[en] The computer code REACT/THERMIX is used for analyses focussed on graphite corrosion processes. A correct reactor shutdown is assumed; the mass of water ingressing into the primary circuit is varied between 1000 and 7500 kg (that means up to hypothetical values). The dependence of accident consequences on parameters as intensity and starting time of the afterheat removal system or kinetic values of the chemical processes is examined. The results show, that also under pessimistic assumptions the extent of the graphite corrosion is relatively low; significant damaging of the fuel elements or of the graphite components does not occur. A primary circuit depressurization in combination with a local burning of water gas would probably not affect the fission product retention potential of the (gas-tight) containment. Summing up, the risk caused by these accidents remains small. (orig./HP)

[en] The AVR pebble bed reactor (46 MW_th) was operated 1967-88 at coolant outlet temperatures up to 990 C. A principle difference of pebble bed HTRs as AVR to conventional reactors is the continuous movement of fuel element pebbles through the core which complicates thermohydraulic, nuclear and safety estimations. Also because of a lack of other experience AVR operation is still a relevant basis for future pebble bed HTRs and thus requires careful examination. This paper deals mainly with some insufficiently published unresolved safety problems of AVR operation and of pebble bed HTRs but skips the widely known advantageous features of pebble bed HTRs. The AVR primary circuit is heavily contaminated with metallic fission products (Sr-90, Cs-137) which create problems in current dismantling. The amount of this contamination is not exactly known, but the evaluation of fission product deposition experiments indicates that the end of life contamination reached several percent of a single core inventory, which is some orders of magnitude more than precalculated and far more than in large LWRs. A major fraction of this contamination is bound on graphitic dust and thus partly mobile in depressurization accidents, which has to be considered in safety analyses of future reactors. A re-evaluation of the AVR contamination is performed here in order to quantify consequences for future HTRs (400 MW_th). It leads to the conclusion that the AVR contamination was mainly caused by inadmissible high core temperatures, increasing fission product release rates, 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 yet be equipped with instruments. The maximum core temperatures are still unknown but were more than 200 K higher than calculated. Further, azimuthal temperature differences at the active core margin of up to 200 K were observed, probably due to a power asymmetry. Unpredictable hot gas currents with temperatures >1100 C, which may have harmed the steam generator, were measured in the top reflector range. After detection of the inadmissible core temperatures, the AVR hot gas temperatures were strongly reduced for safety reasons. Thus a safe and reliable AVR operation at high coolant temperatures, which is taken as a foundation of the pebble bed VHTR development in Generation IV, was not conform with reality. Despite of remarkable effort spent in this problem the high core temperatures, the power asymmetry and the hot gas currents are not yet understood. It remains uncertain whether convincing explanations can be found on basis of the poor AVR data and whether pebble bed specific effects are acting. Respective examinations are however ongoing. Reliable predictions of pebble bed temperatures are at present not yet possible. (orig.)

[en] Accidents which have to be considered are core heat-up, reactivity transients, water of air ingress and primary circuit depressurization. The main effort of this paper belongs to water/air ingress and depressurization, which requires consideration of fission product plateout under normal operation conditions; for the latter it is clearly shown, that absorption (penetration) mechanisms are much less important than assumed sometimes in the past. Source term estimation procedures for core heat-up events are shortly reviewed; reactivity transients are apparently covered by them. Besides a general literature survey including identification of areas with insufficient knowledge this paper contains some estimations on the thermomechanical behaviour of fission products in water in air ingress accidents. Typical source term examples are also presented. In an appendix, evaluations of the AVR experiments VAMPYR-I and -II with respect to plateout and fission product filter efficiency are outlined and used for a validation step of the new plateout code SPATRA. (orig.)

[en] Automatic translation: Safety considerations for high-temperature reactors must include the risks of water ingress into the core as well as air ingress and the penetration of process gas into the core. Water and air react with the reactor graphite at the accident temperatures according to the reaction equations: Water: C + H20 CO + H2 Atmospheric oxygen: C + 02-*C02 C + i/202 - “CO Further reaction of the reaction products with the graphite (H2 reacts with graphite to form methane, C02 according to the Boudouard reaction: C + C02 --* 2 CO) or with the gaseous starting product (CO + H20 C02 + H2) must be taken into account when assessing the overall course of the reaction. Of great importance for the characterization of the safety characteristics of a high-temperature reactor is the answer to the question of whether a significant release of fission product due to corrosive exposure of the fuel area occurs in individual fuel elements in the aforementioned accidents. This risk could exist in particular in the case of the exothermic and quite rapid graphite/air reaction. Other hazards associated with graphite corrosion in HTR are the possibility of the formation of ignitable mixtures in the containment (primarily CO/air and H2/air mixtures), as well as the possible impairment of the mechanical load-bearing strength of some graphite components; it should also be noted that the criticality conditions can also be influenced by corrosive destruction of the boron carbide in the reflector graphite, for example. The special kinetics of graphite corrosion is strongly determined by some structural parameters of the graphite (e.g. BET area, porosity, labyrinth factor) as well as by the change of these structural parameters with the burn-off condition; when considering graphite corrosion, it must also be taken into account that graphite is not a completely homogeneous material, but consists of components of different reactivity (filler, binder, impregnation). The current state of knowledge on the corrosion of reactor graphites is largely limited to corrosion under normal reactor operating conditions (i.e. low corrosion gas concentrations); Extensive material is available for the graphite/H20 reaction under normal operating conditions; in contrast, the graphite/air reaction has hardly been investigated on reactor graphites to date. The aim of this report is to provide an overview of the theory and experimental results on graphite corrosion; the content of this report follows on from the internal report by K.-J. Loenißen /l/. Translated with DeepL.com (free version) Translated with DeepL.com (free version)

