Abstract

Monitoring algorithms, such as the Shewhart and Cusum control charts, are often used for monitoring purposes in the chemical industry or within an environmental context. The statistical properties of these algorithms are known to be highly responsive to measurement errors. Recent studies have underlined the important role played by the twocomponent measurement error model in chemical and environmental monitoring. In the present work, we study the effects of the twocomponent error model on the performance of the X and S Shewhart control charts. Results reveal that gauge imprecision may seriously alter the statistical properties of the control charts. We propose how to reduce the effects of measurement errors, and illustrate how to take errors into account in the design of monitoring algorithms.