Simulating small changes in system response by MCNP

Correlated sampling and perturbation are two developed techniques that enable MCNP to simulate small changes in system response with an acceptable uncertainty at reasonable number of particle histories. In this work the performance of correlated sampling is investigated. It is found that under certain conditions using the default output format of MCNP5 and the original practice of batch statistics developed for correlated sampling May overestimate the uncertainty of the change in system response. The cause of the overestimation is analyzed; correspondingly two improved procedures are proposed for correlated sampling to ensure correct estimation of the uncertainty. In addition, the performance of improved correlated sampling method is also compared with that of the perturbation technique of MCNP5. In terms of precision of the change in system response, the perturbation technique is found to be capable of producing smaller uncertainty than correlated sampling when response change is small; as response change becomes larger, the uncertainty predicted by correlated sampling is comparable with or even better than that predicted by the perturbation technique. In terms of accuracy of the change in system response, however, it is found that the perturbation technique May produce less accurate results than correlated sampling for certain problems that involve intermediate or even small changes in system response. Overall, correlated sampling seems to be a more robust technique than the current perturbation technique of MCNP5 for simulating small changes in system response. (authors)