Science Inventory

Optimal sampling locations to reduce uncertainty in contamination extent in water distribution systems

Citation:

Rodriguez, S., M. Bynum, C. Laird, D. Hart, K. Klise, J. Burkhardt, AND T. Haxton. Optimal sampling locations to reduce uncertainty in contamination extent in water distribution systems. Journal of Infrastructure Systems. American Society of Civil Engineers (ASCE), Reston, VA, 27(3):1-31, (2021). https://doi.org/10.1061/(ASCE)IS.1943-555X.0000628

Impact/Purpose:

Drinking water utilities need to quickly determine which areas of their system might be contaminated during an incident. Sampling is an approach that could help identify these areas. This work describes a framework for optimally identifying sampling locations using a water distribution system model when considering uncertainties in the contamination incident as well as model parameters. The problem formulations developed in the research could help water utilities more effectively select sampling locations during an emergency. Anyone interested in assisting drinking water utilities in responding to contamination incidents would benefit from this work, in particular water utility staff using water distribution models.

Description:

Drinking water utilities rely on samples collected from the distribution system to provide assurance of water quality. If a water contamination incident is suspected, samples can be used to determine the source and extent of contamination. By determining the extent of contamination, the percentage of the population exposed to contamination, or areas of the system unaffected can be identified. Using water distribution system models for this purpose poses a challenge because significant uncertainty exists in the contamination scenarios (e.g., injection location, amount, duration, customer demands, contaminant characteristics). This article outlines an optimization framework to identify strategic sampling locations in water distribution systems. The framework seeks to identify the best sampling locations to quickly determine the extent of the contamination while considering uncertainty with respect to the contamination scenarios. The optimization formulations presented here solve for multiple optimal sampling locations simultaneously and efficiently, even for large systems with a large uncertainty space. These features are demonstrated in two case studies.