Natural Gas Decarbonation with Absorption

Absorption-based CO, capture is the most mature technology currently available for CO, removal from NG process streams. CO, separation from raw NG in Chemical Absorption (CA) and Physical Absorption (PA) occurs through absorption of CO, into a solvent, followed by solvent regeneration. CA regeneration is accomplished by stripping heat at low pressures, whilst a simple pressure reduction can recover the solvent in PA (Al-Mamoori et al.. 2017). The separation mechanism in CA depends on reversible exothermic chemical reactions involving CO,, the alkaline active component (e.g., amine or potassium carbonate) and water, creating weakly bonded intermediates at high CO, fugacities in absorber columns. The intermediates are dissociated at low CO, fugacities in heated stripper columns, which has two finalities: Vaporization of water as the CO, stripping agent and reversal of the exothermic absorption equilibrium to regenerate the solvent (de Medeiros et al., 2013a). In the stripper, the vapor fugacity of CO, is reduced, reversing the chemical absorption reaction and destroying the CO,-amine intermediate, releasing low-pressure CO,. The CO,-lean solvent is recycled to the absorber (Yeo et al., 2012).

Table 1 shows the main CA solvents in use for CO, removal fr om NG. Aqueous amine and aqueous potassium carbonate solvents are presently in commercial scale and widespread in CCS applications, with large-scale demonstration plants worldwide (Araujo and de Medeiros, 2017). In NG applications, CA with aqueous monoethanolamine (MEA) has been the main CO, capture process, with a shift to aqueous methyl-diethanolamine (MDEA) in recent years (Olajire, 2010). Figure 3 depicts a typical CA process, where a CO,-lean solvent (e.g.. aqueous MEA. 25-30%w/w) enters the top of the absorption

Table 1. Main chemical absorption solvents. Adapted from: Polasek and Bullin (2006).

MEA = monoethanolamine; DEA = diethanolamine; DGA = diglycolamme; MDEA = methyl-diethanolamine.

Figure 3. Typical CO, capture by chemical absorption.

tower operating from 40 °C to 50 °C, while the raw NG flows bottom-up in countercurrent (D'Allesaudro et al., 2010). Raw NG must be cleansed of NO_;. and SO_x since they irreversibly react with ethanolamines, forming stable salts that cannot be reclaimed during regeneration (Rochelle, 2009). The CO,-rich solvent flows from the absorber bottom to the stripper, where lean aqueous etlianolamine is regenerated.

The stripper operates at high temperatures (100 °C-140 °C) and low pressures ranging from 1.5 bar to 1.8 bar absolute (Olajire, 2010) in order to reduce the stripping temperature while enhancing the stripping action of boiling water and prer enting thermal degradation of the solvent. Hie CO,-rich amine enters at the second stage of the stripper and flows in countercurrent to hot vapors rising from the reboiler (Songolzadeh et ah, 2017). An important process parameter is the CO, loading of the amine solution, shown in equation (1) for ME A. defined as the ratio between the number of moles of absorbed CO, and the number of moles of the active component in the solvent (e.g., MEA). translating how much CO, is absorbed per unit of active component in the liquid. Lean solvent enters the absorber with loadings ranging from 0.2 to 0.3 mol CO, per mol amine and leaves with loading = 0.5 mol COyinol amine (Kothandaraman, 2010).

Low-pressure CO, leaves the top of the stripper and must be compressed up to 150 bar (Yu et ah,

2012) or beyond (Pinto et ah. 2014) for EOR applications. Ол ег 90% CO, recovery can be obtained with CA in NG applications (Rochelle, 2009), requiring heat-ratios of 3-3.5 GJ/tCO, for solvent regeneration if MEA is used (Harkin et ah, 2010). Process improvements may reduce the heat-ratio to ~ 2.6 GJ/tCO, (Araujo and de Medeiros. 2017).

The main difference between CA and PA is the type of absoiption mechanism. While CA involves reversible chemical reactions, PA occurs through a physical solubility interaction governed by the CO, fugacity in the vapor phase. Solubility increases as CO, fugacity rises and temperature decreases. While CA is better than PA at lower CO, fugacities (Olajire, 2010), a CO, fugacity threshold exists at which PA has a higher CO, absoiption capacity then CA, which depends on the solvent. Usual physical solvents used in PA are methanol (Rectisol®), dimethyl ethers of polyethylene glycol (Selexol®), n-methyl-2- pynolidone (Purisol®) and propylene carbonate (Fluor Solvent™). Differently from CA, PA has low CO,/CH, selectivity resulting in high hydrocarbon losses from NG, giving CA a competitive advantage.

Benefits and shortcomings

CO, capture through PA and CA benefits fr om technology maturity derived from nearly 60 years of use (Yu et al., 2012), with a great variety of solvents and additives that have been optimized for increased efficiency and high CO, selectivity (Olajire, 2010). Absoiption schemes are easily retrofitted to older process plants and accept a wide range of CO, fugacities, with CA being best suited to lower fugacities and PA to higher fugacities.

