Analysis of Adsorbent-Based Warm CO2 Capture Technology for

Article pubs.acs.org/IECR

Analysis of Adsorbent-Based Warm CO2 Capture Technology for Integrated Gasification Combined Cycle (IGCC) Power Plants Zan Liu and William H. Green* Department of Chemical Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Room 66-352, Cambridge, Massachusetts 02139, United States S Supporting Information *

ABSTRACT: Integrated gasification combined cycle with CO2 capture and sequestration (IGCC−CCS) emerges as a promising method for reducing emission of greenhouse gases from coal without reducing efficiency significantly. However, the high capital costs of these plants have limited their deployment. The current solvent-based low-temperature CO2 capture process is energy and capital intensive contributing to the problem. Warm CO2 capture has been predicted to be a key enabling technology for IGCC−CCS. Here, we assessed the applicability of CO2 removal technology to IGCC via a warm pressure swing adsorption (PSA) process based on our newly invented sorbent, which has good cyclic sorption−desorption performance at an elevated temperature. A 16-step warm PSA process was simulated using Aspen Adsorption based on the real sorbent properties. We used the model to fully explore the intercorrelation between hydrogen recovery, CO2 capture percentage, regeneration pressure of sorbent, and steam requirement. Their trade-off effects on IGCC efficiency were investigated by integrating the PSA process into the plant-wide IGCC simulation using Aspen Plus. On the basis of our analysis, IGCC-warm PSA using our new sorbent can produce similar thermal efficiencies to IGCC-cold Selexol. In order to achieve this, warm PSA needs a narrow range of process parameters to have a good balance between the hydrogen loss, steam consumption and work requirement for CO2 compression. This paper provides a rigorous analysis framework for assessing the feasibility of warm CO2 capture by sorbents in an IGCC system.

1. INTRODUCTION Electricity generation from coal-fired power plants is the largest point source for CO2 emissions.1 However, coal is also the most abundant and inexpensive fossil fuel available today. In the near future, coal will play a main role in the energy portfolio, especially in the developing countries, such as China and India.2 To effectively use coal while reducing carbon emissions, new technology is needed. Currently, integrated gasification combined cycle with carbon capture and sequestration (IGCC−CCS) is one of the most promising candidates for achieving this goal.3 All IGCC−CCS systems operate on a similar premise,4 as shown in Figure 1. Coal is gasified with steam and oxygen at high temperature and high pressure to produce syngas, which contains mainly H2 and CO. The CO is then converted to CO2 through a water−gas shift reactor system. The gas mixture is cleaned of sulfur, particulates, and CO2, and burned in a gas turbine for power generation. The heat remaining in the exhaust from the gas turbine can be further used in a heat recovery system to generate steam, which drives a steam turbine, generating additional power. IGCC has certain advantages over the traditional pulverized coal (PC) plant.5 First, because of the combined cycle, IGCC can achieve higher energy efficiency. Second, the gas stream in the IGCC system has much higher pressure, making CO2 capture and other emission control less expensive. The current state-of-the-art CO2 capture process is scrubbing the gas stream with physical or chemical solvents, such as MEA, Rectisol, or Selexol. These solvent-based absorption processes operate at a fairly low temperature. Thus, the gas stream © 2014 American Chemical Society

entering the absorber must be significantly cooled down, as shown in Figure 1, which leads to a high free energy loss5,6 and/or high capital costs for heat recuperation. It would be very desirable to reduce the parasitic energy load and operating costs associated with the traditional low-temperature CO2 capture process. Prior studies have proposed that high temperature CO2 separation processes,7,8 such as H2-permeable membranes,9,10 CO2-permeable membranes,11 and pressure swing adsorption (PSA)12,13 could significantly improve the IGCC efficiency. A more recent study14,15 in our group extended previous analyses to compare a variety of warm separation methods for IGCC−CCS, namely sorbents and membranes, on a unified basis. That study concluded that pressure swing adsorption operated at warm gas temperatures (200−300 °C) could be potentially more efficient compared with other approaches. However, there is a limitation with the research: the simulation was based on an assumed sorbent material whose properties were estimated based on thermodynamics calculation. Also, the PSA process simulated in the study was a five-step cyclic process that did not fully reflect the real PSA process used in commercial operation. To address these limitations, a real sorbent and PSA process need to be used for the simulation. Adsorbent property is the key factor contributing to PSA performance.16,17 In spite of the numerous reports on CO2 Received: Revised: Accepted: Published: 11145

September 24, 2013 May 23, 2014 June 4, 2014 June 4, 2014 dx.doi.org/10.1021/ie4030006 | Ind. Eng. Chem. Res. 2014, 53, 11145−11158

Industrial & Engineering Chemistry Research

Article

Figure 1. IGCC−CCS with cold syngas cleanup.

adsorbents,18 there has been relatively little research on regenerable CO2 sorbents in the temperature range of interest (200−300 °C) for IGCC−CCS. The commercially available sorbents such as activated carbon, zeolites, and alumina lose their CO2 adsorption capacity quickly at temperatures higher than 150 °C. Many other materials have been proposed as potential warm-temperature carbon dioxide sorbents, including super activated carbon,19 basic zeolites,20,21 calcium oxide,22,23 lithium zirconate,24 lithium silicate,25,26 sodium-based sorbent,27 magnesium hydroxide,28 hydrotalcite-like compounds (HTls),29,30 and double salt sorbent.31 However, many of these sorbents cannot meet the requirements for PSA operation. They suffer from slow kinetics, low selectivity, or too high a regeneration temperature. Recently, we have developed a new CO2 sorbent material with good regenerability, fast kinetics, and low heat of adsorption that can be applied in a PSA process at warm temperature range.32 The material has a large surface area and pore size, which facilitates the rapid adsorption of CO2. It has a good balance between working capacity and regeneration temperature. Here, we present our analysis of this promising sorbent and operating conditions for warm CO2 capture via PSA process for IGCC system. We built a 16-step PSA model to fully study the intercorrelation between various factors which affect IGCC efficiency. The PSA model was then integrated into the IGCC simulation in Aspen Plus. The performance of the adsorbent was evaluated based on the plant-wide simulation of IGCC system. This analysis framework is very helpful to evaluate potential sorbent materials that could be used for precombustion CO2 capture in IGCC power plants.

conversion. The purpose of the baseline model is to provide a foundation upon which various new process technologies can be compared on a fair and consistent basis. The simplified flowsheet is shown in Figure 1. The detailed information about the model was described in the manuscript33 and its Supporting Information. It is worth noting that the base case model uses a physical solvent-based process for H2S and CO2 capture. The process flowsheet is shown in Figure 2. Shifted syngas is introduced to a H2S absorber, which uses CO2-rich Selexol to absorb H2S. The H2S-rich Selexol is then introduced to a H2S concentrator, where most of the CO2 will be removed. The H2S-rich Selexol is then regenerated in a Stripper to generate a H2S stream and lean Selexol stream. The H2S stream is fed to a Claus unit, and the lean Selexol is introduced to the CO2 absorber. The H2S-free stream exiting from H2S absorber is fed to the CO2 absorber, which uses both lean and semilean Selexol to absorb CO2. Clean syngas is produced from the CO2 absorber. The CO2-rich Selexol then undergoes a series of pressure drops in three flash drums. The gas stream from the high pressure flash is fed back to the CO2 absorber to increase the H2 recovery. The CO2 generated from the medium pressure and low pressure flashes are sent to the sequestration unit. Overall, the Selexol process can achieve 91.2% CO2 capture (an overall 90.3% CO2 capture for the IGCC plant) and 99.98% H2S capture. The hydrogen recovery is 99.87%. IGCC-Warm PSA Flowsheet. For the warm syngas cleanup process, carbon capture and sulfur removal are accomplished using high temperature sorbent processes.14,15 Thus, an important difference between the warm syngas cleanup flowsheet, Figure 3, and the base case,33 is the removal of both the cooling section, and the reheating unit. In the warm cleanup flowsheet (Figure 3), the gas cleanup is done differently with a sorbent-based sulfur removal unit (RTI/ Eastman process) placed upstream of the CO2 capture process. The sulfur-free syngas is then fed into a sorbent bed for CO2 capture at a temperature above the dew point of the steam in the syngas. The fuel content of the decarbonized syngas is

