France obtained its first milligrams of plutonium in 1949 by reprocessing at Le Bouchet, fuel that had been irradiated in Zoé. After experimentation and operation of a reprocessing pilot at Fontenay, production of plutonium on an industrial scale began at Marcoule in 1959 with the entry into operation of the Usine plutonium 1 (Plutonium plant 1, UP1). A plant called the Atelier pilote de Marcoule (Marcoule pilot, APM) opened shortly afterwards. Then came the large plants at La Hague: UP2-400 (capacity of 400 tons per year), put into service in 1966, then UP3 in 1990 and UP2-800 in 1995, the last two with capacities of 800 tons each.

The production of nuclear weapons was the main reason for the separation of plutonium. UP1 produced its first plutonium ingot in 1959 and France conducted its first nuclear test in 1960.

A second reason was the production of fuel for breeders or fast neutron reactors. The nuclear industry believed that breeders, with plutonium 239 as the fissile material and uranium 238 (in particular in the covers) as the fertile material, would produce more plutonium than they would consume. Thus they would furnish a perpetual source of energy. Moreover, they produced, in their covers, plutonium of excellent military quality. The optimism was international.

In the fifties and sixties, France could not produce enough plutonium to meet the foreseen needs of its civilian and military programs. Foreign countries gave France the ability to begin its “civilian” breeder program by furnishing France with plutonium and enriched uranium. Euratom provided plutonium of American origin for Masurca [LeMo 29.x.66]; the United Kingdom provided plutonium for the first core of Rapsodie [Eurat 64]. Then Canada became a source of plutonium, as it sold France irradiated fuel from its Candu reactors to be reprocessed by Eurochemic [CEARa 68, 69, 70]. Le Monde summarized at the time: France “needs this plutonium if it does not want to assign to civilian uses what it produced to meet military ends” [LeMo 29.x.66].

For its military needs, France increased the plutonium coming from its explicitly military reactors, G1, G2, and G3, as well as from a civilian UNGG plant (Chinon) belonging to EDF. In 1963, the principal of tactical nuclear weapons was retained by General de Gaulle, but the fabrication of weapons had to be postponed, because of a lack of plutonium as well as funds. In 1996, the Conseil de défense decided to take the necessary fissile material from its civilian program [Barrillot 92]. As a result, France has never made a definite separation between its military and civilian programs.

Later, the military obtained plutonium from the covers of the breeder Phénix which went into operation in 1973 and from the two tritium-producing Célestin reactors which began to operate in a plutonium-producing mode around 1978 (all at Marcoule). The Célestin were officially military, but Phénix belongs to the CEA and EDF.

Today France produces more plutonium than it can consume. Each of EDF’s reactors unloads each year about 20 t of fuel, which contain about 200 kg of plutonium. Reprocessing ended at UP1 and at APM in 1997; and now, UP2-800, which reprocesses French fuel in particular, and UP3, which reprocesses foreign fuel, together have a nominal capacity of 1600 t/yr of oxide fuel or about 16 t/yr of plutonium.

The French military does not want more plutonium. If the military again lacks plutonium for weapons, it can recycle the plutonium in the heads of dismantled missiles. As they decided, reprocessing to produce plutonium for explicitly military ends ended at Marcoule in 1994 [Ener 15.xi.93].

The breeder program proved to be a failure. The only French breeder of industrial size, Superphénix (1240 MW), produced only 8.2 TWh of electricity during the ten years that it was coupled to the network or the equivalent of six months of operation at full capacity. In 1998 the government ordered final shutdown operations to begin.

