Analysis of the thermal-hydraulic system response of a PWR during loss of coolant accident and core blockage scenarios

Debris generated during the blowdown phase of a Loss of Coolant Accident (LOCA) is a big concern in the safety of the Light Water Reactors under such accident scenarios. Debris transported by the water flow through the reactor containment floor may reach the sump strainers. During the long-term cooling phase of the accident, the accumulation of debris in the sump screens could impact the Emergency Core Cooling System (ECCS) performance. Small size debris could also pass through the debris bed and sump strainers and be injected in the primary system, reducing the core cooling capabilities (downstream effects). A typical Westinghouse 4-loop Pressurized Water Reactor (PWR) was modeled using RELAP5-3D to simulate the reactor system response during the phases of the accident, under different LOCA scenarios. To account for the multi-dimensional flow behavior, the main regions of the reactor vessel were simulated using multi-dimensional components. The present work provides the results of the simulations performed for a LOCA in two of the main break locations of interest, such as the cold and the hot leg. The main objectives of these calculations are: • To validate a new RELAP5-3D nodalization, used to model the reference nuclear power plant, in which have been entered multi-dimensional components that provide a better estimation of the flow path inside the reactor vessel and through the core during long-term cooling when strong asymmetries are expected due to the different possible break-injection location combinations. • To study the response of the plant during selected accident scenarios and to provide important information and boundary conditions for the development of other project models, such as the break jet boundary conditions and reactor containment calculations. • To analyze the behavior of the reactor system during a hypothetical core blockage, due to the accumulation of small sizes debris at the core inlet that could preclude the passage of water flow into the core. The simulations were repeated for three different break sizes (2 and 6 inch, and Double-Ended Guillotine) as representative cases for small, medium and large break respectively, to provide a complete spectrum of the system behavior. Moreover, since the purpose of the calculation was to provide a best estimate evaluation of the system behavior, all the safety features of the reactor were assumed to be available during the transient and the boundary conditions defined in the input deck (including initial operating conditions, scram, injection and other setup points, ECCS flow rates, auxiliary injection rate and temperature, etc.) were assumed to be realistic. Basic thermal-hydraulic parameters such as primary pressure, break and SI flow rates, core coolant temperatures and maximum cladding temperature were selected to monitor the system behavior during the transient. The liquid inventory in the steam generators tubes, the refilling time and, subsequently the potential circulating flow through the cooling loops were found to play an important role during the phases of the accident. The results obtained also showed important features of the coolant flow paths inside the reactor vessel during the long term cooling phase, when different safety injection configurations (cold and/or hot leg injection) are possible. For a specific break size, the HL break cases showed a higher flow rate through the core since the flow is forced by the SI trains to pass through the core from the injecting cold legs to reach the break. The core mass flow rate for the CL break scenarios was observed to be limited since most of the injected flow left the vessel through the broken cold leg from the upper section of the downcomer. A decrease in the core flow rate was usually predicted at the HL injection switchover, as a result of a different injection path. As a consequence, an increase in the core coolant and peak cladding temperatures is expected during the early stage of this phase. Such features should be considered when studying the effects of the debris transport and deposition inside the reactor vessel during the long-term cooling phase of the accident, when water from the containment sump is pumped by the safety injection system into the primary system. Assuming an excessive accumulation of debris at the core inlet, a full blockage of the bottom of the core and core bypass was imposed at the beginning of the long-term cooling phase. Under such conservative conditions, scenarios that did not lead to potential core damage were identified. The peak cladding temperature was selected as representative thermal-hydraulic parameter to define whether possible core damage was achieved. For all the break sizes analyzed, hot leg break scenarios were found not to lead to core damage due to the alternative coolant flow path through the steam generators (from the cold leg injection site to the SG U-tubes and to the hot leg and top of the core) which guarantees the decay heat removal under core blockage. For medium and large break size, cold leg break was found to lead potential core damage due to the unavailability of alternative flow paths to the top of the core at the time or immediately after the core blockage. It has to be marked that the core damage, when achieved during the simulations, was found to occur within a short period of time from the sump switchover time (approximately 1000 s), while the HL injection switchover occurring much later in the transient. To take into account also of the cross flows existing inside the core, a new nodalization was developed replacing the two pipe components which constituted the core (the average and hot channel) with four multi-dimensional components with which the fuel assemblies were individually simulated. First, preliminary simulations of the blowdown and long-term cooling phases were performed with the same boundary conditions used in the previous simulations to validate the new nodalization. Analyzing the behavior of the 3D core model for small, medium and large break LOCA, in both cold and hot leg, the results obtained with the 1D core model described above were confirmed. After, additional simulations were performed assuming again the total blockage of the bottom of the core and core bypass at the beginning of the long-term cooling, focusing on those scenarios that lead potential core damage and particularly on the 6-inch cold leg break. Also this case showed what was previously found: the peak cladding temperature steadily increases reaching the maximum limit, confirming that the full core and core bypass blockage assumption may lead to core damage