2. MODELING WORK 2.1. IGCC Process Simulation. Base Case: IGCC−CCS Using Selexol. The base case model of IGCC with lowtemperature Selexol process for CO2 capture was developed in Aspen Plus by Field and Brasington.33 The overall design is based on the NETL report34,35 on bituminous coal to electricity 11146

dx.doi.org/10.1021/ie4030006 | Ind. Eng. Chem. Res. 2014, 53, 11145−11158


Article

Figure 2. Two-stage Selexol process for the removal of H2S and CO2.

primarily H2, with small amount of unconverted CO and hydrocarbons present as well. This warm syngas cleanup process can retain a significant amount of inerts, especially steam in the gas stream fed to the turbine. This reduces the requirement for supplying pressurized N2 diluent to the turbine, which is a significant energy cost in the base case. The warm PSA cases simulated here cannot recover more than 98% of the H2. A low temperature oxy-burner burns the H2 and CO in the CO2 waste stream. This heat is used to generate extra steam to propel the steam turbine, and for use as purge gas in the PSA unit. A Note on Warm Sulfur Removal Process. In the IGCCwarm PSA flowsheet, a high temperature ZnO-based adsorbent process developed by RTI36 is simulated for H2S removal (eq 1). The sorbent is then regenerated by oxidation (eq 2) using a stream of O2 supplied from the ASU to form SO2, which is reacted with a slip stream of syngas to yield elemental sulfur (eq 3) via the direct sulfur recovery process (DSRP). ZnO + H 2S → ZnS + H 2O

ZnS +

3 O2 → ZnO + SO2 2

SO2 + 2H 2 → S + 2H 2O

(2) (3)

The operating conditions for this process were estimated using documentation available from Eastman Chemical and RTI.37 This process can also remove COS.37 Sulfur is a salable byproduct, so it is worthwhile to process elemental sulfur from the economical point of view. However, the process is consuming 2 mol of H2 for every mole of sulfur captured, decreasing the power output of the gas turbine, and the ASU unit must supply 3/2 mol of O2 per mole of sulfur removed. This amount of hydrogen loss to sulfur removal is around 2.4% of the total hydrogen generated. To overcome this problem, it was proposed to replace the DSRP step with a simple wet flue gas desulfurization step (FGD) or converting sulfur dioxide to sulfuric acid.14 Previous research indicates air-fed FGD can improve IGCC HHV efficiency by about 1 percentage point (i.e., about a 3% increase in electricity produced) compared with DSRP.14 Other researchers have assumed FGD would be used for sulfur removal in IGCC−CCS.53 However, the FGD

(1) 11147



Article

Figure 3. IGCC−CCS with warm syngas cleanup.

Figure 4. Flowsheet for the 8-bed 16-step pressure swing adsorption process.

4. Pressure swing adsorption has been be widely used for air separation, drying and hydrogen purification.38−42 However, its applicability for CO2 capture is still under investigation and has not been demonstrated in commercial scale.43 The cyclic 8-bed 16-step PSA process, shown in Figure 4 and Table 1, contains three pressure equalization steps, and one steam purge step, similar to the commercial process for hydrogen purification used in the petrochemical industry. The usage of three pressure equalization steps can achieve a good balance between capital cost and hydrogen recovery. The steam purge rather than hydrogen purge is chosen because high

requires limestone, releases CO2 and makes a different product (gypsum), making it difficult to compare this approach with the base case on an equal basis. Here our evaluation of the various warm CO2 removal technologies has been based on the DSRP process making sulfur. 2.2. Pressure Swing Adsorption Process Model for CO2 Separation. Process Description. To determine the hydrogen recovery, CO2 capture percentage and the stream requirement associated with sorbent regeneration, a numerical simulation of the adsorption cycle was developed for a pressure swing process. A simplified process diagram is shown in Figure 11148



Article

Table 1. Switching Sequence for 8-Bed 16-Step PSA Processa time (min) adsorber

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

1 2 3 4 5 6 7 8

A E1R E2R E3R BD E3D E2D E1D

A&P PP I I P I I I

E1D A E1R E2R E3R BD E3D E2D

I A&P PP I I P I I

E2D E1D A E1R E2R E3R BD E3D

I I A&P PP I I P I

E3D E2D E1D A E1R E2R E3R BD

I I I A&P PP I I P

BD E3D E2D E1D A E1R E2R E3R

P I I I A&P PP I I

E3R BD E3D E2D E1D A E1R E2R

I P I I I A&P PP I

E2R E3R BD E3D E2D E1D A E1R

I I P I I I A&P PP

E1R E2R E3R BD E3D E2D E1D A

PP I I P I I I A&P

a

A: Adsorption. A&P: Adsorption and product depressurization. I: Idle. P: Purge. BD: Blowdown. PP: Product pressurization. E1D: Pressure equalization I, cocurrent depressurization. E2D: Pressure equalization II, cocurrent depressurization. E3D: Pressure equalization III, cocurrent depressurization. E1R: Pressure equalization I, countercurrent repressurization. E2R: Pressure equalization I, countercurrent repressurization. E3R: Pressure equalization III, countercurrent repressurization.

Step 11: Pressure equalization III, countercurrent pressurization. The clean bed initially at 1 atm is pressurized by the H2/ steam gas mixture in bed 3. At the end of this step, the pressure increases to 11 atm. Step 12: Idle. The bed maintains current state for 1 min. Step 13: Pressure equalization II, countercurrent pressurization. Similar to step 11 . The bed is connected to and pressurized by bed 5. At the end of this step, the pressure increases to 21 atm. Step 14: Idle. The bed maintains current state for 1 min. Step 15: Pressure equalization I, countercurrent pressurization. Similar to step 13. The bed is connected to and pressurized by bed 7. At the end of this step, the pressure increases to 31 atm. Step 16: Product pressurization. The bed is connected to and pressurized by bed 8. After this step, the bed and gas phase conditions will be back to the original clean high-pressure state at the beginning of step 1, which finalizes the whole cycle. The PSA performance depends on the sorbent properties and operating parameters. In this study, we use the sorbent properties from a material developed in our lab recently.32 We limited our study of process parameters to the residence time, regeneration pressure and steam consumption. We change the residence time by varying the bed length, which will affect the position of concentration front. Regeneration pressure will affect the steam requirement as well as the CO2 compression work requirement. The following regeneration pressures have been studied: 10, 5, 1, 0.75, 0.5, and 0.25 atm. The intermediate pressures of the pressure equalization steps were adjusted appropriately depending on the regeneration pressure. Model Description. Each bed of the PSA system would undergo the same procedure during the cyclic operation but with different phasing, thus the simulation of the cyclic process can be modeled using single-bed approach using Aspen Adsorption. In this study, the singe bed was modeled as a transient, adiabatic fixed bed. To simplify the model, the adsorption process was assumed to be 1-dimensional with no radial gradients, and the Peclet number was considered to be sufficiently high (and the bed sufficiently uniform) that axial diffusion and dispersion could be neglected. Previous work14,15 modeled the bed as adiabatic or isothermal, which effectively create upper and lower bounds for the temperature effects within the bed. However, considering the practical issues regarding the heat transfer, the adiabatic model is chosen, as it is more practical for the very large beds required by a full scale IGCC plant. The resulting model equations14,15 are listed

