Comparison of models for estimation of measurement uncertainties. Determination of BOD5 in wastewaters using a manometric method

Abstract

Biochemical oxygen demand after 5 days (BOD5) is a bioassay frequently used to determine the polluting strength of wastewaters. It quantifies the mass concentration of dissolved O2 consumed during an incubation period of 5 days at 20 °C, in the dark, by biological oxidation of organic and/or inorganic matter in wastewater. In this work, the BOD5 measurand is reported as γ(O2). Reporting analytical results with estimation of uncertainty is an ISO/IEC 17025:2005 requirement and allows results to be compared and their fitness for purpose to be demonstrated. The γ(O2) uncertainty was estimated using two models: the individual components’ model (model 1) and the quality-control-criteria model in which the trueness component is calculated using the equation for recovery from CRMs (model 2.A) and the standard deviation of all recoveries (model 2.B). Model 1 provides a realistic strategy for estimation of γ(O2) uncertainty and for assessment of performance. Models 2.A and 2.B are expeditious and simple procedures for the 100–800 mg L−1 range. When γ(O2) < 100 mg L−1 model 1 must be used because with both Models 2 uncertainty was underestimated, more drastically close to the limit of quantification.