在全球暖化氣候異常的今天,水不再是取用不盡的資源,自來水事業單位應積極進行漏水管理,提升供水效能及降低損漏情形。計量小區(District Metered Area, DMA)是2001年國際水協會(International Water Association, IWA)推動漏水管理策略與方法之一,歐美、日本及臺灣都已有不錯的推動績效。 本論文整理自來水國內外漏水管理的相關文獻,整理說明DMA起源、定義及理論發展,確認計量小區(DMA)是漏水管理重要工作。但管理小區則需要有效率的整合工具,在地理資訊系統(GIS)圖資及監控系統(SCADA)即時資訊外,另一項重要工具就是管網分析軟體,利用水量、壓力及管網拓樸,模擬不同的管網情形評估管理應對措施,可加強小區不確定性的風險管理。 住宅用水量為管網分析來源資料之一,有別於過去的研究,本文利用時間精度較高的小區監測流量,彌補時間精度不佳但空間精度較高的住宅水表抄表量,結合二種尺度流量與Johnson機率分佈,在質量守恆定律下,應用相關性非常態多變數蒙地卡羅法(CNMMCS),模擬計算住宅用水量每小時的用水情形,經由案例研討及小區應用的驗證,確認模擬用水量的模型分佈的可行性。 另外結合模擬水量及蒙地卡羅法,本文提出管網分析軟體EPANET在大量資料模擬增進效率的方法。經由各個節點的壓力變化及水頭損失的統計特性,探討小區內不同漏水情形發生的壓力變化,說明小區管理上的應用。
Climate anomalies in global warming today, the water are no longer the inexhaustible resources. Water Company should be actively managed in leakage control, improve performance of water supply and reduce the loss of water leakage conditions. Since 2001, District Metered Area (DMA) had been promoting to be one of the leakage management strategies and methods in International Water Association (IWA). Europe, Japan and Taiwan already have a good drive performance. We had searched the development of water pipe network, residential water demand, pressure, and leakage management of domestic and foreign literature. Regarding for the definition, theory, and development of the DMA are finishing instructions systematically. Confirm DMA is an important work of leakage management. But we need an efficient management of residential integration tools, geographic information system (GIS) and monitoring system (SCADA), but another important tool is the pipe network analysis software, the use of water, pressure pipe network topology to simulate different situations pipeline management response assessment, uncertainty can strengthen the risk management. As residential water demand for the pipe network analysis, unlike past research on water consumption, this paper use the data of flow meter with high precision, and the data of residential water meters with less time accuracy but the higher spatial precision, combined with this two kinds of flow data and the Johnson probability distributions under the law of conservation of mass, apply with Correlated non-normal multivariate Monte Carlo simulation (CNMMCS) and simulated the case of residential water demands per hour. Residential applications through case studies of verification, validation this simulated model of water distribution is feasible. We also presents a method in combination with water demand and Monte Carlo simulation, to increase efficiency of EPANET analysis in large data process. Through the statistical properties of pressure and head loss of nodes, to discuss the pressure change occurred in different leakage case on the DMA.