hydrogen recovery is essential for energy efficiency in IGCC. It is worth noting that hydrogen recovery can be further increased by adding extra pressure equalization steps, but that would require more sorbent in pressure vessels (i.e., more capital costs). The 8-bed 16-step PSA process cycle for a regeneration pressure of 1 atm is detailed below using bed 1 as an example. Step 1: Adsorption. Gas mixture consisting of mainly H2, CO2, and steam is introduced from one end (feed end) of the packed column, which is initially full of freshly regenerated sorbent. Decarbonized stream, which is mainly H2 and steam, is withdrawn from the other end (product end). When CO2 concentration front moves to roughly 1/3 of the bed, step one stops. The pressure is maintained at 40 atm for this step. Step 2: Adsorption and product pressurization. The CO2 concentration front initially 1/3 of the way down the bed moves further toward the product end. Part of the product stream is used to pressurize bed 2. The pressure of the bed is maintained at 40 atm. Step 3: Pressure equalization I, cocurrent depressurization. The product end of the bed is connected to and pressurizes bed 3. The CO2 concentration front initially about 2/3 of the way down the bed moves to the product end. However, most of the gas exiting the bed is H2 and steam. During this step, the pressure of the bed is decreased to around 30 atm. Step 4: Idle. The bed maintains current state for 1 min. Step 5: Pressure equalization II, cocurrent depressurization. The product end of the bed is connected to and pressurizes bed 5. The bed pressure initially at 30 atm drops to 20 atm during the process. Step 6: Idle. The bed maintains current state for 1 min. Step 7: Pressure equalization III, cocurrent depressurization. The product end of the bed is connected to bed 7. The CO2 initially in the bed moves out of the end of the bed and into bed 7. At the end of the step, the pressure of the bed is around 11 atm. Step 8: Idle. The bed maintains current state for 1 min. Step 9: Countercurrent blowdown. The product end of the bed is closed. The feed end opens and is connected to the waste stream, going to CO2 sequestration. The bed undergoes countercurrent depressurization. The pressure of the bed is dropped to the regeneration pressure of 1 atm. Step 10: Countercurrent desorption with steam purge. Steam is introduced from the product end of the bed. The sorbent undergoes regeneration and the adsorbed CO2 will be released to gas phase and flushed to the waste stream. 11149



Article

represent a porous sorbent dispersed upon a solid support. The isotherm parameters a, b, c, d, kLDF, and ΔHads are the experimental values for the new sorbent developed in our lab.32 2.3. Integration of PSA Models with Aspen Plus. The 16-step pressure swing adsorption process was simulated using Aspen Adsorption and integrated in Aspen Plus to predict its performance for CO2 capture in an IGCC system. A USER2 block in Aspen Plus linked to an Excel file was used to model the warm PSA process. Because the PSA simulation is a dynamic simulation, the cyclic steady state results from the simulation were tabulated in Excel. These main results include the CO2 capture percentage, H2 recovery, steam/feed ratio and steam recovery, which are defined below.

below, assuming that the only adsorbing species in the system is CO2. (1) Bed type: vertical, 1-D model, no internal heat exchanger, adiabatic. (2) Material balance: convection only, no axial or radial dispersion. ε

∂(vgci) dq ∂ci = −ε − ρs i dt ∂t ∂z

(4)

(3) Momentum balance: Ergun equation. 1.75(1 − ε)ρgas vg2 150μ(1 − ε)2 vg ∂P =− − ∂z ε 3d p2 ε 3d p

CO2 Capture Percentage Total moles of CO2 in waste stream per cycle = Total moles of CO2 in feed stream per cycle

(5)

(4) Kinetics: lumped resistance, linear driving force approximation. A constant kLDF is used based on the pseudohomogeneous approximation. dqi dt

= kLDF(qi* − qi)

H 2 Recovery Total moles of H 2 produced in product stream per cycle = Total moles of H 2 in feed stream per cycle

(6)

(10)

Steam/Feed Ratio Total moles of steam fed in regeneration step per cycle = Total moles in feed stream per cycle

(5) Isotherm: temperature-dependent Langmuir model. qi* =

a e(b / T )Pi 1 + c e(d / T )Pi

(7)

(11) Total moles of steam in product stream per cycle Steam Recovery = Total moles of steam in feed stream per cycle

(6) Energy balance: Nonisothermal with no conduction; constant heat of adsorption; no heat transfer to environment. (ερgas Cp,gas

(12)

A total of 36 cases were studied which represent different operating conditions: Pregeneration = 10 atm (cases 1−10); Pregeneration = 5 atm (cases 11−22), Pregeneration = 1 atm (cases 23−33), Pregeneration = 0.75 atm (case 34), Pregeneration = 0.5 atm (case 35), and Pregeneration = 0.25 atm (case 36). Details of each case are shown in Supporting Information. The strategy of using purpose-built unit operation models for adsorption process and integrating them into Aspen Plus flowsheet models of the entire IGCC process as USER2 models has proven to be an effective means of exploring the performance of new technology options.

∂(vgρgas Cp,gasT ) ∂T + ρs Cp,s) =ε ∂t ∂z dqi

− ρs ΔHads

(8)

dt

The above equations and proper conditions were then evaluated using a numerical integrator within Aspen Adsorption. Because we were interested more in the steady-state operation rather than the transient effects of startup in the adsorption beds, we simulated a large number of cycles (150) to attain cyclic steady state. The model parameters chosen for the sorbent bed are shown in Table 2. The values of ε, Cp,s, and dp were all chosen to

3. RESULTS AND DISCUSSION 3.1. Cyclic Steady State Test. PSA is a dynamic process. To incorporate its performance into a steady-state simulator, we need to make sure the PSA operation has reached its cyclic steady state. Here, we monitored the cyclic temperature change and flow rate to determine the required cycle numbers to achieve the cyclic steady state. Temperature changes with time at different positions of the bed are shown in Figure 5. Because of the exothermic nature of the adsorption process and the adiabatic nature of the bed, there are big temperature variations during the PSA cycle. As discussed in previous sections, most of the adsorption takes place in the first half of the bed, thus, the temperature change for the feed end and the midpoint is larger compared with the product end. The temperatures at these three different positions of the bed at the beginning of the each cycle are plotted in Figure 6. Cyclic steady state is achieved after about 50 cycles. For all the results reported here we ran the PSA simulations for 150 cycles.

Table 2. Parameters Used in Adsorption Model param.

value

ε ρs dp kLDF Cp,s D L ΔHads a b c d

0.719 720 kg/m3 0.002 m 1 s−1 1000 J/kg·K 0.055 m 1.2−1.6 m −15 kJ/mol 5.013 × 10−3 mol/(kg·atm) 1.753 × 103 K 1.268 × 10−3 atm−1 1.802 × 103 K

(9)

11150



Article

Figure 7. Total flow of H2 in product stream as the system approaches cyclic steady state.

