Prediction of Thermal-Hydraulic Parameters for Autonomous Load Following LFR Core Using Surrogate Model

Abstract The concept of autonomous load following refers to the reactor’s ability to adjust core power through inherent reactivity feedback, without the need for control rod operation. One promising candidate for this concept is the lead-based fast reactor. Ensuring the safety of the reactor core is a primary concern during the load following process. Given the fixed geometry and similar boundary conditions, surrogate modeling is a suitable approach for calculating maximum temperatures under different quasi-static states. It provides near-accurate results in significantly less time compared to CFD analysis. In this paper, a parametric batch study is conducted using CFD to generate the training dataset with different inlet temperature, mass flow rate, and power. Various surrogate modeling methods are employed for training the model, including the Kriging method, the Gaussian Process method, the Radial basis function method, the Response surface method, and the neural network method. The root mean square error and maximum error of surrogate models are evaluated and compared. Among them, the neural network model demonstrates the best accuracy. Its RMSE is 0.84 K for maximum fuel temperature and 0.24 K for maximum cladding temperature, while the maximum error is 5.65 K for maximum fuel temperature and 0.99 K for maximum cladding temperature.