France has given the IAEA a report listing the quantities of civilian plutonium in France at the end of 1999:

Civilian, unirradiated plutonium

--Separated and unirradiated plutonium stored in reprocessing plants: 55.0 t

--Plutonium separated in the course of fabrication, and plutonium in semi-finished products in installations for manufacturing fuel or other installations: 13.0 t

--Plutonium in unirradiated MOX fuel or in other fabricated products, in power plants or other installations: 8.2 t

--Separated and unirradiated plutonium stored in other installations (this figure includes estimates of plutonium in fuel in the course of reprocessing and of separated plutonium present at research sites): 5.0 t

The plutonium contained in civilian irradiated fuel

--Plutonium contained in used fuel in reactor pools: 80.0 t

--Plutonium contained in used fuel awaiting reprocessing in reprocessing plants: 79.2 t

--Plutonium contained in irradiated fuel stored in other types of sites, in particular research installations: 0.6 t.

In total, 241.0 t of plutonium, of which 37.7 t of separated plutonium belong to foreign entities (Less than 50 kg of separated French plutonium that has not been described above is held in unspecified installations located abroad) [INFCIRC/549/Add.5/3, 19.iii.01]. Of these 241.0 t , which represent the total at the end of 1998, 81.2 t represent separated plutonium, and 159.8 t represent plutonium contained in irradiated fuel. In 1996, the total of the separated plutonium and the plutonium in the fuel was 206 t [Industrie 96]; in 1998, 234.7t.

Between the end of 1994 and the end of 1997, the stocks of separated plutonium in France (French and foreign) went from 42.9 t to 72.3 t, and the stocks of French plutonium from 21.9 to 38.7 t [Takagi 99]. The rate of increase of the French stock has slowed since then; but the stock continues to increase. According to the figures announced by France and the IAEA, the stock was 40.3 t at the end of 1998 and 43.5 tons at the end of 1999.

However, France has invested tens of billions of francs in reprocessing plants. To justify the continuation of reprocessing, the industry has to demonstrate that it masters and can use plutonium. Today, after the premature end of breeders, plutonium is partially used in MOX fuel for the PWRs. In March 2000, EDF put MOX into its twentieth reactor. At the same time, France studies methods of “incinerating” plutonium.

EDF unloads about 1200 t of irradiated fuel per year. Officially, it does not want to obtain more separated plutonium than it can use in the short term in MOX fuel. In reality, EDF is not enthusiastic about the production and use of plutonium.

In private, the directors of EDF admit that reprocessing and the fabrication of MOX are much too expensive and make the operation of reactors more complex in terms of safety. But EDF is not alone in making decisions at this level. The plutonium lobby is powerful, and the several thousand jobs at La Hague carry weight. Consequently, according to information obtained by Wise-Paris, the previous industry minister, Frank Borotra, drew up a letter addressed to EDF in July 1996, demanding in particular an increase in the number of reactors loaded with Mox and maintenance of Cogéma’s industrial tools. Wise-Paris has never succeeded in obtaining a copy of the letter, despite requests to the ministers involved, but the increase in the number of moxed reactors from 8 to 14 between the beginning of 1996 and the end of 1997 seems to be tied to that letter [Takagi 99].

The second report of the Commission Nationale d’Evaluation (CNE), published in mid-1996 states: “in 1996, EDF made known to the Commission its industrial policy based on the reprocessing, starting in 2000 of 850 tonnes a year of fuel (of the 1200 tonnes a year coming from power plants) and the monorecycling [a single passage through a reactor] in Mox fuel of the recovered plutonium. The irradiated fuel not reprocessed, 350 tonnes a year, 215 tonnes of uranium oxide and 135 tonnes of Mox, will be stored under water awaiting a decision as to its final end (delayed reprocessing or disposal).”

Today, the strategy remains the same, but EDF is less reticent than in the past to show its lack of enthusiasm in regard to reprocessing and the separation of plutonium. According to the fifth report of the CNE (that of 1999), “Today, EDF does not justify the policy of monorecycling under the energy angle, but presents it rather as a way of managing the end of the cycle by the concentration of plutonium, aiming thus to limit the quantity of used fuel that is not reprocessed and is put into storage.”

