Associate Professor The University of Texas at Austin
Many people worldwide face intermittent water supply (IWS) conditions in which water is not available 24/7. To overcome the challenges of IWS systems, consumers often use private storage tanks to reserve water for use during periods of non-supply. Consumers supplied by IWS often suffer from insufficient and unreliable water availability and local storage often exacerbate existing disparities in water supply, where some consumers have better access to water depending on their location in the system. The objective of this research is to analyze the improvement in local supply availability and global supply equity through strategically limiting the amount of flow allocated to different regions in the water network. We propose a volume-driven demand approach to modeling consumers with local storage tanks coupled with Bayesian optimization to determine the optimal setting of flow control valves that simultaneously optimize local and global water supply and equity. We assess the potential improvement under different intermittent supply schedules and analyze the mechanisms in which disparate hydraulic conditions cause inequity. The results demonstrate the existence of a hierarchy in supply in which consumers with favorable hydraulic conditions, e.g., low elevations and close proximity to source, are able to fill their tanks any time supply is available, leaving consumers with less-favorable hydraulic conditions to fill their tanks only when an excess in supply exists. Partial control of the network can offset some global disparities and produce more equitable distribution, however, the local supply hierarchy cannot be offset with limited controls, and will require significant investment that might not be feasible. Additionally, intermittency in supply to the network as whole exacerbates the inequity in supply and limits the efficacy of controls, reinforcing the importance of improving the continuity in supply. The proposed approach provides insights into the mechanisms behind inequitable supply in IWS systems, where further investment in research is needed to extend the model time horizon and introduce additional IWS system characteristics such as leakage modeling.
