During the case study the intrusion of raw wastewater was simulated by addition of 1-3% of contaminant to drinking water flow in the pilot scale drinking water supply system which was supplied by the municipal drinking water supply system. To apply the contamination detection algorithms, long term safe drinking water quality monitoring was applied. More than 300 000 readings on water quality were consolidated to be used for AI training. Results showed that the system can detect a wastewater contamination event with a probability of 96.75% and recognizes the clean drinking water with a probability of 99.87%. The presence of biological contamination was validated by continuous flow cytometry measurements and grab sampling with laboratory testing.
