Toxic event detection by respirometry and adaptive principal components analysis

Two methods based on adaptive principal components analysis (APCA) are compared to extract, from primary measurements, information related to the changes of wastewater characteristics induced by variable weather conditions and/or to the presence of toxic substances. The primary measurements are activated sludge respiratory data obtained by short-term experiments in an on-line batch respirometer, combined with indirect information on soluble pollution (UV-visible absorbance, turbidity, pH, etc.) and wastewater flow rate. The Benchmark Simulation Model 1 (BSM1), which simulates the functioning of a large wastewater treatment plant by activated sludge, has been used to obtain large data sets and to test the proposed APCA method, which has then been applied to real wastewater characteristics. Copyright (c) 2005 John Wiley & Sons, Ltd.