Analysis of the Resilience of Intermittent Water Supply Systems and the Disruption-Dynamics of Stakeholders
Millions of households around the world rely on intermittent water supply systems (IWS), where piped water supply is limited to specific hours during the day or on specific days during the week. Households relying on IWS systems, as their primary water source, often adapt to supply intermittency by installing in-house water storage and/or supplying water from non-piped sources (for instance, in the form of water tanker trucks). The piped water distribution network (WDN) in IWS systems is subject to short-term disruptions that cause dynamic behavior and interactions of the system’s stakeholders, including households, vendors of non-piped water, and the water utility. During disruptions of the WDN, households make decisions about obtaining water from different non-piped sources at different prices and wait times. These decisions, made by a large number of households, have an impact on the dynamics (in particular, the prices and availability) of the non-piped water market, which may in turn affect each household decision. Prior studies on the literature of the analysis of IWS systems focused on analyzing each of the components (namely the WDN, households, vendors of non-pied water, and the water utility) of the IWS in isolation, assuming static behavior of the other components.
The overreaching objective of this dissertation is to bridge the gap in knowledge and practice in analyzing the short-term dynamics within each component of the IWS system (focusing on the WDN and the households) and the interactions among all components of the IWS system when responding to physical disruptions of the WDN. First, a new framework for quantifying and analyzing the resilience of intermittent WDNs is presented. The framework incorporates the aspects of intermittent supply (including household storage and supply scheduling) into a hydraulic model that examines the network’s hydraulic performance and its topology to assess three resilience capacities: absorptive, adaptive, and restorative, against various types of physical disruptive events. The evaluation of the model, using the IWS network of a case study city in the Middle East, shows that household storage capacities, timing and length of the disruption, supply inequity, and the supply scheduling are significant factors in determining the resilience of the WDN, and the interactions of these variables result in different combinations of direct and post effects on households. The framework was also used to evaluate the impact of temporary modifications of the supply schedule on the network’s resilience. The results show that this short-term utility adaptive measure can significantly improve the resilience of the network. The proposed framework can assist utilities in the operation of the intermittent WDN under normal conditions and in the evaluation of the impact of different short- and long-term resilience enhancement strategies.
Next, based on empirical data from a survey of households in a city in the Middle East, the households’ decision-making in response to disruptions of the WDN was evaluated using econometric methods. A set of Binary Probit models were developed to model the decision of households regarding their risk attitudes toward running out of water (represented by the timing of their response actions), their willingness to pay for faster delivery of non-piped water, and their willingness to wait in-line to obtain water from a non-piped source. The results show how variables related to household characteristics, wealth, age and occupation of the household’s manager, knowledge of household manager about their households’ water situation, and prior experience with disruptions affect the households’ decisions when the piped-network is disrupted. The outputs of the econometric models can assist the city’s water managers in understanding the behavior of households that affect the demand and prices of different non-piped water sources.
The final component in this research integrates the two previous components into an Agent-Based Model (ABM) to evaluate the dynamics of the stakeholders’ interactions in response to disruptions of the WDN and to evaluate the impact of these interactions on the resilience of the whole system. The ABM examines the interactions between households and vendors of water tankers under utility’s policies that regulate the water tanker market while integrating variables that describe the response of the WDN to the disruption. The demonstration of the model using a representative subset of the IWS system in the case study city shows dynamic behavior patterns in: (a) the dynamics of households, and (b) the performance of the non-piped water market under different deterministic and stochastic scenarios of disruptions of the WDN.
The results of this research address many IWS systems in the Middle East and around the world that are characterized by household storage, as well as households’ dependency on the piped network as the main water supply. The models developed in this dissertation are expandable to adopt various systems’ configurations in terms of types and capacities of household storage, types and attributes of non-piped water sources, and attributes and preferences of households. The results of this doctoral research can assist water managers in cities in understanding the behavior of their IWS system (including the WDN and the system’s stakeholders), evaluating long-term resilience enhancement policies, and planning for short-term response to disruptions of the WDN.