Non-Aqueous Solvent (NAS) CO2 Capture Process Academic research paper on "Chemical sciences"

CS)

CrossMark

Available online at www.sciencedirect.com

ScienceDirect

Energy Procedia 63 (2014) 580 - 594

GHGT-12

Non-Aqueous Solvent (NAS) CO2 Capture Process

Marty Lail, Jak Tanthana, and Luke Coleman *

Energy Technology Division, RTIInternational, Research Triangle Park, NC 27709-2194 USA

Abstract

The Non-Aqueous Solvent (NAS) CO2 Capture Process is being developed as an advanced, next-generation post-combustion capture technology for CO2 removal from power generation and industrial flue gas streams. The core of the NAS process has been the development of a family of CO2-selective NASs that have the potential to substantially reduce the thermal regeneration energy demand associated with solvent regeneration to ~ 2,000 kJt/kg of CO2 captured. This paper discusses our rationale for selecting a NAS platform and provides results from the physical and thermodynamic characterization of the NASs, lab-scale testing results, and an update on our efforts to address specific process design challenges related to NAS processes for evaluation of the NAS CO2 capture process at the (large) bench-scale.

©2014TheAuthors. PublishedbyElsevierLtd.Thisis 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

Keywords: CO2 capture; Non-aqueous Solvent; Novel Systems; Absorption; Process Evaluation; Solvents

1. Introduction

With pending US EPA regulations limiting CO2 emissions for new electricity generating units (EGUs) to ~1,000 lb CO2 / MWh, consistent with natural gas combined cycle (NGCC), which is described by the EPA as the Best Available Control Technology (BACT) for CO2 emissions from fossil fuel-fired EGUs [1], development of CO2 capture technologies that cost-effectively reduce CO2 emissions from fossil fuel-fired EGUs represents an enormous opportunity for mitigating greenhouse gas emissions and ultimately global climate change. The development of technologies that cost-effectively reduce CO2 emissions from coal-fired power plants is very important to retaining coal-fired power plants within the US's power generation portfolio if climate change regulations are enacted. Current state-of-the-art post-combustion CO2 capture technologies are at least an order of

* Corresponding author. Tel.: +1-919-541-6840; fax: +1-919-541-8002. E-mail address: lcoleman@rti.org

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.2014.11.063

magnitude smaller than required for EGUs and are prohibitively expensive. Implementation of current control technologies would result in an increase in the cost of electricity (ICOE) for consumers in excess of 80% [2].

The ICOE and cost of CO2 captured is driven primarily by the high parasitic power load associated with releasing CO2 from the solvent during solvent regeneration and the high capital costs associated with the scale and materials of construction of the process equipment [3]. In conventional aqueous-based CO2 capture processes, the majority of energy consumed is associated with the reboiler heat duty [4]. A recent comprehensive review of solvent-based post-combustion CO2 capture processes provided an overview of worldwide R&D efforts and identified development of novel solvents as having the highest potential for reducing the costs of CO2 capture [5].

For a solvent-based CO2 capture process to achieve a substantial reduction in the cost associated with the capture and compression of CO2 compared to a state-of-the-art (SOTA) CO2 capture process, the energy required for solvent regeneration—the largest contributor to the cost—must be drastically reduced while simultaneously reducing the capital cost and limiting increases in the operating cost. In conventional aqueous-based CO2 capture processes, the energy required for solvent regeneration is delivered to the solvent via a reboiler. The reboiler heat duty consists of three main contributors, specifically 1) the sensible heat required to heat the solvent to the regeneration temperature, 2) the heat of vaporization for the stripping component, and 3) the heat of CO2 absorption as described in eqn. (1) [6] below. This expression provides insight into the contribution of the physical and chemical properties of the solvent system to the reboiler heat duty and is a very useful tool to guide development of better-performing, lower-energy solvent systems. For a lower-energy solvent system, each of these parameters must be adjusted in the directions shown in Table 1. If a system is capable of achieving these criteria, it will have a lower energy penalty than the SOTA CO2 capture processes.

