Dynamic Response of Monoethanolamine (MEA) CO2 Capture Units

Coal-fired power plants need to respond to changes in electricity demand on diurnal, weekly, seasonal, and yearly time scales. The scale and frequency of these responses are forecasted to increase as levels of variable renewable energy sources become a larger part of the electricity supply. To respond to climate change concerns, it is also anticipated that the coal-fired power plants of the future will incorporate carbon capture and sequestration (CCS) technologies. Therefore, it is important to understand the dynamic response of coal-plants with CCS. A variety of engineering studies have been published that investigate the energy penalty and design requirements for coalfired plant running at static power output and carbon capture rate of 90 percent; however, few of these studies explore the ability of the capture unit to respond to changes in load and capture rates. Noting that amine absorption is considered the most advanced near-term technological solution for carbon capture, this paper provides an analysis of the dynamics of a MEA capture unit. This analysis enables the evaluation of the ability of coal-fired CCS power plants to provide load following support at different ramp rates and at varying levels of capture. A dynamic kinetic model of an MEA capture plant was developed using Aspen Dynamics®. The model is used to determine the dynamic characteristics of the capture plant for load following by simulating various ramp rates of flue gas flow from the power plant to the capture unit. These results are used to determine the ability of the capture plant to control power output to the grid and the impact on performance parameters such as the capture rate and energy consumption. The results shown that the capture plant operates on similar time scales of a coal-fired power plant. The capture plant will not prohibit the ability of the coal plant to adjust output. Introduction Electricity generation from fossil-fuel power plants is one of the largest sources of carbon dioxide emisisons (IPCC 2005). To reduce the amount of emissions from these power plants, carbon capture and sequestration is seen as one of the most promising technology options. In particular, chemical absorption is the most widely studied and advanced subset of these technologies. To date, most studies have looked at the steady state, full load, 90% capture operation of the CCS unit (Freguia et al. 2003; Abu-Zahra et al. 2006; Zhang et al. 2009). Current coal-fired power plants need to be able to ramp to respond to changes in electricity demand and prices. If CCS is integrated into the power plant, it is important to determine the ability to respond to the changes in the power plant that may potentially affect the capture operation. Studies investigating load balancing with increasing penetration of renewables show that traditional baseload generation will need to be able to respond to a wide range of operating conditions and ramp rates (IEA 2011; GE Energy 2010). In addition, the operation of a flexible carbon capture system may enhance plant operating economics in by reducing steam required for solvent regeneration (Cohen 2011). Understanding the dynamics of the capture system is important for the operation of the power plant. 2 CMTC CMTC-151075-PP Other papers have considered the part-load, steady state operation of the capture unit integrated with the power plant (Chalmers et al. 2007; Haines et al. 2009), but did not consider the dynamic changes of the capture unit while transitioning to part-load operation. While many studies have investigated individual process units, few have studied the dynamics of an integrated carbon capture system. Lawal et al. (2009) and Kvamsdal et al. (2009) focused on the dynamic modeling of the absorber unit and Ziaii et al. (2009) focused on optimizing the energy usage of a standalone stripper model. This paper considers the dynamic operation of an absorber/stripper post-combustion monoethanolamine (MEA) capture plant of a nominal 500 MWth coal-fired power plant and investigates the time scales and effects of how the capture plant responds to step changes in the various inputs for an integrated absorber/stripper capture model. The cycling of a coalplant is reflected in the changes in the flue gas flow to the capture plant, changes in the capture rate for environmental or economic reasons, and changes in the amount of available steam to the reboiler. This will help determine the deviations from set points and the ability to return to steady state. This study will look at the range of operational issues presented as the capture plant undergoes the changes that are typical to the operation of a coal-fired power plant. Description and Application of Equipment and Processes Process Description A steady-state model of an absorber/stripper post-combustion MEA capture unit was created in Aspen Plus Version 7.3 based on the work of Kothandaraman (2010). The steady-state model was used