ACACIA Project – Development of a Post-Combustion CO2 Capture Process. Case of the DMXTM Process

—The objective of the ACACIA project was to develop processes for post-combustion CO 2 capture at a lower cost and with a higher energetic efﬁciency than ﬁrst generation processes using amines such as MonoEthanolAmine (MEA) which are now considered for the ﬁrst Carbon Capture and Storage (CCS) demonstrators. The partners involved in this project were: Rhodia ( Solvay since then) , Arkema , Lafarge , GDF SUEZ , Veolia Environnement , IFP Energies nouvelles , IRCE Lyon , LMOPS , LTIM , LSA Armines . To validate the relevance of the breakthrough processes studied in this project, techno-economic evaluations were carried out with comparison to the reference process using a 30 wt% MEA solvent. These evaluation studies involved all the industrial partners of the project, each partner bringing speciﬁc cases of CO 2 capture on their industrial facilities. From these studies, only the process using demixing solvent, DMX TM , developed by IFPEN appears as an alternative solution to the MEA process.

aqueuse de MEA a`30 % poids.Ces e´tudes ont e´te´re´alise´es avec les partenaires industriels du projet, chacun apportant des cas concrets de captage du CO 2 sur leurs installations industrielles.De ces travaux, seul le proce´de´liquide de lavage des gaz par solvant demixant, DMX TM , qui est en cours de de´veloppement a`IFPEN apparait comme une solution de rupture alternative au proce´de´de lavage a`la MEA.

INTRODUCTION
The ACACIA project, which was launched by the AXELERA cluster (AXELERA is one of the 71 French "competitiveness cluster" initiated by the French government in September 2004), is dedicated to the development of new processes for CO 2 capture on industrial facilities before geological storage.As considered by IEA [1], Carbon Capture and Storage (CCS) is one of the possible pathway in order to mitigate greenhouse gases emissions; it thus requires the development of high efficiency CO 2 capture technologies.
The ACACIA project partners have chosen to consider only the post-combustion capture pathway for which the CO 2 is extracted directly from the industrial flue gases [2].The main objective of the project was the research and development of new technologies to reduce the cost of capture per ton of CO 2 and the impact of CO 2 capture on the cost of electricity or industrial products (cement, chemicals).It is known from process studies [3] or from pilot demonstration [4], that the energy penalty reduction, especially due to the energy required at reboiler, estimated about 3.7 GJ/ton CO2 for the MEA (MonoEthanolAmine) 30 wt% process, is the key issue for making CO 2 capture an attractive solution for carbon mitigation.
Conventional processes for CO 2 capture are based mostly on absorption by a chemical solvent.Chemical solvents used are primary amines, and in particular MEA.If the MEA can recover up to 98% CO 2 and obtain a purity of 99.9%, its use leads to high operating costs.In the medium term, to make CCS deployment possible, it is necessary to develop new capture processes with lower energy costs further reducing the cost of carbon capture.The purpose of the ACACIA project was to develop new processes with a cost of capture 50% lower than the cost of existing processes while allowing to capture at least 90% CO 2 in the treated gas and obtain a CO 2 purity near 95%.Such a purity level is necessary for transport and storage.
Five types of processes were studied in the ACACIA project: -demixing solvents: use of amine solvents which either for high CO 2 loadings or for high temperature form two non-miscible phases.With this type of solvents, only the heavy CO 2 -rich phase is regenerated which reduces the energy cost of carbon capture [2]; -hydrates: research on thermodynamic additives to improve the operating conditions of CO 2 capture by hydrates; the objective being to capture CO 2 at low temperature and moderate pressure and deliver CO 2 at high pressure with low energy inputs which would reduce the cost of regeneration and CO 2 compression; -enzymes: use of enzymes, which are immobilized in porous materials, to enhance CO 2 absorption with in particular an increase in CO 2 absorption kinetics and an associated investment reduction; -ionic liquids: optimizing the absorption of CO 2 by the use of some ionic liquids offering high solubilities; -innovative chemistry: development of new solvents with innovative chemical routes for CO 2 capture with low enthalpy of formation requiring less energy at regeneration step.Through these lines of research, the ACACIA project aimed to develop solutions applicable in priority to the industry (power plants, cement plants, incinerators, and chemical industry).To validate the relevance of the new processes studied in this project, a benchmark based on a techno-economical study between MEA technology and new processes was undertaken by the industrial partners of the project, each bringing specific cases of CO 2 capture corresponding to an industrial case.The pooling of these cases and appropriate technological solutions was a very important part of this project because it allowed the identification of viable pre industrial study technological solutions: a validation process pilot type being envisaged only after the end of ACACIA depending of the results obtained.
In this paper, we present the results of the technoeconomical study carried out on the DMX TM process and the comparison with the reference process using 30 wt% MEA.For all other original routes studied (hydrates, enzymes, ionic liquid), it has not been possible to obtain sufficient data for performing process evaluation and techno-economic evaluation.
In a first part, a description of the operating conditions for the MEA (1 st generation of chemical solvent) and DMX TM (2 nd generation of chemical solvent) processes is provided.A second part is dedicated to a description of the emission case studies, the study methodology and economic assumptions.In the last part, a comparative analysis between the MEA and DMX TM processes is presented.

