Options
2008
Journal Article
Title
Brittle failure assessment of a PWR-RPV for operating conditions and loss of coolant accident
Abstract
The brittle failure assessment for the reactor pressure vessel (RPV) of a 1300 MW pressurized water reactor was revised according to the state of the art. The RPV steel is 22 NiMoCr 37 (A 508 Cl. 2). The expected neutron fluence at the end of license (EOL) after 32 years of full operation is Phi < 2.3 x 10(18) neutrons/cm(2). The assessment followed a multibarrier concept to independently prove the exclusion of crack initiation, crack arrest, and exclusion of the load necessary to advance the arrested cracks through the RPV wall. Thermal and structural analyses of the RPV were performed both for the reactor shutdown with postulated upset conditions, as the most severe load case at operation, and for loss of coolant accident (LOCA) conditions. For LOCA transients, a leak size screening of different combinations of cold/hot leg injection of emergency core cooling was performed, and the leading leak size was determined. A fracture mechanics based assessment was carried out for extended circumferential flaws in the weld joint between the RPV shell and the flange, as well as for axial flaws in the nozzle corner. These flaw geometries postulated at locations of the highest principal stresses and lowest temperatures under the respective transient conditions are representative for the brittle failure assessment of the whole vessel. For a normal operation, the maximum crack driving force takes place at high temperatures preceding the upset conditions. The transient follows a load path decreasing with temperature, producing a warm prestressing effect, which is considered in the assessment. Thus, a large safety margin against crack initiation can be demonstrated. At LOCA, the most severe conditions are determined for postulated cracks in the nozzle corner. Here, applying the constraint modified master curve, which takes account of the low stress triaxiality in the component, the exclusion of crack initiation is proven. Furthermore, two additional safety barriers are proven, the crack arrest after postulated crack initiation well within the allowable depth, as well as the preclusion of the load necessary to advance the arrested crack through the RPV wall.