(663a) Dynamic Data Reconciliation and Model Validation of a CO2 Capture Process Using Pilot Plant Data  

The post-combustion CO₂ capture systems would be subjected to fluctuating flue gas flow rate and composition as the power plants follow the load. In addition, future CO₂ capture plants can vary the extent of CO₂ capture in response to the electricity price. Therefore optimal servo control and disturbance rejection characteristics are highly desired while designing the control system for CO₂ capture processes. One essential element for designing the control system is a high fidelity dynamic model that is validated with the experimental data . However, there is a lack of validated dynamic models for solvent-based CO₂ capture processes. With this incentive, a rigorous dynamic model of a MEA-based CO₂ capture process is developed and validated using steady-state and dynamic data obtained from the US DOEâ??s National Carbon Capture Center (NCCC) in Wilsonville, Alabama.

Rigorous thermodynamic and transport models with quantified uncertainties are first developed and validated using large number of experimental data. A novel approach is developed to obtain a rigorous mass transfer model by considering the wetted wall column and packed column data simultaneously while regressing the parameters of the interfacial area model, mass transfer coefficient models, and diffusivity models together. In addition, rigorous kinetic model, holdup model, and pressure drop model are developed. The steady-state model is validated using large number of data collected from NCCC.

One issue with the limited dynamic data that are available in the open literature is that they span a limited range of operating conditions and often only contain a single step change. In this work, the dynamic test runs were conducted by designing a test protocol that helps to reflect the process nonlinearity and ensures persistence of excitation approximately. The dynamic data include solvent composition and loading data that were manually collected every 5 minutes and analyzed in the NCCC laboratory. The dynamic data collected from the NCCC pilot plant were found to have significant measurement noise and some inaccurate measurements. In addition, there were a number of key variables that were unmeasured. To handle these issues, a dynamic data reconciliation (DDR) framework is developed to ensure closure of material and energy balances prior to model validation. As the rate-based models are not currently supported in Aspen Plus Dynamics, the software platform of our choice, an approximate method that matches the results from the rate-based Aspen Plus model closely is developed and implemented using scripts in Aspen Plus Dynamics. It is observed that the results from the dynamic model match very well with the experimental data that span more than three days. The dynamic model is then leveraged to perform a number of transient studies by simulating expected disturbances and setpoint tracking problems.