An efficient and fast method to calculate integral experimental correlation coefficients – S2Cor

•A new method to calculate integral experimental covariance data is presented and discussed.•The combined Latin Hypercube and Random Monte Carlo sampling approach leades in combination with a scaling factor to a significantly faster and more economic calculation of integral experiment correlation coefficients.•The method and results are compared to the ones from a Random Monte Carlo approach.•Two series of critical benchmark experiments were chosen and the Pearson’s correlation coefficients of the neutron multiplication factors are calculated. Generating covariance matrices, for example for integral experimental data or a set of experiments in a validation suite, can become a heavily time and infrastructure consuming process. We propose a new Monte Carlo based method to derive covariances and the corresponding correlation coefficients. The method is based on a combination of Latin Hypercube sampling, Random Monte Carlo sampling and a scaling factor and reduces the calculation time significantly. We describe the method and compare exemplary results for some critical experiments with results from the Random Monte Carlo sampling only method. The proposed method is not limited to application to critical experiments but can generally replace any Monte Carlo approach to generate covariance matrices.