TY - JOUR
T1 - Exploring the key facets of leakage dynamics in water distribution networks
T2 - Experimental verification, hydraulic modeling, and sensitivity analysis
AU - Zaman, Dina
AU - Gupta, Ashok Kumar
AU - Uddameri, Venkatesh
AU - Tiwari, Manoj Kumar
AU - Sen, Dhrubajyoti
N1 - Publisher Copyright:
© 2022 Elsevier Ltd
PY - 2022/8/15
Y1 - 2022/8/15
N2 - Hydraulic solvers are extensively used for the real-time operational analysis and asset management of water distribution networks. However, the influence of leakage on network hydraulics is still not fully understood, and limited experimental verification concerning leakage modeling and calibration efficacy of hydraulic solvers is presently available. To address these limitations, this research develops a novel strategy to quantitatively assess the efficacy of the popular hydraulic solver, EPANET 2.0 in emulating the leakage dynamics through an extensive experimental study conducted on a realistic sensor-equipped network prototype followed by comprehensive leakage modeling and analysis. The network response to diverse leak scenarios is experimentally analyzed, and the inferences derived are utilized to configure the leak models. A stage-wise calibration routine of the base and leak models is executed, which exhibits reasonable accuracy with mean absolute errors of 0.011 and 0.060, respectively. The calibrated leak models manifest satisfactory performance concerning the simulated system pressure and effectively predict the leak flow trends at all leakage levels. The sensitivity matrix of the leak models also captures the variation in the localized and global impact of leak levels on system flow and pressure concerning different leak scenarios. Thus, this research experimentally verifies the efficacy of leakage modeling using EPANET 2.0 and delineates the potential challenges concerning efficient calibration and performance analysis of leak models using the predefined leak parameters. The implications of this study can be instrumental for utility managers and researchers in configuring improved leak models using hydraulic solvers and strategizing model-based leakage management plans.
AB - Hydraulic solvers are extensively used for the real-time operational analysis and asset management of water distribution networks. However, the influence of leakage on network hydraulics is still not fully understood, and limited experimental verification concerning leakage modeling and calibration efficacy of hydraulic solvers is presently available. To address these limitations, this research develops a novel strategy to quantitatively assess the efficacy of the popular hydraulic solver, EPANET 2.0 in emulating the leakage dynamics through an extensive experimental study conducted on a realistic sensor-equipped network prototype followed by comprehensive leakage modeling and analysis. The network response to diverse leak scenarios is experimentally analyzed, and the inferences derived are utilized to configure the leak models. A stage-wise calibration routine of the base and leak models is executed, which exhibits reasonable accuracy with mean absolute errors of 0.011 and 0.060, respectively. The calibrated leak models manifest satisfactory performance concerning the simulated system pressure and effectively predict the leak flow trends at all leakage levels. The sensitivity matrix of the leak models also captures the variation in the localized and global impact of leak levels on system flow and pressure concerning different leak scenarios. Thus, this research experimentally verifies the efficacy of leakage modeling using EPANET 2.0 and delineates the potential challenges concerning efficient calibration and performance analysis of leak models using the predefined leak parameters. The implications of this study can be instrumental for utility managers and researchers in configuring improved leak models using hydraulic solvers and strategizing model-based leakage management plans.
KW - Calibration
KW - EPANET 2.0
KW - Emitter coefficient
KW - Emitter exponent
KW - Pipe roughness
KW - Sensitivity analysis
UR - http://www.scopus.com/inward/record.url?scp=85130562374&partnerID=8YFLogxK
U2 - 10.1016/j.jclepro.2022.132236
DO - 10.1016/j.jclepro.2022.132236
M3 - Article
AN - SCOPUS:85130562374
SN - 0959-6526
VL - 362
JO - Journal of Cleaner Production
JF - Journal of Cleaner Production
M1 - 132236
ER -