System of Systems Approach to Develop an Energy-Water Nexus Model Under Uncertainty

An energy-water nexus model is an interdependent systems approach to study two infrastructure subsystems, energy and water, simultaneously. In the first step of this thesis, a novel integrated model for water system planning which includes both water consumption planning and water post-consumption planning is proposed. The integrated water planning model has been developed in a way that it can be integrated with a proper power model in the second step of this thesis. Using the new water model, it was possible to find more cost-effective solutions to integrate the water consumption and post-consumption planning into a holistic model since they are interdependent. In the second step, a new integrated quantitative model to investigate the interdependence of energy system planning, water consumption planning, and water post-consumption planning is developed and analyzed. Outputs of the model include individual decisions for all the existing subsystems in the energy-water nexus. This is a data-driven model, meaning that it is developed based on categories of existing data available for energy and water systems. When adequate data are not available, the model relies upon ranged estimation. As a result, a new two-stage adjustable robust approach is also developed. The energy-water nexus model, which is categorized as an NP-hard problem, is solved using a novel heuristic day-to-day planning approach for long-term decision-making. The model developed is a robust, linear, mixed-integer, multi-period model, appropriate for recommending tactical, operational, and strategic decisions. In this model, hourly operational decisions are generated for all agents involved. This new robust integrated energy-water nexus model indicates decisions that result in lower power cut-off rate and lower total system cost compared to a separated management approach. Also, a new robust approach is used to make decisions that are valid given a reasonable amount of uncertainty, most importantly uncertainty in demand and efficiency. Additionally, managers using this robust approach will be able to adjust the system robustness to make different trade-offs between robustness and cost based on possible scenarios, budget constraints, and system reliability. The results of the day-to-day heuristic planning model are shown to be close to those from the exact solution method.