Abstract
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.
Highlights
One of the major contributions to CO2 emissions is the flue gas from coal-fired power plants, accounting for over one third of total carbon emissions [1]
For multi-input multi-output (MIMO) nonlinear process, the model with an appropriate mathematical expression can greatly reduce the complexity of the advanced controller design and the computational cost of the control law. Considering these requirements for multi-model system development and the flexible operation of the CO2 capture plant, this paper investigates the nonlinearity distribution of the process to select proper local linear models and studies the construction of the multi-linear model
The accuracy of the local linear models is tested first, because it is the foundation for the establishment of the multi-linear
Summary
One of the major contributions to CO2 emissions is the flue gas from coal-fired power plants, accounting for over one third of total carbon emissions [1]. Monoethanolamine (MEA)-based post-combustion CO2 capture technology has been extensively studied to capture the CO2 from the coal-fired power plants [2,3,4,5,6,7,8,9,10,11,12,13,14,15,16]. The MEA-based CO2 capture technique has several distinguishing features, such as a high CO2 capture level, easy integration to power plants without much reformation of the existing plants, and relatively low construction cost, which makes it the most promising technique for commercial use [1,17]. The CO2 capture rate should be reduced rapidly during the peak time of electricity consumption, when the cost of a high CO2 capture rate overruns its
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call