UP2-800 and UP3 are under Euratom control; and the IAEA inspects the irradiated fuel storage pools. Moreover, Cogéma has installed computers designed to prevent the stealing of plutonium. Nevertheless, because of problems belonging to every reprocessing plant, control of nuclear material is never one hundred percent effective. Some examples of the problems follow.

The most obvious difficulty is the lack of knowledge of the exact quantity of plutonium in the plant. Arriving plutonium and uranium cannot be measured with exactness before the assemblies have been sheared, dissolved, clarified, and transferred to a control vat. The quantity of plutonium in each assembly can be estimated by taking account of the level of irradiation and the length of the cooling period, but this figure is never exact enough for control purposes.

The question of quantity is made still more difficult by the fact that plutonium always remains fixed to the pipes and other equipment within the plant. One can try to measure this plutonium from a distance, but the uncertainty is at least 5% and can be 30% [Leventhal 94].

Therefore in a reprocessing plant there is always a difference between the measured quantities that enter and the measured quantities that leave-the definitive difference or MUF (materials unaccounted for), materials for which the company cannot give the destination. Whether they have been “lost” in the plant, have been stolen, or have been mixed with the wastes is not known.

The MUF in regard to uranium and plutonium at La Hague between 1984 and 1991 was 0.07% and 0.32 % respectively. (Cogéma’s 1994 data included no figures for MUF.) Wise-Paris calculated the value of the MUF in absolute figures at the nominal capacity of UP2 -800 and UP3 and determined that at least 537 kg of uranium and 45.3 kg of plutonium per year could not be accounted for.

But, apart from the problem of the determination of quantity, there is the factor of time. Can the authorities detect a loss of fissile material before the thieves are able to fabricate a nuclear device? After discovery of a possible loss, the IAEA has to do much supplemental research to verify the loss before announcing a problem. This research takes time. According to the IAEA, only about three weeks are necessary to convert PUO2 powder into a weapon [Leventhal 94].

                                                                                                           Revised 10/8/01