A Rate-Based Process Modeling Study of CO2 Capture with Aqueous Amine Solutions using aspenONE Process Engineering™ Chau-Chyun Chen, David Tremblay, Chai Bhat Presented at the Clearwater Coal Conference, Clearwater Florida, June 4th 2008 ABSTRACT Rising energy prices coupled with a higher environmental consciousness have generated renewed interest in abundant fossil fuels such as coal, and ways to successfully capture and store the Carbon Dioxide (CO2) produced during coal gasification. Mono Ethyl Amine (MEA), Methyl Diethyl Amine (MDEA) and activated Tri Ethyl Amine (TEA) solutions are being considered by the industry as a means of capturing CO2. In this study we used Aspen RateSep to simulate the U. Texas-Austin experiments (Dugas, 2006[1]) for CO2 capture with aqueous MEA solution. Aspen RateSep is a second generation rate-based absorption and stripping unit operation model that works with Aspen Plus. The model predictions against the MEA pilot plant data are presented. The superiority of ratebased model predictions is shown. The impact of simulation results of various rate-based modeling constructs, i.e., film discretization, flow mixing; interfacial area correlations, etc are examined. DESIGNING GASIFICATION PROCESSES USING aspenONE PROCESS ENGINEERING™ aspenONE Process Engineering enables engineers to model the entire gasification process in one integrated environment. aspenONE Process Engineering has been successfully utilized by plant owners/operators, engineering and construction companies, and technology providers to improve yields, increase plant efficiency and quality, and reduce capital and operating costs. Designing a gasification facility can be challenging because of the different options available such as gasifier type, feed type (different types of coal, cellulose, industrial waste, biomass, petroleum coke, etc), end product demand (power, chemicals, fuel, etc). It is imperative that each sub-process within a gasification facility, such as the sizing unit, the air separation unit, the gasifier, gas cleaning unit (absorption and stripping of the acid gas), the CLAUS unit, etc be modeled precisely so as to derive optimum conversion efficiency in the gasifier and clean out any acid gas generated. Aspen Plus is used by many companies to design every sub-process in one integrated environment; Aspen Plus is an integral part of aspenONE Process Engineering. In fact, Aspen Plus was born out of the 1970’s energy crisis and the need for simulation of unconventional solids such as coal. For over 25 years Aspen Plus has been used to model advanced power processes, and all the major types of gasifiers used in the industry. Aspen Plus has unique out-of-box capability of handling coal, while the equation oriented capability is ideal for simulating large scale integrated coal plants. Figure 1 illustrates the overall engineering workflow and life cycle process of designing a gasification plant. Based on the process concept and business objective, one can establish the performance of the concept and then improve the concept using conceptual design methodologies in Aspen Plus. The base case and the improved case economic feasibility can be compared using standard cost analysis environment such as Aspen Icarus Project Evaluator. It is

also important to establish a detailed performance model for critical equipment. This helps identify a practical design option during the conceptual phase. The basis for these detailed equipment models needs to be consistent between the base case and the improved case. Detailed models of the gasifier can be developed using Aspen Custom Modeler. After developing an improved process concept, Aspen Dynamics can then test the process for safety, operability, and controllability issues; this defines the key control loops and the instrumentation for the process. Once the process control strategy and key instrumentation is defined, the definition of the process intent for the design is complete. Aspen Zyqad can then be used to develop the FEED package, incorporating the PFD or process P&ID for the improved concept, equipment designs, data sheets, summary sheets and basic control loops and instrumentation. The process design information can then be further transferred into detailed P&ID and instrumentation environments. The performance models developed for the process can be re-used for operational monitoring and improvement of the plant using Aspen Simulation Workbook. The performance model can also be deployed to non-expert users who may not have background in simulation to perform “what if analysis” studies over the web using Aspen Online Deployment. Traditionally gas cleaning has been modeled using equilibrium stage models, which have not provided the accuracy needed for real world industrial applications. RATE BASED DISTILLATION Aspen Rate Based Distillation (Aspen RateSep) is part of the aspenONE Process Engineering solution, and extends the functionality of Aspen Plus RadFrac distillation model with second-generation rate-based technology which accurately predicts simulation over a wide range of operating conditions. Aspen Rate Based Distillation uses sate-of-the-art mass- and heat transfer correlations to predict column performance, without the need of efficiency factors. This added degree of rigor is especially critical for modeling gas scrubbers, sour water strippers, azeotropic systems, reactive distillations, nitric acid absorption columns, narrow-boiling separations, and other highly non-ideal separation processes. The rate-based modeling approach is superior to the traditional equilibrium-stage modeling approach that has been employed extensively in the process industries over the decades. The rate-based models assume that separation is caused by mass transfer between the contacting phases, and use the Maxwell-Stefan theory to calculate mass transfer rates[2]. Conversely, the equilibrium-stage models assume that the contacting phases are in equilibrium with each other, which is an inherent approximation because the contacting phases are never in equilibrium in a real column. The rate-based modeling approach has many advantages over the equilibrium-stage modeling approach. The rate-based models represent a higher fidelity, more realistic modeling approach and the simulation results are more accurate than those attainable from the equilibriumstage models. The rate-based modeling approach can reduce the risk of inadequate designs or off-spec operation because the rate-based models explicitly account for the actual column configuration which affects column performance. Designed to model reactive multistage separation problems rigorously and accurately, Aspen RateSep balances gas and liquid phase separately and considers mass and heat transfer resistances according to the film theory by explicit calculation of interfacial fluxes and film discretization. The film model equations are combined with relevant diffusion and reaction kinetics and include the specific features of electrolyte solution chemistry, electrolyte thermodynamics, and electroneutrality where appropriate. The hydrodynamics of the column is accounted for via

