Model Analysis for the Implementation of a Fast Model Predictive Control Scheme on the Absorption/Stripping CO₂...

Area of Science:

• Chemical Engineering
• Process Control
• Carbon Capture Technologies

Background:

• Implementing advanced control strategies like Fast Model Predictive Control (MPC) in chemical systems, particularly CO2 capture plants, faces challenges due to complex dynamics and model sensitivity.
• Previous implementations of Fast MPC on absorption/stripping CO2 capture systems have been limited by significant control offsets and performance issues.

Purpose of the Study:

• To investigate the feasibility and performance of Fast MPC for CO2 capture plants.
• To identify the most suitable model for Fast MPC implementation to achieve fast responses, minimal offsets, and reduced errors.
• To compare different modeling approaches for Fast MPC in CO2 capture processes.

Main Methods:

• Designed steady-state and dynamic simulation models of a CO2 capture plant using Aspen PLUS.
• Identified multiple state-space models using the System Identification Toolbox to create an accurate model for Fast MPC.
• Implemented and evaluated the Fast MPC controller under setpoint tracking conditions with varying step changes (±5% and ±15%).

Main Results:

• The Fast MPC controller utilizing a prediction-focused state-space model demonstrated superior performance compared to a simulation-focused model.
• The prediction-focused model resulted in an average of 7.9 times lower offset.
• Integral absolute error values were reduced by an average of 10.4 times with the prediction-focused model.

Conclusions:

• Fast MPC is an efficient control strategy for CO2 capture plants utilizing absorption/stripping systems.
• Prediction-based focus state-space models are optimal for implementing Fast MPC, leading to minimized offsets and errors.
• The proposed Fast MPC approach enhances the control performance of CO2 capture processes.