The AVR pebble bed reactor (46 MWth) was operated 1967–1988 at coolant outlet temperatures up to 990°C. Also because of a lack of other experience the AVR operation is a basis for future HTRs. This paper deals with insufficiently published unresolved safety problems of AVR and of pebble bed HTRs. The AVR primary circuit is heavily contaminated with dust bound and mobile metallic fission products (Sr-90, Cs-137) which create problems in current dismantling. The evaluation of fission product deposition experiments indicates that the end of life contamination reached several percent of a single core inventory. A re-evaluation of the AVR contamination is performed in order to quantify consequences for future HTRs: The AVR contamination was mainly caused by inadmissible high core temperatures, and not — as presumed in the past — by inadequate fuel quality only. The high AVR core temperatures were detected not earlier than one year before final AVR shut-down, because a pebble bed core cannot be equipped with instruments. The maximum core temperatures were more than 200 K higher than precalculated. Further, azimuthal temperature differences at the active core margin were observed, as unpredictable hot gas currents with temperatures > 1100°C. Despite of remarkable effort these problems are not yet understood. Having the black box character of the AVR core in mind it remains uncertain whether convincing explanations can be found without major experimental R&D. After detection of the inadmissible core temperatures, the AVR hot gas temperatures were strongly reduced for safety reasons. Metallic fission products diffuse in fuel kernel, coatings and graphite and their break through takes place in long term normal operation, if fission product specific temperature limits are exceeded. This is an unresolved weak point of HTRs in contrast to other reactors and is particularly problematic in pebble bed systems with their large dust content. Another disadvantage, responsible for the pronounced AVR contamination, lies in the fact that activity released from fuel elements is distributed in HTRs all over the coolant circuit surfaces and on graphitic dust and accumulates there. Consequences of AVR experience on future reactors are discussed. As long as pebble bed intrinsic reasons for the high AVR temperatures cannot be excluded they have to be conservatively considered in operation and design basis accidents. For an HTR of 400 MWth, 900°C hot gas temperature, modern fuel and 32 fpy the contaminations are expected to approach at least the same order as in AVR end of life. This creates major problems in design basis accidents, for maintenance and dismantling. Application of German dose criteria on advanced pebble bed reactors leads to the conclusion that a pebble bed HTR needs a gas tight containment even if inadmissible high temperatures as observed in AVR are not considered. However, a gas tight containment does not diminish the consequences of the primary circuit contamination on maintenance and dismantling. Thus complementary measures are discussed. A reduction of demands on future reactors (hot gas temperatures, fuel burn-up) is one option; another one is an elaborate R&D program for solution of unresolved problems related to operation and design basis accidents. These problems are listed in the paper.
Initial prediction of dust production in pebble bed reactors