Desktop version

Home arrow Engineering

  • Increase font
  • Decrease font


<<   CONTENTS   >>

Moving bed chemical looping reforming for syngas production

Some commonly used metal oxide oxygen carriers in CLC systems, such as iron oxide and copper oxide, pose challenges when used for fluidized bed CLRs, as they are thermodynamically favorable of frilly oxidizing methane to CO, and H,0 instead of CO and H,. To overcome this problem, the CLR systems have to be operated under conditions with a fuel-to-oxygen carrier ratio higher than the value of full combustion by stoichiometry in order to increase syngas generation. However, other problems, such as low fuel conversion, low syngas purity, and carbon deposition, will occur under such operating conditions. There is another metal oxide, NiO, that is suitable for both CLC processes and fluidized bed CLR, owing to the high syngas yield resulting from its excellent reactivity. However, the toxicity and price of NiO may limit the feasibility of its use in large-scale CLR processes.

Alternatively, when the reducer in a CLR for syngas process is designed in a moving bed mode, it is possible to have a high performance in both fuel conversion and syngas purity when an noil-based oxygen carrier is used (Fan, 2010; Zeng et al.. 2012; Fan et al., 2015). This makes the moving bed reducer design an attractive configuration option for CLR processes from both economical and operational standpoints.

Similar to that of the fluidized bed CLR process, as described in section 2.1.3, the economic analysis of a moving bed CLR process using ITCMO as oxygen carrier particles was also conducted via an evaluation of a NGTS-GTL process (Fan, 2017). The overall process diagram is the same as that shown in Figure 7, except that instead of the fluidized bed CLR, the NGTS system uses a cocurrent moving bed CLR process to convert natural gas to syngas. The NGTS system uses steam-natural gas co-feeding scheme.

Since the flow in a moving bed reactor can be considered as a plug flow with a constant residence time (Barelli et al., 2008), a single stage RGibbs reactor block in Aspen Plus* can be used to simulate its fluid flow behavior. To ensure that the syngas production of the packed moving bed NGTS process is equivalent to the baseline syngas production from the ATR (Gollener et ah, 2013), an optimized NGTS reducer configuration was developed, with the adjustment on the parameters consisting of H,0:C molar ratio, Fe,0,:C molar ratio, temperature swing along the reducer reactor, natural gas pre-heat temperature, and steam pre-heat temperature.

Table 5 listed the performance results for the NGTS-GTL process using moving bed CLR, as compared to the baseline GTL process with a 50,000 bpd liquid fuel output. The natural gas feed required by the process with moving bed CLR is 11% less than that of conventional process with ATR for generating an equivalent amount of liquid fuel. This benefit, combined with the elimination of the energy intensive ASU, decreases the parasitic energy requirement for syngas production by 60%, resulting in doubled net power output compared to a conventional plant. Hie moving bed CLR can achieve a high syngas selectivity with less than 3% CO, produced and no carbon deposition.

The results of economic analysis of the NGTS-GTL process using moving bed CLR are listed in Table 6. The NGTS process shows a 72.5% reduction on the capital cost for syngas production, leading to a 31.6% reduction on the total plant cost. As a result, the NGTS-GTL process using moving bed CLR reduces the liquid fuel production capital cost to S65,000/(bbl/day). a 25% decrease compared with the value of approximately $86,000/(bbl/day) of the baseline GTL process. The substantial reduction on the capital cost enables the NGTS-GTL process using moving bed CLR to be competitive when the crude oil price is over $48/bbl, a significant advantage over the baseline GTL process and the NGTS-GTL process using fluidized bed CLR which was described in section 2.1.3.

2.2.1 Iron-based oxygen carriers

Iron-based oxygen carriers are particularly attractive in chemical looping systems as they are abundantly available and low in cost. Certain types of non ores, like ilmenite, can be used directly as oxygen earners in chemical looping processes (Deshpande et al., 2015); they are also non-toxic.

There are three possible redox pans for non oxide when used as oxygen carriers, Fe,03-Fe304, Fe304-Fe0. and FeO-Fe. The Fe,03-Fe304 redox pan is thermodynamically the most oxidative as Fe,03 can frilly convert the hydrocarbon fuel into CO, and H,0, while being reduced to Fe304. As a result, the amount of unconverted CO/H, is significantly lower when compared to the case using NiO-Ni as the oxygen carrier redox pair. The Fe,04-FeO and FeO-Fe redox pairs are comparatively less oxidative. As a

Table 5. Overall performance of the NGTS process using moving bed CLR in a 50,000 bpd GTL plant (Gollener et al.,

2013).

