October 23, 2016

France – country of the atom

If nuclear power seems cheap in France, it is because half the costs have been ignored. An accurate accounting of costs, direct and indirect, reveals France’s massive nuclear electricity programme as a ruinously expensive folly.

Written with Peter Bunyard, co-editor of The Ecologist. Published in The Ecologist Vol. 11 No. 6, December 1981.

France’s nuclear power prog­ramme is the most ambitious in the western world; held up by the nuclear lobby as a shining example for others to emulate. It is no coincidence that Mrs Thatcher decided on her own ambitious programme of nuclear construction in the UK after a visit to various French nuclear installations in 1979. What particularly impressed her, it seems, was the sheer size and breadth of the French nuclear programme and claims for its incredible cheapness.

Such blind acceptance makes it particularly important that the French nuclear costings should be looked at closely – a difficult undertaking since Electricite de France (EDF) has shown itself extremely reluctant to divulge information other than for PR purposes. Nevertheless by piecing together bits of information that have been made available over the last few years, it is possible to get an idea of the true cost of nuclear electricity in France.

Particularly useful in this respect is an unpublished report on the subject, undertaken by a group working in the field of alternative energy, Ecologie, Energie et Survie, headed by Ferone de la Selva. From the information contained in this report we offer the following comments on the French nuclear costings.

The cost of delivered electricity

According to the chairman of EDF, Marcel Boiteux [Liberation 23/09/80] the price of nuclear generated electricity was then 13 centimes/Kwh and thus considerably lower than that of coal or oil-fired generating plant. Yet the figures published in the EDF’s Annual Report for 1979, hence for the preceding year, indicated a much higher figure for nuclear power. Thus total production from nuclear generation was then 36,200 million kWh at a cost of 6,813 million francs; that puts the cost at 18.82 centimes/kWh.

Given that inflation in France is running at approximately 13 percent, Boiteux’s figure of 13 centimes/kWh becomes 11.5 centimes/kWh for 1979. Thus the cost of nuclear generated electricity in France appears to be underestimated by Boiteux to the extent of 7.3 centimes/kWh. On that basis the cost of nuclear electricity comes out at 63 percent higher than the cost quoted by the EDF chairman.

The EDF Annual Report for 1979 also indicates the cost of electricity generated from fossil fuel fired plant to be 14.72 centimes/kWh (total production 96,000 million kilowatt-hours; total cost 14,134 million francs). Similarly, hydroelectric electricity will have cost 7.05 centimes/KWh, having taken the output of 61,700 million kWh at a cost of 4,352 million francs. With those figures in mind it is clear that nuclear generated electricity during 1979 was considerably more expensive than that generated by either fossil fuel or hydroelectric plant.

On 5 December 1980, Mr Bergogneaux, assistant director of general economic studies of the EDF, stated that the price of nuclear electricity in 1990 would be 15 centimes/kWh, taking into account financial charges of 8.10 centimes, fuel 4.00 centimes and running costs of 2.90 centimes/kWh. Meanwhile, he said, the price of electricity derived from coal would be 25.80 centimes/kWh comprised 6.2 centimes for financial charges, 14.00 centimes for fuel, 2.90 centimes for running costs and 2.70 centimes/kWh for de-sulphurisation.

The EDF has used various expediencies to mask the true cost of nuclear electricity; when such expedients are taken into account, the cost of nuclear electricity turns out to be much higher than admitted and far higher than the price of electricity derived from coal.

Back to top

The Cost of Research and Development

The budget of the Commissariat de l’Energie Atomique for research and development into civilian, as opposed to military, uses of nuclear power was 4,980,400,000 francs in 1980 and was increased to 5,674,900,000 francs in 1981.

Of those sums, 3,048,600,000 francs for 1980 and 3,478,100,000 for 1981 were paid for by government subsidies. Moreover the CEA does not bill the EDF for the results of its research and development; these are provided free, with the result that the costs of research and development are not incorporated in the cost of nuclear electricity. We have here an obvious hidden subsidy paid by the French taxpayer to the nuclear industry.

Back to top

Capital costs

In its report of 1980, entitled Energy in Developing Countries, the World Bank estimated the capital costs of large nuclear reactors at $1,600/kW. On the basis of a rate of exchange of 4.75 francs/dollar the World Bank figure translates into 7,600 francs/kW. The World Bank estimated a coal-fired plant to cost $1,000/kW. EDF expects the cost of a PWR to be exactly one half that of the World Bank’s. Its coal-fired plant costs 40 percent less than the World Bank estimate.

Undoubtedly EDF’s estimate is for the perfect PWR – one that is built exactly according to specification and schedule with no account taken for likely cost overruns.

Back to top

Cost overruns

Most nuclear power stations take longer to build than was originally foreseen and budgeted for. Moreover, according to the EDF, each extra day’s delay in commissioning a nuclear plant costs a million francs [Valeurs Actuelles 7th July 1977]. The Fessenheim PWRs took 22 months more to build than predicted, leading to added costs of 68 million francs in 1977 values. Since contractors normally pay only a small indemnity for each day’s delay, EDF absorbs most of the extra cost; yet it makes no allowance for cost overruns in its estimates of the cost of nuclear electricity.