Figure 5. Temperatures change at different positions of the bed: feed end (blue), midpoint (green), product end (red).

different steam flow rate (steam/feed ratio, defined in eq 11) were used and the corresponding CO2 capture percentage and H2 recovery were recorded. For low pressure regeneration scenarios (0.25 atm, 0.5 atm, and 0.75 atm), there is no need for steam purge. At each condition the bed length was adjusted to achieve a CO2 capture in the range 92.5% ± 1%, and the corresponding steam recovery and H2 recovery were recorded. The relationship between CO2 capture and steam consumption at different residence times and regeneration pressures are shown in Figure 8 (Pregeneration = 1 atm), Figure

Figure 6. Temperature in the beginning of each cycle at different positions of the bed: feed end (blue ◊), midpoint (green □), product end (red ○).

The flow rate of H2 in product stream is also monitored, as shown in Figure 7. The production of H2 has a large value at step 1, which is because the bed is initially saturated with a pure H2 in this study. The fluctuations at later times might arise from the numerical integration. Despite the fluctuation, at steady state, the mass closure for all species is within 0.5%. 3.2. Intercorrelation between H2 Recovery, CO2 Capture Percentage, Steam Requirement, and Regeneration Pressure. The main factors contributing to the overall efficiency of IGCC−CCS process are the amount of steam required for sorbent regeneration, the H2 recovery, and the regeneration pressure (which affects works requirement for CO2 compression). Here, we study the intercorrelation between these factors in the PSA process. First, the regeneration pressure was fixed as one of the following numbers: 10 atm, 5 atm, 1 atm, 0.75, 0.5, and 0.25 atm. At high pressure scenarios (10 atm, 5 atm, and 1 atm), we tried several different bed lengths which affect the residence time and the position of concentration front during pressure equalization steps. At each fixed pressure and fixed bed length,

Figure 8. Relationship between CO2 capture and steam requirement at constant purging pressure at 1 atm, with different bed lengths: (○) 1.6 m; (□) 1.5 m; (◊) 1.4 m; (☆) 1.38 m; (Δ) 1.36 m.

9 (Pregeneration = 5 atm) and Figure 10 (Pregeneration = 10 atm). It is clear that at a fixed regeneration pressure and fixed residence time, the CO2 capture percentage increases with the steam/feed ratio. That is because more purging steam dilutes the gas phase CO2 concentration which facilitates desorption of CO2. If we fixed the purging pressure and CO2 capture percentage, we noticed that shorter beds need larger steam/feed ratio to achieve the same CO2 capture. That is because during the pressurize equalization step III, the concentration front would move out of the bed and enter another bed if the bed was too 11151



Article

H2 separation at the fixed sorbent performance and PSA configuration.

Figure 9. Relationship between CO2 capture and steam requirement at constant purging pressure at 5 atm, with different bed lengths: (○) 1.6 m; (□) 1.5 m; (◊) 1.4 m; (☆) 1.35 m; (Δ) 1.3 m. Figure 11. Best achievable H2 recovery vs the CO2 capture percentage at different regeneration pressures: (○) 1 atm, (□) 5 atm, (◊) 10 atm.

Figure 10. Relationship between CO2 capture and steam requirement at purging pressure of 10 atm, with different bed lengths: (○) 1.6 m, (□) 1.5 m, (◊) 1.4 m, (☆) 1.3 m, (Δ) 1.25 m.

There is an obvious trade-off between H2 recovery and CO2 capture. To achieve high H2 recovery, the bed length needs to be short enough that the concentration front will reach the end of bed during the pressure equalization steps. The side effect of the short bed is that a significant amount of CO2 will accompany the H2 during this step, which results in a lower CO2 capture percentage. The feasible region shrinks when the regeneration pressure is larger. That is because a larger regeneration pressure means a larger initial pressure during blowdown step. At high pressure, there are more moles of hydrogen in the gas phase, which blows down into the waste stream, resulting in a lower H2 recovery. If we fix the CO2 capture to be 92.5%, the relationship between average H2 recovery and regeneration pressure is shown in Figure 12. Summary of Trade-offs. To make the IGCC-warm PSA comparable with IGCC-Selexol on the same basis, we fixed the

short. If that happened, a large amount of CO2 would bypass the desorption step and be kept in the bed. Then, more steam would be needed in the regeneration step to force more CO2 to be released from the sorbent surface, in order to reach the same CO2 capture as the longer bed case. As for the effect of purging pressure on steam requirement, it is pretty straightforward. As indicated in the three figures, in order to achieve same CO2 capture, the number of moles of steam required increases with the regeneration pressure. That is because high pressure regeneration would need high pressure steam, requiring a larger molar flow rate to maintain a similar volumetric flow rate. In addition, the higher residual CO2 partial pressure at an increased purge pressure also contributes to the increased steam purge requirement. Effect of Carbon Capture on H2 Recovery. When the steam flow rate during the regeneration step is large enough, a maximum CO2 capture percentage would be reached. The relationship between CO2 capture and H2 recovery at different regeneration pressures is shown in Figure 11. In the figure, the area below the line is the feasible region for possible CO2 and

Figure 12. Average hydrogen recovery at different regeneration pressure, with CO2 capture percentage constant at 92.5%. 11152



Article

CO2 capture percentage as 92.5% ± 1% for the PSA process, which corresponds to an overall 90% ± 1% of carbon capture for the whole IGCC process, defined in eq 13. A total of 36 cases which have 92.5% ± 1% CO2 capture were studied. The 36 cases with similar CO2 capture have sharply different H2 recovery, steam requirement, and regeneration pressure. These parameters have very different influence on the IGCC efficiency. Higher hydrogen recovery increases the hydrogen supplied to the gas turbine and thus increases the thermal efficiency. A lower steam requirement would increase the steam flow rate fed to the steam turbine and thus be beneficial to IGCC efficiency. A higher regeneration pressure reduces the work requirement for CO2 compression, and thus improves efficiency. However, as discussed above, high H2 recovery, low steam requirement, and large regeneration pressure are not simultaneously achievable. The trade-offs are summarized in Figure 13. The trend of the line indicates a positive correlation between hydrogen recovery

Carbon Capture =

Flow rate of CO2 fed to sequestration Flow rate of carbon‐containing species exiting gasifier

(13)

Specific CO2 emission =

CO2 emission flow rate Net power output

(14)

As for the specific CO2 emission, defined in eq 14, the range is 0.0931 ± 0.0116 kg/kWh for IGCC-warm PSA, which is very similar to the base case as 0.0931 kg/kWh. These two specifications confirmed our simulation replicates the requirement for carbon capture and carbon emission regulation as the base case. It is worth noting that a slightly higher CO2 capture is needed for IGCC-warm PSA to achieve a similar specific CO2 emission with IGCC-Selexol. That is because the overall efficiency of most IGCC-warm PSA cases is slightly lower than base case, as illustrated in the next sections. Thus, for the same amount of CO2 emission, IGCC-warm PSA would produce slightly smaller amount of electricity. More capture is needed to compensate this efficiency decrease. Hydrogen Usage. Hydrogen recovery is the main factor contributing to the IGCC efficiency. For the base case, a twostage Selexol process is configured to recover up to 99.87% of hydrogen. However, in a hydrogen purification PSA process, the hydrogen recovery is normally less than 92%. To recover more hydrogen, steam purge rather than hydrogen purge is used, but the overall hydrogen usage in gas turbine is still significantly lower than the base case, as shown in Table 1 of the Supporting Information. There are several sources for the hydrogen loss. First, the PSA process will lose 6%−10% of H2 for the high pressure regeneration cases (cases 1−10). Even though the hydrogen would be burned in an oxy-burner, but the electricity recovered would be lower than that from hydrogen which goes through the combined cycle. Second, the sorbent-based warm desulfurization process would consume a constant 2.36% of H2. As discussed in the previous sections on desulfurization, (eqs 1, 2, and 3), two moles of H2 are consumed to capture one mole of sulfur. Because of the coal simulated is relatively high-sulfur coal (approximately 3 wt %, Illinois No. 6 bituminous coal), this amount of H2 is significant. Third part of the loss is that extra H2 is needed in the burner to generate enough steam for sorbent regeneration (denoted as extra H2 burning in Table 1 of the Supporting Information). These requirements only exist for certain high pressure regeneration cases but could be a significant number in those scenarios. Gas Compression Work Requirements. CO2 Compression Work Requirement. Gas compression consumes significant power in IGCC systems. For the base case IGCC cold capture process, Selexol produces two CO2 streams, which have a high pressure 1.1 MPa and a low pressure 0.15 MPa with a molar flow rate ratio of 2:3. As shown in Table 2 of the Supporting Information, the CO2 compression work is 28.35 MW for this base case. One potential advantage of IGCC-warm PSA process is the customized regeneration pressure; increased pressure would reduce the power requirement for the CO2 compression. As shown in Table 2 of the Supporting Information, the CO2 compression work is much lower for high pressure regeneration (cases 1−10). Case 10 needs more CO2 compression work than the other cases, because there is an extremely low hydrogen recovery in this case. To keep the H2 flow to the gas turbine constant, we require a much larger coal flow rate,