Component

Base case

NGTS (10 atm) moving bed

Natural gas flow, kg'hr

354,365

317,094

H.O/C in natural gas

0.68

0.25

H,:CO

2.19

2.178

Stoichiometric number (S#)

1.59

1,96

Butane feed flow, kg'hr

18,843

18,843

Diesel fuel, bbl/day

34,543

34,543

Gasoline, bbl/day

15,460

15,460

Total liquids, bbl/day

50,003

50,003

Electncal Load (kWe)

Total Gross Power

303,700

303,700

Net Plant Power

40,800

179,050

Table 6. Cost summary for NGTS-GTL plant producing 50,000 bpd liquid fuel (Gollener et al., 2013).

Component (2011 $)

ATR

NGTS

Total plant cost ($x 1000)

2,750,000

1,880,000

Total as-spent cost ($xi000)

4,310,000

3,250,000

Capital costs ($/(bbl day))

86,000

65,000

Capital costs ($/bbl)

51.5

41.8

O&M costs (S/bbl)

27.29

27.23

Fuel costs ($/bbl)

16.9

12.6

Electricity costs ($/bbl)

-1.15

-5.1

Required selling price* ($/bbl)

94.54

76.53

West Texas intermediate crude oil competitive price (S/bbl)

60.5

48.9

*: calculated when the required return on equity equals the internal rate of return for 30 years of operation with assumed financial structure and escalations.

result, CLC processes using Fe-based oxygen carriers in a fluidized bed reducer use only the Fe,0,-Fe,04 redox pair with no Anther reduction of the oxygen carrier. For CLR applications, however, Fe-based metal oxide needs to be reduced to the FeO-Fe redox pair, where it falls in the “syngas production” region, as shown in the modified Ellingham diagram of Figure 2(b). This indicates that the FeO-Fe redox pair thermodynamically inhibits the full oxidation of syngas into CO,/H,0. In addition, adding support materials, such as TiO, and A1,03, to an Fe-based oxygen carrier may result in the formation of complex materials (FeTiO, and FeAl,04) that are even less oxidative than FeO, resulting in an even higher syngas purity when compared to a purely iron oxide oxygen carrier. Therefore, if the Fe-based oxygen carrier is sufficiently reduced to form a mixture of FeO and Fe in the CLR reducer reactor, a high syngas purity with low C'0,/H,0 concentration can be obtained. When iron-based metal oxide is used as oxygen carrier, the reducer reactor suitable for such a CLR reducer reaction scheme has to be designed in a moving bed mode, such that a high performance in both Aiel conversion and syngas purity can be obtained.

A composite oxygen carrier consisting of 60% Fe,O}/40% Al.O, by weight was tested in a thennogravimetric analyzer (TGA) at 900 °C using pure methane. The oxygen carrier had a particle diameter of 1.5 mm in order to accommodate the moving bed operating conditions. The weight change of the oxygen carrier during methane reduction process was shown in Figure 14. Hie process can be divided into two stages of Fe,03 reduction, the stage where weight loss was observed and the carbon deposition stage where weight gain was observed. In the Fe,03 reduction stage, the weight of the oxygen carrier decreased rapidly, owing to the loss of oxygen from oxygen carrier. The weight loss was up to approximately 11% of its original value. Then, the weight eventually increases sharply, when carbon deposition is catalyzed by reduced metallic iron.

TGA studies showed that two oxidation states of oxygen earner particles are undesirable for syngas production, under-reduced states of Fe,0,/Fe,04 and Fe.OyFeO. and over-reduced state of Fe. The

Weight change of 60% Fe,O/40% Al.Oj oxygen earner m methane reduction process

Figure 14. Weight change of 60% Fe,O}/40% Al.Oj oxygen earner m methane reduction process.

particles with under-reduced states frilly oxidize the reactants to CO, and H,0, while those with overreduced states cause carbon deposition. In a moving bed reducer, as the particles are in a plug flow mode, the residence time of the oxygen earner particles can be precisely controlled. By selecting a suitable residence time of the oxygen carrier particles, both over-reduction and under-reduction of oxygen carrier particles can be avoided, and only FeO/Fe particles can present at the exit of the reducer. Additionally, with a cocurrent configuration where both syngas and oxygen carrier particles are fed from the top of the reactor and flow downwards together, the syngas is in contact with only reduced oxygen earner particles consisting of FeO/Fe, which thermodynamically inhibits full oxidation of H, and CO. Therefore, producing syngas with a high purity can be achieved.