correlations for interfacial area, hold-up, pressure drop, and mass transfer coefficients. Figure 2 illustrates the basic picture of CO2 transfer across the vapor and liquid films.

SIMULATION OF THE UNIVERSITY OF TEXAS - AUSTIN PILOT PLANT EXPERIMENTS WITH ASPEN RATESEP The pilot plant facility is a closed-looped absorption/stripping system for CO2 removal from a flue gas by means of a 32.5% aqueous MEA solution. In this study the pilot plant absorber was simulated using Aspen RateSep. Aspen RateSep automatically selects the correlations to execute the various calculations. Aspen RateSep offers several options for modeling film resistance, the film discretization option was chosen for this study. Aspen RateSep also provides four different flow models to determine the bulk properties used to evaluate the mass and energy fluxes and reaction rates of a stage. In this study the “Countercurrent” flow model was used as the base calculation method. The Aspen RateSep simulation results are summarized in Table 1, along with the experimental performance data. The feed stream data is shown in the “Lean” and “Rich” columns. “Lean” column contains values of CO2/MEA mole ratio of the lean stream, while the “Rich” column contains the mole ratios of CO2/MEA mole ratio of the rich stream. These values indicate the performance of the absorber. Note that the equations used to calculate the “Experimental” mole ratio of CO2/MEA ignore the influence of MEA concentration; however in reality the solution density is sensitive to MEA concentration especially for the rich stream with the high CO2 loading. The CO2 loading was determined by taking sample solution density at five locations on the absorber. Table 2 contains the MEA mass concentrations of the five samples. There is a great deal of variability in this data, this confirms that the density of the liquid is sensitive to the MEA concentration especially when the CO2 concentration is high. A correction can be applied to the density so that the estimated CO2 loading absorber will decrease and the simulation results will be much closer to the experimental results. An absorber temperature profile was obtained by temperature sensors located on the absorber. When CO2 reacts with MEA the heat of reaction produces a temperature bulge in the column. This temperature bulge can significantly affect the absorption rates in the column since the kinetics of the absorption reaction, the phase equilibrium composition of the system, and the fluid transport properties depend on the temperature. This temperature bulge occurs at the top section of the absorber until a pinch point where the temperature is the highest. In the section below this point, the temperature will decrease. Three distinct types of temperature profiles were evident in the shape of the temperature bulge. All the cases in this study can be categorized in one of the three temperature profiles. It was observed that the temperature profiles obtained from each case were related to the ratio of gas flow rate to liquid flow rate. When the ratio of gas flow rate to liquid flow rate was relatively high the absorption reaction occurred at the top section of the absorber, in which case the temperature profile had a pinch point near the top of the absorber. When the liquid flow rate is high almost all the CO2 is absorbed at the bottom section of the absorber, this leaves very little CO2 to react with the MEA in the top section and the pinch point is near the bottom of the absorber. When the CO2 absorption takes place throughout the absorber a very broad pinch is visible in the temperature profile. The temperature changes sharply at both ends, one near the