The mam drawback of absorption-based technologies is the high temperature and pressure swings applied in order to regenerate the CO,-rich solvent, requiring external energy inputs despite the high heat- integration in NG processing (Al-Mamoori et al., 2017). Another aspect is the large footprint of PA and CA. normally hindering its use in CO,-rich NG processing in offshore platforms that are constrained in area and weight (Ar aujo et al., 2017). One example of the large footprint is Schlumberger's amine-based process for NG decarbouation, which requires a 17 m absorber and 16 m stripper (Schlumberger, 2016).

In CA, solvent degradation occurs through salt formation with sulphur- and nitrogen-based impurities and oxidation due to the presence of small amounts of O,. Losses via salt formation are minimized by NG pre-treatment to remove NO_x and SO_x impurities (Forsyth et al., 2017). Oxidative amine degradation is minimal as NG has low O, (< 0.2%mol) contents (Yu et al., 2012). A critical operational parameter is the feed CO, content (Rochelle. 2009), which affects solvent recirculation rate, heat consumption and equipment size. Time-varying systems—e.g., CO, removal from CO,-rich NG dining long-term EOR (Figure 2)—inevitably result in suboptimal operation of the designed absoiption system, resulting in reduced CO, capture efficiency and increased heat and solvent consumption over time. Salvindera et al. (2017) used a composition controller to show that, for CA-MEA (C A with ME A 20%w/w in water), CO, content in the lean and CO,-rich gases would greatly vary along operation tune.

Recent focus

Research focus on CO, absoiptiou consists of new process configurations to minimize heat demand (e.g., absorber intercooling, split-flow and heat integration). Moullec and Kanniche (2011) evaluated process flowsheets for CA-MEA, concluding that process modifications must be coupled with new solvents and heat-integration with the power plant. Shanna et al. (2016) used multi-objective optimization of NG decarbonation to reduce the energy penalty at high CO, absoiptiou requirements, demanding energy mtegration/cooling schemes.

Research into new solvents aims to increase CO, equilibrium loading, solvent chemical resilience and decrease stripping heat duty and corrosiveness. Adeyami et al. (2016) used a hybrid PA/CA mechanism allowing for higher CO, loading and reducing heat requirements for stripping. Chowdbury et al. (2011, 2013) focused on how side groups in ethanolamines (e.g., methyl, ethyl, propyl) affect the CO, absoiptiou rate, cyclic capacity and enthalpy of reaction. Blended amines, additives, and special amines (Bouzina et al., 2015) are another flout of solvent development. Zhang et al. (2012) studied phase-changing homogeneous amine blend that exhibits phase separation upon CO, absoiptiou and only the CO, rich-phase enters the shipper, reducing reboiler duty. Adeyami et al. (2016) synthesized three amine-based deep eutectic solvents for CO, capture with enhanced environmental performance. Another attempt to minimize heat demand is the use of uon-aqueous solvents (NAS), aiming to reduce the heat of vaporization, partially responsible for the heat demand in the stripper (Lail et al., 2014).

Although no commercial application in CO, capture exists and, despite their high cost, ionic liquids are a promising alternative due to their thermal and chemical stabilities (Breunecke and Gurkau. 2010). Research is currently underway as CO, capture may occur through physical and or chemical absorption, and innovative schemes explore its high degradation temperature, allowing high-pressure stripping, reducing compression costs (Barbosa et al., 2019).

Concerning sustainability aspects, Marx et al. (2011) reviewed the literature on lifecycle assessment (LCA) of CCS aud noticed that most studies considering the NG industry focus ou post-combustion capture from flue-gas rather than NG upgrading. Karl et al. (2009) studied the worst emissions cases using common aqueous ethanolamine solvents (MEA, MDEA aud DEA) at a Norway gas-fired power plant, concluding that insecure CA processes can negative affect human health and are devastating to local vegetation and aquatic algae. Zakuciova et al. (2015) investigated CCS technologies in pre-combustion, post-combustion and oxy-combustion applications, concluding that freshwater eutrophication is the main CA environmental issue due to amine toxicity.

Kahn et al. (2017) studied hybrid solvents for CO, stripping at high pressure which could reduce compression costs in geological storage applications. Mazzetti et al. (2014) analyzed the economics of treating North Sea NG with CA-MEA, concluding that, under the Norwegian CO, taxes, the technology is adequate for EOR. Araujo et al. (2017) comparatively analyzed separation technologies for processing CO,-rich NG in ultra-deepwater oil-gas fields. In then analysis, CA exhibited the lowest hydrocarbon losses aud the lowest specific electric power consumption at the expense of the highest footprint for all investigated scenarios.