Unreasonably long test time and test cost forces the utilization of test compaction methods in production line. Test compaction methods reduce the test cost at the expense of degrading the test quality. When test compaction is used, it is essential to estimate the resulting test quality. Traditional Monte-Carlo simulation devotes most of the effort sampling the median region of the process parameters. However, defective escapes are generally marginal and accurate estimation of defective parts per million (DPPM) requires extensive simulation, especially when DPPM level is low. In this work, we aim at reducing the number of simulations required to estimate DPPM accurately through a two-step methodology exploiting the layered structure of process variation. In the first step, we generate an essential experiment set using a modified version of Taguchi's design of experiment method. We optimize this experiment set for accuracy in order to get a minimal set of experiments. In the second step, we emulate the low level process variation on the optimized essential experiment set. Instead of using traditional Monte-Carlo sampling method, employing layered sampling of process parameters enable us to achieve an accurate DPPM value for a substantially reduced number of simulations.