Figure 13. Relationship between H2 recovery and steam requirement at different regeneration pressures: (◊ 10 atm, (○) 5 atm, (□) 1 atm. CO2 capture percentage is fixed at 92.5%.

and steam requirement. The position of the three lines in the figure clearly shows higher regeneration pressure is linked with a lower hydrogen recovery and higher steam consumption. Among the factors contributing to IGCC efficiencies, any pair of the factors has trade-offs, and only a plant-wide simulation can reveal their effects on IGCC efficiency and identify the potentially ‘optimal’ operating conditions, which will be discussed in the following sections. 3.3. IGCC-Warm PSA vs IGCC-Cold Selexol. Carbon Capture and Specific CO2 Emission. By integrating the PSA model into the IGCC process, we can compare the different aspects of IGCC performance for both IGCC-warm PSA and IGCC-Selexol. To make the comparisons on a fair basis, all 36 cases that represent warm CO2 capture with different operating conditions have a carbon capture percentage (defined in eq 13) in the range of 90.8 ± 1.7%, which replicate the carbon capture percentage of the base case, which is 90.3%. The slight variation comes from the fact that it is hard to accurately control the CO2 capture in PSA simulations. 11153



Article

Table 3. Comparisons between Cold and Warm CO2 Capture Technology power summary (MW)

cold cleanup

regeneration pressure total power generated gas turbine steam turbine auxiliary total power consumed N2 compression O2 compression CO2 compression air compression auxiliary net power output thermal input (as coal) HHV thermal efficiency (HH((HHV)

737.01 464.32 264.46 8.23 −187.16 −34.80 −11.27 −28.35 −71.32 −41.42 549.85 1688.00 32.58%

warm cleanup 10 atm

5 atm

1 atm

0.75 atm

768.58 460.81 286.69 21.08 −155.16 −10.27 −12.94 −21.31 −84.47 −26.17 613.42 1936.39 31.68%

760.28 461.63 270.21 28.44 −157.01 −12.08 −12.34 −26.69 −80.55 −25.35 603.27 1848.74 32.63%

762.81 462.01 266.18 34.62 −174.32 −14.90 −12.02 −44.02 −78.48 −24.90 588.49 1800.23 32.69%

761.41 462.12 265.62 33.68 −175.91 −14.49 −11.95 −46.68 −77.99 −24.80 585.50 1789.87 32.71%

gas turbine to be constant, all the PSA cases require more coal (and so more H2 input to the CO2/H2 separator) than the Selexol base case. The correlation between coal input and the percentage of H2 actually burned in gas turbine is shown in the Supporting Information. The flow rate of oxygen required scales with the coal flow rate, so the O2 compression work requirement for IGCC-warm PSA cases is larger than in the IGCC-Selexol base case, as shown in Table 2 of the Supporting Information. Overall Thermal Efficiency (HHV). The thermal efficiencies for all IGCC-warm PSA cases and the cold Selexol base case are summarized in Supporting Information. There are several IGCC-warm PSA cases which can achieve similar energy efficiency to the IGCC-Selexol base case (32.58%). Table 3 summarizes the best cases achieved by warm PSA for different regeneration pressures, which are case 4, case 14, case 25, and case 34 respectively (details for the cases are in the Supporting Information). There are several trends observed from the table. Compared with Selexol, warm PSA provides a significant energy saving in N2 compression for all cases, and reduces energy needed for CO2 compression as well in some cases. But the reduced hydrogen recovery for PSA requires a much larger coal flow rate and related work requirement for air separation and oxygen compression. The auxiliary power generation for IGCC-warm PSA cases is larger and arises from the expansion turbine in the warm desulfurization process and also energy from steam generated from the rejected H2 in the oxy-burner. The auxiliary power consumption for IGCC-warm PSA is lower than the base case, because the PSA process virtually requires no power compared with Selexol process. Even though the IGCC-warm PSA cases produce more electricity and consumes less energy for N2 compression than the base case, the required thermal input (as coal) is much larger because of the lower H2 recovery. The final result is that the IGCC-PSA and IGCC-Selexol efficiencies are comparable. We acknowledge that due to the limitations of the numerical models, a lack of optimization of the overall IGCC flowsheet and some uncertainties surrounding many of the model parameters, the absolute HHV efficiencies of the IGCC processes calculated in this work are not exact. However, because many of these limitations are the same in each model studied, we feel that the efficiency differences between the different technologies captured within this study are correct.

resulting in an abnormally large amount of CO2 in the waste stream. For cases with Pregeneration = 5 atm (cases 11−22), the compression work requirement is similar to the base case. It is not surprising that for the low-pressure regeneration cases (cases 34−36) or for the Pregeneration = 1 atm (cases 23−33), there is a significantly larger CO2 compression work requirement than the IGCC-Selexol. N2 Compression Work Requirement. For both IGCCSelexol and IGCC-warm PSA, the N2 diluent is compressed and added to the gas turbine to maintain the appropriate heating value of the syngas entering the gas turbine burner. The purpose for the diluent is to reduce the flame temperature and therefore achieve a lower NOx emission. For the cold Selexol base case process, almost all of the steam was removed during the cooling. Thus, there is a need for large flow rate of N2 diluent to the gas turbine, which results in a significant work requirement for N2 compression (34.80 MW). The warm syngas cleanup process not only eliminates the need for the cooling down and reheating step (as shown in Figure 3), but more importantly, it can significantly reduce the requirement for N2 diluent as a large amount of steam is retained in the gas stream. As shown in Table 2 of the Supporting Information, the N2 compression work reduction is significant (>50%). The N2 compression work is lower for PSA processes with higher pressure regeneration (cases 1−10). The reason is linked with the fact that PSA with high pressure regeneration has higher steam recovery (defined in eq 12, and shown in Supporting Information). High pressure regeneration needs more steam in the PSA process, and some of this steam enters the product H2 stream during step 1 of the PSA cycle. Regardless of the different steam recovery, for all IGCCWarm PSA cases, there is a significant amount of saving on N2 compression. This is the main efficiency advantage of warm PSA over cold Selexol process. O2 Compression. Throughout the simulation, the amount of coal fed to the gasifier is adjusted to a system designed around actual size of two GE 7FB gas turbines, each with a volumetric flow rate of 4.60005 × 1012 cc/h.34,35 The gas turbine power output is maintained around 462 MW for all cases simulated. However, because of the variation of the hydrogen recovered during CO2 capture in different cases, the coal flow rate (“thermal input”) is variable for each case simulated in this work. Since the PSA process rejects more of the H2 than Selexol, and we are constraining the amount of H2 fed to the 11154



