Design of load sharing algorithm in a compression station by MPC method and real-time optimization in response to variable demand

To transport the natural gas from the producer sources to the consumers, gas pipelines are used. The extension of the pipelines reduces gas pressure and compression stations are considered to compensate for the loss of pressure in the transport path. At these stations, several compressor units usually work in parallel to share the gas inlet flow. The purpose of the present study is to investigate the load distribution method between parallel compressors in an optimal manner according to the capacity of each compressor. In this paper, a load sharing algorithm is proposed with the help of a model-based predictive controller (MPC) to achieve a stable and efficient operation at the compression station. In this algorithm, real-time optimization is used by an adaptive modifier method when faced with a variable demand from the consumer. The optimization is done according to the capacity of each compressor and using the approximated efficiency curve. In this way, the coordinates of the optimal working points for each compressor are obtained. The proposed algorithm is implemented in MATLAB software on the dynamics of centrifugal compressors in parallel arrangement. The results show that the load distribution using the proposed method has a much better performance than the old methods such as equal load balancing. This method results in the optimal operation of each compressor and thus storing and saving the natural gas.