Assessment of statistical methods for converting biochemical oxygen demand and carbonaceous biochemical oxygen demand to total organic carbon in wastewater

Abstract

The 5-day biochemical oxygen demand (BOD 5 ) and carbonaceous BOD 5 (CBOD 5 ) tests are widely used parameters for monitoring wastewater. Total organic carbon (TOC) has many advantages over these tests. Wastewater utilities have conducted studies to modify National Pollutant Discharge Elimination System (NPDES) permits to allow TOC analysis; however, statistical methods vary across studies. This study examined parametric and nonparametric correlation and parametric, nonparametric, and nonlinear regression methods for analysing BOD 5 , CBOD 5 , and TOC concentrations collected for 1 year from seven wastewater treatment plants from the Metropolitan Water Reclamation District of Greater Chicago. Spearman ρ correlation and Theil-Sen regression on log-transformed concentrations were the most appropriate methods. Correlation coefficients were 0.83 or greater and regression residuals were as small as or smaller than the other two methods. This study demonstrated that nonparametric methods performed best for analysing non-normal data in seeking to incorporate TOC analysis into NPDES reporting.