Reboiler Heat

Cp(TR-Tp) M

Mco? xsol

Sensible Heat

Heat of Vaporization

abs,C02

Heat of Reaction

Table 1. Approach to Identifying Lower Energy Solvent Systems

Solvent cp [J/g K] AHabs [kJ/mol] AHVap [kJ/mol] xsolv [mol solvent/ mol solution] Aa [mol CO2/ mol solvent] Reboiler Heat Duty [kJt/kg CO2]

30wt% MEA-H2O 3.8 85 40 0.11 0.34 3,750

Lower Energy Solvent System

Based on this assessment, a lower-energy solvent system would have a combination of a low heat of absorption (AHabs), a high CO2 working capacity (Aa), a low specific heat capacity (CP), and low heat of vaporization (AHvap), and would be concentrated (xsolv). It must be noted that several of these factors are not independent, and therefore, a holistic approach that simultaneously considers all contributing factors is necessary. However, the AHabs remains the largest contributor to the reboiler heat duty, 56% in the 30wt% MEA-water system [4], and therefore, to substantially reduce the solvent regeneration energy, the heat of absorption must be substantially reduced without adversely affecting the remaining factors—most importantly the working capacity.

Our approach to developing a lower-energy solvent system was to consider solvent formulations that holistically addressed all factors contributing to the reboiler heat duty. To this end, non-aqueous (organic) CO2-selective solvents that exhibit low heats of absorption (AHabs) while maintaining a high CO2 working capacity (Aa) under flue gas conditions and low regeneration temperatures were considered. Selecting an NAS platform is very promising since many of the factors in Table 1, specifically the specific heat capacity (CP), the heat of vaporization (AHvap), and solvent concentration (xsolv) cannot be adjusted in the required direction if the solvent is water based. Of significant importance for moving to a NAS platform is that promising CO2-absorbing components meeting many of the criteria described in Table 1 but having sparing water solubility and therefore a low xsolv with water could be considered. Another consideration in the selection of a NAS platform versus aqueous solvents was the decoupling of the solvent working capacity and the maximum CO2 product pressure from the vapor-liquid equilibrium of water, thus enabling the potential for lower temperature solvent regeneration. As such, using steam to strip the solvent and distribute heat in the regenerator couples the vapor-liquid equilibrium of water and the CO2 isotherm of the solvent. This has led to the development of aqueous-amine solvents that have very specific CO2 isotherms and ultimately to solvents that are regenerated above 100°C. However, if water, and more specifically a boiling agent, is removed from the solvent, solvent regeneration and the CO2 product pressure is determined by the regenerator temperature and the CO2 isotherm of the solvent.

An additional rationale for transitioning to a NAS platform can be explained by considering the CO2 isotherm for a solvent, which can be described by the temperature-dependent Gibbs free energy expression. While the solubility of CO2 in solvents is typically driven by the enthalpy change, the influence of temperature is dependent on entropy in addition to enthalpic changes including specific heat capacity. Although van't Hoff suggests that the effect of temperature on equilibrium (i.e., CO2 isotherm) is related to the enthalpy change (i.e., AHabs), the basic assumptions for this theory are that 1) the enthalpy is relatively constant over the temperature range and 2) the entropy change is negligible and independent of temperature. As such, entropically driven absorption reactions can be reversed with small changes in temperature and also have low enthalpies (i.e., heat of absorption). This has implications on the process design and can be used favorably to reduce the energy for solvent regeneration. If the solvent can be regenerated to produce a high-pressure CO2 product gas at lower temperatures, then lower-quality low pressure steam or alternative sources of waste heat can be used for solvent regeneration.

2. RTI's Non-Aqueous Solvents (NASs)

Since 2009, we've been developing a family of CO2-selective NASs with the potential to reduce the thermal regeneration energy demand to ~2,000 kJt/kg of CO2 captured. A comprehensive elimination-based solvent screening and characterization program was employed to down-select the most-promising NAS formulations. As part of the rapid screening stage, >200 NAS formulations were screened and several very promising candidates identified. A majority of the proposed NAS formulations were rejected in the first stage of screening due to negligible CO2 uptake, precipitate formation, or solvent gelation leading to a highly viscous gel or waxy solid. Evaluation of NASs in the presence of water was a critical step because although the solvent and the CO2 absorption mechanism do not require or involve water, water is omnipresent in the process as it is introduced via the flue gas. Therefore, NASs must be stable and the CO2 absorption must remain selective in the presence of water. CO2 uptake experiments were performed in the presence of water, in many cases forming a second phase with NAS formulation. Many NAS formulations exhibited critical issues when water was present, including precipitation, indicating formation of a bicarbonate as shown in Figure 1A and formation of a solid scale in the overhead condenser during regeneration (Figure 1B). Formulations exhibiting critical issues with water during CO2 absorption or solvent regeneration were eliminated leaving a select few NAS formulations that met the criteria and were stable and selective in the presence of water (Figure 1C).

