Available online at www.sciencedirect.com ScienceDirect Energy Procedia 63 (2014 ) 1084 1090 GHGT-12 Improving the efficiency of a chilled ammonia CO 2 capture plant through solid formation: a thermodynamic analysis Matteo Gazzani, Daniel Sutter, Marco Mazzotti* ETH Zurich, Institute of Process Engineering, Sonneggstrasse 3, 8092 Zurich, Switzerland Abstract Post-combustion chemical absorption is regarded as the state-of-the-art commercially-available CO 2 capture process. The adoption of aqueous ammonia as solvent, leading to the so-called Chilled Ammonia capture Process (CAP), has long been considered one of the most promising alternatives to amine-based for post-combustion carbon capture. This work investigates the development of a second generation CAP where the capture efficiency is improved by making use of a crystallizer to form solids in the process. The reference standard CAP and the advanced crystallizer-based CAP are simulated in Aspen using the Extended UNIQUAC thermodynamic model. The two CAP solutions are compared in term of the different energy penalties introduced applying the capture process to a conventional Ultra Super Critical (USC) power plant. Thanks to the solid formation, the CAP with the crystallizer features a lower energy penalization with a decrease of about 10% compared to the total penalty of the standard CAP. 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). 2013 The Authors. Published by Elsevier Ltd. Selection Peer-review and under peer-review responsibility under of the responsibility Organizing Committee of GHGT. of GHGT-12 Keywords: Chilled Ammonia Process, CO 2 post-combustion capture, CCS, CO 2 capture energy penalty CAP FGD LHV USC Chilled Ammonia Process Flue Gas Desulfurizer Lower Heating Value, MJ/kg fuel Ultra Super Critical * Corresponding author. Tel.: +41 44 632 2456; fax: +41 44 632 1141. E-mail address: marco.mazzotti@ipe.mavt.ethz.ch 1876-6102 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Organizing Committee of GHGT-12 doi:10.1016/j.egypro.2014.11.116
Matteo Gazzani et al. / Energy Procedia 63 ( 2014 ) 1084 1090 1085 1. Introduction The power plant flue gas decarbonisation via conventional amine scrubbing is regarded as the state-of-the-art commercially-available CO 2 capture process. This is mainly due to the large number of existing plants for acid gas treatment and the associated maturity of the technology. On the other hand, the large energy penalty, the solvent degradation and the difficulty of handling corrosive solutions have prompted the research towards advanced processes. The use of chilled ammonia as solvent is a promising solution which has long been regarded as a possible alternative to amines. The main advantages of this technology, also known as Chilled Ammonia capture Process (CAP), include: i) low cost and large availability of the solvent, ii) chemically stable solution, iii) high stability to oxygen, iv) regeneration at medium pressure and v) high CO 2 carrying capacity. The performance of the CAP has been assessed in only a few works, without reaching a consistent and thorough assessment. Darde et al [1], Valenti et al [2] and Versteeg and Rubin [3] report significant thermodynamic advantages over the conventional amine solutions whilst the contrary is shown in Mathias et al [4]. Recently, Valenti et al [5] reported substantially equivalent performance between amine and CAP. The performance evaluation is complex because the chilled ammonia plants entail several interdependent energy intensive requirements: i) heat for rich solution regeneration, ii) chilling duty for lean solution and flue gas cooling, and iii) heat requirement for the ammonia wash section. It is well known that the plant operating conditions strongly affect the chemical behaviour of the process: solid phases, primarily consisting of ammonium carbonate and bicarbonate, may form in the absorber and in related components. In the existing CAP plants the solid formation has been strictly avoided due to the complexity of handling solids. On the other hand, solid formation offers different advantages from a thermodynamic point of view: i) reduction of the mass flow per tonne of CO 2 of the rich solution sent to the stripper, thus resulting in a significant decrease of the heat requirement; ii) reduction of the stripper dimension; and iii) reduction of the ammonia slip from the absorber. Accordingly, the performance of the next generation CAP can be improved when exploiting the formation of solids. This work investigates the possibility of using solids in the CAP with a dedicated solid formation unit and compares the results of the proposed concept with the standard CAP. 2. Plant layout 2.1. Standard CAP without solid formation The overall plant layout of the standard CAP is shown in Figure 1. The flue gas exiting the Flue Gas Desulfurizer (FGD) (at about 40-60 C, in saturated conditions) enters the direct contact cooling section where most of the water and residual contaminants are removed along with the gas cooling. The flue gas enters the CO 2 capture section at 18 C; the capture scheme is a standard layout with two adsorption/desorption towers, a regenerative heat exchanger to harness the energy content of the CO 2 lean stream and a solution pump to match the different operating pressure of the absorber/desorber. The treated flue gas exiting the CO 2 absorber are sent to the ammonia abatement section where the NH 3 slip is lowered to few ppm. The bulk NH 3 removal is carried out in the water wash through a conventional absorber/desorber process and therefore similar to the cycle for CO 2 capture. The flue gas, with a lower amount of CO 2 but containing ammonia, is chilled and enters the bottom of the absorber, where it is contacted with water flowing from the upper stages of the column [6]. The ammonia content in the flue gas leaving the upper part of the column is reduced to environmentally acceptable level in a final acid wash. The liquid exiting the bottom of the absorber is regenerated in the desorption column providing almost pure water back to the absorber. The ammonia recovered from the flue gas is recycled to the absorber in the CO 2 capture island.
