Determining the spatio-temporal relationship between water quality monitors in drinking water distribution systems

Abstract

A novel method to both assess the strength of connectivity and determine hydraulic transit times between water quality monitors from time series data is reported. It was developed using a network of over 50 mobile multi-parameter sensors deployed for 18 months across a UK drinking water distribution system, and then validated using a network of 18 sensors from a different UK utility. Correlation coefficients are calculated at different time shifts for each possible sensor pair, with strength of connectivity represented by the highest correlation coefficient, and with the temporal lag of this highest correlation also designates the transit time. The results demonstrate the potential to derive valuable spatio-temporal information, with potential to increase understanding of system performance and connectivity. This information can be used to assist with further analytics such as tracking water quality events and improving hydraulic and disinfection residual decay modelling.