IIIT Hyderabad Publications |
|||||||||
|
GIS Framework for Water Pipeline Leakage Detection using Sensor InputsAuthor: Nancy Nancy Date: 2020-06-02 Report no: IIIT/TH/2020/54 Advisor:K S Rajan AbstractWater is one of the most essential resources on this planet, of which the potable water for human consumption and use is quite limited. This valuable resource is distributed around to the population though a system of interconnected pipelines, mostly situated underground. There is a high possibility of leaks in such pipelines, which is difficult to detect, given that the system is mostly underground. Since water is a very scarce resource and has to be used judiciously, leaks in the pipeline system will prove detrimental to the world. Thus, there is a pressing need to devise solutions that can help detect leaks in such underground pipelines. Though impossible for the naked eye, advancements in technology have paved the way for development of effective monitoring mechanisms. With the advancements in sensor technologies, flow parameters can be measured without manual intervention even when these are relatively small. Leakage in pipes, however minor can cause huge loss of water if not detected early. Since a leak causes change in the flow rate and pressure in the pipeline, these changes, however small, can be picked up by some of these advanced sensors. While monitoring systems help in the identification of changes and there might be ways at addressing it, it is important to devise methods that can help detect and identify the location of water pipeline leaks. The proposed method takes inputs from water flow sensors for a given water distribution network to estimate changes to the hydraulic conditions, if any, at equilibrium and leak induced conditions. This method leverages the electronic sensor data to detect anomalies and flags the corresponding pipeline in the network. The sensors are installed as part of the water distribution network. The entire network is loaded in a Geospatial platform, like QGIS, to help visualize both the distribution network and leak locations. From a users ease point of view, a plugin was developed over the open source GIS platform of QGIS. The plugin helps trigger leak detection based on the sensor input over the spatial water distribution network. The proposed method has been tested over two lab-scale test beds of water distribution network with varying number of joints and pipes. In addition, using these two networks, the system performance was evaluated for different leak locations and varying number of leaks. Overall these cases consisted of different sensor locations, variable number of sensors, different leak locations and variable number vi vii of leaks. The leak detection system can be run in two modes - one being the on-demand mode and the other being the continuous monitoring mode. In both the modes the accuracy of the calibrated flow rate was compared with the sensor data at equilibrium and was found to be satisfactory. It is expected that the method proposed in this work will help water-pipeline managers to detect and locate the leaks in the pipeline networks, enabling them to take necessary measures to avoid wastage of water. Full thesis: pdf Centre for Spatial Informatics |
||||||||
Copyright © 2009 - IIIT Hyderabad. All Rights Reserved. |