1086 Matteo Gazzani et al. / Energy Procedia 63 ( 2014 ) 1084 1090 Figure 1: Plant layout of the standard CAP In order to limit the energy penalty associated with the ammonia slip abatement, the CO 2 absorber column is designed such that the first stages of the column contribute to lower the ammonia vapour pressure. This is achieved recycling part of the CO 2 rich solution from the bottom of the column which is cooled and chilled before re-entering the column [7]. The combination of high CO 2 concentration and low temperature prevents the further CO 2 uptake while favouring the ammonia absorption. A schematic representation of the column is shown in Figure 2. 2.2. Advanced CAP with crystallizer for solid formation Figure 2: Details of the absorber configuration [7] From a qualitative point of view the overall plant layout of the crystallizer-based CAP (as introduced in [8] and shown in Figure 3) is similar to the standard CAP but for the CO 2 capture island. Provided that handling solids in a packed column is not feasible from an engineering point of view, the operating conditions of the plant have been tuned in order to carefully avoid any solid formation inside the absorber and desorber. The solid formation in the rich solution is restricted to a crystallizer downstream of the absorber. In order to reduce the mass flow sent to the stripper as much as possible, all the solid material exiting the crystallizer is separated in a hydrocyclone together with a part of the liquid solution. The obtained slurry is sent to a solid dissolution reactor and then to a regenerative heat exchanger; therefore all the solids are dissolved before entering the desorption column. The liquid stream separated in the hydrocyclone is sent back to the absorber to control the absorber temperature and the ammonia slip.
Matteo Gazzani et al. / Energy Procedia 63 ( 2014 ) 1084 1090 1087 The rich solution exiting the absorber must be cooled to 5 C to favour the solid formation. No advanced heat integration, for example the regenerative heat exchange between the separated slurry and the rich CO 2 liquid, has been considered in this work. The chilling duty for the crystallization has been calculated cooling the solution from 18 to 5 C. 3. Methodology and approach Figure 3: Plant layout of the crystallizer CAP The use of an appropriate thermodynamic model that correctly describes the interaction among NH 3, CO 2 and H 2 O in the system and the solid-liquid-vapor equilibria is of a paramount importance for obtaining reliable results in the simulations of the CAP. Accordingly, this work uses the Extended UNIQUAC model developed by Thomsen [9] and upgraded by Darde et al [10]. The model considers five different solid phases: 1. Ammonium bicarbonate (BC) NH 4 HCO 3 2. Ammonium carbonate (CB) (NH 4 ) 2 CO 3 H 2 O 3. Ammonium sesqui-carbonate (SC) (NH 4 ) 2 CO 3 2NH 4 HCO 3 4. Ammonium carbamate (CM) NH 2 COONH 4 5. Ice H 2 O The solubility data for solids 1-4 used in the parameter fitting procedure are based on Janecke data [11]. The gasphase fugacities are calculated with the Soave-Redlich-Kwong equation of state. Both the CAP solutions with and without crystallizer have been simulated with Aspen Plus. The vapour-liquid equilibrium in the absorber/desorber of the CO 2 section and the NH 3 water wash has been modelled with the rigorous Aspen RadFraq approach for multistage vapor-liquid systems assigning stagewise Murphree efficiencies (only in the absorbers) and checking for salt precipitation. The crystallization step is simulated with a continuous stirred reactor working at thermodynamic equilibrium. The plant simulation is a closed loop with connected streams between the desorber and absorber section. The convergence is guaranteed by making use of tear streams and calculating the ammonia and water make-up. The solid formation along the process is carefully checked by making use of ternary phase diagrams for the CO 2 - NH 3 -H 2 O system: critical points and the CO 2 /NH 3 absorber profiles are reported on ternary diagrams to control that no solid formation takes place outside the crystallizer reactor. More details about solid formation are reported in Sutter et al [12] The main assumptions for the CAP simulations are reported in Table 1.
