The U.S. NRC is currently evaluating nuclear plant responses to Generic Letter (GL) 96-06, “Assurance of Equipment Operability and Containment Integrity During Design-Basis Accident Conditions” [1]. GL 96-06 is concerned with potential two-phase flow and water hammer conditions that could be present in the cooling water systems of nuclear power plants during design-basis accidents. Nuclear power plants rely on large capacity service water pumps to supply cooling water flow, via an extensive pipe network, to heat exchangers such as room coolers, pump lube oil coolers, and containment air coolers (CACs), for normal and abnormal plant operation. Following a postulated a loss of offsite power (LOOP) event, the normal electrical power supply to the service water pump would be lost resulting in a 20 to 30 second cooling water flow interruption while a diesel generator is started and the service water pump load is sequenced onto the diesel generator. In power plants, such as the Davis-Besse Nuclear Power Plant with open service water systems that draw from a lake or a river and supply safety-related CAC heat exchangers located 30 to 40 feet above the pump outlet, this could lead to cold water column separation in the heat exchanger supply and return piping. If a loss of coolant accident (LOCA) occurs coincident with the LOOP, then boiling in the CAC heat exchanger tubes could occur, as well. Upon restoration of the cooling water flow, dynamic loading could be expected as steam condenses and water columns rejoin. The TREMOLO computer program [2,3] has been used to calculate dynamic thermal hydraulic response and reaction forces in service water piping systems for several nuclear power plants in response to GL 96-06. A consistent result obtained in each of these GL 96-06 analyses is that the LOOP + LOCA scenario produces the bounding loads rather than the LOOP-only scenario. This result seemingly contradicts current industry thinking which suggests that because the water columns are colder and the void fraction lower during LOOP-only scenarios, the LOOP-only loads should be bounding [4,5,6]. While the physics supports the conclusion that the rejoining of colder water columns will generally yield the largest water hammer pressure rise, when actual plant geometry and credible accident scenarios are analyzed, a different picture emerges. This paper couples insights obtained from the GL 96-06 TREMOLO analysis of the Davis-Besse Nuclear Power Plant with independent hand calculations and experimental evidence to support the conclusion that the LOCA+LOOP scenario will produce the bounding loads in service water piping systems.

This content is only available via PDF.
You do not currently have access to this content.