MEA Process Description
To separate CO 2 from the flue gas (low pressure, low CO 2 content), the reference process is a chemical absorption process using 30 wt% MEA as solvent.It is widely admitted that this process is the reference for CO 2 capture on flue gases [3].Within the CASTOR and CESAR FP7 European projects, this process has been demonstrated at pilot plant scale on real power plant flue gas [4][5][6] and some companies are able to commercialize such a process with already some large scale references existing in the food industry [7,8]. Figure 1 shows a typical process flow diagram for a first generation process such as the reference MEA 30 wt% process.The capture process is composed of five main sections: -a cooling tower which purpose is first to cool down the flue gas issued at 140°C and second to perform a preconditioning of the flue gas (washing of ash, impurities, etc.); -an absorber, operated at ambient pressure and moderate temperature, where CO 2 is separated from the flue gas by being contacted with the solvent; -a washing section which ensures that the decarbonized flue gas sent to the stack does not contain any unwanted pollutants (amines, degradation products or any other volatile compounds); -a regenerator operated at moderate pressure and high temperature, where CO 2 is separated from the solvent, the latter being regenerated; -a compression section needed to deliver high-pressure pure CO 2 ready for storage.All these five sections are specific to a given process and are interconnected.As an example, the cooling tower, using a first washing section, can be more or less important depending on the solvent sensitivity towards impurities contained in the flue gas such as SO x or NO x .In the same idea, a process using a volatile solvent may require a large washing section downstream the absorber, while a small section may be enough for others.Similarly, the operating conditions in the regeneration section, in particular in terms of pressure may impact the compression section.It is thus mandatory to consider all the needed sections for the process at constant boundary limits, inlet flue gas and outlet CO 2 and treated gas as shown in Figure 1, when making comparison.

DMX TM Process Description
The DMX TM process has been developed and patented by IFPEN (see [9] or [10] for process or physical and chemical basic information respectively).It has been described with further details in [11,12] and only a quick description is given hereafter.The main objective being to present the techno-economic comparison with the MEA process.
The DMX TM process is based on the use of very specific solvents which, for given loading and temperature conditions, can form two immiscible liquid phases.These phases have sufficient density differences that they can be separated by decantation.The light liquid phase is such that it contains almost no CO 2 , the latter being concentrated in the heavy phase.This result is similar to what could be obtained with a high capacity.The DMX TM solvent is also characterized by an easy separation which can be performed in a standard decanter placed downstream the lean/rich heat exchanger, downstream the absorption column, as can be seen in Figure 2. The decanter is preferably positioned after the amine/amine heat exchanger and before the regenerator in particular to make decantation easier via the reduction of liquid viscosity associated with the increase of temperature.Only the CO 2 rich loaded heavy phase is sent to the stripper, the CO 2 lean light phase being directly sent back to the absorber.Note that depending on the operating conditions chosen at stripper and at decanter, one may observe an important CO 2 gas release at decanter.The compression section is then modified turning into a possible supplementary energy reduction when the decanter is operated at a pressure higher than the pressure at stripper.
Figure 3 shows a picture of a transparent decanter that has been used on a mini-pilot at IFPEN.The flow goes from left to right as indicated by the plain arrow.As can be seen in the close view, the inlet flow contains CO 2 gas bubbles and droplets of the light phase dispersed in the heavy phase.On the right-hand-side, one observes that, very quickly, a clear separation of the phases is reached, the interface being indicated with a dashed line.
Such a process presents a significant decrease of solvent mass flow and of captured CO 2 sent to the regeneration column requiring less energy input.It can thus offer a significant cost reduction compared to the reference case that is the MEA 30 wt% based process.
The choice for the formulation of the demixing solvent DMX-1, was firstly based on its thermodynamic capacity which comes in addition to demixtion for reducing the solvent flow rate going to the stripper.Secondly, we paid a particular attention towards degradation performances.As shown in Raynal et al. [11], the DMX-1 degradation performances are much better than those of MDEA (MethylDiEthanolAmine), a commercial amine known as being much more stable than MEA.As discussed by Raynal et al. [12], degradation impacts many costs and not only solvent make-up; it makes possible the operation of the stripper at higher pressure/temperature operating conditions enabling CO 2 compression cost reduction.Last, kinetics performance and operability issues were considered.

