Background of energy performance and related absorber-desorber configuration

https://doi.org/10.1016/j.seppur.2021.118959

”

Additionally, despite the fact that NH₃-based capture processes may reach specific reboiler energy demands as low as 2.1 MJthkgCO2captured−1 for solvent regeneration, the energetic optimization of the CAP that avoids solid formation applied to coal-fired power plants and to cement plants has shown that approximately 2/3 of the overall energy demand of the process, expressed as specific equivalent work, is associated with the heat requirements in the reboiler of the CO₂ desorber [14]. The remaining 1/3 is provided by means of electrical power. Aiming at decreasing the energy required for the regeneration of the solvent, advanced process configurations originally developed for capture technologies using aqueous amine solvents [29], [30] have been implemented in NH₃-based capture processes. Among them, the RSS, whose implementation has minor effect on the capital costs and operability of the process, has been proven to decrease significantly the energy requirements of NH₃-based capture processes when the split fraction of cold CO₂-rich stream bypassing the rich/lean heat exchanger is optimized together with the pressure of the CO₂ desorber and with the feed stage of the hot CO₂-rich stream to the CO₂ desorber [14], [27]. Li et al. assessed the performance of the Inter-heated Desorber (IHD) [2] and of the Cold Rich Solvent Preheating (CRSP) [31], alone and in combination with the RSS, thus obtaining specific reboiler duties above 2.4 MJthkgCO2captured−1 in all cases. Liu [23] assessed the performance of the IHD in mild-temperature NH₃-based capture processes and obtained energy savings of almost 22% with respect to the simplest absorber-desorber configuration. Nevertheless, the IHD achieves negligible energy savings when implemented together with the RSS at optimal conditions [27]. Jiang et al. [4], [21] evaluated the performance of the Advanced Flash Stripper (AFS), originally developed by Rochelle [32] for CO₂ capture with aqueous piperazine, in combination with the RSS, which led to specific reboiler duties for the regeneration of the CO₂-lean NH₃ solvent as low as 1.9 MJthkgCO2captured−1. Although very competitive in terms of reboiler duties, the results obtained by Jiang et al. [21] can be questioned because their simulations: (i) neglected the additional energy requirements associated with the solvent recovery from the solvent stream purged to avoid water accumulation, (ii) did not constrain the NH₃ concentration in the CO₂ stream to avoid corrosion in pipelines and equipment during CO₂ compression and transportation [33], and (iii) modelled the counter-current rich/lean heat exchanger by setting the cold side temperature approach to 3 °C and neglecting the internal vaporization of the hot CO₂-rich solution, which does not allow to detect infeasible pinch point temperatures within the rich/lean heat exchanger below 3 °C or even temperature crossovers [14]. As a result, the heat recovered in the rich/lean heat exchanger is overestimated and the reboiler duty underestimated. In addition, the AFS was only assessed in combination with the RSS so that the stand-alone effect of the feed stage of the hot CO₂-rich stream to the CO₂ desorber cannot be determined; on the contrary, independent investigations have found stage 7 out of 10—including the reboiler—from the top of the column as the optimal feed stage for the hot CO₂-rich stream to the CO₂ desorber, which minimizes the energy consumption required for solvent regeneration [14], [27]. Finally, Ullah et al. [34] assessed the performance of the Lean Vapour Compression (LVC) and of the Rich Vapour Compression (RVC), and reported energetic savings of 15% and 3%, respectively, with respect to the simplest absorber-desorber configuration in NH₃-based capture processes. Similarly, Nguyen et al. [35] reported more than 43% energy savings for the LVC. Nevertheless, when implemented together with the RSS at optimal conditions, Liu et al. [27] reported negligible energy savings for the LVC and even negative values for the RVC, with respect to the process with the RSS. The LVC and the RVC decrease the energy required for solvent regeneration at the cost of increasing electrical demand for vapour re-compression. Therefore, their performance depends strongly on the features of the electricity accessible in the environment in which the capture plant is available, i.e. cost and associated CO₂ emissions. In addition, other authors have investigated other advanced process configurations specific to NH₃-based capture processes that aim at improving the energy consumption by decreasing the reboiler duty of the CO₂ desorber. Namely, Liu [36] proposed to compress the vapour exiting the top of the NH₃ desorber and inject it to the bottom of the CO₂ desorber, in combination with the LVC.