top and the other near the bottom, and the temperature profile between the ends is relatively flat. This is a transition type of temperature profile. Both the vapor and the liquid phase temperature profiles are obtained from the Aspen RateSep simulation. There was not much temperature difference between the two phases, this is evident in Figure 3. For all other cases only the liquid phase temperature profile is shown. As illustrated in Figure 4 the simulated temperature profile matches the absorber experimental data very well, this was observed in all the cases simulated with Aspen RateSep. CONCLUSION In this study we established that the rate-based simulation results using Aspen RateSep conform to the pilot plant absorber experimental data generated at U. T. – Austin. The Aspen RateSep model accurately predicts the temperature profile in the pilot plant. The Aspen RateSep model with proper model parameters provides excellent 1st principles-based predictive capability and it should be very helpful for industrial process and plant design of the CO2 capture processes. With Aspen RateSep’s far superior rate-based modeling capabilities, ability to rigorously model the liquid film with industry leading film discretization capability, and seamless integration with aspenONE Process Engineering engineers can simulate plant performance accurately and design plants with confidence, minimizing contingency factors, lowering capital and operating costs, and avoid potential expensive start-up issues. REFERENCES [1] Dugas, R.E., Pilot Plant Study of Carbon Dioxide Capture by Aqueous Monoethylamine. Master thesis, Chemical Engineering, the University of Texas at Austin, 2006. [2] Taylor, R., R. Krishna, and H. Kooijman, “Real-World Modeling of Distillation,” Chem. Eng. Prog., July 2003, 28-39.

FIGURES Figure 1: Integrated Process Engineering Workflow Diagram Reuse Process Models for Operations Decisions / OTS

Operations/ Maintenance

Aspen Simulation

Safety & Controllability Analysis

Aspen Dynamic Conceptual Design / R&D

Reuse Process Models for Planning Decisions

Aspe n

Aspe n

Process Performance Modeling & Analysis

Aspen Plus

Aspe n

Detailed Equipment Design Develop Basic Design Package

Basic Engineering

Analyze Economic Performance

Aspen HTFS+

Aspe n

Analyze Detailed Costs

Aspe n

Develop Detailed Plant Design

Detailed Engineering

Figure 2: Discretized film concept is combined with the countercurrent flow configuration in Aspen RateSep

Figure 3: Absorber temperature profile of Type A, Case 7: Experimental data from Dugas (2006)[1], Liquid temperature calculated by Aspen RateSep, Vapor temperature calculated by Aspen RateSep

200

Temperature, F

180 160 140 120 100 80 -1

1

3

5

7

9

Height from Bottom , m

Figure 4: Absorber temperature profile of Type A, Case 24: Experimental data from Dugas (2006)[1], Liquid temperature calculated by Aspen RateSep

200

Temperature, F

180 160 140 120 100 80 -1

1

3

5

Height from Bottom , m

7

9

TABLES Table 1: Performance of the absorber Case 5 6 7 8 9 10 11 12 13 14 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Lean (molCO2/molMEA) Experimental 0.180 0.182 0.177 0.170 0.150 0.148 0.147 0.143 0.323 0.329 0.268 0.274 0.271 0.274 0.282 0.277 0.275 0.278 0.275 0.284 0.287 0.285 0.284 0.281 0.279 0.283 0.282 0.280 0.284 0.282 0.281 0.228 0.229 0.235 0.232 0.231 0.231 0.285 0.286 0.281 0.285

Rich (molCO2/molMEA) Experimental Simulation 0.525 0.485 0.523 0.478 0.496 0.457 0.493 0.456 0.532 0.475 0.533 0.476 0.537 0.481 0.546 0.482 0.507 0.459 0.508 0.459 0.506 0.459 0.495 0.456 0.538 0.476 0.540 0.477 0.554 0.482 0.557 0.483 0.506 0.451 0.386 0.346 0.376 0.341 0.413 0.372 0.412 0.373 0.448 0.428 0.453 0.439 0.426 0.419 0.428 0.407 0.422 0.405 0.420 0.406 0.415 0.388 0.425 0.405 0.404 0.386 0.402 0.379 0.367 0.322 0.371 0.325 0.433 0.399 0.430 0.401 0.491 0.438 0.492 0.435 0.433 0.386 0.426 0.388 0.539 0.466 0.537 0.466

Table 2: Mass ratios of MEA/(H2O+MEA) of five samples for Case 22 Sample Lean into the top of the absorber Middle of the absorber Rich out of the bottom of the absorber Middle of the stripper Lean out of the bottom of the stripper

Mass ratios of MEA/(H2O+MEA) 0.325 0.328 0.334 0.300 0.399