A novel monitoring and decision support system to manage water loss in urban water distribution systems

城市配水系统中的水损失是当今城市面临的主要问题之一。在2015年冬季，安大略的圭尔夫市经历了近30次管道爆裂，预计多伦多市的这一数字将超过1000次。二月，安大略基奇纳市的管道爆裂，导致圣玛丽医院的手术取消了好几天。据估计，15-25%的供水在城市供水系统中损失。老化的管道基础设施，加上市政预算下降、环境压力和气候变化，使加拿大及其他地区的供水基础设施面临重大失败风险。监测管网以检测水损失，能够预测管网中哪些管段故障风险最高，并能够做出实时数据支持的决策，这些都是使我们能够最大限度地减少水损失并维护城市供水系统中老化基础设施的关键因素。在这项研究中，我们将解决如何：（i）最佳地监测大型配水网络，以使用永久传感器网络管理水损失，（ii）预测可能会经历结构完整性损失的管道脆弱部分，并相应地计划干预措施，以及（iii）开发决策支持系统以帮助传感器数据可视化，泄漏检测以及在大型网络内最佳规划传感器位置。还提议在实验室和实地对这一系统进行试验性测试。该系统将在寒冷气候下全年提供数据并支持决策，这对加拿大尤为重要。拟议的研究绝不会解决配电系统中的水损失问题;然而，使用我们在这里提出的智能监测技术和分析将产生一系列新的丰富的科学证据，工具和理解如何更好地管理水损失和脆弱的基础设施问题。

Loss of water in urban water distribution systems is one of the major issues facing municipalities today. During the winter of 2015, the City of Guelph, Ontario, experienced nearly 30 pipe bursts and this number is expected to be over 1000 in the City of Toronto. The February pipe burst in the City of Kitchener, Ontario, resulted in the cancellation of surgeries for several days at the St. Mary's Hospital. It is estimated that 15-25% of supply water is lost in urban water distribution systems. Aging pipe infrastructure coupled with declining municipal budgets, environmental stressors, and climate change have put water distribution infrastructure in Canada and beyond at significant risk of failing. Monitoring the network to detect water loss, being able to predict which pipe sections of the network are at the highest risk of failing, and being able to make real-time data enabled decisions are all key factors that will allow us to minimize water loss and maintain the aging infrastructure in urban water distribution systems.In this research, we will address the problems of how to: (i) optimally monitor a large water distribution network to manage water loss using a permanent sensor network, (ii) predict vulnerable sections of the pipe which are likely going to experience a loss of structural integrity and plan interventions accordingly, and (iii) develop a decision support system to aid in sensor data visualization, leak detection and to optimally plan sensor locations within a large network. Laboratory and field-testing of this system on a pilot scale are also being proposed. The system will provide data and support decisions throughout the year in cold climates, which is especially important for Canada. The proposed research will by no means solve the problem of water loss in distribution systems; however, using smart monitoring technology and analytics we propose here will produce a new and rich body of scientific evidence, tools, and understanding on how to manage the problem of water loss and vulnerable infrastructure better.