Estimating the disposal costs of spent fuel

In its 2008 white paper Managing Radioactive Waste Safely, the UK government set out a framework for managing higher activity radioactive waste in the long-term through geological disposal, coupled with safe and secure interim storage and ongoing research and development to support its optimised implementation. It also invited communities to have discussions, without commitment, with government on the possibility of hosting a geological repository. So far, two areas in Cumbria (Allerdale and Copeland) have expressed interest in hosting the facility.

Current cost models of this planned £12 billion UK deep geological disposal facility (GDF) for radioactive waste (see box, p46) exclude spent fuel from new nuclear reactors. The government is planning to charge electrical utilities for the disposal of this new spent fuel. Although it had originally planned to charge a fixed price, this was changed in 2011 to a variable, but capped, Waste Transfer Price (WTP).

The Waste Transfer Price will increase over time, as the final outturn costs of actually siting, building and operating the deep GDF are better understood. The coalition government has also proposed that the Waste Transfer Price should be deferred for a period up to 30 years after the start of new nuclear reactor power generation (assumed to be around 2020). By 2050 the government will know the true outturn capital cost of siting and constructing the repository and will also have had 10 years’ practical operating experience running the repository, which is planned to be fully operational by 2040.

Ian Jackson, author of Nuclear Engineering International’s 2008 special publication Nukenomics, developed a software model of the economical issues affecting the spent fuel disposal costs for new nuclear reactors (called ‘FUPSIM’ after the earlier fixed price unit plan), in a research commission for Greenpeace UK.

In a March 2011 Greenpeace UK report, ‘Subsidy Assessment of Waste Transfer Pricing for Disposal of Spent Fuel from New Nuclear Power Stations’, Jackson argues that the cap, £978,000/tU, may be too low to cover the government’s costs. Jackson says the government assumes nuclear disposal costs will rise at only 3.3% per annum above inflation. But past experience shows that nuclear costs typically escalate at between 4.2–4.5% above inflation. Over the past decade the costs of similar major nuclear projects such as NDA liabilities, Yucca Mountain, Olkiluoto-3 and most recently Flamanville-3 have all risen at around this rate. At that rate, spent fuel disposal costs will breach the government price cap much sooner than expected (around 2047), he says, and so are likely to have been underestimated. In addition, he argues that the government may have underestimated the costs of disposition. Both issues would require a government subsidy to cover costs.

His software FUPSIM simulates waste disposal liabilities for any size of new nuclear power station via 21 adjustable input parameters (such as power station output, generating lifetime, load factor, spent fuel burn-up, storage period, discount rates) and displays 31 calculation results. It calculates the approximate repository cost and disposal price for spent fuel disposal for a new nuclear power station project, the funding needed by energy companies to meet these spent fuel liabilities, and any potential funding shortfall that may need to be subsidised by the taxpayer. It also calculates liability costs of spent fuel disposal for energy companies.

In the report, Jackson writes, “The total cost of a GDF repository is very sensitive to the total quantity of spent fuel it contains. For example the NDA’s historic legacy of spent fuel represents only 2% of the total volume of waste in a repository but is responsible for around 50% of the gross repository cost. This makes it difficult to reliably model unit disposal costs because even small changes in the spent fuel inventory can significantly increase the total lifecycle cost of the repository.”

He also points out that unit disposal costs (£k/tU) for new build PWR spent fuel will be only half the cost of AGR spent fuel disposal from Britain’s existing reactors, because it assumes that AGR fuel will not be able to be packed as tightly as modern PWR fuel. AGR fuel assemblies are bulkier than PWR fuel assemblies, but that should not stop size-reduction work before canister placement, Jackson argues. The disposal price difference “may perhaps give the appearance of favouring new nuclear build,” says Jackson.

The software program uses a conservative approach to calculate unit disposal costs for new nuclear reactors by scaling upwards from known NDA costs based on the total mass of legacy AGR and PWR uranium fuels needing disposal.

FUPSIM calculates the marginal cost (£m/tU) of increasing the repository size, the basic minimum unit cost of the extra spent fuel space. FUPSIM then calculates the full share cost (£m/tU) of increasing the repository size, which combines the unit cost of disposal of both new and legacy spent fuel.

In the report, Jackson says that as the NDA repository expands with new nuclear build fuel added, spent fuel unit disposal costs may be approximated through iterations of the six-tenths rule cost estimating formulae: CB = CA x (SB/SA)SF.” In this formula, CB is cost B, CA is cost A, SB is size B, SA is size A, and SF is scaling factor 0.6. Using this rule, he concludes that the shared spent fuel unit disposal cost for a new 10-reactor PWR fleet is £473,000/tU, which is about £280,000 higher/tU than UK government predictions. The fewer reactors are built, the higher costs per-tonne rise, up to £628,000/tU for spent fuel from just one new PWR added to the current repository.

A cost underestimation would mean that the NDA will not fully recover all of its disposal costs for new build reactor fuel, and so the NDA will require an indirect government subsidy to make up the shortfall.

The report makes two recommendations: remove the price cap, and link the price to actual repository costs, and set the same costs for PWR, AGR and MOX fuel disposal.

FUPSIM used UK government generic reactor modelling assumptions based on a 1.35 GWe PWR reactor operating for 40 years lifetime, generation start-up 2020, end of generation 2060, average load factor 90%, with a lifetime generating output of 424,000 GWh over 40 years.

This article was first published in the October 2011 issue of Nuclear Engineering International (p45-46)