2.2.2 Lab-scale fixed bed tests

The feasibility of moving bed CLR using an noil-based oxygen earner was first investigated with fixed bed experiments (Luo et al.. 2014). A fixed bed reactor with a height of 38.1 cm and an internal diameter of 1.27 cm was used to investigate the performance of the cocurrent moving bed reactor. To mimic the oxidation state of the cocurrent moving bed reducer, two layers of iron-based oxygen carriers with different oxidation states were used to pack the fixed bed reactor with reduced oxygen earner particles at the lower section and frilly oxidized particles at the top. The mass ratio of the two different oxygen carrier particles was 2.78:1. The ratio was determined to represent the specific residence time for various oxidation states of non obtained from TGA experiments for the kinetic study of an oxygen carrier. N, diluted CH4 was injected as the reductive gas and N, diluted CO, was injected as the oxidative gas into the reactor from the top using digital mass flow controllers. The fixed bed reactor was put in an electrical furnace and was operated at 990 °C. The composition of the product at the gas outlet of the reactor was analyzed by both a nou-dispersive infrared (NDIR) gas analyzer and a gas chromatograph (GC).

The composition of the gas stream from the outlet of the fixed bed reactor is shown in Figure 15(a). In the first 2500 seconds, 50 ml/min CH, and 50 ml/min N, were fed together to the reactor. The methane concentration increased at the beginning, maintained near constant levels in the middle, and then decreased at the end; while the CO concentration decreased at the beginning, maintained constant levels in the middle, and increased in the end. It indicated that the gas composition was regulated by different oxidation states of the ITCMO particles, including Fe, FeTi03, and/or Fe,Ti04. The intermediate oxidation state of Fe304 has a slower reaction rate with CH4, compared to Fe,TiOs and Fe,Ti04/ FeTi03, which is shown with the TGA reactivity test results. When the gas feeding rate was reduced to 60 ml/min (50% CH4 balanced with N,) at tune point (1), as marked in Figure 15(a), an immediate decrease in the methane concentration was observed. In the meantime, the CO:CO, ratio increased to approximately 10, resulting in a very high syngas selectivity. This is mainly due to the longer gas

residence time as gas velocity is reduced. Slight improvements on the methane conversion and syngas selectivity were observed again when the reactor temperature was increased to 1050 °C at time point (2). The thermodynamic analysis using the modified Ellingham diagram (Figure 2) indicates that methane could be fully converted under all the experimental conditions. However, the incomplete methane conversion and its improvement towards long residence time and higher reactor temperature indicated that there is a kinetic restriction for the chemical reaction between ITCMO particles and methane.

The kinetic limitation of syngas production was further explored by oxidizing the above reduced ITCMO particles in the fixed bed with CO, at a flow rate of 30 ml/min and diluted by 30 ml/min N, for about 1000s, subsequently with a pure CO, stream at 60 ml/min. As shown in Figure 15(b), the reduced ITCMO particles were oxidized by the CO, stream which was converted to CO. The CO:CO, ratio reached the thermodynamic equilibrium shown in the modified Ellingham diagram (Figure 2) and had no change when CO, concentration in the feed increased at time point (1), marked in Figure 15(b), indicating the product gas composition was dictated by thermodynamics rather than kinetics. A sharp change in the product composition, where CO and CO, concentration flipped over near the end of the CO, oxidation experiment, was observed. Hie change illustrated the dependence of this reaction on oxygen carrier composition. When the oxygen carrier is in the oxidation state of Fe/FeO, the CO concentration is greater than CO, and a high purity syngas stream is generated when methane is fed, as shown in Figure 15(a). However, when the oxygen earner is oxidized to a state above FeO, the CO:CO, ratio in the gas steam immediately drops to a low value.

The experiments involving CH4 reduction and CO, oxidation indicated that the reaction of methane with iron-based oxygen earners was relatively slow and, hence, rate limiting. The product gas composition, the syngas purity, was determined by thermodynamics rather than kinetics (Luo et al., 2014).

2.2.3 Bench-scale moving bed reducer tests

A bench-scale moving bed reactor system with an internal reactor diameter of 0.05 m and a height of 0.9 m was used to further investigate the performance of a moving bed reducer for CLR applications using ITCMO particles. The moving bed was operated in a cocurrent mode with all the feedstock and ITCMO particles introduced from the top of the reactor, moved downwards and exited the reactor from the bottom. The solids flow rate was controlled by a screw feeder at the bottom of the reactor. Gaseous products were sampled along the reactor to quantify the concentrations of CO, CH4, CO,, O,, and H, at different stages of the reaction. The reactor was operated under ambient pressure.

The feedstock with a flow rate of 2 1/min and a gas composition of 90% CH4 and 10% N, was tested at a temperature of 1000 °C, with a Fe,0,:CH4molar ratio of 0.8. As shown in Figure 16, stable syngas generation was achieved with a CO:CO, molar ratio higher than 9 and H,:CO ratio of around 2. The CH4 conversion was around 95%, with a syngas purity higher than 85%.

