Determining pipe groupings for water distribution networks

Mathematical modelling of a water distribution system allows comparisons between design and operation alternatives. However, to make meaningful comparisons, the true field system must be represented within the model. Water distribution systems contain a large number of pipes with unknown roughness coefficients. Even with an extensive data collection effort, pipe roughnesses for all links cannot be determined exactly. Therefore, the system is simplified by assuming sets of pipes have the same roughness coefficient. The impacts of such simplification have not been examined in a quantitative manner. This work develops a methodology to quantify impacts introduced by system simplification and identify the best number of pipe groupings for a network.