wet storage conditions. This corrosion can be minimised by carefully controlled pond water chemistry or an appropriate dry storage regime.
Uranium corrosion in damp air produces a mixture of uranium dioxide, hydrogen and (UH3) uranium hydride. Uranium hydride is generally unstable and can decompose to form pyrophoric uranium. This decomposition can be initiated either mechanically or by a change in the hydrogen partial pressure.
There has been some experience with corroded uranium and the effect of inert gas atmospheres. Humidity and an inert gas atmosphere could influence the corrosion of the uranium present leading to very difficult mid and long-term handling and safety problems.
The rates of corrosion of the Magnox cladding is much reduced in a controlled inert atmosphere, however eventually corrosion could expose large areas of uranium metal to the store atmosphere. Over long time periods, the chemically reactive nature of the Magnox cladding and uranium fuel would prove significant problems even for fuel which had been discharged direct to a dry store, without a wetted transfer mechanism.
A long-term dry store would therefore have to protect against hazards from the problems of hydrogen evolution, pyrophoric uranium, secondary wastes from corroded cladding material and handling difficulties from swollen and corroded fuel rods. Such a facility would make great use of remote handling and maintenance; worker doses may be estimated as comparable to current reprocessing facilities. (Magnox reprocessing 1.9 millisieverts per year, THORP is 0.6 millisieverts per year)
The permanent disposal of Magnox fuel is a major technical problem. Current NIREX requirements are that waste to be placed in a repository is in a chemically stable form. Clearly, reactive Magnox fuel poses a significant risk of migration of soluble material into the far field environment if consigned to a repository in its untreated form.
The fuel would ultimately need to be prepared for disposal if it is not reprocessed. If this is the case, it may be necessary to process the fuel to turn it into a suitably stable waste form for disposal. Currently, no process has been demonstrated to achieve this, however it would seem that any process would need to be similar in scope to reprocessing.
In the U.S. there has been some experience with the use of ‘molten salt’ conditioning to treat reactive uranium fuels from early reactor designs. The molten salt process is similar in scope to reprocessing and has never been pursued in an industrial scale trial. BNFL has considered that the process was less inherently safe than reprocessing with no advantages in terms of discharges, secondary wastes or economics.
A single interim dry store exists at Wylfa power station. The store exists to provide short-term dry storage for Magnox fuel from that station. The Wylfa experience has demonstrated however the vulnerability of a dry store to water ingress and subsequent fuel corrosion. Long term dry storage with no possibility of an alternative disposal route for this material (i.e. reprocessing), is a very different proposition and it is doubtful that the NII would allow such a store.
Considering the reactivity of the fuel, it would be difficult to prepare a suitable post closure safety case for a repository containing significant quantities of unreprocessed Magnox fuel. It is doubtful that such material could be guaranteed not to enter the human environment over the significant time periods such a facility would have to operate.
Significant facilities would be required for drying, testing and long term storage of Magnox fuel. A public enquiry would be necessary to construct such facilities. Given the safety issues the regulator may have difficulty licensing a novel facility, particularly in the absence of alternative reprocessing facilities. Safety case preparation, construction and commissioning would also take a significant period of time. It is envisaged that the ensuing time delay would make the facilities economically non-viable, since a significant delay in operating the facilities would reduce the amount of fuel to be treated by this route and damage the economic case for its operation.
To summarise, it is judged that prompt reprocessing is a more appropriate option than prolonged storage of Magnox fuel. Long term storage would ultimately have to be followed by conditioning for disposal. Conditioning would depend on the development of a new process that may ultimately be similar in scope to reprocessing in terms of wastes, discharges and costs.
From a practical viewpoint, reprocessing has the advantage of being proven technology with no unknown downside risks. Facilities already exist to treat the fuel under a robust safety case. No significant additional capital spend is required.
The BNFL statement argues that reprocessing of Magnox spent fuel is unavoidable for a number of reasons: short term technological difficulties; regulatory and planning hurdles; economic reasons; and long-term conditioning problems. Although RWMAC requests for information on the feasibility of dry storage of spent fuel asked for quotable references, BNFL responded that much of the material that they draw on was not available in published form. Given the sensitivity of Magnox reprocessing, RWMAC considers this lack of published information to be unfortunate. Nevertheless, the Committee did use the statement provided by BNFL and other published information to arrive at the following view:
|Page published 14 November 2000; last modified 3 November, 2002|