The Costs of Nuclear Power
Introduction
One of the main concerns about building another generation of nuclear power stations is the cost and uncertainties surrounding those costs.
‘In 1990, the nuclear power plants were unattractive to private investors and so uneconomic that they had to be withdrawn from the privatisation process and required a consumer subsidy of about £1 bn per year simply to cover their running costs. Six years later, the efficiency of operation of the plants had improved sufficiently for all except the very oldest plants to be privatised and for 2-3 years, the privatised nuclear company, British energy, was highly successful. Only three years on in 2002, British Energy had collapsed and was rescued by the government at a cost to taxpayers of several billion pounds.’ (1)
In 2002 the Performance and Innovation Unit undertook a study (2) of the relative costs of electricity generated from different sources.
This produced the following results:-
Technology 2020 cost
Large CHP Under 2 p/kWh
Micro CHP 2.5-3.5 p/kWh
PV (Solar) 10-16 p/kWh
Onshore wind 1.5-2.5 p/kWh
Offshore wind 2.0-3.0 p/kWh
Energy crops 2.5-4.0 p/kWh
Wave 3.0-6.0 p/kWh
Fossil generation with Carbon Dioxide C&S 3.0-4.5 p/kWh
Nuclear 3.0-4.0 p/kWh
CCGT 2.0-2.3 p/kWh
Coal (IGCC) 3.0-3.5 p/kWh
Since then oil and gas prices have increased significantly. Moreover, the case currently being made for new nuclear capacity is that it produces minimal carbon dioxide emissions, so the relevant comparators for judging whether nuclear is economic are renewables and energy efficiency measures to reduce demand.
One point sometimes made against the use of renewables is that it is more unreliable and cannot necessarily be used at periods of peak demand. It is worth noting that the economics of nuclear power make it suitable only for “baseload” production, in other words it too lacks flexibility for peak-load production.
This chapter looks at the components of nuclear costs:- building power stations; running power stations; decommissioning power stations and storing wastage and at a range of assumptions underlying these calculations which would be reasonable. Claims of costs outside these would be over optimistic or pessimistic or indicate that policy had been changed to facilitate implicit subsidies.
Construction Costs
Approximately two thirds of the cost of nuclear power are fixed costs relating to construction and decommissioning. In order to allow comparisons between reactors with different output capacities, costs are measured in £/kw.
Estimated construction costs in £/kw vary form £500 to £3,000/kw. The PIU estimated that the cost was £840/kw. In the light of experience both here and abroad this looks like an under estimate. The actual cost of Sizewell B (the most recent nuclear station built in Britain) was estimated by the NAO in 1998 as £3,000/kw. Last year the Finnish government issued a construction licence for a Generation III PWR from Framatome for Olkilnoto. It will cost about £1,250/kw and is expected to be built in under 5 years, though it may have been a loss leader.
Factors affecting the costs are:-
(a) Speeding up the planning process and construction time from about 10 to 5 years and from 8 to 4 years respectively. This reduces the “deadtime” during which no income is earned but when the financing cost of capital must be met.
(b) In a competitive electricity market like the UK’s capital must be rased at up to 15%. This can double the initial outlay over 5 years, so halving the period of (a) has a significant impact. A regulated market or one where the government offers guarantees to underwrite nuclear electricity can bring the cost of capital down to 5-8%. However, Ministers have made it clear that they do not want to give nuclear generation privileged access to capital by offering guarantees in this way.
(c) Buying on “turnkey” term moves the risk from buyer to seller.
(d) Reducing the amount of on-site engineering cuts costs and building power stations as part of a series rather than to one-off designs reduces R + D costs. In the case of Sizewell these were estimated to be £750/kw, so stripping out this element brings the ‘tme’ cost down to £2,250/kw.
(e) Extending the life of the plant spreads the cost of construction over a greater production run. Top of the range estimates by the industry are for 60 years. 40 years would seen to be more realistic, but the effect of discounting (see (b) above) means the impact of extending plant life by 50% has little more than a 5% impact on electricity prices.
(f) Similarly increasing plant efficiency spreads capital costs. Top of the range life time average load factors are 90%. Actual experience in the UK varies between 35% and 85%. So a range of 80 – 85% would be realistic. This raises prices/kw hr by 12 % over the design capacity price (i.e the price if the station always operated at full tilt).
Taking all these points into account, a reasonable estimate of the costs of construction is £1,000 to £1,500/kw.
Operating Costs
Operating and maintenance costs are lower for nuclear than they are for coal, gas and oil fired stations. The cost of uranium is not currently the most significant part of this, though if many countries went down the nuclear route it would undoubtedly rise. However, the UK might then use reprocessed magnox fuel which is currently twice as expensive as uranium. Security of supply is an issue with uranium as with oil and gas. Current operating costs in UK nuclear stations run at 2p/kw hr, though this might be reduced to 1p/kw hr for new plant.
Insurance is not available on the capital markets to cover risks above £140m (the limit in the Treaty of Vienna 1963). These are met by HMG although again Ministers have said they would not continue to provide this implicit subsidy to new power stations.
