A Mixed-Integer Optimization Model for Compressor Selection in Natural Gas Pipeline Network System Operations

This paper presents a Mixed-Integer Linear Programming (MILP) model to optimize the compressor selection opera- tions in natural gas pipeline network system. The objectives of natural gas pipeline network system operations are to minimize the operation costs and provide sufficient gas to the local customers. A pipeline network system is the most cost effective way for mov- ing fluid products over long distances. In this case, it is used for transmitting natural gas from a producer to customers. To ensure de- mand for natural gas can be met, a dispatcher turns on or off compressor(s) in order to increase or decrease the amount of natural gas in the pipeline system. Compressor selection is one of the most critical operations in the natural gas pipeline network system because the costs associated with turning on or off the compressor make up a large percentage of the total operating costs. In order to minimize the operating costs of the pipeline system, the three most crucial factors that affect the costs are integrated into the MILP model. The three factors include the capacities of compressors, the energy used to turn on the compressors, and the energy used to turn them off. The MILP model provides the decision support in determining the optimal solutions for controlling the compressors. It was developed and verified using the operation data supplied by a gas pipeline company in Saskatchewan, Canada.