Figure 1. Screening NASs for CO2 uptake and release in the presence of water. A. Formation of a precipitate (insoluble bicarbonate). B. Bicarbonate scale formed in overhead condenser during solvent regeneration. C. Biphase solvent exhibiting no adverse interaction with water.

Two families of promising NASs have been discovered and the general reaction pathways with CO2 are depicted in Figure 2. These promising NAS formulations have been described in several RTI patent applications [7,8,9] and form the basis of the NAS CO2 Capture Process. The NAS family that a majority of our development work has focused on can be described as hydrophobic, sterically hindered, carbamate-forming amines (Pathway 1) with low water solubility solubilized in an organic diluent having low vapor pressure, low viscosity, and low water solubility. NAS formulations that follow Pathway 2 can be described as hydrophobic, ionic liquid mixture consisting of an amidine/guanidine and an acidic species that reacts with CO2 to form a carbonate ester.

Figure 2. Reaction pathways for RTI's NASs.

To date, a number of promising NAS formulations have been identified and the most promising have been thoroughly characterized including measurement of thermodynamic properties and key physical properties. Details are provided below.

2.1. Thermodynamic Properties of RTI's NASs

CO2 isotherms and heat of CO2 absorption measurement were collected for the most promising NAS formulations using an automated vapor-liquid equilibrium and reaction calorimetry system. A simple comparison of the CO2 isotherm for the NAS and 30 wt% MEA-water solvent (Figure 3) suggests that the NAS does not absorb as much CO2 at a given CO2 partial pressure as the 30 wt% MEA-water solvent, particularly in the CO2 partial pressures common in flue gases; however, increasing the temperature has a much larger effect on reducing the CO2 loading. NAS shown here has a dynamic capacity that is >2.5-times that of the 30 wt% MEA-water solvent under the conditions presented. Therefore, it can be concluded that NASs achieve larger working capacities (Aa) with smaller changes in temperature, which allows these solvents to be regenerated at lower temperatures and thus use lower-quality steam. In the example shown here, the NAS is capable of achieving a CO2 partial pressure of 200 kPa at 90°C that compares favorably to 30 wt% MEA that requires regeneration above 120°C to achieve a similar CO2 product pressure.

CD —

CD —

-»-30wt% MEA [Shen and Li (1992)]

-■-30wt% MEA [RTI Data]

-*-NAS

90°C Aa y* / Aa.y

2^80°C / 30wt%MEA/Water

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70

CO2 Loading, a, [mol CO2 / mol amine]

Figure 3: A comparison of CO2 Isotherms for RTI's NAS and 30 wt% MEA.

The heat of CO2 absorption was measured for select NAS formulations using an automated vapor-liquid equilibrium and reaction calorimetry system. A comparison of the heat of absorption for a select NAS formulation and 30 wt% MEA-water solvent indicate that the NAS formulation has a lower heat of absorption than 30 wt% MEA-water (Figure 4). At 40°C, which is consistent with CO2 capture conditions, the NAS formulation was found to have a heat of absorption of 70-75 kJ/mol CO2 while 30 wt% MEA has a heat of absorption of 80-85 kJ/mol of CO2, indicating that the NASs have a 6%-18% lower heat of absorption under absorption conditions. However, at higher temperatures, consistent with the NAS regeneration temperature, the heat of absorption for the NAS reduces to 55-70 kJ/mol of CO2 whereas the heat of absorption for 30 wt% MEA-water, and many of the conventional aqueous-amine solvents, increases from 85 kJ/mol CO2 to 107 kJ/mol CO2 at 120°C. As such, the heat of absorption is approximately 50% lower for the NASs than 30 wt% MEA-water at their respective regeneration conditions.

Considering that the heat of absorption accounts for 56% of the energy for solvent regeneration for 30 wt% MEA-water, reducing this contribution by 50% will significantly reduce the regeneration energy and ultimately the energy penalty of the CO2 capture process.

Figure 4. A comparison of Heats of Absorption for RTI's NAS and 30 wt% MEA.

