Nonlinearity Analysis and Multi-Model Modeling of an MEA-Based Post-Combustion CO2 Capture Process for Advanced Control Design

The monoethanolamine (MEA)-based post-combustion CO2 capture plant must operate flexibly under the variation of the power plant load and the desired CO2 capture rate. However, in the presence of process nonlinearity, conventional linear control strategy cannot achieve the best performance under a wide operation range. Considering this problem, this paper systematically studies the multi-model modeling of the MEA-based CO2 capture process for the purpose of (1) implementing well-developed linear control techniques to the design of an advanced controller and (2) achieving a wide-range flexible operation of the CO2 capture process. The local linear models of the CO2 capture process are firstly established at given operating points using the method of subspace identification. Then the nonlinearity distribution at different loads of an upstream power plant and different CO2 capture rates is investigated via the gap metric. Finally, based on the nonlinearity investigation results, the suitable linear models are selected and combined together to form the multi-model system. The proposed model is validated using the measurement data, which is generated from a post-combustion CO2 capture model developed in the go-carbon capture and storage (gCCS) simulation platform. As the proposed multi-linear model has a simple mathematical expression and high prediction accuracy, it can be directly employed as the control model of a practical advanced control strategy to achieve a wide operating range control of the CO2 capture process.