Integrated Thermodynamic Digital Twin for Energy Performance Monitoring of Large Compression Units

Abstract

The compression units installed on sales gas network face wide range of operating modes owing to the varying supply & demand scenarios and the associated network dynamics. It is very challenging to ascertain the real performance in such applications due to changing specific energy consumption. The paper presents development of a novel and robust monitoring system, enabling realtime energy performance monitoring of dynamic compression and revealing realistic opportunities for energy savings. The methodology comprised review of design and OEM data for the compression units, followed by review of operating envelope. Subsequently, developed a thermodynamic model of compression units encompassing all the operating modes. Then embedded the actual performance curves from OEM in the thermodynamic digital twin and validated the model with actual operating data. Carried out site visits and held discussions with technical teams as part of the comprehensive approach. A mathematical model additionally developed, in addition to the thermodynamic model, to enable operators for off-line monitoring of the compressors performance during unavailability of thermodynamic digital twin. Detailed performance analysis of centrifugal compressors is essential to ascertain their condition and functioning. A decrease in performance can be an indication of internal wear or fouling, which if allowed to continue, may result in reduced throughput or excessive energy consumption or even unscheduled outages. Thus, the performance is not just an indication of energy or operating cost but also reflects other vital aspects like reliability. The integrated thermodynamic digital twin developed for large sales gas compression units, with total throughput capacity of more than 500 MMSCFD, has enabled and demonstrated effective energy performance monitoring even with changing operating scenarios. It facilitated real-time comparison of actual performance (specific energy consumption) with model based expected performance. It also aided real-time trending of polytropic efficiency as well as real-time display of potential energy savings opportunities. The digital twin has proven to be a reliable and low cost tool to predict compressor performance for various operating modes on real time basis. The data from the compression digital twin can be tied into process simulation models for process optimization. The model can complement supervisory capabilities, diagnostics, control capabilities and even facilitate in predicting failures ahead of time.