The nuclear power industry is increasingly identifying applications of machine learning to reduce design, engineering, manufacturing, and operational costs. In some cases, applications have been deployed and are realizing value, in particular in the higher volume and data rich manufacturing areas of the nuclear industry. In this paper, we use machine learning to develop metamodel approximations of a computationally intense safety analysis code used to simulate a postulated loss-of-coolant accident (LOCA). The benefit of an accurate metamodel is that it runs at a fraction of the computational cost (milliseconds) compared to the LOCA analysis code. Metamodels can therefore support applications requiring a high volume of runs, for example optimization, uncertainty analysis, and probabilistic decision analysis, which would otherwise not be possible using the computationally intense code. We first generate training data by running the safety analysis code over a design of experiment. We then perform exploratory data analysis and an initial fitting of several model forms, including neighbor-based models, tree-based models, support vector machines, and artificial neural networks. We select neural network as the most promising candidate and perform hyperparameter optimization using a genetic algorithm. We discuss the resulting model, its potential applications, and areas for further research.