Depending on the actual energy flux on the Be PFCs in ITER during ELMs, melt damage may or may not occur. For Type I ELMs, which are compatible with the ITER divertor lifetime (-10 MJ ELMs16,18), the expected energy flux on the main chamber in ITER will be in the range of 2-3 MJ. The area of the wall over which this flux will be distributed is —30-60 m2, for a toroidally symmetric energy deposition. This leads to ELM energy fluxes —0.02-0.08 MJ m~2 on the main chamber wall, which will cause no Be melting at all. If toroidal asymmetries and/or poloi — dal structures dominate the ELM energy deposi­tion on the first wall, a substantial reduction of the first-wall effective area for energy deposition is expected (by a factor of —5). In this case, the ELM energy fluxes on the first wall would be 0.1-0.4MJm~2, which can cause up to 18 pm of
melting, lasting ^300 ms.209 Figure 21 shows the results of an analysis carried out with the code described in Raffray and Federici.21 , The erosion lifetime, expressed in number of ELMs or corresponding ITER full power pulses (approxi­mately 700 ELMs/pulses for a Be target initially 10 mm thick) is found to sharply decrease above a certain ELM energy threshold. Depending on the duration of the ELM event, the threshold energy density varies between 0.2 and 0.7 MJ m~2. For com­parison, the results of a W wall are also shown. More recently, analysis has been carried out using more sophisticated modeling tools and the results are described elsewhere.212

From the JET Be divertor experience, we expect that only a very small part of the melt layer produced during each ELM will be mobilized (typically <5%) and may lead to a Be influx into the plasma. Larger ELM energy fluxes onto the Be wall in ITER are indeed possible and would lead to serious pro­blems for the use of Be as main plasma PFC in ITER, both because of lifetime issues and because of plasma contamination. However, for the argu­ments explained above, a regime with repetitive ELM energy loads which are not compatible with the lifetime of the Be main chamber wall in ITER is not compatible with the ITER divertor lifetime either and will not be the reference regime of ITER operation. The development of techniques that can either eliminate or greatly reduce ELM energy losses
without significantly degrading confinement have therefore been recognized to be critically important for successful operation of ITER and have stimulated further worldwide research on ELMs mitigation.213