Online Monitoring Of Breaks In Urban Water Distribution Systems

A break in urban water distribution systems, with strong unexpected characteristics, not only will result in waste of water resource, but also may affect ordinary operation of the city. Up to now, break position and extent in water distribution systems are found mainly based on if there is water overflow (by residential reporting or artificial inspection) in large and moderate cities in our nation. Although existing SCADA systems can perform simple warning function, because the density of water pressure/flowrate stations is very low in water distribution systems in large and moderate cities in our nation, it is difficult to diagnose the break position and extent quickly and accurately.Based on the existing research work, this dissertation systematically studied online monitoring and diagnosis of breaks in urban water distribution systems, which includes hydraulic analysis in break status, online monitoring of breaks, optimal monitoring of flowrate stations, etc. The main innovations of the dissertation are followed:(1) A method for hydraulic analysis of an urban water distribution system in multiple breaks status is developed, in which leakage, low supply pressure, and pumps can be considered. The effect of position and extent of a break on node pressure was discussed.(2) A method for online monitoring of a break in an urban water distribution system is proposed. The method diagnoses break position, extent and influence area using neural network technique based on online monitoring of water pressure variation at several nodes in the water distribution system before and after the break. The method can be applied to online monitoring of breaks and other unexpected events in a water distribution system.(3) The second method for online monitoring of a break in an urban water distribution system is proposed. The method quickly and accurately diagnoses break position, extent and influence area based on water pressure variation at all non-monitoring nodes in the water distribution system, which are estimated by online monitoring of water pressure variation at several nodes before and after the breaks using neural network technique.(4) A method for optimal monitoring of folwrate in an urban water distribution system is proposed. The method establishes sensitivity equation of flowrate in the water distribution system by perturbation analysis, and resolves the optimal monitoring of flowrate by sensitivity analysis. The method not only can find the most sensitive pipe segment, but also can zone the water distribution system based on the sensitivity order of flowrate.