Technical Bulletin No. 0688: A Critique of Statistical Distribution Assumptions used by USEPA to Describe Paper Industry Effluent and Process Streams

Concentration and production-normalized mass loading data sets from paper industry facilities have been evaluated using probability plot regression techniques with conventional plotting positions. The results indicate that the single 2-parameter log-normal statistical distribution assumption is more ‘appropriate’ than the bimodal distribution assumption employed in the USEPA’s D-LOG procedure. Combined with the results of previously published studies, the implication of this finding is that the use of the D-LOG procedure for determining estimates of means, standard deviations, quantiles and variability factors will lead to biased judgments involving effluent guidelines development or permit compliance monitoring.