Currently, centralized, large-scale systems are the dominant form of water and energy supply in the US. While these systems provide essential functions for human and societal well-being, they have often been cited for issues related to infrastructure aging, high vulnerability to natural and manmade threats, and lack of adaptability to intermittent but renewable water and energy resources. On the other hand, decentralized, household water and energy systems have been increasingly implemented across the nation and the world. Such implementations, however, were often not conducted in a coordinated fashion, which can result in sub-optimal outcomes in terms of sustainability and resiliency at the city scale. Motivated by these limitations, this project studies how we can best integrate decentralized, household water and energy systems into the existing centralized water and energy supply networks to improve sustainability and resiliency in cities. A modeling framework that combines choice experiment, agent-based modeling, system dynamics modeling, and spatial optimization was proposed to tackle the complexities of integrated water and energy management consider the interactions among infrastructure, humans, and natural resources.

A choice experiment was first conducted to understand how different people make choices on different types of household water and energy systems based upon their attributes. We identified six attributes that are important to people?s choices: system type, ownership (own or share with neighbors), upfront cost, annual cost savings, environmental benefits, and neighbor?s choice (or peer pressure). Through a crowdsourced platform called Amazon Mechanical Turk, we asked people to make choices between two systems based upon hypothetical, yet realistic descriptions of the six attributes for each system. We collected responses from two testbed areas: Boston, MA and Atlanta, GA. We then conducted a latent-class analysis to identify social groups with similar preferences of decentralized systems and socioeconomic characteristics. The analysis also allowed us to develop models to predict the likelihood of a certain individual to adopt a certain type of decentralized system. Through this analysis, we found six different social groups have distinct preferences on decentralized water and energy systems in each city. The social groups are unique to each city. Overall, Boston was found to have a larger population that belongs to the undiscerning early adopters than Atlanta, suggesting Boston would be expected to adopt decentralized water facilities at a faster rate once introduced. In both cities, households are more likely to adopt a decentralized water facility if their neighbors already install one, and if households have the perception of water scarcity, they are willing to share the investment of a decentralized water facility within the community. Our findings suggest support from the public sector to help initialize the adoption of decentralized water facilities will accelerate the diffusion. The spatial visualization of the distribution of different classes highlights the areas of early demand for decentralized water facilities appear in downtown Atlanta and southeastern Boston.

We also examined the scientifically optimal sizing and siting of the decentralized water and energy systems in the two cities. This is achieved through developing integrated, dynamic water-energy system models at both centralized and decentralized scales and applying optimization algorithms to the models for each household in the two cities. Overall, we found greywater recycling (GWR) systems offered greater cost, energy, and carbon savings than the rainwater harvesting (RWH) systems throughout Boston if adopted. However, optimal system sizing and the maximum achievable savings for the same household setting can vary significantly across cities. The lowest cost and energy consumptions are achieved in northwest Boston, where population density is the highest. Southeastern Boston, on the other hand, is one of the least preferred areas for decentralized water system adoptions. This finding highlights a discrepancy between the ?preferred? adoption scenario based upon people?s choices and the scientifically optimal adoption scenario. Hence, policies and incentives might be necessary to guide the future adoption of decentralized systems for the best sustainability and resiliency outcomes.

This work has resulted in the two collaborating institutions. The project will inform decision making about possible outcomes and tradeoffs in different decentralized water and energy adoption scenarios, and critically guide policies and incentives design. The proposed approach can be used to facilitate the planning and design of decentralized systems and create a more sustainable and resilient infrastructure system for urban communities. Through this project, 2 PhD students, 2 MS students, and 1 undergraduate student have been recruited and trained. 4 out of the five students are women. Knowledge and survey instruments that have been generated through this project have been incorporated in two undergraduate courses with over 130 enrollments combined. A learning module has been developed based upon the system dynamics model used in this study, which will be tested in a graduate course at UNH and be disseminated to systems analysis community. Outcomes from this project have been disseminated through 4 journal publications and 9 conference presentations.

Last Modified: 12/27/2019
Modified by: Weiwei Mo