2.2. NAS Viscosity and Foaming Tendency

Solvent viscosity is a critical parameter for gas absorption processes as it affects essentially all aspects of the process, including the rate of absorption/ desorption via film thickness and diffusion lengths, heat transfer rates via film thickness, pressure drop in piping, and column performance via column wetting/flooding characteristics and pressure drop. As such, the viscosity of the NASs must be within a reasonable range to utilize conventional gas absorption process equipment. To this end, the viscosity of CO2-lean and CO2-rich NAS formulations was measured using a Brookfield DVI-Plus viscometer at relevant process temperatures (40°C and 80°C). Measured viscosities were found to range between 2.5 and 38 cP at the selected test conditions (Table 2). The viscosity of 30 wt% MEA under the same conditions is approximately one order of magnitude lower, particularly for the CO2-rich solvent [10]. The NASs, and in particular the second-generation NAS formulations, have very low viscosities compared to other NASs proposed for post-combustion CO2 capture applications and in fact are comparable with solvents used in numerous gas treatment applications [11], including glycols in gas dehydration processes (e.g., diethylene glycol, triethylene glycol) and glycol-amine mixtures in combined acid gas scrubbing-dehydration processes (e.g., diethylene glyol-monoethanolamine).

The foaming tendency of a solvent is an important parameter in gas absorption processes as foaming leads to poor process performance, including the early onset of column flooding, fluctuations in column AP, excessive

solvent loss by foam carryover, reduced capture efficiency, and insufficient solvent regeneration, and the onset of foaming requires immediate corrective actions to return process to optimal operating conditions. Foaming tendency was measured for the NASs and 30 wt% MEA-water (Figure 5) using the procedure described in ASTM D1881-73 and it was found that the NASs do not exhibit foaming issues as can be seen by a lack of foam head. 30 wt% MEA-water was found to be susceptible to foaming, especially when the solvent was rich in CO2. In fact, 30 wt% MEA-water failed the foaming test at 40°C and 80°C by expanding in volume greater than 3-times and overflowing the test vessel. As such, many acid gas scrubbing processes rely on anti-foaming agents to minimize foaming. The anti-foaming nature of the NASs and the broader NAS family is most likely related to their very low surface tension, reasonable viscosity, and high density.

Table 2. Comparison of Viscosities for Select NASs and Aq.-Amine Solvents.

Sample Name Viscosity (cP) Temp. (°C)

7.2-7.3 40

2.5 80

NAS2, Rich 27.1 40

Gen2 NAS, CO -rich 2 9.34 40

CO BOL, CO -rich [12] 22 ~200 40

1.7 40

0.77 80

30 wt% MEA-Water, Rich 2.7 40

8m PZ, CO-rich [13] 10.34 40

Figure 5. Foaming tendency observations for CO2-loaded NAS (left) and 30 wt%MEA-water (right).

2.3. Reboiler Heat Duty Estimation for NASs

The solvent screening program identified several NASs that specifically address each of the physical and chemical properties described by the reboiler heat duty calculation (Eqn.1) in the direction required and as such have the potential to substantially reduce the energy requirement as shown in Table 3.

Table 3. Potential for NASs to lower the energy for solvent regeneration.

Solvent cp [J/g K] AHabs [kJ/mol] AHVap [kJ/mol] xsolv [mol solvent/ mol solution] Aa [mol CO2/ mol solvent] Reboiler Heat Duty [kJt/kg CO2]

30 wt% MEA-H2O 3.8 85 40 0.11 0.34 3,750

RTI's NASs

1.28-1.49

0.4-0.5

0.2-0.4

The minimum thermal regeneration energy demand (i.e., reboiler heat duty) and the optimal solvent recirculation rate for the NASs were estimated using a validated "short-cut" method developed by Notz et al. [14]. Although the "short-cut" method is not as robust or reliable as a detailed electrolyte-based, gas-liquid equilibrium model, it is a versatile tool for estimating the minimum thermal regeneration energy for novel solvent formulations with minimal experimental data. The experimental data required by the "short-cut" method:

• vapor-liquid equilibrium curve of the solvent system at Absorber and Regenerator conditions,

• heat of CO2 absorption,

• specific heat capacity of the solvent, and

• vapor pressure of the solvent

The minimum thermal regeneration energy was estimated using the "short-cut" method for several select NAS formulations (Figure 6). Calculations were performed using the following process conditions:

• Flue Gas Composition (mole %): N2: 66.90; O2: 2.35; H2O: 16.68; CO2: 13.26

• Percent CO2 Captured: 90%

• Temperature: Absorber: 40°C; Desorber: 120°C

• Cossover Exchanger Approach: 10°C

• Pressure: Absorber: 101.3 kPa; Desorber: 200 kPa

• Note: NAS2 can achieve a CO2 product pressure of 780 kPa at 120°C but was not considered in this analysis.