Decommissioning
Decommissioning has three parts:- (1) Fuel removal – done immediately
(2) Demolishing lightly contaminated structures - planned to be done after 40 years.
(3) Removal of reactor core – planned to be done after 135 years.
The undiscounted costs of these phases is estimated by British Energy to be.
£m
(1) 167
(2) 333
(3) 500
Total 1000
However there must be considerable doubt about these figures since the estimate for cleaning up Magnox fuel from the 4 Magnox power stations is already £12.6 bn (part of the £56 billion total) suggesting a bill of £3 billion per plant and CORWM estimate that the volume of spent fuel which would be produced is four times the current level. If the actual costs do rise above forecasts, the tax payer will undoubtedly bear the burden as by then there will be no income generating asset to offset the costs. Below the £1billion figure is used, but the risk is that it could be three times higher.
The costs and benefits of building and decommissioning power stations fall over a long time period and so they have to be discounted to ‘present values’ so that they can be compared with other options on a like basis. Discounting is based on the principle that generally people prefer to receive goods and services now rather than later and the recommended rate up to 30 years is 3.5%. This can have a radical impact on costs and benefits: for example discounting a cost of £100 at 3.5% over 20 years gives a cost of £50.26. So if something would cost £100 in 20 years time its present value cost would be only £50.26.
In the latest edition of the Treasury Green Book – Annex 6 set out the reasons why beyond a 30 year time horizon a lower discount rates should be applied.
Thus:-
at 50 years – discounted rate – 3%
at 90 years – discounted rate – 2.5%
at 185 years – discounted rate – 2.0%
beyond 300 years – discounted rate – 1.0%
Discounting the British Energy estimates of decommissioning cost, according to the Treasury schedule gives a present day cost of £67 m per plant. In other words, to pay for end of life decommissioning it would be necessary to have a fund to pay for this of at least £1 billion. This adds some £70 m to today’s costs per plant on the assumption that this sum can be invested (elsewhere) and grow to cover the £1 billion which will be needed. This adds 0.1p/kw hr to prices. With the more pessimistic clean up costs set out could rise to 0.3 p/kw.
The current policy assumption is that after “final” storage there are no further costs. This is not credible. Any society would need to invest a small sum to (a) guard the site (b) monitor the geological and seismic conditions and (c) maintain a nuclear expertise. Already, this last is proving problematic. Taking a conservative estimate of these costs to the nation overall suggests an annual cost of around £1 billion, which spread across all plants and waste hitherto would add an annual charge of £20m per plant for the next 60,000 years. In other words by Year 200 a fund to cover annual costs of £20 m pa would also be needed. Currently a fund of £500m would cover this. To build up such a fund adds £5 million to today’s costs (also discounted as above).
Using the long term discount rates recommended by the Treasury, something costing £100 in 200 years time would be costed at 62 pence: if it cost £100 in 500 years time, it would be costed at 2 pence in present value terms.
One of the problems in securing a rational discussion of the costs of nuclear power is that those opposed to nuclear power point to the risks and hence potential costs over a 60,000 year period to which we are committing future generations. Meanwhile those in favour argue for discounting these costs on traditional lines. This technique and the figures used are based on time preference curves built over short-term periods, which suggest that current generations place no (or only a very tiny) value on costs to future generations. In other words the technique itself closes off the possibility of being concerned for future generations, which is precisely what the opponents refuse to do.
Thoughtful writers understand this. Amartyra Sen in 1982 wrote: ‘The cost and utility calculus cannot begin to convey the complexity of choices surrounding investment in nuclear power (because) we are capitalizing on the arbitrary fact that we could get at the resources before the future generation could.’ And 300 years earlier, John Locke said ‘Each man is entitled to the fruits of his labour, as long as as much and as good is left for the next.’
TABLE 1 – How the costs impact on prices - See orginal SERA publication for details
A useful way of estimating likely nuclear costs is to look at independent studies which use assumptions in the realistic range.
TABLE 2 - See orginal SERA publication for details
(1) Ibid
Conclusion
This suggests that on realistic assumptions the cost of nuclear generated electricity will lie between 3.6 and 3.9p/kwh.
Costings and proposals which lie outside the “reasonable range” of assumptions are either unrealistic or imply significant policy changes (eg the provision of government guarantees or subsidies.)
What this means is that other low carbon dioxide generating options which cost less than 3.5p/kwh are a more cost effective and economic way of meeting our climate change objectives. Wind power, CHP, energy crops and some wave power meet this target. Currently, photo voltaics do not, but over the timescale under consideration these costs might come down since they are still in a relatively early stage of development.
Notes
(1) Steve Thomas – University of Greenwich – 2005 Paper for the Environment Agency
(2) Performance and Innovation Unit – 2002
(3) Treasury Green Book – 2003
(4) Sen: Choice of Discount Rates for Social Benefit Cost Analysis in Discounting for Time and Risk in Energy Policy–Lind-1982
Published by SERA as part of ‘What’s in the Mix: The Future of Energy Policy’, 29 March 2006
