There is a tendency to take modeling results at face value. The model reached a steady state; therefore, the solution is good. Even veteran modelers can be complacent and not notice that things have gone wrong with the model even though a solution has been obtained. Occasionally, simulators produce mathematically correct solutions, which can never be duplicated by the actual piping network or be operationally feasible. Looking at the graphical results my not indicate that there is a problem. The modeler must probe into the model results to determine the problem. The reason for the answers being unrealistic can be caused by model data, by the internal logic of modeling software, or by selecting the wrong type of simulator. Case studies will be presented to demonstrate possible pitfalls such as: poor compressor operation, unreasonable regulator operation, overriding pressure or flow constraints at supplies or demands, unrealistic line pack distribution. The case studies will demonstrate the results seen and the steps used to determine the operational feasibility of the solutions. The studies will be commercial product independent. The authors will present a checklist for modelers to use when verifying the validity of a steady state hydraulic solution.
Today in the pipeline simulation industry, the user is offered a multitude of products for problem solving. The user can choose from on-line, leak detection, multi-phase flow, transient, optimizer, flow assurance etc…. Even with all these options, the authors have found that the majority of users still work with off-line steady state models to do the majority of their simulations. Steady-state models have progressed to tools that contain state of the art graphical interfaces, menu options of solution techniques, and detailed equipment models. The line-by-line text entry and output interpretations of long ago (maybe 5 to 10 years!) are gone. With detailed pipeline schematics, data input prompts, solution relaxation techniques, warning flags and error messages, steady-state models have never been easier to use. With all these checks and balances, accurate data into the model will net the user a steady-state solution. However, are the results workable solutions for the pipeline? Often users can become so focused on a particular area of the pipeline solution, that they often fail to review all of the outputs from the simulation. Advancements in software allow users to build extremely large and complicated models. Users will concentrate on the particular area of the pipeline system where the flow adjustments are being made. Once a steady state is achieved, the user often accepts the answers as valid and does not review the entire system to see what effects these changes may have had through out the system. Users may have inadvertently constrained a part of the system that is a correct mathematical solution but not a valid operational mode for the pipeline. Models allow so much more flexibility for solving simulation problems that users need to review the data very carefully.