Latent heat of CO2 absorption using water lean solvents

“Most investigators who claim that water-lean solvents might be able to deliver lower regeneration heat duties employ the following equation to sustain their argument.Q≈Cp·ΔTΔq+LDMCO2·pDpCO2+ΔHCO2absMCO2

On the right-hand side of the equation above, the first term refers to sensible heat, the second term refers to latent heat and the last term refers to absorption (or desorption) heat. In literature, one will find investigators claiming that each one of these terms can be reduced by shifting from aqueous to water-lean solvents. Therefore, we shall look at each one separately, starting from the right and coming back to the left.”

“Many researchers such as Huang et al. [73], Babamohammadi et al. [89] and Wanderley et al. [59] have proposed that the low volatility of water-lean blends is proof that these solvents can be regenerated while incurring in less parasitic heat losses to diluent vaporization than aqueous amines do. In fact, the same has been proposed by Rivas and Prausnitz [34], who nevertheless advise that CO2 desorption should be carried with stripping gases in hybrid solvents. Tan et al. [86] suggest that their TEG–MEA solvent can be recovered at 10 kPa of CO2 with neither MEA nor TEG evaporation. However, it is worth considering how this would be feasible, as some pressure is certainly required if one does not want to operate the desorber at vacuum or with the aid of stripping gases.

Data obtained by Guo et al. [80] at 120 °C for water-free mixtures of 2-methoxyethanol and 5 mol∙l−1 MEA (≈ 31 %wt. MEA) is compared to data for aqueous 30 %wt. MEA and MEG + 30 %wt. MEA [59] in Fig. 15. Two distinct options are presented. The first one is for regeneration of the solvent in a reboiler operating at 100 kPa. In this scenario, the nonaqueous solutions could be regenerated with loadings of 0.26 mol CO2∙mol MEA−1 (for MEG) and 0.05 mol CO2∙mol MEA−1 (for 2ME). Aqueous MEA cannot be regenerated at 100 kPa and 120 °C because the partial pressure of the unloaded solvent is already above 100 kPa. The second one is for regeneration at 200 kPa, in which the nonaqueous solutions would be possibly regenerated with loadings of 0.33 mol CO2∙mol MEA−1 (for MEG) and 0.19 mol CO2∙mol MEA−1 (for 2ME), while the aqueous one would achieve the loading of 0.22 mol CO2∙mol MEA−1.”


Fig. 15. Loading of aqueous MEA and MEG + MEA [59] compared to that of 2ME + MEA [80] against total pressure at 120 °C.”

Fig. 15 shows that water-lean solvents are not alike. Both MEG and 2ME have very low partial pressures when compared to water. Both allow for regeneration at 100 kPa, whereas water does not. However, a crucial point often made is that it is advantageous for CCS separation processes to operate the regeneration step at higher pressures [198]. The reason is that, after its separation, CO2 must necessarily be compressed for storage or injection. Compression work is energetically more costly when done with electric power than when done with heat, since there are inherent losses in the heat-to-electricity gas turbine cycle. Therefore, though it is possible to recover CO2 at 100 kPa with both these water-lean solvents, it would be more interesting for industrial applications to recover CO2 at 200 kPa.

At 200 kPa and 120 °C, the lean loading obtained with 2ME is marginally smaller than that obtained for aqueous 30 %wt. MEA, and both are smaller than that obtained with MEG. The determinant factor in this case is clearly the slope of the VLE curves. This is a point that is often overlooked when searching for diluents with low volatility. Although the boiling point of 2-methoxyethanol (124 °C) is lower than that of ethylene glycol (198 °C), 2-methoxyethanol will enable a lean loading smaller than that of aqueous MEA when operating in a conventional reboiler at 200 kPa and 120 °C, while MEG will not. Unless, of course, that CO2 desorption is performed with a stripping gas or under vacuum. A similar discussion has been carried out in Wanderley et al. [199].

Evidently, higher lean loadings such as the one found for MEG + MEA will directly impact every aspect of the CO2 capture plant, as more solvent becomes necessary to perform the same absorption task. Counter-intuitively, picking the diluent with lowest volatility might backfire in terms of avoiding high reboiler duties.

At pressures of 300 kPa and above, both nonaqueous solvents are indeed more competitive than aqueous MEA in the sense that they should deliver leaner loadings at 120 °C. Under this perspective, shifting from aqueous to water-lean solvents might be a solution for operating the desorption at higher pressures without necessarily having to increase temperatures, while still achieving reasonable lean loadings. This possibility has been pointed out by Barzagli et al. [5].”

Leave a Comment