for model development and verification. The absorber and stripper were assumed to have 30 equilibrium stages and 20 equilibrium stages, respectively. The sizing of the columns was done to accommodate the flow from a nominal 500 MWth coal-fired power plant using only one absorber/stripper train. The major process components that were modeled are shown in Figure 1. Figure 1. Process flowsheet for absorber/stripper post-combustion capture process For this study, the electrolyte non-random two-liquid (ELECNRTL) property method was used. The absorber/stripper chemistry was represented with seven reactions. SO2 and NOx were assumed to be inert gases for the purposes of the model. H2O + MEA+ <--> MEA + H3O+ (1) 2 H2O <--> H3O+ OH(2) HCO3+ H2O <--> CO3-+ H3O+ (3) CO2 + OH<--> HCO3(4) HCO3--> CO2 + OH(5) MEA + CO2 + H2O --> MEACOO+ H3O+ (6) MEACOO+ H3O+  MEA + CO2 + H2O (7) The dimensions of the absorber are 17 m in height and 12 m in diameter and the dimensions of the stripper are 15 m in height and 7 meters in diameter. The absorber and stripper pressure were set to 1 atm and 1.75 atm, respectively. The CMTC CMTC-151075-PP 3 condenser was modeled externally from the RadFrac column due to convergence issues, but his did not affect the results of the model. To achieve water and amine balance for the model due to process losses, the makeup was added in the surge tank. The water makeup was used to maintain to 30-wt% of the MEA solution, while the amine makeup was added primarily for model convergence. Temperature was maintained through the system of heat exchangers shown above. The lean solvent heat exchanger and the condenser were modeled to maintain a constant exit temperature, while the cross heat exchanger was modeled to allow for changes in process conditions allowing the exit temperatures to be free variables in the model. Dynamic Simulation and Control Strategy The steady-state model was exported to Aspen Dynamics and the flow-driven type model was chosen. Figure 2 illustrates the control structure of the post-combustion MEA capture. The primary objective of the system is to maintain a specified capture rate of 90%. To achieve this level of capture, the lean solvent circulation rate is manipulated. The capture rate is determined by measuring the incoming mass flow of CO2 to the absorber and the mass flow exiting the absorber through the vent stream. To control the lean solvent rate, the flow from the storage tank is manipulated. Figure 2. Absorber/stripper process flow diagram with control structure The lean solvent loading is one of the key parameters in CO2 capture. To regenerate the rich solvent from the absorber, heat is applied in the form of steam to the reboiler to drive off the CO2 from solution, resulting in an energy intensive process. The amount of reboiler duty drives the equilibrium of the lean solvent at the fixed pressure of the column. From the steady-state simulations, it was determined that the energetic optimum for the lean concentration is 0.20 on a dry molar basis. The study by Panahi et al. (2011) stated that a proper indicator for lean loading is the column stage temperature. A direct measurement of the lean loading would be costly and time intensive. The reboiler temperature was maintained at a temperature of 124°C to achieve the desired lean loading. For the purposes of the model, the steam temperature and pressure were assumed to be constant throughout the operating range. This paper does not consider the issue of steam extraction from the crossover piping of the steam turbine. Given the arrangement of the turbine system, the steam may deviate from full load temperature and pressure at different operating conditions. The other process controls were used to control the operating pressure and sump level of the absorber and stripper. To maintain the operating level of the columns, the gas flow exiting the column was manipulated. To maintain the sump level of the columns, the liquid flow was manipulated. The heat exchangers were modeled as instantaneous processes. Presentation of Data and Results Three disturbances were tested using the dynamic simulation. These disturbances were used to determine the dynamic behavior of the process variables, the response time of the system and time to achieve steady state, and the performance of the system in the new steady state. The three disturbances were to the flue gas flow rate to simulate a change in plant load, a 4 CMTC CMTC-151075-PP change in capture rate to simulate flexible operation for economic reasons, and a change in the reboiler temperature to decrease steam extraction from the turbine and increase power output. Disturbance to flue gas flow To analyze the dynamics of the system for a change in fuel consumption by the base coal plant, a negative 10% step change was made to the incoming flue gas flow, steady state was achieved again, and a positive 10% step change was made to return to the original inputs to the capture plant. The system was run for 2 hours before the