The NASs have the potential to reduce the thermal regeneration energy requirement compared to conventional CO2 capture processes by 40%-50% (Figure 6). In fact, two of the NASs appear to be capable of achieving a regeneration energy of between 1,700 and 2,000 kJt/kg CO2 captured, which represents a 45%-55% reduction compared to 30 wt% MEA-water. This substantial reduction in the regeneration energy is the result of addressing each of the parameters in the reboiler heat duty calculation (Eqn. 1) in the appropriate direction, which resulted in reducing the contribution of the heat of absorption (AHabs), essentially eliminating the heat of vaporization contribution due to the very low vapor pressure of the NASs at the regenerations temperatures, and reducing the sensible heat contribution by lowering the specific CP and increasing the working capacity (Aa) under process conditions. It should be noted that the NASs require higher L/G ratios due to a higher solvent densities, 1,300-1,400

kg/m3.

1 s-i <D Ö W

o • ^

s-i <D Ö <D M <D

3,680 kJ/kg CO2

L/G [kg/kg]

30 wt% MEA NAS 1 NAS 2 NAS 3 NAS 4

Figure 6. Comparison of estimated minimum thermal regeneration energies.

2.4. Preliminary Technology Assessment

A preliminary technical assessment of the NAS CO2 Capture Process has been performed to estimate the net power loss associated with CO2 capture from a supercritical pulverized coal-fired power plant (DOE/NETL-2010/1281 - Case #12) [15] to demonstrate the potential for the NAS CO2 Capture Process to reduce the cost of electricity and CO2 capture. A simple block flow diagram of the Case 12 supercritical pulverized coal (SCPC) plant integrated with a post-combustion capture (PCC) unit showing the major process units and streams is provided in Figure 7. The approach used in this technical assessment was to replace the CO2 capture plant described in the Case 12 study (Fluor's Econamine FG PlusSM CO2 Capture Process) with an adequately sized NAS CO2 Capture Process to capture the same amount of CO2 from the power plant flue gas and generate a purified CO2 product stream having the same specifications. An in-house-developed Aspen Plus® process model of the nominal 550 MWe Supercritical PC power plant was used to perform mass and energy balances. The in-house-developed model was validated against data provided for each case with (Case 11) and without (Case 12) CO2 capture, and the mass and energy balances of all streams were reproduced with >99.5% accuracy.

A summary of the power plant generating capacity, net efficiency, and efficiency point loss associated with CO2 capture and compression is provided in Table 4 for three cases: 1)No Capture (Case 11), 2) Case 12 (2010) with Econamine FG PlusSM, and 3) Case 12 (2010) with NAS CO2 Capture Process. The total power generated by the NAS case is significantly greater than the Econamine FG PlusSM process case, since the NAS process has a much lower thermal regeneration energy requirement (NAS: <2,000 kJt/kg CO2 vs. Econamine FG Plus: 3,556 kJt/kg CO2 [15]) and therefore extracts less steam from the IP/LP crossover and also requires a lower steam temperature and

hence steam quality. As a result, the SCPC power plant equipped with the NAS CO2 capture process has a net plant efficiency of 32.5%, which compares very favorably to the net plant efficiency of 28.4% for the SCPC plant equipped with an Econamine FG PlusSM CO2 capture process. The parasitic power load associated with CO2 capture and compression shows that the NAS CO2 capture process consumes approximately 50% less energy than the Econamine FG PlusSM process. Overall, it is evident that the NAS CO2 capture process vastly outperforms the SOTA Econamine FG PlusSM CO2 capture process in terms of power consumption.

Figure 7: Block flow diagram showing location of NAS CO2 Capture Process in a coal-fired power plant.

Table 4. Summary of power plant performance efficiencies for different PCC technology options.

