Sustainable energy management of water distribution systems through optimised pump scheduling

• Main Author: Menke, Ruben
• Other Authors: Stoianov, Ivan ; Parpas, Panos
• Published: Imperial College London 2017
• Subjects: 333.79
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.724194

Climate change, caused by anthropogenic greenhouse gas emissions, is one of the most important issues of our time. It is often overlooked that up to 5% of a cities electrical energy consumption is required to pump water in the water distribution system. This thesis describes how pump scheduling can contribute to a more sustainable energy management of water distribution systems. The analysis is enabled though a new convex formulation of the pump scheduling problem, which provides theoretical optimality bounds, a requirement for the comparison of different energy management strategies. The problem formulation is then expanded to include variable speed pumps. The first energy management method investigated is the provision of demand response to the grid by adapting pump operating schedules. Demand response from water distribution systems is shown to provide reserve energy at lower cost and with less than half of the greenhouse gas emissions 240gCO2e/kWh of the competing technologies. Installing variable speed pumps in a network is shown to increase the possible applications of demand response and increase the revenue potential. Next, the trade-offs between cost and emission reduction are analysed in a range of energy supply scenarios in the UK, in the current grid as well as possible future scenarios. It is found that greenhouse gas emission cost must rise significantly, above £100 t/CO2, before affecting scheduling decisions. Furthermore, it is shown that the pump utilisation rate, the share of time the pump operates, is key to profit from possible changes in the energy supply. Finally, the operating schedule of pumps powered in part by an on-site wind turbine is optimised. This analysis calculates the cost of uncertainty in the wind power forecast in renewable energy generation. A stochastic optimisation method is presented that reduces the cost of uncertainty.