(189ae) Optimal Sampling Locations to Reduce Uncertainty in Contamination Extent in Water Distribution Systems

Water facilities rely on samples collected from the distribution system to provide assurance of water quality and meet regulatory requirements. If there is a water quality concern, additional sampling can be conducted to determine the extent of the water quality event and the source of the concern. Finding the location of the contaminant injection is important for long-term decontamination and recovery, while determining the extent of contamination is important for short-term response. Using models for this purpose poses a challenge because significant uncertainty exists in the system hydraulics, contaminant reaction dynamics, and incident characteristics. Determining the overall extent of contamination is important to allow utilities to prioritize their recovery efforts.

This presentation outlines an optimization formulation to identify strategic sampling locations in the water network that can quickly reduce uncertainty in extent of contamination. We present two mixed-integer linear programs (MILP) that seek to identify the best location or locations in the network to gather additional measurements to quickly determine the characteristics of the contamination incident. Similar approaches have been proposed for solving the source identification problem, however, the novelty of this work is that it incorporates probability metrics within a rigorous optimization formulation. In addition, the mixed integer programming formulations presented in this work have greatly improved computational complexity over the previous formulations. The new formulation solves within seconds, and is applicable for larger water network models.

Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000