Sludge settleability detection using automated SV30 measurement and comparisons of feature extraction methods

Abstract

The need for automation and measurement technologies to detect the process state has been a driving force in the development of various measurements at wastewater treatment plants. While the number of applications of automation & measurement technologies to the field is increasing, there have only been a few cases where they have been applied to the area of sludge settling. It is not easy to develop an automated operation support system for the detection of sludge settleability due to its site-specific characteristics. To automate the human operator’s daily test and diagnosis work on sludge settling, an on-line SV30 measurement was developed and an automated detection algorithm on settleability was developed that imitated heuristics to detect settleability faults. The automated SV30 measurement is based on automatic pumping with a predefined schedule, the image capture of the settling test with a digital camera, and an analysis of the images to detect the settled sludge height. To detect settleability faults such as deflocculation and bulking from these images, two feature extraction methods were used and their performance was evaluated.