WEF/WERF Cooperative Study of BNR Plants Approaching the Limit of Technology: II. Statistical Evaluation of Process Reliability

The Water Environment Federation (WEF) and the Water Environment Research Foundation (WERF) cooperated in a comprehensive study of nutrient removal plants designed and operated to meet the “Limit of Technology” (LOT) for nutrient removal. LOT is loosely defined as plants meeting either 3 mg/L TN and 0.1 mg/L TP. This effort was focused on maximizing what can be learned from existing full scale plants. In the first year of the project, eleven exemplary plants were studied. The objective of this paper was to describe the common statistical methods employed to analyze final effluent data from each treatment plant and to describe alternative methods for assessing process reliability and the relationship between reliability and permit limits. The degree to which nutrient removal efficiency is related to effluent quality variability was also considered. It is clear from this work that calculating the probability or reliability of achieving a given permit limit is useful. If the data reasonably follow a log-normal distribution, these calculations are relatively straightforward and can be accomplished in a standard spreadsheet package. If not, it is a relatively trivial matter to calculate probabilities (percentiles from the data) and to develop log-transformed probability plots from which the reliability of complying with a given treatment objective can be determined. Equally useful is evaluating process performance under an appropriate set of averaging conditions (e.g. daily data, 30-day moving average, annual moving average) at a selection of probability values that have important meaning with respect to permit compliance (e.g. 50, 91.8, 95%). This concept immediately lends itself to employing a statistical basis for permitting as well as defining the Limit of Technology or Technology Achievable Limits (although this terminology may need further refinement). Although water quality protection must be the focus of point source nutrient permitting efforts, nearly all discharge permits applied to treatment plants in the United States require near-100% reliability; the consequence of not achieving this level of reliability is a permit violation. Based on this limited study of 11 plants approaching the LOT, deterministic permit limits may not be appropriate for plants achieving very low nutrient limits, particularly when the limit is technology- rather than water quality-based.