Measurement and prediction of VOC emission rates from wastewater treatment plants

Volatile organic compound (VOC) emissions from wastewater treatment plants have attracted attention from both the public and regulatory agencies regarding health concerns associated with plant workers and potentially affected residents. The 1990 Clean Air Act Amendments (CAAA) started regulating wastewater treatment plants. According to the CAAA, facility that emits 25 tons of a combination of two or more toxic pollutants per year or 10 tons of a single toxic pollutant per year in a severe non-attainment area is classified as a major source of toxic pollutant emissions.;Conventional methods for determining the source status use the sample average, usually obtained from a small sample size, to estimate an annual emission rate. The method compares this value with the criteria of major source to determine the source status. This conventional method may not reflect the true population values. A different approach is developed for this study of water pollution control plant (WPCP) emissions; the new approach is based on inferential statistics and it estimates the population values using sample values of measured VOC emission rates from 12 WPCP in New York City. This method establishes a large safety margin when determining source status of wastewater treatment plants.;A new predictive statistical model of VOC emission rates was developed using measured VOC emission rates and plant operating data (POD). The statistical model for predicting VOC emission rates is a simple and efficient method for estimating VOC emission rates from wastewater treatment plants. The model was developed using 75% of 181 emission rate measurements from 12 wastewater treatment plants. These measurements and corresponding plant operating data, routinely and easily measured by plant engineers, are used to formulate a multivariable model that predicts VOC emission rates of the 12 wastewater treatment plants in New York City. The remaining 25% of the measured emission rates were used to evaluate the model with a database that was not used in formulating the model. Evaluation of the model supports a robust model prediction ability for 60 to 80 percent of the emission rate variation that can be explained by the variation of the predictive variables.;It is concluded that none of the 12 wastewater treatment plants falls into the criteria of major source even with the extreme estimation UCL...