Abstract
A water distribution system model may be developed for different purposes and the required prediction accuracy is not necessarily the same for each goal. As such, the number of data necessary to adequately calibrate a system for different purposes varies depending upon the intended use. To examine the impact of model purpose on the calibration process, a real system is modelled for two objectives: daily pump scheduling and system expansion design. Results indicate that an operation model can be calibrated with limited data since the sensitivity to decisions is small. A system expansion model requires more data since the design demand condition causes critical low pressures and model pressure predictions are strongly influenced by calibrated parameters.
Acknowledgement
The authors appreciate the constructive input from the reviewers and the editor to improve the clarity and presentation of this manuscript.