Pump Scheduling for Optimised Energy Cost and Water Quality in Water Distribution Networks

Abstract

Delivering water to customers in sufficient quantity and quality and at low cost is the main driver for many water utilities around the world. One way of working toward this goal is to optimize the operation of a water distribution system. This means scheduling the operation of pumps in a way that results in minimal cost of energy used. It is not an easy process due to nonlinearity of hydraulic system response to different schedules and complexity of water networks in general. This thesis reviewed over 250 papers about pump scheduling published in the last 5 decades. The review revealed that, despite a lot of good work done in the past, the existing pump scheduling methods have several drawbacks revolving mainly around the ability to find globally optimal pump schedules and in a computationally efficient manner whilst dealing with water quality and other complexities of large pipe networks. A new pump scheduling method, entitled iterative Extended Lexicographic Goal Programming (iELGP) method, is developed and presented in this thesis with aim to overcome above drawbacks. The pump scheduling problem is formulated and solved as an optimisation problem with objectives being the electricity cost and the water age (used as a surrogate for water quality). The developed pump scheduling method is general and can be applied to any water distribution network configuration. Moreover, the new method can optimize the operation of fixed and variable speed pumps. The new method was tested on three different case studies. Each case study has different topography, demand patterns, number of pumps and number of tanks. The objective in the first and second case studies is to minimise energy cost only, whereas in the third case study, energy cost and water age are minimized simultaneously. The results obtained by using the new method are compared with results obtained from other pump scheduling methods that were applied to the same case studies. The results obtained demonstrate that the iELGP method is capable of determining optimal, low cost pump schedules whilst trading-off energy costs and water quality. The optimal schedules can be generated in a computationally very efficient manner. Given this, the iELGP method has potential to be applied in real-time scheduling of pumps in larger water distribution networks and without the need to simplify the respective hydraulic models or replace these with surrogate models.