Another nuclear station at Gravelines comprising two reactors of 900 MW each incurred a cost overrun estimated at 7 billion francs; a sum which the EDF again failed to take into account, despite it involving on its own figures an extra capital expenditure of 3,889 francs/KW.

Such cost overruns are particularly significant in a period of high inflation, when much of the money to meet the massive capital costs of nuclear installations has to be borrowed. In fact EDF is heavily subsidised by the government; the extent of the subsidy coming to light in a debate in the House of Commons on 21 January 1981 (see HC 78.1 Pricing Policy 21.1.81 pp.82-3).

Mr Stoddart MP, asked the Secretary of State for Energy if it were true that the French government had written off £1.4 billion of EDF’s debts relating to capital expenditure and had suspended, until 1985, interest charges on borrowings for the construction of nuclear installations. The Secretary of State for Energy, Mr David Howell, admitted that “Broadly similar things have been done”.

Back to top

Correction for inflation

The EDF applies a correction for inflation on its capital costs (taux d’actuellisation) of 10 percent. That value underestimates the inflation rate, which in France over the last few years has been closer to 13 percent. Use of the latter, more realistic, figure would lead to a considerably higher figure for the capital costs of a nuclear power station. Thus, on a starting value of 100, a 10 percent inflation rate over ten years would yield a figure of 259.36: a 13 percent inflation rate over ten years would yield a figure of 339.38 – more than 30 percent higher.

Because of the much lower capital costs of fossil-fuel fired plant, that underestimate of the true cost of inflation does not have such a pronounced effect on its overall generating costs, compared with the effect on nuclear power generated electricity.

Back to top

Load factor

A further technique for underestimating the capital costs of nuclear electricity is to calculate them on the basis of an unrealistically high load factor. The World Bank assumes that a nuclear power station will function for 7,000 hours per year; the EDF is marginally more realistic in assuming a figure of 6,600 hours per year, therefore a load factor of 75 percent.

Mr Paul Quilles, a nuclear specialist, who acts as energy spokesman for the French Socialist Party, stated on 23 December 1980, in Liberation, that nuclear power stations in France functioned for an average of 5,000 hours per year. In its Annual Report for 1979 [Rapport d’Activite Compte de Gestion] the EDF admits that in 1979 its PWRs functioned with a load factor of 54.4 percent (similar, according to Komanoff, for large PWRs in the US). Thus the EDF has assumed a load factor which is 20 percent too high. The effect of that poorer performance is to raise the capital costs/kW installed of nuclear electricity by a corresponding amount.

Back to top

Insurance against accidents

In France, EDF pays nothing for insuring its nuclear power stations. The reason is that nationalised industries are supposed to insure themselves – auto-assurance – and the insurance companies are themselves nationalised. Thus losses incurred by nationalised companies, such as EDF, as a result of accidents, would be paid for by the same taxpayers, whether or not an insurance had been taken out against the risks involved.

In essence, therefore, the French taxpayer is landed with the entire risk of operating the nuclear power stations. That hidden subsidy is not revealed in the price of nuclear-generated electricity.

Back to top


On the EDF’s own figures, hydroelectricity provides by far the cheapest electricity (7.05 centimes/Kwh 1979 value) in France; moreover as a renewable energy source it has no fuel costs whatsoever. Consequently hydroelectricity must take pride of place in the merit order, being operated exclusively on base load.

Yet, apparently in France, that situation no longer wholly prevails, the reason being the need to justify the enormous investment in nuclear plant by running it as hard as possible and the technical difficulties of running it otherwise. Thus some hydroelectric plants have been relegated further down the merit order to make way for the new nuclear plant coming on stream.

But there is another reason for that relegation. Because of breakdowns in the nuclear plant and poorer than expected performance, the EDF must keep certain plant on hand that it can operate with great flexibility in case of sudden demand. Hydroelectricity fits the bill, as does gas turbine plant. The latter, in fact, is extremely costly to operate because of high fuel costs – hence a preference to use hydroelectricity.

Certain hydroelectric plants are therefore kept in a state of readiness, with the reservoirs behind the dams full. Consequently EDF is losing a certain proportion of power available to it, through enhanced evaporation from the reservoirs. The holding in reserve of hydro-electricity, together with the loss of power from evaporation, both add to the cost of EDF’s electricity, yet no account is taken of such an increase.

Back to top

New coal-fired plant

Gas turbines and hydropower are in themselves insufficient insurance against nuclear inflexibility and unreliability. Furthermore the time – and cost -overruns on the nuclear programme have led to less generating capacity being available than expected. The need for extra fossil fuel fired stations has become apparent. Thus on 18 May 1977, Paul Delouvrier, then chairman of EDF, declared, at Grenobles, that the EDF would ask the government’s permission to build two 1,400 MW fossil fuel fired plants, that had become necessary because of delays in the nuclear programme.

