Evaluation of spatially clustered ordnance when using compliance sampling surveys after clean-up at military training sites

After site clean-up teams have removed all of what they believe to be UXO within a specific impact area, statistical compliance sampling is a possible method for verifying with a specified probability that this area has been cleaned to specifications. Schilling [J Qual Technol 10(2):47-51, 1978, Acceptance sampling in quality control. Marcel Dekker, Inc., New York, 1982] developed a compliance sampling methodology based on the hypergeometric distribution. Bowen and Bennett (1987) also use compliance sampling where they provide an approximation for estimating the number of samples (n) required to state with desired probability that the entire population of sample units (N, where n < N) are in compliance with cleanup goals. This article describes two methods (anomaly and transect) for applying the Schilling [J Qual Technol 10(2):47-51, 1978, Acceptance sampling in quality control. Marcel Dekker, Inc., New York, 1982] compliance sampling method to military training sites. After describing these methods, a simulation study is presented which demonstrates the performance of transect compliance sampling calculations based on varied degrees of clustered UXO within a specific impact area and different types of sampling routines.