Article

electric power produced by the steam cycle. These losses are not well compensated by the energy savings in the CO2 compression. Second, there is a strong trade-off between hydrogen recovery and steam consumption. As mentioned earlier, the increase of steam consumption can have a positive effect on hydrogen recovery, but there are diminishing returns. As shown in Figure 13, a small increase in steam consumption can significantly improve hydrogen recovery when the hydrogen recovery is low. However, as the hydrogen recovery increases, the marginal benefit of increasing steam consumption drops, and eventually, the energy lost by using the steam outweighs the energy gained by capturing more of the H2. Thus, it is not surprising that we observe an ‘up-and-down’ trend in Figures 14 and 15. Our simulation indicates that there exists an optimal operating condition which can balance the hydrogen recovery and steam consumption, leading to a maximum efficiency at fixed regeneration pressure. In addition, as shown in Figured 14 and 15, there exists an efficient operating window from about P = 0.75 atm up to about P = 5 atm, with various levels of steam and corresponding bed sizes and H2 recoveries. Inside this window one can choose which design one prefers (e.g., to minimize capital cost or maintenance costs, or to increase reliability). To the accuracy with which we can perform the simulations, the HHV efficiency is approximately the same across this window, and it is similar to the efficiency of the IGCC-Selexol base case process. Moreover, a rough estimation of the capital costs for the PSA process along with the sorbents indicates a cost around $15 M, which is significantly cheaper than the Selexol process, which is estimated to be around $110 M based on NETL report.34,35 Warm PSA process can achieve a similar separation performance as the Selexol process but it appears to have a much lower cost, so this is a promising technology. However, we note that changes in the total capital cost of the entire IGCC−CCS plant may outweigh the direct cost difference between Selexol and PSA quoted here, since each scenario will have a different coal throughput, require different compressors, etc. Even though the warm PSA in this work is not outperforming the Selexol process significantly, it is worth noting that there is a lot of room for improvement in the IGCC-warm PSA process. First, earlier work identified additional gains of at least 1 HHV% point could be achieved through better heat integration.44 Second, as discussed in the section on desulfurization process, an alternative sulfur recovery process, such as FGD can significantly reduce the H2 consumption and so increase the overall thermal efficiency.14 Third, the analysis in this paper is based on a not-so-perfect real-world sorbent. With the development of new warm CO2 sorbent materials, IGCC-warm PSA is very promising to outperform the base case significantly. As shown in Figure 16, an imaginary sorbent with 10 times larger kLDF (Green line) or with 3 times larger capacity per unit bed volume (red line), can shift the steam requirement-H2 recovery curve to the right, which means with the same amount of steam consumption, a better sorbent can achieve a higher H2 recovery. (We note that these increases in sorbent performance may be possible by increasing the surface area or altering the morphology of the sorbent support, or using a structured packing.45,46) Full process simulation of these imaginary sorbents indicates a thermal efficiency improvement of 0.2 and 0.9 HHV% points, respectively. In addition, our simulation indicates the capacity improvement can also decrease capital costs significantly by

Because of the complex trade-offs among hydrogen recovery, steam consumption, and regeneration pressure, there exists a narrow operating window for warm PSA to beat Selexol in thermal efficiency. There are some trends which can be observed from our simulation. First, as shown in Figures 14 and

Figure 14. Thermal efficiency of IGCC-PSA vs steam consumption at different regeneration pressures: (◊) 10 atm, (○) 5 atm, (□) 1 atm, (▽) 0.75 atm, (Δ) 0.5 atm, (☆) 0.25 atm. The peak efficiencies are comparable to the IGCC-Selexol base case efficiency (32.58%).

Figure 15. Thermal efficiencies of the IGCC-Selexol base case and IGCC-PSA vs hydrogen recovery at different regeneration pressures: (●) IGCC-Selexol, (◊) 10 atm, (○) 5 atm, (□) 1 atm, (▽) 0.75 atm, (Δ) 0.5 atm, (☆) 0.25 atm.

15, those cases with too high a regeneration pressure have lower thermal efficiencies. Previous research12 reported an efficiency improvement of 1.5% HHV using high-pressure sorbent regeneration. However, that report is not validated in this research. As discussed in the previous sections, high pressure regeneration is linked to higher H2 rejection through the PSA unit, and extremely large steam requirement for sorbent regeneration. At 10 atm regeneration, hydrogen recovery is less than 92%, and the steam needs to be high pressure steam (high quality), which significantly reduces the 11155



Article

sorbent, the PSA process, and the sulfur capture processeach of these could significantly improve the energy efficiencyand a more detailed study of the capital cost differential between the competing technologies.

4. CONCLUSION In this work, we have evaluated the suitability of sorbent-based warm CO2 capture technology in IGCC. A 16-step warm PSA process is simulated using Aspen Adsorption based on the real properties of our new sorbent which has good cyclic adsorption−desorption performance. We used the PSA model to fully explore the trade-offs between hydrogen recovery, regeneration pressure of sorbent, and steam requirement, which are the main factors affecting the overall thermal efficiency. To study their influence on IGCC efficiencies and identify the potentially optimal operating conditions, we integrated the PSA model with IGCC flowsheet in Aspen Plus. Based on the plantwide simulation, we noticed a significant energy savings achieved by PSA process in N 2 compression and CO2 compression (for certain cases). However, the relatively large hydrogen loss makes the total thermal (coal) input and related oxygen compression very large. In terms of the overall efficiency, IGCC-warm PSA can produce similar thermal efficiencies to IGCC-cold Selexol. In order to achieve this, warm PSA needs a narrow operating range of process parameters to balance the trade-off between the hydrogen loss, steam consumption and work requirement for CO2 compression. This work provides a framework for assessing the feasibility of sorbent-technology for warm CO2 capture in IGCC system. New sorbents could be tested and evaluated using this framework. Our finding indicates conducting the gas cleanup at high temperature is only slightly more advantageous than the cold Selexol process. But it is worth noting that all our analysis is based on a warm desulfurization process consuming 2.3% of the hydrogen, a new sorbent which still has lots of room for improvement, and a PSA process without process optimization. Further research is needed toward synthesizing new sorbent materials with higher working capacity and improved mass transfer, designing a better PSA configuration with higher H2 recovery and less steam consumption, inventing a new warm desulfurization process with reduced H2 consumption, and better heat integration of the whole IGCC plant. These developments could help make IGCC−CCS more affordable and acceptable.