According to EDF, the 8 to 8.5 t of plutonium extracted each year from used French fuel would furnish enough Mox for 22 reactors [NucF 1.v.00]. The 16 reactors in the CP1 class (900 MW) have been authorized to use MOX fuel since they went into service. In 1998, EDF was authorized to use MOX also in four reactors of the Chinon power plant, reactors of the CP2 class (900 MW). EDF has announced its intention of seeking authorization to use MOX in twelve other CP2 reactors in order to have the necessary flexibility in case one or more reactors does not operate for a period of time. Nevertheless, Christian Pierret, secretary of state for industry, indicated in October 1997, “We do not envisage loading MOX fuel into the 28 reactors that could receive it” [Takagi 99]; and for two years the environment minister has blocked the authorizations that EDF has demanded for the Gravelines C5 and C6 reactors. The number of authorized reactors thus remains 20 for the time being.

Unlike EDF, Cogéma is enthusiastic about Mox as a source of energy. To profit to the maximum from its experience in manufacturing Mox, the company turned to foreign utilities some ten years ago. Cogéma’s CFCa at Cadarache works for German companies. Cogéma has built a second line at the Melox plant intended for the fabrication of fuel for boiling water reactors. This fuel will necessarily be used by German and/or Japanese clients, since EDF only owns pressurized water reactors.

The extension in question, which Cogéma calls “l’aménagement Melox” (the Melox adjustment), can manufacture 30 to 80 tons per year of Mox, depending on the mode of operation. The entry into service authorization does not allow Cogéma to increase the nominal capacity of the plant. Nevertheless, Cogéma has indicated that it could, once an authorization has been granted, produce 250 t/yr of Mox operating 5x8 (5 teams, 8 hours per team) at the basic installation and the extension. If a new authorization is not granted, Melox will have to reduce production for EDF in order to produce Mox for export.

With the present authorization (115 tons of oxide or 101.3 tons of heavy metal per year), Cogéma is not capable of producing enough MOX to stabilize EDF’s stock of separated plutonium even if Melox produces MOX only for EDF.  The 850 t of fuel that EDF plans to process each year contain 8 to 8.5 t of plutonium.  In 2000 with Mox fuel at about 6% plutonium, it would be necessary to produce about 140 tons of MOX heavy metal to absorb the annual flow of 8.5 tons of plutonium, resulting from the reprocessing of 850 tons of EDF’s used fuel. “At the maximum level of 7.08% of plutonium in MOX authorized since the end of 1998, the total production of Melox, 101.3t, uses no more than 7.2 t of plutonium a year” [Wise ATPu 00]. Nevertheless, Cogéma carried out in Melox the “first fabrication of Japanese MOX pour the Kansai electric utility” in 1999 [CLIGard 27.vi.00]. Furthermore, Cogéma is asking the government to allow it to transfer from ATPu at Cadarache to Melox the production of MOX for Germany, in return for Cogéma’s shutting down ATPu as demanded by DSIN since 1995 because of the earthquake risk in the region (see the chapter on Cadarache). Cogéma has set up a situation in which it apparently believes that it will be able to force the government to authorize the production of more than 115 t/yr of MOX [Wise ATPu 00].

Only nations that must find means of using separated plutonium are likely to ask Cogéma to fabricate MOX. The production of MOX (6-7 MF per t) is clearly more expensive that the production of standard fuel (2-3 MF per t).In their study for prime minister Lionel Jospin, Etude économique prospective de la filière nucléaire, Jean-Michel Charpin, Benjamin Dessus and René Pellat stated that the option of stopping reprocessing in 2010 would cost less than the options that include reprocessing and the production of MOX [Charpin 00]. The plutonium is perfectly usable for the fabrication of bombs, and the transportation of plutonium and of fuel with plutonium create significant risks in the event of an accident on the road as well as in the event of theft. Consequently, its transportation and use necessitate surveillance and protection measures that are expensive to put into effect and that are likely to infringe on civil liberties. Moreover, MOX makes the organization of fuel handling and even reactor safety more complex.

