A case study on the effect of smart meter sampling intervals and gap-filling approaches on water distribution network simulations

Water usage data collected from smart meters at the end user can improve the accuracy and applicability of water distribution network models. Collecting and storing large amounts of data across hundreds or more smart meters is costly, which makes it important to consider what constitutes a sufficient sampling interval. This paper explores the effect of varying sampling intervals in smart meter data on model performance in regard to flow, pressure and water age simulations. Furthermore, the effect of using linear interpolation, a demand pattern or a network-inflow-weighted approach to fill gaps when data are sampled coarsely, is investigated. The study was based on real data from 525 smart meters in a district metered area in Denmark. The results show that smart meter data can improve modelling results, and if the sampling intervals are coarser than 2 h, then a weighted gap-filling approach markedly outperforms linear interpolation and models with coarse bi-annual demand data.