[en] Analyses on air ingress in the pebble bed reactors PNP-500, THTR-300, HTR-Modul, AVR-II and AVR process heat plant are outlined; in addition, some results for the VHTR with block type fuel are given. Air ingress requires primary circuit depressurization and large leak(s) to reactor buildings and environment and belongs therefore to highly hypothetical events in the sense of classical safety analysis. One accident class examined is air ingress with forced flow by emergency cooling: For this case, the range of mass flow/air content in cooling gas has been evaluated, in which safe core cool down is possible resp. long term core burning occurs; for highest available emergency cooling flow, a safe cool down of the THTR-300, which has no reactor building, is possible for up to 20 vol-% of air in the cooling gas, wheras low flow allows only for about 5 vol-%. If the amount of available air is restricted to the content of a reactor building, as is examined for the PNP-500, relevant consequences have not to be expected; this remains also true for forced convection flow, if burning of CO, formed by graphite oxidation, within the building is considered. For the second accident class examined, air ingress with natural convection flow by chimney draught as studied for the HTR-Modul and some other concepts, the time span until significant fission product release begins has been determined; in case, that the bottom reflector is hot at accident start (> 600 C) and therefore consumes most of the ingressing oxygen, this time span is at least several hours and leak tightening counter measures may be possible. (orig./HP)

[en] This report makes the attempt to summarize the state of knowledge with respect to the modeling of fission product release in HTR, and to define a suitable proceeding for the pre-phases of the safety analyses for HTR-100, HTR-Modul, and HTR-500. The behaviour of relevant fission product nuclides during normal operations and its release trails during the 'classical' HTR accident scenarios core heatup, water ingress, and depressurization (including the accidental impact on the surroundings) are considered. For describing most of the essential stations of the release trail, calculation models and corresponding input data are existent. Incomplete knowledge in this field is specified as a point of further research. (orig.)

[en] Automatic translation: With regard to the adsorption processes, the SPATRA code is based on that of Kress and Neill / 1 / developed the approach. Volume effects can be based on a diffusive transport from the adsorbed state into the bulk phase become; a comprehensive description of the different approaches to Consideration of volume effects can be found in 121. Chemical reactions of Fission products can be modeled with one another with SPATRA Code / 3 /. In older Versions of the code was also the 'penetration model' used to describe Volume effects built in.

[en] A reevalution of experimental data for the kinetics of the reaction between oxygen and nuclear graphites at temperatures 970 K - 1170 K is presented. This reevalution mainly covers an improved handling of the influence of boundary layer diffusion (mass transfer) on the measured rates. It is demonstrated, that a remarkable influence of boundary layer diffusion exists down to rather low temperatures. Rates for the in-pore diffusion controlled regime, which have been presented earlier, are therefore corrected to higher values which increase with increasing temperature. The maximum correction factor is 1.6. (orig.)

[en] In the course of analyses of severe air ingress accidents in high temperature gas-cooled reactors the present report deals with the kinetics of the reactions between oxygen and the graphitical fuel element matrix materials A3-3 and A3-27. Kinetic experiments covering a temperature range of approx. 950 to 1200 K (approx. 680 to 930⁰C) and a range of oxygen partial pressures between 1500 and 15000 Pa (0,015 and 0,15 bar) at a constant total pressure of 1,5 bar (carrier gas Helium) are presented. Their results are comprehended to Hinshelwood-Langmuir-like correlations. These correlations give maximum reaction rates, as they are valid for that graphite burn-off field, in which the reaction rate no longer increases with burn-off. The experiments, however, give some evidence about the burn-off dependence. A clear influence of the temperature on the increase factor of the reaction rate can be realized: the increase factor declines with rising temperature. A comparison of the experimental results with literature data, which, however, touch the above mentioned parameter range only at its periphery and thus had to be extrapolated, established satisfatory results. (orig.)

[en] The AVR pebble bed reactor (46 MWth) was operated 1967-1988 at coolant outlet temperatures up to 990 deg. 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 pre-calculated. Further, azimuthal temperature differences at the active core margin were observed, as unpredictable hot gas currents with temperatures > 1100 deg. 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 and 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 deg. 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 and D program for solution of unresolved problems related to operation and design basis accidents. These problems are listed in the paper. (authors)