The use of meters to handle demands for water utilities was discussed in my last blog. The recommended approach is to use our meter connection element to represent the large users. Residential customers which normally use around 80% or more of the total demand are handled by identifying the number of residential connections for appropriate pipes and defining an average demand for each connection. For example, if a water utility has 10,000 residential connections and the average demand is 0.8 gpm/connection then 8000 gpm will be distributed throughout the distribution system for the average demand model. This demand distribution will be based on the number of residential connections for each pipe and is divided equally to the end nodes for each pipe. Of course, determining the number of residential connections for each pipe in the model could be a very a time consuming task. As always, the task of demand distribution is one of the daunting task facing the modeler.
Following several years of experience in handling residential demands as described above we see a pattern for the location of residential demands. It is relatively easy to identify residential areas and, generally, smaller pipes are supplying most of the demand to the various subdivisions. There is a definite inverse relationship between the pipe size and the density of connections in residential areas. Large pipelines usually have no direct residential connections. Also, it is easy to identify residential regions of the water distribution system. We utilized these patterns to formulate an algorithm to distribute the residential demands based on pipe size and location. The user need only to input the number of residential connections and average residential demand and Pipe2000 will distribute the connections to the distribution system pipes based on pipe size and location. Setting up this Demand Distribution module takes only minutes and provides an accurate demand distribution.
How well does this work? We tested this on a number of good sized distribution system models and always got pressures for the average demand conditions that agreed within 1-2 psi. to the results based on time consuming detailed demand distribution. These models had the residential demands distributed by the usual tedious process of inspecting the distribution system maps. So the bottom line is that you can now use our new Demand Distribution feature (included in Professional Version) to accurately and effortlessly distribute the majority of the demand in your water utility model with just a few minutes of effort. You then will only need to account for the few large users using a metered connection for these customers to complete the demand distribution for your water utility. Thisis yet another very useful and time saving Pipe2000 development