The difficulties in using MOX in pressurized water reactors are of different types and make operation of the reactors more complicated than in the case of uranium oxide fuel:

-- constitution of three different levels of plutonium (trizonage), the level being higher in the elements at the center of the assembly than in the elements on its periphery;

-- noticeably and systematically larger releases of fission gas from MOX, which increase greatly with burnup;

--variation in the isotopic composition of plutonium from lot to lot;

--more rapid changes in the power level (ie of the reactivity) of the reactor

-- reduction of the power of absorption of the control rods;

--higher residual power after shutdown.

The increase in the rate of burnup of Mox to make the use of Mox more profitable will lead to "a major increase in the average level of plutonium. In 1999, DSIN authorized use of MOX with 7.08% plutonium on average per assembly, instead of the 5.03% authorized previously. EDF would now like to obtain the authorization to use MOX with an average level of 8.65% plutonium, and 9.8% on average in the central zone of the assembly. These changes can only make the use of MOX still more delicate.

As noted above, the 103.1t heavy metal of MOX that Melox is authorized to manufacture will not use up the 8 to 8.5 t of plutonium that result from the reprocessing of 850 t of fuel each year for EDF. Therefore, for the foreseeable future, the stock of separated plutonium will continue to grow. If Melox receives an authorization to use in MOX all the 8 to 8.5 t of plutonium separated annually for EDF, EDF’s stock of separated plutonium will stabilize but not shrink. Moreover, the stock of irradiated fuel will continue to increase. EDF stores around 10,300 t of irradiated fuel (7000 t of it at La Hague as of May 31, 2001). As we have already mentioned, the reprocessing of 850 t/yr of fuel leaves 300 to 350 t of irradiated fuel, including irradiated MOX, and these quantities are to be added to the existing stock. EDF has no intention, in the year 2000, of reprocessing the irradiated MOX.

Since 1987, the CEA has conducted studies at Fontenay and at Marcoule to evaluate the “reprocessability” of MOX/REP. Cogéma management in fact announced in 1995 that the company was “able to propose reprocessing services” for MOX in UP2-400 [Gillet 95]. The UP2-800 and UP3 installations are not the right size for reprocessing MOX fuel, but Cogéma is seeking authorization to reprocess MOX in UP3 as well as in UP2. Only small quantities of MOX for pressurized water reactors have been reprocessed. APM at Marcoule reprocessed 2.1 t of MOX in 1992. During the same year, Cogéma reprocessed 4.7 t of German MOX in UP2-400 at La Hague [Gillet 94]. In 1998, Cogéma reprocessed 4.9 t of German Mox in that plant [Takagi 99].

In comparison with U02 fuel, the reprocessing of Mox poses supplemental problems, including an increase in the risk of a criticality because of the increase in the concentrations of plutonium, the degradation of solvents because of the increase in alpha emissions, the difficulty of decontaminating plutonium because of the presence of products from the degradation of tributylphosphate, and the difficulty of manipulating plutonium containers because of the heat emitted by the plutonium 238 [Baetslé 94]. Moreover, “the increase in the quantities of actinides mixed with fission products would lead to an increase in the number of vitrified packages, for the same quantity of energy produced” [CNE 00].

If one reprocesses light water Mox fuel, the plutonium that leaves the reprocessing plant will not be easily usable as materials for light water fuel. Moreover the irradiation of Mox degrades the quality of plutonium in the fuel.

During irradiation, the proportion of plutonium 239 in Mox decreases and the proportion of even-numbered isotopes increases. These isotopes are neutron poisons, i.e. when slow neutrons strike these isotopes, the isotopes are more likely to absorb the neutrons than to be divided by them. The irradiation of Mox also increases the inventory of other long-lived bodies, such as americium 241 and curium. With each recycling the quality of plutonium decreases.

The CNE notes that the recycling of plutonium “in a depleted uranium fuel necessitates an increase in the level of plutonium in each recycling, in order to compensate for the penalizing effect of these even-numbered isotopes on reactivity. For reasons of both safety and cost, there can hardly be more than one or two recyclings in these conditions” [CNE 96]. In other words, whatever the scenario, it will end up being necessary to manage un-reprocessed irradiated fuels.

