Transient performance of combined cycle power plant with absorption based post-combustion CO2 capture: dynamic simulations and pilot plant testing

Abstract

The thesis presents transient performance analysis of chemical absorption processes for reducing CO2 emissions from natural gas combined cycle power plants (NGCCs), which can contribute to reduce greenhouse gas emissions to the atmosphere and mitigate climate change. Objectives focused on understanding process dynamics of NGCC with post-combustion CO2 capture (PCC) with an amine based chemical absorption process. Contributions comprised development and validation of high fidelity dynamic process models and evaluation of process dynamics of commercial scale NGCC with PCC power plant, including analysis of the performance of decentralized control structures for PCC units. In addition, experimental transient testing was conducted at a large-scale state-ofthe-art pilot plant for the evaluation of control structures applied to the chemical absorption process. Additional contributions included design of validation cases for dynamic process models of chemical absorption processes. The main methods employed were dynamic modeling and simulation and experimental transient testing. The thesis results include five peer-review research articles. High efficiency thermal power plants using novel solutions for operational flexibility improvements and CO2 emission reductions will be needed now and in the future to balance the variable renewable energy within decarbonized power systems. Carbon capture and storage (CCS) technologies can significantly reduce the carbon intensity of thermal power plants. The carbon intensity of state-of-the-art combined cycle power plants is around 365 g CO2/kWh, while for NGCC with PCC it is calculated to be around 50 g CO2/kWh. A 600 MW NGCC with PCC was designed and evaluated. The process con- figuration selected included one heavy-duty gas turbine and a triple-pressure reheat heat recovery steam generator in the combined cycle power plant, and a chemical absorption post-combustion CO2 capture unit with 30 wt% MEA as chemical solvent. The resulting net LHV electric efficiency of the integrated process was 52.8% and the specific reboiler duty at design point was 3.73 MJ/kgCO2. In order to identify scenarios for flexible operation of thermal power plants with CCS, a study on power markets and technical requirements was conducted. Technical grid requirements and frameworks for power units to provide ancillary services and bidding in balancing markets in four different power areas in EU were identified. In order to assess the transient performance of the NGCC with PCC, it was required to develop high fidelity physical dynamic process models. The selected tool for dynamic process modeling was the open physical modeling language Modelica. The focus of the study was to evaluate flexible operation of the chemical absorption process when integrated with power plant operations, with focus on power plant load variations. A detailed literature review proved necessary to validate dynamic process models of post-combustion chemical absorption of CO2 with large-scale pilot plant data for flue gas with low CO2 content characteristic of GT flue gas. However, the availability of suitable data sets for validation was scarce. Therefore, a set of validation cases for dynamic process model of the post-combustion CO2 capture process with chemical absorption using 30 wt% MEA was designed with data from operations of the large-scale amine plant at Technology Centre Mongstad. The plant can capture 80 ton CO2/day when operated iv with flue gas with a CO2 content of around 3.7 vol%. The data consisted of ten data sets representing a wide range of steady-state operating conditions with a slipstream of flue gas from a natural gas fueled power plant. The data included three transient tests for dynamic process model validation under transient conditions representing the main disturbances applied to the process. The validation results of a dynamic process model of the pilot plant showed capabilities of dynamic process modeling applied to large-scale experimental tests of the chemical absorption process with aqueous MEA. The validation of the thermal power plant model was conducted with steady-state design and off-design data from simulations. The software-to-software validation showed the proper implementation and development of the dynamic process model of the thermal power plant. The evaluation of process dynamics of a state-of-the-art PCC pilot plant was done via dynamic process model simulations and experimental transient testing. Results showed that, when the plant was operated at part load, it took a longer time to stabilize the main process variables in response to open-loop step changes in the main inputs of the process, namely solvent flow rate, flue gas flow rate and reboiler duty. Circulation times and solvent hold-up distribution through the equipment of the chemical absorption process showed to be a key aspect for process dynamics. It was found that the desorption rate stabilized faster than the absorption rate for set-point step changes in solvent flow rate and reboiler duty. An evaluation of performance of decentralized control structures of the PCC pilot plant was done via dynamic process model simulations. Simulation results showed that the best performance was obtained with the control structure in which capture rate is controlled by manipulating reboiler duty, and stripper bottom temperature controlled by manipulating solvent flow rate. Experimental transient tests for fast load change scenarios were conducted at the pilot plant. Testing results revealed that the process can reject fast disturbances in flue gas flow rate and could bring the process towards desired off-design steady-state conditions within 60 min by means of decentralized control structures. These tests provided empirical evidence at large-scale that combined cycle power plants with post-combustion CO2 capture can keep similar operational procedures as equivalent unabated power plants, considering fast load changes driven by GT load change. However, fast and large changes in solvent flow rate as a control measure can cause instabilities due to the interaction between the stripper temperature and the capture rate control loops. The transient performance of NGCC with PCC was studied by co-simulating and linking the dynamic process model of the power plant and the dynamic process model of the scaled-up PCC unit. Tests on load change driven by changes in GT load were conducted for variable ramp rates and for different control structures in the PCC unit. Based on these simulations, it was concluded that the addition of the PCC unit to the NGCC plant should not impose any constraint on, or problem for, stable power plant operation under scheduled load changes, even for aggressive ramp rates. The control structure where liquid-to-gas ratio in the absorber column was kept constant and reboiler temperature controlled by the steam throttle valve, showed similar part-load off-design performance as found in control structures with controlled capture rate. This control structure resulted in relatively faster total stabilization time of the steam turbine power output and CO2 product flow rate. v Areas for future work include: i) studying the transient performance of the system with higher levels of process integration; ii) assessment of transient performance of PCC with other chemical solvents; iii) development of reduced order models for faster numerical solution of dynamic process model simulations; iv) economic evaluation of flexible operation strategies including lifetime reduction due to thermal stresses in critical components of the process; v) optimization of start-up sequence of the integrated processes.