This study aims to systematically evaluate abandonment risk for a main control board (MCB) fire in nuclear power plants (NPPs). To achieve this goal, fire modeling for the MCB fire of the reference NPP using Fire Dynamics Simulator (FDS) was conducted and abandonment probability was estimated.
A decision tree was developed and the characteristics of main control room (MCR) and MCB were identified to systematically evaluate the abandonment risk for the MCB fire and to determine fire scenarios for it. Two types of fire scenarios were considered for fire simulations: non-propagating fire scenarios occurring within a single MCB panel and propagating fire scenarios spreading from one control panel to the adjacent panels. Further, the fire scenarios were classified into fires with and without heating, ventilation, and air conditioning system (HVACS). Three MCB panels for the fire simulations were selected as the fire initiation places.
The basic fire simulations results showed that the optical density was the major abandonment impact factor for the MCB fire of the reference NPP. The minimum heat release rate (HRR) causing MCR abandonment for the propagating fire was smaller than that for the non-propagating fire. Only the propagating fire contributed to the abandonment risk for the MCB fire of the reference NPP. Compared with the previous industry study results for the reference NPP, the MCB fire risk has decreased by approximately 66%. The previous study did not conduct fire simulations. Consequently, this study shows that the abandonment risk due to the MCB fire of the reference NPP can be reduced using the fire simulations. The sensitivity study results on validations for the basic fire modeling confirmed that the fire modeling approaches used for this study were reasonable.
The decision tree for evaluating fire risk proposed in this study can be confirmed that it was appropriately applied to the estimation of abandonment risk for the MCB fire through the basic fire modeling and the sensitivity study on the fire propagation time. For the case where the fire propagation time is ten minutes, only the propagating fire scenario is to be considered for the estimation of abandonment risk for the MCB fire. However, for the case where the fire propagation time is fifteen minutes, both non-propagating and the propagating fire scenarios are to be considered for the estimation of abandonment risk for the MCB fire.
The heat release rates used for fire modeling of this study came from NUREG/CR-6850. These are based on expert judgment with few fire experiments. Therefore, more fire experiments are needed to realistically estimate the HRR profiles of the MCB for the fire simulations. In addition, further studies on the development of a fire spread model for the propagating fire within the MCB are needed to reliably estimate the MCB fire risk.