It is well known that algorithms exist for reducing pipeline operating costs. These algorithms are exact for ideal pipelines and need to be modified to provide solutions for the real world. The issues include pipeline configurations, utility cost structures, and quantification of hydraulic safety. Successful modification requires understanding of the pipeline operating environment (on-line operations) and must be linked to pipeline operating conditions.
Many of the optimization tools available to the pipeline industry today are based upon a dynamic programming algorithm attributed to Bellman. The costs of unit operations are balanced with the energy absorbed in heat due to frictional and other losses. This is carried out in such a way as to reduce the massive computational effort of an exhaustive solution search to a manageable level. For a pedagogical treatment of the problem, this is adequate. However, there are many significant factors which need to be added into and around this basic calculation. First, an algorithm with electrical cost factors only cannot evaluate penalties associated with poor hydraulics choices. Demand grouping, parallel pipelines, large amplitude pressure cycles, look ahead, and unit cycling also can and should be included in a full analysis.
A modification to Bellman’s algorithm for non-linear pipeline configurations and electrical contracts will be developed and discussed in the context of a real-world petroleum pipeline operation.