Design and Optimization of Industrial-scale Compression system for Its Efficient and Robust Operation

Abstract

Compression systems are one of the essential units in chemical processes. A compression system plays an essential role but consumes a significant amount of power. Also, this system is primarily employed to maintain a constant discharge pressure and needs to stay protected against surge phenomena. The surge phenomena causes back flow and vibration, which are damaging to the bearings, seals, and other parts of the compressor. Therefore, operating an efficient and robust compression system is the most important issue in plant design and management. A compressor needs to be operated within an operable range, which should be considered both at the design stage and at the operation stage. <br/> First, the authors propose a new process design method that improves the operability of the compression system, away from the design approach that considers only the economics. The suggested approach differs from a traditional one in that it performs design and optimization with several steady-state operation regimes depending on the load variation. The proposed design approach makes a loss in the compressor equipment cost, but it reduces the operation cost over a wide range of operations, leading to the overall improvement of economics and operability of compressor. <br/> Secondly, the author suggests the Nonlinear Autoregressive eXogenous Neural Net model (NARX NN)(Park, 2015) based real- time optimization for more efficient operation of industrial-scale multi-stage compression system in a commercial terephthalic acid manufacturing plant. NARX model is constructed to consider time-dependent system characteristics using actual plant operation data. The prediction performance is improved by extracting the thermodynamic characteristics of the chemical process as a feature of this model. And a systematic RTO method is suggested for calculating an optimal operating condition of compression system by recursively updating the NARX model. <br/> Finally, the author proposes an advanced control system for robust operation of a parallel compression system. Control of a parallel compressor system has proven to be challenging because the control targets usually exhibit control interactions between the different control loops. To decouple this control interference, Mitsubishi Heavy Industries has developed an advanced feed-forward control structure for parallel fixed-speed compressor systems. However, operation in the presence of an unpredictable disturbance presents a few technical challenges for this structure. Most of these problems result in poor load sharing and then operation in the recycle mode in order to protect the system from surge conditions. Moreover, an anti-surge control delay occurs when operating under a low load. To overcome these problems, an improved control structure that incorporates an additional discharge flow controller signal and a nonlinear signal calculator for anti-surge valve control is proposed.