CO2 Capture Process Gross Power [kWe] Aux. Power [kWe] Net Power [kWe] Net Efficiency [%] Efficiency Point Loss

No Capture Case 11 (2010) 827,647 42,947 784,700 39.1% -

Econoamine FG PlusSM Case 12 (2010) 662,800 112,830 549,970 28.4% 10.7

NAS CO2 Capture Process Case 12 (2010) 765,172 113,093 652,079 32.5% 6.6

The NAS CO2 Capture Process has shown great potential for dramatically improving the energetic efficiency of capturing and compressing CO2 from flue gases. This is primarily due to the superior thermodynamic characteristics of the NASs, specifically their low heat of absorption, low specific heat capacity, absence of a low boiler (at relevant process conditions), large working capacity, and favorable vapor-liquid equilibrium enabling low temperature (< 80 - 100°C) regeneration while producing a high quality CO2 product. It must be noted that numerous assumptions were made to perform this preliminary assessment, and these assumptions need to be validated with process data collected from a small pilot system. Of specific importance are the verification of the regeneration energy demand, the lack of a detailed process design and an associated capital cost for the process, and a confident estimate of the solvent makeup costs, which are the largest contributor to the operating costs of the CO2 capture process. The goal of our future work is to address these unknowns/uncertainties.

2.5. Lab-Scale Evaluation of the NAS CO2 Capture Process

Lab-scale evaluations were performed in a small, continuous-flow, gas absorption system (Figure 8). Testing was performed to demonstrate that the NASs could achieve stable operation in a conventional gas absorption process arrangement, evaluate water balancing strategies, and develop an operational understanding of the process. Over 2,000 hr of on-stream testing has been performed using the NASs with much of the focus on understanding the effect that water entering the NAS process via the flue gas has on the performance of the process specifically in terms of system operability, CO2 capture efficiency, and energy consumption.

Results from a typical lab-scale system test with a 'best-candidate' NAS formation are provided in Figure 9. These results clearly demonstrate that the NAS process is capable of stable operation with respect to CO2 capture performance in a conventional process arrangement. Compositional analysis of the NAS indicated that upon the introduction of the feed gas to the system the CO2 content stabilized very quickly as exhibited by the stable CO2 concentration in the treated gas and CO2 product gas streams. However, the water content of the NAS was found to increase slowly and stabilize after ~60 hr indicating that the system had a neutral water balance. A neutral water balance was also verified by performing a water mass balance around the system. This experiment demonstrates that the NAS CO2 capture process can be operated such that a neutral water balance can be achieved with the bulk of water leaving the system via the treated gas stream, much like aqueous-based solvent systems without requiring dehumidification of the feed gas, without the additional heat load required to revaporize accumulated water, and without production of a liquid waste stream requiring treatment. Detailed additional experimentation has demonstrated that the water balance in the process can be controlled through specific choice of the operating parameters of the process.

Figure 8. Lab-scale Continuous Flow Gas Absorption System

Preliminary experimental measurements of the thermal regeneration energy for the NASs have been performed. The NASs were found to require substantially less thermal energy for regeneration, on the order of 40% less, than 30wt% MEA-water. Accurate estimation of the regeneration energy is very difficult in small systems, which experience significant heat losses, however, two independent approaches were evaluated. In the first approach, separate experiments were performed with 30 wt% MEA-water and a selected NAS in the lab-scale system in which the energy (heat) input to the system was adjusted to maintain similar process conditions and the same CO2 capture performance (~90% CO2 capture). In this approach, the NAS was found to require ~32% less energy than 30wt%

MEA-water under the process conditions and this specific process arrangement. In the second approach, the amount of energy (heat) input to the system was maintained at a constant setting for experiments with 30 wt% MEA-water and a selected NAS and the CO2 capture efficiency was measured. In this approach, the NAS achieved a significantly higher CO2 capture efficiency (85.6% vs 69.9%) with essentially the same amount of heating being delivered to the regenerator. Overall, these experiments provide strong evidence that the NASs have much lower regeneration energy requirements than conventional aqueous amine solvents. Much more comprehensive experiments are necessary to gain a better estimate of the regeneration for the NASs.

0 60 0 10 20 30 40 50 60 70 80 90 100

Time-on-Stream [h]

Figure 9. Time-on-stream data for experiment evaluating effect of water on NAS CO2 Capture Process. Feed Gas Composition: 13.3% CO2, 5.65°% H2O (sat @ 35°C), Bal. N2; Absorber Solvent Feed Temperature: 40°C; Regenerator Temperature: 90°C.