CASE DESCRIPTION STUDY METHODOLOGY AND
ECONOMIC EVALUATION: MEA VERSUS DMX TM PROCESS

Case Description
Each industry has defined the gas to be treated by the CO 2 capture process considered within the ACACIA project.
The information on these gases, include: -the flow rate, the density at standard conditions, the pressure and temperature; -the molar composition; -the expected impurities (dust, SO x , NO x ).
This information is given for a nominal flowrate case of with a range of expected changes to account for the flexibility of the units.
The emitting industries concerned were the following: -electricity production by gas-fired power plant (GDF SUEZ) and coal-fired power plant (Electrabel GDF SUEZ); -production of cement (Lafarge); -chemistry (Rhodia Operations and Arkema); -incineration of household waste (Veolia Environnement).Table 1 summarizes the characteristics of the flue gases to be decarbonized, which were given by these different industries.
For geological storage application, the CO 2 delivery pressure at battery limit was at 110 barg.
Compression energy to this pressure level was of course taken into account within the techno-economic evaluation.

Evaluation Methodology
The economic assessment methodology implemented in the ACACIA project allowed to establish a strong synergy between academic and industrial partners.
The stakeholders were as follows: -For the MEA process: IFPEN conducted the process studies and economic evaluation for Veolia Environnement, Arkema and Rhodia cases; GDF SUEZ has completed the design of facilities for the collection for PC and NGCC plants; Rhodia generated sizing and quantification of the Lafarge plant; Lafarge got his own experience of the design of the system and has estimated the cost of the MEA solution and proposed the cost estimation derived from that produced by Rhodia, for the purposes of Solvay (without compression step); -For the DMX TM process: Among the available demixing solvents IFPEN, proposed to consider the DMX-1 system which is currently the best solvent for the DMX TM process.IFPEN provided to partners who have chosen the DMX TM process solution (GDF SUEZ, Veolia Environnement, Lafarge), the mass and energy balances as well as the sizing of the main equipments: -IFPEN provided balance sheets and equipment sizing devices to GDF SUEZ and Lafarge; -Lafarge and GDF SUEZ made their own economic evaluation based on data provided by IFPEN for their respective cases; -IFPEN performed the entire study for the Veolia case.

Study Basis and Economic Assumptions
Emissions flows are described in the basis for studies cited above.
The economic assumptions are detailed in Table 2.The cooling water is available on site in sufficient quantities to ensure the capture units needs.Without any previous specification, the temperature of the cooling water was taken equal to 15°C (sea water) and the maximum elevation of the cooling water was set at 10°C to reach a final temperature of 25°C.For the power plant cases (GDF SUEZ cases) the Low Pressure (LP) steam needed for the reboilers comes from thermal power, which reduces the production of electricity from the LP turbine.The power consumption of the various equipment is provided by the power plant.
For the other cases, steam comes from a steam generator dedicated to the CO 2 capture plant.
(*) The power consumption of the capture units/compress CO 2 is considered a loss for the thermal power plant.
(**) TEG unit is devoted to dry CO 2 stream before transportation and injection.

COST EVALUATION AND ANALYSIS
The overall CO 2 cost, expressed in 1/t CO 2 , is obtained by combining CAPEX and OPEX costs in a complete economic analysis.The obtained value corresponds to the minimum price of CO 2 on market for which a CCS project is profitable.That is the minimum price for which it is more interesting to invest in a CCS project rather than buying CO 2 emissions rights on the market.In the present analysis, we split the different costs for the main contribution in the overall CO 2 cost to emphasize the advantages and weaknesses of a given process.

