EPANET Calibration Checklist v3.2

Calibration of an EPANET model is the iterative process of refining model inputs—such as pipe roughness (Hazen-Williams C or Darcy-Weisbach f), nodal demands, pump curves, tank levels, and valve settings—until simulated outputs (e.g., pressure, flow, chlorine residual) align within acceptable tolerances with observed field data. Version 3.2 of the checklist emphasizes traceability, uncertainty awareness, and stakeholder alignment: it mandates documentation of all assumptions, data sources, measurement uncertainties, and calibration decisions at each step. The checklist is organized into four sequential phases—Data Verification & Preparation, Network Topology Validation, Hydraulic Calibration, and Water Quality Calibration—with built-in quality gates requiring sign-off before progression. It integrates best practices from AWWA M31, EPA guidance, and ISO/IEC 17025 principles for computational modeling, including sensitivity analysis, parameter identifiability assessment, and goodness-of-fit metrics (e.g., Nash–Sutcliffe Efficiency, RMSE, R²). Crucially, v3.2 introduces explicit guidance on handling temporal variability (e.g., diurnal demand patterns), sensor placement bias, and the use of automated calibration tools (e.g., EPANET-MSX with PEST++ or EpanetR) while retaining manual verification requirements to prevent overfitting.