The model automatically performs all of the four steps mentioned, resulting in outputs as a prioritized list of likely high leakage locations.ĭescription of Selected DMA The selected service area is an established DMA called DMA 15-03-02 in Prachachun Area of the Metropolitan Bangkok. A unique feature of the model used in this paper is its ability of leakage hotspots optimization (Wu and Sage 2007) that accounts for pressure-dependent flow or leakage based on the FAVAD principle2.
In this paper, the model1 is created using off-shelf software that is user-friendly for user to undertake all the modeling tasks. Steps required for hydraulic modeling are: (i) model development, (ii) model calibration, (iii) model verification, and (iv) model application. The physical laws instantiated with data inputs are generally referred to as a “Mathematical Model”. Its required inputs include piping configuration of the service area (how are the pipes connected in the area), and water consumption (both in space and time). Water Network Analysis is an attempt to simulate flow in pipe networks based on mass balance and energy conservation laws. As step testing requires careful planning and is expensive to deploy, this paper provides a first attempt to assess the potential of water network analysis (WNA) software for leak localization. “Sounding”, an acoustic method of leak detection, is subsequently performed to pinpoint leak locations for excavation and repair. In modern practice of water leakage reduction, the observation of disproportionately high minimum night flow (MNF) initiates the execution of a step test, the result of which indicates pipe sections that may contain leaks. As the test involves closing of valves, continuous supply of water to consumers is interrupted and hence is carried out during the night hours to minimize service interruption. Conventionally, it is implemented by a field process known as a step test.
Leak Localization Leak localization is a process to identify and prioritize pipe sections with comparatively high leakages. Keywords: Leak Localization, Water Network Analysis
The accuracy of leak localization using the WNA software depends on the following factors: (i) Quality of model calibration (ii) Quality of input data Quantifications of these effects are the subject for further investigations. The results show that WNA modeling is a promising alternative method for leak localization. The model-predicted leakage hotspots are compared with actual step testing results. This paper reports initial experiences in using WNA modeling software for leak localization for a DMA in the Metropolitan Bangkok area. Hundreds of thousands of possible leakage locations are evaluated and the most likely leakage hotspots are found such that the model-simulated flows and pressures match as closely as possible with the field observed values. The field observed data are then imported into a state-of-the-art leakage detection optimization model to predict the most likely leakage hotspots. In addition to metering the inflow into the DMA, pressure loggers are also deployed into the field within the study DMA to collect the field pressures along with hydraulic boundary conditions.
Firstly, the water distribution system is partitioned into a DMA and a hydraulic model is constructed for the DMA. Recently, leak localization using Water Network Analysis (WNA) modeling software has been proposed as an alternative, so as to circumvent the aforementioned deployment issues of steptesting. Bangkok 10700, Thailand ** Bentley Systems, Incorporated, Watertown, CT 06795, USA *** Metropolitan Waterworks Authority of Thailand, Bangkok 10210 ,ThailandĪbstract Leak localization using traditional step testing involves extensive resource and equipment deployment, and creates water supply service interruptions during the step testing period. Areekul*** * Technology Service and Consulting 1656 Co., Ltd.
Experiences Using Water Network Analysis Modeling for Leak Localization S.