Figure 16. Relationship between steam requirement and H2 recovery at regeneration pressures 1 atm for (○) current sorbent, (□) imaginary sorbent with faster kinetics (kLDF increased by 10 times), and (◊) imaginary sorbent with larger capacity (capacity increased by 3 times).

reducing the volume of a single bed from ∼530 m3 to 180 m3. Fourthly, the PSA cycle used in this paper is not specifically optimized for CO2 capture.47,43 Considering the requirement for a higher hydrogen recovery, more pressure equalization steps could be added to the cycle, which could further enhance the hydrogen recovery and IGCC-PSA thermal efficiency. In addition, the inclusion of a CO2 reflux step would help CO2 enrichment. Several recent papers have shown PSA with heavy reflux can achieve very high CO2 recovery and purity.45−48 However, many of the improved separation methods require more sorbent, increasing the capital cost. A good balance between capital investment and PSA performance could be achieved based on the recent progress49−51 on the study of CO2 capture using PSA. Fifthly, if properly chosen, the warm PSA sorbent can also work as a catalyst for the water− gas shift reaction, and also capture the sulfur. Thus, the warm PSA process can combine with a water gas shift reactor to form a new process called sorption enhanced water gas shift (SEWGS). The reactor is configured like the PSA cycle described here but can achieve enhanced CO conversion to CO2, and CO2 and H2S capture at the same time. Process simulation52,53 indicates IGCC-SEWGS based on K2CO3promoted HTls54−56 combined with removing the sulfur as gypsum by FGD (rather than as elemental sulfur via the RTI/ Eastman process as assumed here) can improve the thermal efficiency significantly (by about 2.5 HHV% points52). Because the IGCC-warm PSA process discussed here already has performance competitive with the conventional IGCC− CCS (Selexol) technology, even modest improvements could make this a preferred technology option. Further research is needed. To summarize: this paper shows that IGCC-warm PSA using an existing sorbent is competitive with conventional IGCC− CCS with cold solvent CO2 capture, and there is significant room for improvement in efficiency of the PSA-based process. It also appears that the IGCC-warm PSA process could be significantly cheaper than the conventional approach. We recommend further research in this area to improve the CO2

■

ASSOCIATED CONTENT

S Supporting Information *

Information about the PSA models in Aspen Adsorption, IGCC model in Aspen Plus, as well as a necessary Excel file for replicating the calculations, and results for all the cases considered. This material is available free of charge via the Internet at http://pubs.acs.org/.

■

AUTHOR INFORMATION

Corresponding Author

*Email: [email protected]. Notes

The authors declare no competing financial interest. 11156



■

Article

Technological and economical analysis. Energy Convers. Manage. 2007, 48, 2680−2693. (10) Chiesa, P.; Kreutz, T.; Lozza, G. CO2 Sequestration from IGCC Power Plants by Means of Metallic. J. Eng. Gas Turbines Power 2007, 129, 123−134. (11) Grainger, D.; Hägg, M.-B. Techno-economic evaluation of a PVAm CO2-selective membrane in an IGCC power plant with CO2 capture. Fuel 2008, 87, 14−24. (12) Alptekin, G.; Jayaraman, A.; Copeland, R. A low cost, high capacity regenerable sorbent for pre-combustion CO2 capture. In 2011 NETL CO2 Capture Technology Meeting, Pittsburgh, PA; Department of Energy/National Energy Technology Laboratory: Pittsburgh, PA, 2011. (13) Ito, S.; Makino, H. Carbon dioxde separation from coal gas by physical adsorption at warm temperature. Greenhouse Gas Control Technol. 1998, 131−136. (14) Couling, D. J.; Prakash, K.; Green, W. H. Analysis of membrane and adsorbent processes for warm syngas cleanup in integrated gasification combined-cycle power with CO2 capture and sequestration. Ind. Eng. Chem. Res. 2011, 50, 11313−11336. (15) Couling, D. J.; Das, U.; Green, W. H. Analysis of hydroxide sorbents for CO2 capture from warm syngas. Ind. Eng. Chem. Res. 2012, 51, 13473−13481. (16) Yang, R. T. Adsorbents Adsorbents: Fundamentals and Applications; John Wiley & Sons, Inc.: New York, 2003. (17) Ruthven, D. M. Principles of Adsorption and Adsorption Process; John Wiley & Sons, Inc.: New York, 1994; pp 204−218. (18) Choi, S.; Drese, J. H.; Jones, C. W. Adsorbent materials for carbon dioxide capture from large anthropogenic point sources. ChemSusChem 2009, 2, 796−854. (19) Kyaw, K.; Shibata, T.; Watanabe, F.; Matsuda, H.; Hasatani, M. Applicability of zeolite for CO2 storage in a CaO−CO2 high temperature energy storage system. Energy Convers. Manage 1997, 38, 1025−1033. (20) Díaz, E.; Muñoz, E.; Vega, A.; Ordóñez, S. Enhancement of the CO2 retention capacity of X zeolites by Na- and Cs-treatments. Chemosphere 2008, 70, 1375−82. (21) Díaz, E.; Muñoz, E.; Vega, A.; Ordóñez, S. Enhancement of the CO2 retention capacity of Y zeolites by Na and Cs treatments: Effect of adsorption temperature and water treatment. Ind. Eng. Chem. Res. 2008, 47, 412−418. (22) Beruto, D.; Botter, R.; Searcy, A. W. Thermodynamics and kinetics of carbon dioxide chemisorption on calcium oxide. J. Phys. Chem. 1984, 88, 4052−4055. (23) Filitz, R.; Kierzkowska, A. M.; Broda, M.; Müller, C. R. Highly efficient CO2 sorbents: Development of synthetic, calcium-rich dolomites. Environ. Sci. Technol. 2012, 46, 559−565. (24) Nakagawat, K.; Ohashi, T. A novel method of CO2 capture from high temperature gases. J. Electrochem. Soc. 1998, 145, 1344−1346. (25) Bretado, M. E.; Guzmán Velderrain, V.; Lardizábal Gutiérrez, D.; Collins-Martínez, V.; Ortiz, A. L. A new synthesis route to Li4SiO4 as CO2 catalytic/sorbent. Catal. Today 2005, 107−108, 863−867. (26) Gauer, C.; Heschel, W. Doped lithium orthosilicate for absorption of carbon dioxide. J. Mater. Sci. 2006, 41, 2405−2409. (27) Siriwardane, R. V.; Robinson, C.; Shen, M.; Simonyi, T. Novel regenerable sodium-based sorbents for CO2 capture at warm gas temperatures. Energy Fuels 2007, 21, 2088−2097. (28) Siriwardane, R. V.; Stevens, R. W. Novel regenerable magnesium hydroxide sorbents for CO2 capture at warm gas temperatures. Ind. Eng. Chem. Res. 2009, 48, 2135−2141. (29) Ding, Y.; Alpay, E. Equilibria and kinetics of CO2 adsorption on hydrotalcite adsorbent. Chem. Eng. Sci. 2000, 55, 3461−3474. (30) Yong, Z.; Rodrigues, A. E. Hydrotalcite-like compounds as adsorbents for carbon dioxide. Energy Convers. Manage. 2002, 43, 1865−1876. (31) Mayorga, S. G.; Weigel, S. J.; Gaffney, T. R.; Brzozowski, J. R. Carbon Dioxide Adsorbents Containing Magnesium Oxide Suitable for Use at Hight Temperature; U.S. Patent No. 6280503 B1, 2001.

ACKNOWLEDGMENTS We thank Dr. David J. Couling for the first-version IGCC flowsheet with warm syngas cleanup. The authors also thank Mr. Randall Field and Dr. Srinivas Seethamraju for helpful discussions. We gratefully acknowledge financial support from BP. We also thank AspenTech for the use of Aspen Plus and Aspen Adsorption in this research.

