Safety study for HTR conceptual designs under German siting conditions. Phase I B, specialized volume I

The basic methodology for determining sequences of events and their frequencies (events and fault trees) does not differ significantly from that of other risk studies. This applies analogously to the treatment of statistical data uncertainties and the description of results in the form of expected value with uncertainty factor. System unavailabilities are determined by means of failure rates, most of which originate from the German Risk Study, and consecutive test intervals. Unlike in other risk studies, common mode failures of components of the same kind are being considered by a mostly 10% fraction of the overall failure of the multi-train system (β-factor). A multitude of planned or unplanned operator actions are identified in the study. They are assessed using models from AIPA and according to WASH-1400. HTR-specific aspects allow mitigating operator actions in the range of days, which are approximately covered by subjective estimates, and extensive reversibility of human errors. British experience with gas-cooled reactors proved to be useful for HTR-specific components. Rates of 0.2 to 1 for small leaks and 1.5 x 10-3 per reactor-year for larger leaks (tube ruptures) are derived on the basis of 2000 steam generator operating years. Failures of the main blowers (0.1 per blower-year) are covered by other transient events. The behaviour of structural components is of great significance for the progression of core heatup accidents. The liner of the reactor pressure vessel and the concrete located behind will fail over a large area due to decreasing strength at temperatures above 8000C. A rupture of closure plugs may be virtually precluded. This also applies to a failure of the reactor containment at internal design pressure. The ultimate strength will only be reached at pressures of more than 14 bar. (orig.)