3. Current Efforts and Future Developments

Current efforts are focused on addressing and overcoming specific challenges related to solvent makeup cost, improving the fidelity of the process design, and ultimately demonstrating the feasibility and substantially lower regeneration energy of the NAS CO2 capture process at the bench-scale. The most significant technical challenges that must be addresses during this development stage are: 1) mitigating solvent make-up costs and 2) developing/identifying an industrially-relevant and scalable solvent regenerator design. Although these challenges are primarily technical in nature, the implications are much broader and affect the capital and operating costs, the ability to meet emission regulations, scalability, and ultimately the viability of the process.

To continue the development of the NAS process, RTI has constructed a versatile, small pilot unit to evaluate the performance of our NASs and compare them to the state-of-the-art aqueous amine solvents for the removal of CO2 from a variety of simulated flue gases. The small pilot unit has a conventional gas scrubbing process arrangement and was designed to be a highly instrumented and controlled experimental system capable of

performing detailed evaluations of solvents for the removal of CO2 from simulated flue gas streams and to operate for extended periods to support long-term performance evaluation. A 3D drawing and a picture of the unit are provided in Figure 10 and the specifications are provided in Table 5. A process model of the small pilot unit has been developed using Bryan Research & Engineering's ProMax® gas absorption simulation software package to validate that the absorber and regenerator columns were designed with adequate height to achieve >90% CO2 capture at realistic gas throughput (superficial velocity in the absorber), near optimal L/G ratio, and near optimal thermal regeneration energy.

Figure 10. RTI's Bench-scale System.

Table 5. Small pilot testing unit specifications

Simulated Flue Gas CO2 Absorber Solvent Regenerator

FG Flow Rate: 100 to 485 SLPM Diameter: 3" Sch. 10 SS316 Diameter: 3" Sch. 10 SS316

CO2 Feed Rate: 1.8 to 8.6 kg/h Height: 8.5 m (28') Height: 7.1 m (23'-4")

Feed Temp.: 30 to 50°C Packing: 8.05 m (26'-4") Packing: 6.7 m (22')

Target Comp: CO2: 13.3%; H2O: 6.1%; Mellapak 350.X Mellapak 350.X

O2: 2.35%; N2: bal. Temp.: 30 to 55°C Temp.: up to 150°C

CO2 Content: up to 15 %vol Pressure: ambient to 200 kPa Pressure: ambient to 1 MPa(g)

Water Content: ~0 to 12.3%vol Gas Vel.: 0.33 to 1.5 m/s

SO2 Content: up to 20 ppmv L: 15 -75 kg/h

Starting in early 2015, a testing campaign will be initiated to thoroughly evaluate the NASs under realistic process conditions in a representative process configuration consistent with the envisioned full-scale process to:

• Evaluate the effectiveness of the proposed process improvement/modifications including the addition of an NAS Recovery/Wash section and a novel Solvent Regenerator design specific for NASs without a low boiling point component.

• Experimentally demonstrate that the NASs are capable of achieving 90% CO2 capture and generating high-purity CO2 product (>95% CO2) at an 'optimal' L/G ratio (or working capacity) as determined from process modeling with a thermal regeneration energy to ~2,000 kJt / kg CO2, a 30%-50% reduction compared to aqueous-amine processes.

• Develop a detailed understanding of the operational and performance differences between the NAS and conventional aqueous-amine CO2 capture processes.

• Develop and tune an engineering-level process models of the NAS CO2 capture process by building upon the rate-based, equilibrium-limited reaction model developed in Year 1.

This effort will ultimately conclude with a detailed technical and economic assessment to understand potential to substantially reduce the cost associated with CO2 capture from coal-fired power plants and an environmental health and safety (EH&S) risk assessment to understand the potential risks associated with the proposed process and novel solvents.