Moreover, the storage and stocking of irradiated MOX fuels creates problems because of their “very high temperature”. “A long period of under-water storage” would be necessary “before it would be possible to envisage dry cooling in long term storage” [CNE 00]. Moreover, the MOX would occupy a substantial part of the area of likely general stocking” [CNE 00}.

The present reprocessing contract between EDF and Cogéma ends in 2001, and at the end of July 2000, EDF had not yet signed a contract with Cogéma for the following years. Consequently, we cannot be certain that EDF will pursue its current strategy, in particular because the opening up of the electricity market in Europe has modified the context in which EDF operates.

In 1999, the CNE presented EDF’s strategy to 2070, based “on the hypothesis that nuclear energy represents a long-term option in France and is a majority component of the national electricity supply.” According to this strategy, EDF will maintain a park of reactors “corresponding to a constant production of 400 TWh of nuclear-generated electricity, and a policy of monorecycling of plutonium in the form of Mox fuel, corresponding to reprocessing annually on the order of 850 t of used uranium oxide fuel.” “EDF is planning today on technical lives of forty years for its 900 MWe reactors and of fifty years for its 1300 MWe series.” As these reactors age, EDF would replace them by reactors “of evolutionary type from the series called REP-2000, which would probably consist of EPR reactors. Use of Mox, at an average15% level, in 19 reactors of a 35 EPR park would guarantee equilibrium in production-consumption of plutonium. The use of Mox in supplemental reactors would allow the absorption of a part of the stock of unreprocessed uranium oxide fuel, according to EDF.

EDF “excludes explicitly all options of multirecycling plutonium, like Mix, which consists of homogeneous recycling of plutonium at a low level, on a support of enriched uranium” [CNE 00]. In the Mix option, studied by the CEA, each assembly contains plutonium. “Burnup rates of 55,000 MWd/t are possible with 2% plutonium and 3.8% uranium 235 [Bataille 98].

The CNE notes that, “over the 70-year period considered in the presentation of EDF, the adoption of multirecylcing of plutonium in fuel of the Mix type, or other types [ ]. . . , would have the effect of greatly reducing the quantities of unreprocessed used fuel (in principal all the plutonium from the park is recycled) and the inventory of plutonium in reactors and in the cycle would stabilize at around 200 t, according to the CEA source cited in report no. 3, in comparison to 600 t announced by EDF in 2070” [CNE 99].

However, C. Bataille and R. Galley indicates the problems with this method. “A first problem is that all the pressurized water reactors would have to be adapted to use this fuel.” “moreover, the plutonium stocks would be stabilized with Mix. But that would only happen after fifty years, while, in the interval, the net stock of plutonium would grow.” Moreover, “according to the evidence, the Mix technique would stabilize the plutonium but in parallel would inevitably increase the proportion of minor actinides in the irradiated fuel [Bataille 98].

EDF’s strategy to 2070 does not take the RNR (rapid neutron) reactors into account. In 1993, the CEA had launched the Capra project (Consommation accrue de plutonium dans les rapides [Increased plutonium consumption in rapid reactors]) to perfect future neutron reactors, designed to “consume” plutonium. Pressurized water reactors using Mox fuel can “transform” fissile plutonium isotopes into other materials, but the CEA prefers the RNR’s, because, theoretically), because they can transform plutonium of any isotopic quality in large quantities, as well as other actinides. The official shut-down of Superphenix has badly hurt the Capra program. Indeed, this program cannot “be totally carried out”, although certain studies may be made in the Phenix reactor [Bataille 98].

Nor does EDF’s strategy to 2070 include any reference to applications deriving from “axis 1” [CNE 00]. That means (except for MOX) that EDF does not presently plan to apply any of the transmutation-separation scenarios studied by CEA and other organisms.