Operators of long natural gas pipeline networks with multiple gas compressor stations are looking for ways to optimize operations, reduce fuel consumption and costs, and maximize producer and consumer throughput in real time. This paper introduces a unique method to optimize these complex natural gas pipeline total value streams and achieves the stated objectives by integrating turbomachinery algorithms into a high-fidelity simulator to improve total model and simulation accuracy. Through much trial and error, the authors discovered how to account for the physical network size and how the combination of turbomachinery and gas pipeline algorithms as well as advances to auto-tuning methods may be used to solve the challenge of simulating complex natural gas pipeline networks in real time. Auto-tuning algorithms and the simulator results were validated with real field data from natural gas pipelines to ensure model and simulation accuracy. Technological advances incorporated into the pipeline simulator included a new parsing engine, advanced regulatory control, and incorporating turbomachinery algorithms for mixed mode operations (gas compressors driven by gas and electrical turbines as well as reciprocating engines). This paper describes our method and these advances and demonstrates how they may be applied to all natural gas pipelines to significantly improve operations, reduce costs, and improve the operating envelope.
Natural gas transmission pipelines transport large quantities of gas across long distances and deliver it to major consumers (local distributors, large industrial end users, electrical generation facilities). Natural gas is introduced into a pipeline transmission system at various points, such as LNG terminals, processing facilities near supply fields, and interconnections with other gas pipelines. This gas is transported in high-pressure pipelines and a series of compressor stations. Compressor stations provide the power required to transport the gas in the pipeline from one location to another and usually contain more than one compression unit. A unit is defined as a combination of a compressor and its engine. A gas compressor station may have a diverse combination of units, resulting in a more complex operational envelope not possible to solve using traditional optimization techniques. The methodology presented in this paper is unique in that it was designed to model diverse unit configurations and to generate optimization results based on multiple objective functions.