In the meantime existing thermal power stations are used to the maximum of their capacities. Indeed the load factor of those ageing plants actually went up from 69 percent in 1978 to 72 percent in 1979 (as against the 54 percent of the PWRs). In its annual report for 1979, the EDF boasts of such high performances, which were achieved,

“in spite of a greater use of the equipment, of the mediocrity of the fuel burnt, and of the age of the present population of coal-fired stations – more than 40 percent of which have already more than 100,000 functioning hours to their credit.”

That over-use of equipment must inevitably reduce the plants’ lifespan, thus a further hidden cost of nuclear power that should be reflected in its generating costs.

Back to top

Sale of electricity at a loss

In France new nuclear stations are coming on stream each year as a consequence of the construction programme. To keep nuclear stations operating solely on base load is clearly important if costs are to be kept down, and to achieve that, EDF must boost sales. But sales are unlikely to grow if the product is dear, and EDF has therefore begun selling at a loss. Not that it admits freely to such a practice; on the contrary it employs a strategy to convince the public how cheap nuclear electricity is.

Thus, in January 1980, when still President, Giscard informed Frenchmen that, because nuclear electricity was cheaper than other forms of electricity, those living within a 20 kilometre radius of nuclear stations would have their electricity bills slashed by 15 percent. No such enticement was given to those in the vicinity of hydroelectric schemes.

That large sectors of industry are buying their electricity below cost, is admitted by Paul Delouvrier. In an interview with Agence France Presse in 1976, he stated that

“The reduction in industrial production has had the effect of reducing EDF’s deficit predicted in 1975. Since each kilowatt sent out is sold below its cost price, every kilowatt not sold actually reduced the deficit.”

EDF will soon embark on a major sale loss to Eurodif, which has established a massive uranium enrichment plant at Tricastin. The site is served by four 900 MW PWRs which will feed the enrichment plant with some 15,000 million kilowatt-hours per year. The electricity is at 11.97 centimes/kWh. Since EDF’s annual report for 1979 indicates the delivered cost of electricity to be 18.82 centimes/kWh, the loss of revenue is 6.85 centimes/kWh leading to a total loss of 1,030 million francs per year.

Moreover the 106,000 employees of the EDF actually get their electricity almost free (they pay no more than 4 centimes/kWh – presumably in lieu of extra wages). At the same time the EDF is making great efforts to force French industry into staggering hours and thus into returning to a three-shift system so as to even out electricity consumption during each 24 hour period.

Back to top

Fuel costs

Jean Bergougnoux, EDF’s economist, expects the price of nuclear fuel to fall by more than 5 percent, while that of coal will increase by nearly 17 percent.

In reality, nuclear fuel costs have turned out to be very much more expensive than predicted, largely because of the unexpected high cost of uranium enrichment and the intransigent problems associated with reprocessing spent fuel. The extra cost of those processes was already admitted by the EDF in its Annual Report for 1979. Thus it indicated a 34.3 percent increase in a single year in the costs of “contract work, equipment and external services”.

The same report discloses that the provisions for losses “have been increased by 28.3 percent in a single year, thus by 2,325 million francs”. Hence the total amount put into those provisions in 1979, amounted to 8,215,547,700 francs – a truly stupendous sum. EDF acknowledges that the provisions are to pay for the cost of nuclear fuel and also for the decommissioning (declassement) of nuclear power stations. It has failed to make clear how it can be putting aside vast sums for meeting additional costs on the one hand and on the other, can be enjoying falling nuclear costs.

Back to top

Breeder reactors

Both the French nuclear industry and British are aware that the nuclear adventure must inevitably be short-lived, unless there is a shift to fast reactors. In France, Phénix 250 MW has been in operation for a number of years and Super Phénix, a 1200 MW fast reactor, is currently under construction at Creys Malville. It should come on stream by 1983. According to Paul Quilles, the Socialist Party nuclear expert [Liberation 23/09/80] already 8 to 10 billion francs have been spent on it.

Marcel Boiteux, the present President of the EDF, is concerned that Super Phénix is turning out to be more expensive than foreseen:

“Although it is premature to put Super Phénix in the same boat as Concorde, it is questionable whether it will be competitive since the price of the kilowatt-hour furnished by Super Phénix is likely to be closer to that furnished by heavy oil” (33.52 centimes).

From the available material it is clear that the cheapness of France’s nuclear electricity is a myth. To have any basis of reality, the costs of France’s generated electricity would have to include:

  • the costs of research and development of both reactors and of other installations necessary for the operation of nuclear power;
  • cost overruns;
  • the 20 percent poorer performance than budgeted for;
  • the need to keep hydropower as a back-up system;
  • the need to build coal-fired plant for the same purpose and to cater for delays in the commissioning of nuclear plants;
  • a rate of increase in construction costs to take account of the real fall in value of the franc, as a result of inflation.

The inclusion of those costs alone will lead at least to a doubling of EDF’s price of nuclear electricity and take it above the 25.80 centimes/kWh given for future coal-fired generation. That likely doubling of costs is still an underestimate, since the costs of dealing with the decommissioning of power plants and of coping with wastes generated in spent reactor fuel will undoubtedly be greater than the provisions made for them.


  • Twitter
  • Facebook
  • Digg
  • Reddit
  • StumbleUpon
  • Diaspora
  • Identi.ca
  • email
  • Add to favorites
Back to top