A Two-Phase Model to Detect and Localize Water Distribution System Leakages
| Main Authors: | Min, Kyoung Won, Kim, Taewook, Choi, Young Hwan, Jung, Donghwi, Kim, Joong Hoon |
|---|---|
| Format: | Report |
| Terbitan: |
, 2020
|
| Subjects: | |
| Online Access: |
https://zenodo.org/record/3921773 |
| ctrlnum |
3921773 |
|---|---|
| fullrecord |
<?xml version="1.0"?>
<dc schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd"><creator>Min, Kyoung Won</creator><creator>Kim, Taewook</creator><creator>Choi, Young Hwan</creator><creator>Jung, Donghwi</creator><creator>Kim, Joong Hoon</creator><date>2020-06-29</date><description>Water Distribution Systems (WDSs) is to supply the required quantity of customer’s water demand under adequate pressure and acceptable water quality. Leakage in WDSs due to excessive pressure, pipe aging, and earthquakes leads to problems such as repair costs, disruption of water supply and economic losses. Adding the volume of water loss to customer’s demand increases overall pipe flow rates and head losses throughout the system, which finally result in a low pressure at withdrawal point and the degradation of system functionality. The goal of the Battle of the Leakage Detection and Isolation Methods (BattleDIM) is to propose a method to detect and pinpoint the leakage events in L-town in 2019, as fast and accurately as possible. The SCADA measurements of flow and pressure sensor is given with the repaired date of leakages events during 2018. This study presents a new approach of two-Phase: (1) detecting the period of the individual leakage events, (2) pinpointing leak locations. In Phase 1, the data (e.g., pipe flow, tank level, nodal pressure) selected from correlation analysis is provided to the K-means clustering algorithm and Western Electric Company rules, by which normal and abnormal period of times are determined. The leakage events are assumed as the previous period for the repair completion, by which, the performance is compared with respect to the detection results of the two techniques. In Phase 2, the sensitivity analysis of applying an emitter to each node is performed in the calibrated L-town network with the pipe roughness and demand pattern. The leakage location is identified with the minimum flow variation between the 2019 SCADA measurements and the results of the calibrated network applied an emitter.</description><identifier>https://zenodo.org/record/3921773</identifier><identifier>10.5281/zenodo.3921773</identifier><identifier>oai:zenodo.org:3921773</identifier><relation>doi:10.5281/zenodo.3921772</relation><relation>url:https://zenodo.org/communities/battledim2020</relation><rights>info:eu-repo/semantics/openAccess</rights><rights>https://creativecommons.org/licenses/by/4.0/legalcode</rights><subject>Leakage detection, Correlation analysis, K-Means clustering, Western Electric Company, Sensitivity analysis</subject><title>A Two-Phase Model to Detect and Localize Water Distribution System Leakages</title><type>Report:Report</type><type>Report:Report</type><recordID>3921773</recordID></dc>
|
| format |
Report:Report Report |
| author |
Min, Kyoung Won Kim, Taewook Choi, Young Hwan Jung, Donghwi Kim, Joong Hoon |
| title |
A Two-Phase Model to Detect and Localize Water Distribution System Leakages |
| publishDate |
2020 |
| topic |
Leakage detection Correlation analysis K-Means clustering Western Electric Company Sensitivity analysis |
| url |
https://zenodo.org/record/3921773 |
| contents |
Water Distribution Systems (WDSs) is to supply the required quantity of customer’s water demand under adequate pressure and acceptable water quality. Leakage in WDSs due to excessive pressure, pipe aging, and earthquakes leads to problems such as repair costs, disruption of water supply and economic losses. Adding the volume of water loss to customer’s demand increases overall pipe flow rates and head losses throughout the system, which finally result in a low pressure at withdrawal point and the degradation of system functionality. The goal of the Battle of the Leakage Detection and Isolation Methods (BattleDIM) is to propose a method to detect and pinpoint the leakage events in L-town in 2019, as fast and accurately as possible. The SCADA measurements of flow and pressure sensor is given with the repaired date of leakages events during 2018. This study presents a new approach of two-Phase: (1) detecting the period of the individual leakage events, (2) pinpointing leak locations. In Phase 1, the data (e.g., pipe flow, tank level, nodal pressure) selected from correlation analysis is provided to the K-means clustering algorithm and Western Electric Company rules, by which normal and abnormal period of times are determined. The leakage events are assumed as the previous period for the repair completion, by which, the performance is compared with respect to the detection results of the two techniques. In Phase 2, the sensitivity analysis of applying an emitter to each node is performed in the calibrated L-town network with the pipe roughness and demand pattern. The leakage location is identified with the minimum flow variation between the 2019 SCADA measurements and the results of the calibrated network applied an emitter. |
| id |
IOS16997.3921773 |
| institution |
DEFAULT |
| institution_type |
library:public library |
| library |
DEFAULT |
| collection |
DEFAULT |
| city |
DEFAULT |
| province |
DEFAULT |
| repoId |
IOS16997 |
| first_indexed |
2022-06-06T06:51:44Z |
| last_indexed |
2022-06-06T06:51:44Z |
| recordtype |
dc |
| merged_child_boolean |
1 |
| _version_ |
1739477202416697344 |
| score |
17.609869 |