1088 Matteo Gazzani et al. / Energy Procedia 63 ( 2014 ) 1084 1090 Table 1: Main assumptions for the CAP calculation in Aspen CAP process specifications CO 2 capture, % 90 CO 2 lean loading (entering the absorber) 0.3-0.35 CO 2 rich loading (exiting the absorber) > 0.5 CO 2 purity before storage > 0.99 Ammonia slip in the flue gas exiting the CO 2 absorber, ppm < 8000 Ammonia slip after the water wash, ppm 200 Flue gas temperature entering absorber, C 18 Adsorption/Desorption pressure, bar 1.01/10 Flue gas composition: CO 2 15.6%, N 2 66.0%, O 2 17.4%, Ar 1.0% Utilities Chilling water temperature, C Cooling water temperature, C Heat exchanger ΔT min, C or K 2 15 3 In order to calculate the overall energy penalty of the two CAP solutions, a base case USC plant without carbon capture has been considered as proposed by EBTF [13]. The power balance of this plant is reported in Table 2. Table 2: Power balance of the considered USC plant without CO 2 capture as in [13] Net power output, MW Fuel input, MW Net LHV efficiency, % Flue gas mass flow, kg/s Specific emissions, g CO2 /kwh el 758.6 1676.6 45.2 740 772 The energy penalties introduced with the CAP process are due to the need of thermal and electric energy in five main processes: i) thermal energy for the CO 2 capture reboiler, ii) electric energy for the chilling duty, iii) thermal energy for the NH 3 wash reboiler, iv) electric energy for the CO 2 compression and v) electric energy for the pumps. The penalty associated with the use of steam in the plant reboilers has been calculated considering the decrease in the steam turbine net power. The resulting difference in the power required to handle the condenser duty has been as well considered. The electric energy associated with the chilling duty has been computed using the Coefficient of Performance (COP) of the cooling cycle. (1) The COP has been calculated considering the ideal Carnot COP derived from the temperature of the evaporator and condenser in the inverse Rankine cycle and a second-law efficiency as following: Where: T eva = 273 K, T cond = 298 K and η II = 0.6. (2) The energy required in the CO 2 compression has been simulated in Aspen Plus as reported in [14].
Matteo Gazzani et al. / Energy Procedia 63 ( 2014 ) 1084 1090 1089 4. Results and Discussion The results in terms of energy penalties of the standard CAP and the crystallizer CAP are shown in Figure 4. Compared to the state-of-the-art CAP without solid formation, the mass flow sent to the stripper is reduced by about 40% thanks to the higher CO 2 loading, thus decreasing the heat required for regenerating the rich solution. The penalization on the steam turbine power output reduces accordingly. In the standard CAP the decrease in the steam turbine power due to the reboiler duty accounts for 70% of the total energy penalty whilst in the crystallizer CAP it decreases to about 51.1% (based on the total energy penalty of the standard case). On the other hand, in order to trigger the precipitation in the crystallizer, the entire solution exiting the absorber has to be cooled thus entailing a large energy consumption because of the chilling cycle. The associated energy penalty increases from 3.9 to 19.7 percentage points. The pump and waste heat management play a secondary role; anyhow, the crystallizer CAP allows reducing significantly the pump power thanks to the smaller circulating flow. When the desorber pressure is constant in the two CAP solutions, no differences arise in the CO 2 compression section. However, it is worth mentioning that an increase in the desorber pressure of the crystallizer CAP does not suffer of a higher pump power thanks to the reduced solvent flow rate. The additional design cost would also be limited thanks to the smaller desorber size. Another important feature of the crystallizer CAP is the reduction in the consumption of the ammonia slip process. In fact, the CAP with solid formation can further reduce the ammonia slip in the top part of the absorber thanks to the