MEA Evaluation
The techno-economic studies confirmed the very high cost of CO 2 capture for the reference 30 wt% MEA process, whatever the considered case.The cost of capture by amine scrubbing (Tab.3) ranges from 39 1/tCO 2 to 239 1/tCO 2 .This is related, in the first analysis, to the scale effect.Indeed, Arkema case deals only 2.5 tCO 2 /hour of CO 2 while the coal-fired plant emits 582 tCO 2 /h.These two extreme costs define the minimum and maximum value for capture costs.A case as small as the Arkema case which could correspond to a demonstration case is associated to a very high CAPEX in particular due to building, instrumentation and control costs almost as expensive as a very large case.Otherwise, for comparable emission flow rate, costs range from 63 1/tCO 2 for Veolia to 91 1/tCO 2 for Lafarge.The explanation here comes from the fact that Veolia has lower operating costs related to the integration of the production of steam for regeneration.

DMX TM Evaluation
The GDF SUEZ coal-fired plant case, the Veolia and Lafarge cases were considered for the DMX TM process.
It appears that the DMX TM process could be a very interesting technology.The evaluation of this process showed significant gains on the cost of CO 2 capture as one can observed in Table 4. Indeed when compared to the MEA process, it appears that the DMX TM process can offer reduction of À20% and up to À50% in CO 2 capture cost.So, with this breakthrough technology, it is possible to meet part of the initial goal of the ACACIA project: 50% of the cost of CO 2 capture.Some comments can be made on these results: -About Lafarge case (cement plant): The cost of CO 2 capture is halved with DMX TM compared to the reference MEA process, which was the objective of the project.It is interesting to discuss how such a gain can be explained.Three main reasons can explain this result: a small part is due to the investment, a little lower for DMX TM : À4%.(for the Lafarge case, it is necessary to build a boiler in order to generate the steam necessary for the solvent regeneration/size of this boiler is reduced for the DMX TM process); the most important part corresponds to variable costs and especially steam: À20%; the low possibility of heat integration between the cement plant and the DMX TM process explains the better performance of the process DMX TM , by a significant reduction of steam consumption from utility device; -About GDF SUEZ case (coal power plant): The gain for the plant performance related to the use of DMX TM process is 1.3% for the thermal overall efficiency (see Tab. 4 above).IFPEN expects to have more than 2 points performance gain with an innovative heat integration with the power plant steam cycle.The cost of captured CO 2 is estimated at 37.  However, the cost reduction compared to MEA is not as important, from 63 1/tCO 2 to 52 1/tCO 2 .
The thermal integration is here already done for MEA case by steam extraction available on the incineration plant, the transition to DMX TM is less profitable even if it allows a significant reduction in cost.

CONCLUSIONS
With the DMX TM process patented and developed by IFPEN, it is possible to have significant energy savings compared to the reference MEA.This gain was 1.3% on energy penalty for the coal power plant but studies show that it is even possible to achieve a gain of 2%.Gains on operating costs (OPEX) enable cost reduction CO 2 capture 15 to 50% depending on the cases.
In order to go one step further in terms of process development, it is now necessary to perform industrial demonstration of the DMX TM process.This is one of the goal of the European FP7 OCTAVIUS project, which started on March 1 st 2012.Tests at large scale are scheduled to be performed in 2015-206 on the ENEL pilot plant in Brindisi which treats 10 000 Nm 3 /h of flue gases issued from a coal fired power plant (2.5 t/h CO 2 captured equivalent), and for which a important revamp is planned.

Figure 1
Figure 1 Simplified process flow diagram of the MEA post-combustion capture process.

Figure 2
Figure 2 Simplified process flow diagram of the DMX TM post-combustion capture process.

Figure 3
Figure 3 Picture of the decanter inlet of the mini-pilot of IFPEN.The three-phase flow, G/L/L, enters the decanter on the left-hand-side, the decantation being achieved in the large diameter section on the right-hand-side.

TABLE 2
1 1/tCO 2 for the reference MEA, and 31.4 1/tCO 2 for DMX TM process, that implies a decrease of 15.4% of the capture cost.
-About Veolia Environnement case (central waste incineration): The emissions flow rates are comparable to the cement plant and the cost of treatment with DMX TM is the same order of magnitude (52 1/tCO 2 captured).