Investigating the role of decentralized water systems as strategies for urban water and wastewater management




Berhanu, Bruk M.

Journal Title

Journal ISSN

Volume Title



In the face of population growth and climate change impacts, water supply planners and utilities have increased interest in decentralized water systems as a means to meet potable water demands and improve water supply resilience in times of drought. Decentralized water systems (DWS) refer to technologies that capture, treat, and recycle alternative water supplies (rainwater/stormwater, graywater/blackwater, foundation drainage, air-conditioned condensate, etc.) to supply non-potable water for uses including toilet/urinal flushing, clothes washing, and irrigation on a building site. These systems offer benefits to individual users including water/wastewater rate savings, stormwater reductions, and resilience to water/wastewater service interruptions. Benefits can also extend to centralized water and wastewater networks by reducing water and wastewater treatment needs, preserving water supplies, and reducing net energy demands for water/wastewater treatment and delivery. However, these systems have not seen widespread adoption across the United States nor internationally. For utilities and water providers wishing to promote adoption of these technologies, little information is available regarded expected efficacy and cost metrics that can be used to design policies tailored for their service area. Moreover, current methods to assess these systems at-scale typically rely on assumptions of homogeneous single-family residential service areas and/or adoption rates that do not consider individual-level differences in land use characteristics, customer types, or adoption criteria. This research provides a framework by which to integrate these factors into assessments of decentralized water systems at the service area scale. The research is presented in three steps; an investigation into the drivers of indoor and outdoor water demands for residential and non-residential customers, integration of these demands via customer type into daily time-scale simulations of DWS performance under uncertainty, and application of psycho-social influences combined with technical and economic performance criteria to the decision to adopt DWS over long-range planning horizons. The non-residential sector has not seen the same level of investigation as the residential sector, and so a classification and statistical analysis model using mixed-effects linear regression was developed to relate parcel-level attributes such as industry type, conditioned area, and employment level to total monthly water demands. Results from this model suggest that parcel-level water demand differs most across industry classifications, while estimates for effect of parcel attributes on demand are more similar between classifications. Moreover, the predictor variables studied accounted for over 90% of variance seen in monthly demand values, after accounting for individual-level random effects. The results from the statistical analysis were then incorporated, along with residential demand estimates, into a daily water balance simulation model of DWS performance. Results suggest that tradeoffs exist between capital cost and cost-effectiveness of DWS adoption and technical performance via non-potable demand reductions. Hybrid rainwater-graywater systems and graywater for indoor and outdoor uses were both found to meet 100% of non-potable demand across uncertainty ranges in demands and alternative water supplies, though hybrid systems had higher capital costs primarily due to the need for a rainwater storage cistern. All systems incorporating graywater were capable of realizing net benefits under certainty conditions while solely rainwater based configurations almost never saw such outcomes. The results from the first and second analyses were then integrated into a hybrid agent-based and discrete-event sequence model to simulate adoption of DWS technologies by customers in a synthetically generated neighborhood. The model applies Systems-of-Systems analysis concepts to integrate individual-level decision making and behavior with aggregate-scale exogenous factors to understand how technical, economic, and psycho-social factors might affect adoptions rates of DWS and corresponding impacts to centralized water/wastewater infrastructure networks. Results suggest that non-residential customers are more heterogeneous in their adoption preference, with respect to DWS configuration, than the residential sector which primarily adopts graywater-based systems solely for outdoor use. Additionally, the tradeoff between capital cost and non-potable offsets for higher performing DWS configurations typically results in low adoption rates of these systems, but high per-system demand reductions that are commensurate with more widely adopted but lesser performing systems, such as sole rainwater for irrigation. These results highlight the importance of tailored DWS adoption decisions to adopter preferences and provide insight into possible policy instruments such as capital cost incentives, dual-reticulation mandates for new construction, and water demand offset requirements (without a technology prescription) that might promote adoption of these systems. When the results from these analyses are considered in context of a particular study area, this research provides a comprehensive set of methods by which to evaluate the opportunities for and potential impacts from parcel-level DWS adoptions from the perspectives of both the customer and the centralized water/wastewater service provider. Such a set of methods can be incorporated into long-range centralized water/wastewater infrastructure planning efforts to leverage the benefits of DWS adoption while managing the potential downsides


LCSH Subject Headings