CALCULATING REMOVAL EFFICIENCIES - CONFIDENCE-LIMITS AND HYPOTHESIS TESTS

Emerging environmental control technologies are frequently characteriz ed in terms of their removal efficiencies. Based on a limited number o f tests, the environmental engineer must typically predict the true re moval efficiency along with its level of uncertainty. This paper provi des a ''road map'' (based on the number of non-detects, the underlying distribution of data, and whether data are paired or unpaired) for ca lculating such uncertainty. Both confidence intervals and hypothesis t ests are considered.