• Aucun résultat trouvé

7. ADDITIONAL WASTE MANAGEMENT

7.3. Impact of transmutation

Once separation–conditioning and target fabrication processes have been developed and demonstrated on a pilot scale, transmutation by irradiation in dedicated facilities can theoretically be envisaged. Since this implies also the

simultaneous development of dedicated irradiation facilities (FRs or ADSs), it will take several decades before this option could become available. However, the impact of the transmutation option starts as soon as the targets and/or fuels are ready for irradiation, which may be much earlier than the implementation of dedicated irradiation facilities. The safety and security of this option requires much attention.

(a) Once through transmutation of neptunium targets in PWRs results in a limited depletion of the initial content, by roughly 40–45%, and an overwhelming production of 238Pu and some higher plutonium isotopes.

Depending on the irradiation period (800–2460 EFPD), only limited (<18%) fissioning occurs. Transmutation in PWRs makes separated neptunium less accessible from the safeguards point of view and unusable in a nuclear explosive device. From the radiotoxic point of view this operation drastically increases (103–104) the radiotoxicity (238Pu) and generates (~8%) a long lived source term made of 239,240,241,242

Pu. The net reduction of the neptunium inventory depends on the total neptunium target depletion in the dedicated reactor and on the generation of neptunium in the uranium bearing fuel. An LWR MOX fuelled irradiation reactor is preferable to an LWR UO2 core. The irradiated targets need to be stored together with the spent MOX fuel.

(b) Multiple recycling transmutation of neptunium targets can substantially deplete the initial load, at the cost of a high neutron consumption and a correspondingly higher enrichment. Separate processing facilities for the partially depleted targets are necessary to limit the 238Pu concentration in the main stream of the reprocessing plant. About three cycles are required to reduce the inventory by ~95%. The separated fission products follow the vitrification route. The quantitative depletion of the 237Np inventory by transmutation reduces considerably the very long term dose to a human. However, in the short and medium term (up to 800 years) the transmutation increases the radioactive burden and the heat load of the waste. Increasing the heat load has a direct effect on the engineered storage technology, on the prolongation of the intermediate storage time and on the size of the disposal facility. During these recycling operations secondary alpha waste is produced, which also has to be disposed of.

(c) A fast neutron spectrum leads to a higher yield of fission products and a somewhat lower 238Pu and higher actinide formation, which is favourable from a radiotoxic point of view. The neutron economy (number of neutrons per fission) is much better in an FR or ADS than in a PWR, since the transmutation chain to achieve fission is much shorter. In other words, transmutation in a fast neutron spectrum is cleaner and does not

generate additional long lived actinides. However, the irradiation time to reach the same level of depletion is much longer.

(d) Transmutation of separated neptunium makes its handling and storage operations in engineered storage facilities more difficult. It increases the heat load of the waste and generates secondary waste. These drawbacks have to be weighed against the very long term benefit of decreasing the dose to humans in the very distant future (millions of years). However, it gives the present generation an ethical guarantee that the present nuclear legacy will not unduly burden future civilizations. By transmuting neptunium the very long term hazard will be decreased, but it still has to be compared with the natural actinide decay chains of uranium (depleted and reprocessed).

(e) Once through transmutation of americium targets in an LWR MOX fuelled reactor results in a similar composition pattern with respect to plutonium radionuclides (30% 238Pu and ~10% higher plutonium isotopes) but shows a much higher 244Cm generation (~8%) from 243Am.

The disappearance rate of 241Am is very high: about 80% in one single irradiation cycle. Since this radionuclide is the parent isotope of neptunium, transmutation in an LWR MOX reactor decreases signifi-cantly the very long term radiotoxic inventory. However, in the short and medium term the transmutation leads to a very hot target, which decays with the half-life of 244Cm (i.e. 18 years). Prolonged storage of the irradiated americium targets has to be envisaged and the radiotoxic burden will be determined by the generated plutonium isotopes. Disposal of such targets will require special conditioning techniques with a high heat removal capacity. Transmutation of americium and curium targets does not improve their safeguarding quality, since this is intrinsically a strongly irradiating mixture that cannot be manipulated without effective precautions.

(f) Multiple recycling of americium targets in LWRs is unlikely because the irradiated target is too hot for aqueous reprocessing and recycling. The presence of increased quantities of 244Cm is a major obstacle for this option. Pyrochemical multiple recycling of irradiated americium targets is a theoretical alternative that could be considered for a TRU mixture.

Secondary waste generation would be a serious issue because of the increasing curium content of the mixture after each recycle.

(g) Once through irradiation of americium in a moderated subassembly of an FR has been considered as a transmutation possibility that would not need further reprocessing. Elimination of 98% of americium would require a very long irradiation period (20 years) and leads to a post-irradiative mixture of 80% fission products and ~20% actinides, primarily

plutonium and curium. From a waste management point of view this option reduces the americium inventory on an atomic basis by a factor of five, but the 241Am (half-life of 432 years) is partially substituted by 239Pu (half-life of 24 400 years) and the long lived 243Am by 244Cm. The neutron economic advantage is preserved because of the fast reactor neutron economy, but the feasibility of operating a kind of EFR power reactor with many 241,242,243Am targets (2500 kg) needs further safety studies. The post-irradiation storage and disposal would be very similar to very high burnup FR MOX fuel.

(h) Multiple recycling of plutonium and MAs (TRUs) in FRs is theoretically possible. Irradiating TRUs in FRs has the advantage of reducing the inventory in proportion to the achieved burnup, independent of the TRU composition. A burnup of 18 at.% requires multiple recycling over at least five cycles to reach equilibrium between the in and out flux of TRUs.

The recycling operations would be carried out in small pyrochemical units associated with an IFR reactor fleet. As fissioning would prevail on capture, a smaller amount of higher actinides would be produced and the overall mean half-life of the residual TRU inventory would decrease.

Other waste forms would be introduced: an iron–zirconium matrix for metallic (fission and fissile) radionuclides together with the hulls and a sodalite–borosilicate glass mixture for the fission products. Very little has been published and most of the processes are still laboratory or pilot scale investigations.

(i) Multiple recycling of TRUs through ADS irradiation and pyrochemical fuel cycles is the latest of the proposed options. It implies the construction of a large ADS capacity (about ten times higher than for MAs). The advantage of this conceptual system is the relatively small size of the facilities required compared with those needed for aqueous reprocessing, the resistance against criticality risks and the proliferation resistance of the TRU mixture. These advantages are balanced by a lack of technical maturity of the irradiation system at the industrial scale and of processing facilities. The very high decay heat load of the treated spent fuel (192–455 kW/t HM) is compensated for by the small throughput, but the industrial feasibility of the process will have to be demonstrated.

Much R&D will be required to make the process comparable with FR technology.