Most of the literature about network simulation or network analysis techniques, both steady state and transient has been devoted to discussions of the mathematical and programming techniques which are necessary to arrive at results. Some of the discussions have related to the various forms of output which are necessary to give the user a grasp of what can truly be called the mathematical model of the piping system. Network Simulation studies using digital techniques, got their start in 1955 with Sickafoose's paper on the application of the Hardy Cross method using the IBM 604 punched card calculator. As stored computer programs evolved more and more programs were developed and currently, for steady state analysis, there are available, very flexible programs for simulating all real conditions.
The initial thrust, considering the problems of data assembly, was the testing of the adequacy of what in distribution terms were called feeder or medium pressure systems which in turn fed district regulator stations feeding low pressure systems. These systems had little tolerance in terms of pressure and during the period of rapid growth of distribution systems, the question always "was do we need to reinforce?". Parallel with that were the attempts at describing the large low pressure networks fed from the district regulator stations. Again, the basic use was for design, to answer the question, "do we need to reinforce?". Following that came the question "what is the optimum reinforcement?" Reinforcement studies are studies to determine adequacies of existing systems. Generally tests assumed design loads which occur very infrequently. In order to be conservative the designer would normally test to somewhat higher loads than he expects. In the 70's and 80's as computer resources became more readily available, many other uses of network simulation for distribution systems became economically attractive since the cost of the solution and the elapsed time were both radically reduced. The same "do I need to reinforce question?" is still to be solved as loads change. Next in line comes the frequent question "can I take this line out of service?" either temporarily, during some foreign construction or are possibly permanently? Third, "I have the line out and the weather forecast is such and such, do I need to work overtime", "or can I leave the line out of service for a few more days?" More recently in many companies, especially on feeder systems, contingency studies are made. These are to test the consequences of each meter regulator station, pipeline segment or other major component being out of service. We attempt to find what are the limiting loads which a partial system can tolerate. In all these non-peak load cases, the engineer needs to know a lot more about loads at less than design conditions. The simplifying conservative assumption of assuming more load than is expected for the reinforcement calculation is no longer valid for the short term. Therefore, more and more questions are asked about the validity of the model.