RTI has partnered with Linde LLC to address specific process design challenges related to the NASs and to perform detailed technology assessments. Linde is a leading global industrial gases and engineering organization, which provides extensive process engineering, technology demonstration, and technology assessment expertise to the project. RTI has also partnered with SINTEF, the Norwegian independent research institute, to perform specific project tasks related to solvent degradation and emission studies for the NASs. SINTEF has more than 20 years of experience within research and development of amine based absorption processes and is a world-renowned leader in conducting solvent degradation and emission monitoring studies for amine processes. As part of this project, SINTEF will perform detailed solvent degradation studies including speciation and quantification of solvent degradation products and characterization of the emissions in the treated gas streams for the NASs. In addition, SINTEF will conduct extended exposure tests with the NASs using simulated flue gas including contaminants (i.e., SO2, NO, NO2) at realistic process conditions using SINTEF's continuous flow solvent degradation test rig (SDR). The SDR is a well-instrumented, continuous flow gas absorption regeneration system capable of generating flue gas compositions with common contaminants including SO2, NO, and NO2, and operating continuously for long periods. It enables measurement of the NAS degradation rates on timescales consistent with industrial processes and allows determination of degradation pathways and estimation of the rate of emissions in the treated gas stream.

Acknowledgements

RTI would like to acknowledge the United States Department of Energy/Advanced Research Project Agency -Energy (DOE/RPA-E) and the Department of Energy/National Energy Laboratory (DOE/NETL) for funding the development of this technology under co-operative agreements DE-AR0000093 & DE-FE0013865 respectively. Additionally, RTI would like to acknowledge partnership with BASF Corp, BASF SE, Linde LLC, and SINTEF.

References

1. ENVIRONMENTAL PROTECTION AGENCY, 40 CFR Part 60 [EPA-HQ-OAR-2013-0495; FRL-9839-4] RIN 2060-AQ91. Standards of Performance for Greenhouse Gas Emissions from New Stationary Sources: Electric Utility Generating Units.

2. DOE/NETL Carbon Dioxide Capture and Storage RD&D Roadmap. U.S. Department of Energy, National Energy Technology Laboratory. December 2010.

3. Rochelle, G. T. Amine Scrubbing for CO2 Capture. Science 2009, 325, 1652-1654.

4. Oexmann, J.; Kather, A., Minimising the regeneration heat duty of post-combustion CO2 capture by wet chemical absorption: The misguided focus on low heat of absorption solvents. Int. J. Greenh. Gas Control 2010, 4 (1), 36-43.

5. Wang., M., Lawal, A., Stephenson, P., Sidders, J., Ramshaw, C. Post-combustion CO2 capture with chemical absorption: A state-of-the-art review. Chemical and Engineering Research and Design 89 (2011) 1609-1624.

6. Oexmann, J.; Kather, A., Minimising the regeneration heat duty of post-combustion CO2 capture by wet chemical absorption: The misguided focus on low heat of absorption solvents. Int. J. Greenh. Gas Control 2010, 4 (1), 36-43.

7. WO/2012/031274 (PCT/US2011/050442). Regenerable Ionic Liquid Solvent for Acid-Gas Separation. Filing Date: 09/03/2010. Applicants: Marty Lail and Luke Coleman.

8. WO/2012/031281 (PCT/US2011/050452). Regenerable Solvent Mixtures with Nitrogenous Base for Acid-Gas Separation. Filing Date: 09/03/2010. Applicants: Marty Lail and Luke Coleman.

9. WO2013130997 A1 (PCT/US2013/028660). Regenerable Solvent Mixtures for Acid-Gas Separation. Filing Date: 03/03/2012. Applicants: Marty Lail and Luke Coleman.

10. Amundsen, T. G. CO2 Absorption in Alkaline . MSc. Thesis. Telemark University College 2008.

11. Kohl, A., Riesenfeld, F. Gas Purification. 4th Ed. Gulf Publishing Co. Houston, 1985.

12. Heldebrant, D. CO2-Binding Organic Liquids, Enhanced CO2 Capture Process With a Polarity-Swing-Assisted Regeneration. NETL CO2 CAPTURE TECHNOLOGY MEETING JULY 30, 2014.

13. Dugas R. and G. Rochelle. CO2 Capture by Aqueous Absorption Summary of 2nd Quarterly Progress Reports 2008.

14. Notz, R.; Tonnies, I.; Mangalapally, H. P.; Hoch, S.; Hasse, H. A short-cut method for assessing absorbents for post-combustion carbon dioxide capture. Int. J. Greenhouse Gas Control 5 (2011) 413-421.

15. Cost and Performance Baseline for Fossil Energy Plants, Volume 1, Bituminous Coal and Natural gas to Electricity, Revision 2, November 2010, DOE-NETL- 2010/1397.