Nevertheless, all the aforementioned advanced configurations of the CO₂ desorber have been analysed in the literature by means of single-variable sensitivity analyses that do not take into account the interdependencies among different process variables, and have been compared to reference processes whose configuration and operating conditions have not been optimized previously, thus hindering comprehensive assessments and conclusive results.

As far as the source of heat required for solvent regeneration is concerned, the most cost efficient strategy is the use of saturated steam whose temperature must be above—at least by 10 °C—the reboiler temperature of the CO₂ desorber. In the case of NH₃-based capture processes, the optimal regeneration of the aqueous NH₃ solution that allows for the minimal energy consumption takes place at pressures around 20 bar and temperatures between 140 and 150 °C [14]. Similar steam temperatures are required by other amine-based capture processes, where the reboiler temperature in the CO₂ desorber ranges between ca. 120 °C, when using aqueous monoethanolamine (MEA) solutions as absorbent [37], and 150 °C, for aqueous piperazine [38]. When applied to power plants, the steam required for the CO₂ capture process is extracted from the low pressure turbine of the electricity generation island. Increasing reboiler duties and temperatures penalize the electricity throughput thus worsening the energetic performance of the capture process. In the case of applying the capture process to a steelworks facility, the steam required for solvent regeneration can also be extracted from the existing power island. While the electricity required can be imported from the grid if there does not exist a power plant in the vicinity, the need of steam hinders the application of solvent-based CO₂ capture processes to other industrial CO₂ intensive point sources where the steam required might not be available, for example in the case of cement plants, thus leading to additional capital costs related to on-site steam production [39]. In such cases, importing steam from a neighbouring Combined Heat and Power (CHP) plant has been found to decrease, between 10% and 20%, the cost of CO₂ avoided in cement plants when using solvent-based post-combustion technologies for CO₂ capture [40]. Unfortunately, the vast majority of existing cement plants are not located in the vicinity of a power plant [39], i.e. less than 10% of the existing European cement plants [33]. The competitiveness of solvent-based CO₂ capture processes can also be favoured by the possibility of integrating the excess heat available in the CO₂ point source. In the case of a typical European cement plant using the Best Available Technology (BAT), it has been reported that approximately between 5 and 10% of the steam required for solvent regeneration could be generated at limited cost and complexity using the excess heat of the clinker cooler exhaust air [39], [40]. Norcem Brevik is an extreme, still rare, case of a cement plant where up to 50% of the steam required for solvent regeneration in the capture plant could be generated by the excess heat available in the cement plant [41]. Other CO₂ intensive industries such as refineries and iron and steel plants also have process streams whose excess heat can be used for solvent regeneration. The amount of excess heat that can be integrated with the capture plant increases for lower solvent regeneration temperatures: On the one hand, the number of process streams whose heat cannot be used internally increases and, on the other hand, the pinch point temperature with respect to the capture process increases [42], [43]. In fact, a solvent regeneration temperature around 90 °C or below has been reported to increase significantly the use of excess heat in the iron and steel industry [43]. Arasto et al. [44] simulated a hypothetical solvent with regeneration temperature of 70 °C for CO₂ capture in an iron and steel plant in order to show how low solvent regeneration temperatures enable the application of solvent-based capture processes even for relatively high solvent regeneration duties. In case of application to the power sector, the condenser of the power plant might supply heat for solvent regeneration at temperatures around 50 °C or below [3]. Therefore, the performance of the advanced configurations that affect the energy needs of the capture plant and the comparison among them depend not only on the steam and electricity requirements, but also on: (i) the temperature of the steam required, thus on the solvent regeneration temperature, (ii) the features of the available electricity and steam, i.e. the primary energy source and the associated CO₂ emissions, and (iii) the amount and temperature of the excess heat that can be recuperated from the CO₂ point source and used for solvent regeneration.

“