Operational Optimization Of Urban Water Distribution Systems Based On Combined Control Of Pumps And Valves

In 2020,China’s total urban public water supply was 58.645 billion m3,while the water leakage was 7.854 billion m3,with an overall leakage rate of 13.39%.The energy consumption of pumps in urban public water supply accounts for 90%～98% of the whole water distribution network,and the energy consumption of the water distribution network accounts for 40%～70% of the water supply cost.The conjunctive operation of pumps and valves can reduce the network pressure,and pumps’ operational parameters can be optimized to solve the problem of leakage and energy consumption in water distribution networks.However,using pressure-reducing valves in the network may cause water quality problems.In order to reduce the energy consumption and the background leakage of water distribution networks and ensure water supply safety,this thesis proposed an operational optimization method for water distribution networks based on the joint regulation of pumps and valves.The method was mainly divided into three parts.(1)A water distribution network is partitioned through three clustering algorithms:the K-means clustering algorithm,the Hierarchical clustering algorithm and the Spectral clustering algorithm.It is determined that the K-means clustering algorithm performs the best,and the best partition is obtained.At the same time,the pressure-reducing valve is installed at the connected locations of each partition within the network.(2)The optimization model of the combined regulation of pumps and valves of a water distribution network is established.The objective functions are to minimize the pipe network background leakage,the pump energy consumption and the weighted average water age of the network.The variable frequency pumps’ hourly speed ratios and pressure-reducing valves’ hourly setting values are taken as decision variables.The constraints include the hydraulic balance equations,the lowest pressure at each demand node,the opening and closing settings of each pump,the range of each pump’s speed ratio,the load distribution mode of parallel pumps,the range of pressure-reducing valves’ settings and the upper and lower levels of each tank are taken.The Pareto-optimal solutions to the operational optimization model are obtained via the Non-Dominated Sorting Genetic Algorithm-Ⅲ(NSGA-Ⅲ).(3)For the obtained Pareto-optimal front,the good and bad results are ranked according to the predefined weighting coefficients through the Technique for Order Preference by Similarity to an Ideal Solution(TOPSIS).It provides decision-makers with optimal network operation schemes under different weights.The operational optimization method for water distribution networks proposed in this thesis was applied to the Anytown and the KY3 models with different water supply modes and scales.The results show that the weighted average water age of the Anytown pipe network was reduced by 20% on average,the energy consumption was reduced by4%,and the background loss was reduced by 23%.At the same time,it was found that the energy consumption in the Anytown network was inversely proportional to the weighted average water age and the background leakage.There was no clear trade-off between the background leakage and the weighted average water age.The weighted average water age of the KY3 network was reduced by 36%,energy consumption was reduced by 46%,and background leakage was reduced by 10%.At the same time,it was found that the background leakage of the KY3 network was directly proportional to energy consumption.By contrast,the background leakage was inversely proportional to the weighted average water age,and the weighted average water age was inversely proportional to the energy consumption.The operational optimization method of urban water distribution networks based on the conjunctive regulation of pumps and valves proposed was verified by different pipe networks.It can save energy,control leakage and ensure water supply safety,and clarify the complex trade-off among the operation cost of water distribution networks,background leakage and network water quality.It provides the basis for making the best regulation scheme with the most significant comprehensive benefits.