Confidence interval procedures for Monte Carlo transport simulations

Monte Carlo simulations performed with radiation transport codes such as MCNP are extensively used to model radiation transport. Resulting sampled tally distributions can be extremely right-skewed. This skewness hampers valid confidence interval (CI) formation, lengthening the time needed to obtain a sample representative of the entire underlying distribution. The relevant concern is whether the number of histories n is large enough so that the standardized sample mean follows an approximately Gaussian distribution; i.e., when the conditions of the central limit theorem are satisfied. When these conditions are satisfied, CIs are valid in that they cover the true mean at the proper rate