■

NOMENCLATURE

Symbols Used in the Adsorption Model

Cp,gas = gas phase capacity, J/mol·K Cp,s = solid phase heat capacity, J/mol·kg ci = concentration, mol/m3 dp = sorbent particle diameter, m D = bed diameter, m ΔHads = heat of adsorption, J/mol a = isotherm parameter, mol/(kg·atm) b = isotherm parameter, K c = isotherm parameter, atm−1 d = Isotherm parameter, K kLDF = linear driving force rate constant, s−1 L = bed length, m P = pressure, Pa Pi = partial pressure of species i, atm qi = sorbent loading of species i, mol/kg qi* = sorbent loading of species i at equilibrium, mol/kg T = remperature, K t = rime, s vg = gas phase velocity, m/s Greek Symbols

μ = dynamic viscosity, Pa·s ε = void fraction ρgas = gas phase molar density, kg/m3 ρs = adsorbent bulk density, kg/m3

■

REFERENCES

(1) Katzer, J.; Ansolabehere, S.; Beer, J.; Deutch, J.; Ellerman, A. D.; Friedmann, S. J.; Herzog, H.; Jacoby, H. D.; Joskow, P. L.; McRae, G.; Lester, R.; Moniz, E. J.; Steinfeld, E. The Future of Coal; Massachusetts Institute of Technology: Cambridge, MA, 2007. (2) International Energy Outlook 2011; Department of Energy/Energy Information Administration: Pittsburgh, PA, 2011. (3) Frey, H. C.; Zhu, Y. Improved system integration for integrated gasification combined cycle (IGCC) systems. Environ. Sci. Technol. 2006, 40, 1693−9. (4) Esber, G. S. Carbon dioxide capture technology for the coalpowered electricity industry: A systematic prioritization of research needs. Master’s Thesis, Massachusetts Institute of Technology, Cambridge, MA, 2006. (5) Fossil Energy Power Plant Desk Reference; Department of Energy/ National Energy Technology Laboratory: Pittsburgh, PA, 2007. (6) Cost and Performance Baseline for Fossil Energy Plants; Department of Energy/National Energy Technology Laboratory: Pittsburgh, PA, 2007; Vol. 1. (7) Current and Future Technologies for Gasification- Based Power Generation; Department of Energy/National Energy Technology Laboratory: Pittsburgh, PA, 2010; Vol. 2. (8) Schlather, J.; Turk, B. Desulfurization process versus traditional scrubbers for a warm syngas cleanup technologies. Gasification Technologies Conference; Department of Energy/National Energy Technology Laboratory: Pittsburgh, PA, 2007. (9) Amelio, M.; Morrone, P.; Gallucci, F.; Basile, A. Integrated gasification gas combined cycle plant with membrane reactors: 11157



Article

and H2S capture for use in sorption enhanced water−gas shift. Int. J. Greenhouse Gas Control 2011, 5, 505−511. (56) Walspurger, S.; Boels, L.; Cobden, P. D.; Elzinga, G. D.; Haije, W. G.; van den Brink, R. W. The crucial role of the K-aluminium oxide interaction in K-promoted alumina- and hydrotalcite-based materials for CO2 sorption at high temperatures. ChemSusChem 2008, 1, 643− 650.

(32) Liu, Z.; Green, W. H. Experimental investigation of sorbent for warm CO2 capture by pressure swing adsorption. Ind. Eng. Chem. Res. 2013, 52, 9665−9673. (33) Field, R. P.; Brasington, R. Baseline flowsheet model for IGCC with carbon capture. Ind. Eng. Chem. Res. 2011, 50, 11306−11312. (34) Cost and Performance Baseline for Fossil Energy Plants. Volume 1: Bitumminous Coal and Natural Gas to Electricity; Final Report Rev 1; Department of Energy/National Energy Technology Laboratory: Pittsburgh, PA, 2007; Vol. 1. (35) Cost and Performance Baseline for Fossil Energy Plants. Volume 1: Bitumminous Coal and Natural Gas to Electricity; Final Report Rev 2; Department of Energy/National Energy Technology Laboratory: Pittsburgh, PA, 2010; Vol. 1. (36) Gupta, R. P.; Turk, B. S.; Portzer, J. W.; Cicero, D. C. Desulfurization of syngas in a transport reactor. Environ. Prog. 2001, 20, 187−195. (37) Gupta, R.; Turk, B.; Lesemann, M.; Schlather, J.; Denton, D. Status of RTI/Eastman warm gas clean-up technology and commercialization plans. Gasification Technologies Conference; Washington, DC, 2008. (38) Stöcker, J.; Whysall, M.; Miller, G. Q. 30 Years of PSA Technology for Hydrogen Purification 1998, 1−25. (39) Crittenden, B.; Thomas, W. J. Adsorption Technology & Design; Butterworth-Heinemann: Oxford, U.K., 1998; pp 189−201. (40) Ruthven, D. M.; Farooq, S.; Knaebel, K. S. Pressure Swing Adsorption; Wiley-VCH Publisher: Weinheim, 1995; pp 67−93. (41) Suzuki, M. Adsorption Engineering; Kodansha, Ltd.: Tokyo, 1990; pp 245−275. (42) Yang, R. T. Gas Separation by Adsorption Process; Imperial College Press: London, 1997; pp 126−133, 237−268.. (43) Grande, C. Advances in pressure swing adsorption for gas separation. ISRN Chem. Eng. 2012, 1−13. (44) Botros, B. B.; Brisson, J. G. Targeting the optimum steam system for power generation with increased flexibility in the steam power island design. Energy 2011, 36, 4625−4632. (45) Ritter, J. A. Bench-scale development and testing of rapid PSA for CO2 capture. 2013 NETL CO2 Capture Technology Meeting; National Energy Technology Laboratory: Pittsburgh, PA, 2013 (46) Ritter, J. A.; Rahman, A.; Ebner, A. D. Carbon dioxide capture from flue gas by high feed throughput pressure swing adsorption processes. 2012 AIChE Meeting 2012, 331. (47) Reynolds, S. P.; Ebner, A. D.; Ritter, J. A. Stripping PSA cycles for CO2 recovery from flue gas at high temperature using a hydrotalcite-like adsorbent. Ind. Eng. Chem. Res. 2006, 45, 4278−4294. (48) Reynolds, S. P.; Mehrotra, A.; Ebner, A. D.; Ritter, J. a. Heavy reflux PSA cycles for CO2 recovery from flue gas: Part I. Performance evaluation. Adsorption 2008, 14, 399−413. (49) Agarwal, A. Advanced strategies for optimal design and operation of pressure swing adsorption processes. PhD Thesis, Carnegie Mellon University, Pittsburgh, PA, 2010. (50) Dowling, A. W.; Vetukuri, S. R. R.; Biegler, L. T. Large-scale optimization strategies for pressure swing adsorption cycle synthesis. AIChE J. 2012, 58, 3777−3791. (51) Ritter, J. A.; Ebner, A. D.; Reynolds, S. P. New Adsorption Cycles for Carbon Dioxide Capture and Concentration; National Energy Technology Laboratory: Pittsburgh, PA, 2009 (52) Manzolini, G.; Macchi, E.; Gazzani, M. CO2 capture in Integrated Gasification Combined Cycle with SEWGS. Part B: Economic assessment. Fuel 2013, 105, 220−227. (53) Gazzani, M.; Macchi, E.; Manzolini, G. CO2 capture in integrated gasification combined cycle with SEWGS. Part A: Thermodynamic performances. Fuel 2013, 105, 206−219. (54) Van Selow, E. R.; Cobden, P. D.; Verbraeken, P. A.; Hufton, J. R.; van den Brink, R. W. Carbon capture by sorption-enhanced water− gas shift reaction process using hydrotalcite-based material. Ind. Eng. Chem. Res. 2009, 48, 4184−4193. (55) Van Dijk, H. A. J.; Walspurger, S.; Cobden, P. D.; van den Brink, R. W.; de Vos, F. G. Testing of hydrotalcite-based sorbents for CO2 11158


Analysis of Adsorbent-Based Warm CO2 Capture Technology for

Recommend Documents