lower temperature and higher flow rate of the pump around. Accordingly, the consumption decreases from 10 to 5.7 percentage points for the standard and crystallizer CAP respectively. The resulting overall energy penalization is reduced by about 10% adopting a second generation CAP with the crystallizer. Moreover, the chilling duty can be further reduced by improving the heat integration within the whole plant. Figure 4: Comparison of the energy penalties between the standard and crystallizer CAP; bar length is normalized based on the standard CAP penalty = 100, whilst numbers within the bar refer to the total penalty of each layout. 5. Conclusions This work discussed the development of an advanced CAP layout where a crystallizer for solid formation is adopted to reduce the capture energy penalization. Two plant layouts were simulated in Aspen Plus with the Extended UNIQUAC model and compared using a reference USC power plant without CO 2 capture. Compared to
1090 Matteo Gazzani et al. / Energy Procedia 63 ( 2014 ) 1084 1090 the standard CAP, the advanced solution has shown a reduction in the energy penalization of about 10%. Further energy saving can be pursued increasing the crystallizer heat integration with other components. Nevertheless, in order to further develop this concept more work has to be done: i) the solid formation kinetics have to be investigated, ii) the crystallizer design has be optimized and iii) the full process energy requirements have to be optimized by manipulating the plant operating variables. References [1] Darde, V., Thomsen, K., van Well, W.J.M., Stenby, E.H., 2010. Chilled ammonia process for CO 2 capture. Int. J. Greenhouse Gas Control, 4 (2) 131-136. [2] Valenti, G., Bonalumi, D., Macchi, E., 2009. Energy and exergy analyses for the carbon capture with the chilled ammonia process. Presented at Greenhouse Gas Technologies 9. Energy Procedia 1, 1059 1066. [3] Versteeg, P., Rubin, E.S., 2011. Technical and Economic Assessment of Ammonia Based Post-Combustion CO 2 Capture. Presented at Greenhouse Gas Technologies 10, Amsterdam, the Netherlands. [4] Mathias, P.M., Reddy, S., O Connel, J.P., 2010. Quantitative evaluation of the aqueous ammonia process for CO 2 capture fundamental data and thermodynamic analysis. Int. J. Greenhouse Gas Control 4, 174 179. [5] Valenti, G., Bonaumi, D., Fosbol, P., Macchi, E., Thomsen, K., Gatti, D., 2013. Alternative layouts for the carbon capture with the chilled ammonia process. Greenhouse Gas Control Technologies 11. Energy Procedia 37, 2076-2083. [6] Gal, E.; Jayaweera, I.; Chilled ammonia based CO 2 capture system with water wash system. US patent no. 2010/0083831 A1 [7] Black, S., Dube, S., Muraskin, D.J., Kozak, F., 2013. Method and system for removal of carbon dioxide from a process gas. US patent no. 2013/0028807 A1. [8] Sutter D., Gazzani M., Mazzotti M.: Ternary diagrams for the CO2-NH3-H2O system and their application to the Chilled Ammonia Process, IEAGHG Post Combustion Capture Conference 2013, Trondheim, Norway [9] Thomsen, K., Rasmussen, P., 1999, Modeling of vapor liquid solid equilibrium in gas aqueous electrolyte systems, Chem. Eng. Sci. 54 (12) 1787-1802 [10] Darde, V., Thomsen, K., van Well, W.J.M., Bonaumi, D., Valenti, G., Macchi, E. 2012 Comparison of two electrolyte models for the carbon capture with aqueous ammonia. Int. J. Greenhouse Gas Control, 8 61-72. [11] Jänecke, E. 1929. Uber das System H 2O, CO 2 und NH 3. Zeitschrift fuer Elektrochemie, 35, 332 334#716 728 [12] Sutter, D.; Gazzani, M.; Mazzotti, M.: Kinetics of solid formation in the chilled ammonia system and implications for a 2nd generation process, Greenhouse Gas Technologies 12, Austin, USA. [13] European Benchmark Task Force 2011: European best practice guide for assessment of CO 2 capture technologies. [14] Manzolini, G.; Macchi, E.; Binotti, M.; Gazzani, M. 2011: Integration of SEWGS for carbon capture in natural gas combined cycle. Part B: reference case comparison. Int J Greenhouse Gas Control 5(2) 214 25.