CH4 can be со-fed with coal to the moving bed chemical looping systems as a H,-rich feedstock for gasification purpose. Syngas with higher H, concentrations can be obtained, as compared with CLG process with coal only. As shown in Figure 17(a), when methane was со-fed with coal, the H,:CO ratio in the syngas increased to 1, compared to around 0.65 for the coal only case. The syngas purity is above 95% on a diy basis with the CO:CO, ratio greater than 10 and minimal unconverted CH4. When steam was also added together with coal and methane, the H,:CO ratio in the syngas was further increased, as shown in Figure 17(b). When the CH4:steam:coal mass ratio was about 1:0.9:1, the produced syngas had a H,:CO ratio of around 1.8, a CO:CO, ratio of around 6, and a syngas purity of greater than 85% on a dry basis.

2.2.4 Sub-pilot moving bed CLR system

A 25 kWth sub-pilot scale moving bed CLR system using ITCMO as the oxygen carrier and CH4 as feedstock, as shown in Figure 18, was tested (Fan et al., 2013). The reactor had a height of 1.52 m and an inner diameter of 0.1 m. The solids flow rate was controlled by a rotary solids feeder at the bottom of the reactor system. External heaters with a PID control program for maintaining the reactor temperature were used to heat the reactor system. A solids reservoir section was placed above the reducer such that the solid level was maintained above the reaction section and solids were at the required operating temperature

Syngas product distribution for CLR of methane (Luo et al., 2014)

Figure 16. Syngas product distribution for CLR of methane (Luo et al., 2014).

Syngas product for coal and methane co-feeding CLG processes

Figure 17. Syngas product for coal and methane co-feeding CLG processes.

when entering the reaction section. Various parameters, including Fe,0,:CH4 molar ratios from 0.5 to 1.4, temperatures from 900 to 1050 °C, and steam:CH4 molar ratios from 0 to 0.4, were tested.

A typical product gas composition from the sub-pilot CLR unit, with an operating temperature of 975 °C, a Fe:03:CH4molar ratio of 0.73 and a CH4 flow rate of 101/min, was shown in the Figure 18 (Fan et al., 2013). When changing the operating conditions, such as reactant residence time or temperature, achieving a higher CH4 conversion and a higher syngas purity is possible. The methane conversion rate was over 99.9%, with a H;:CO ratio of 1.97 and a syngas purity of 91.3%. Comparing to the bench- scale unit described in section 2.2.3, the sub-pilot unit achieved a higher methane conversion and a higher syngas purity, even though the sub-pilot unit was operated under a condition with a slightly lower operating temperature and a lower Fe,OrCH4 molar ratio. This was mainly due to the higher reactant residence time (20%) compared to the bench-scale unit. Thus, it can be concluded that the reactant residence time is a key factor for optimization of a moving bed CLR process design. This can be adher ed by using a non-mechanic valve, like an L-valve, which can control the solids flow rate by adjusting the injected aeration gas flow rate (Wang and Fan, 2015), as the reducer is operated in a packed bed mode and the oxygen carrier particles have a large diameter.

kW sub-pilot CLR process

Figure 18. 25 kW4 sub-pilot CLR process: Expenment unit (left); flow schematic (middle); product gas composition (light)

  • (Fan et al., 2013).
  • 2.2.5 Other natural gas reforming schemes using moving bed CLR

The moling bed CLR process can have various configurations applied for different feedstocks and applications. In addition to natural gas/steam co-feeding for the STS process, as described at the beginning of section 2.2, co-feeding natural gas with other source of carbon, such as solid fuels like coal and biomass, as well as gaseous CO, is possible. When CO, is used as a feedstock, co-feeding with natural gas. the relationship between the CO, flow rate and H,:CO ratio in the product is nonlinear, which enables a modular approach to the reducer design of the moving bed CLR process for product yield enhancement. Cases with several different feedstock and modular design are described below.

Coal and natural gas co-feed

NaUiral gas is a good feedstock for supplemental hydrogen and carbon in chemical production processes relying on coal-derived syngas. Natural gas has a hydrogen-to-carbon ratio of about 4, a value much higher than that of coal, which is around 1. Wien the natural gas fraction in the co-feed feedstock increases, the carbon efficiency of the syngas generation step and the overall process increases. The economic feasibility of a coal and natural gas co-feeding moving bed coal to syngas (CTS) process was investigated via an evaluation of the coal and natural gas co-feeding CTS process for syngas generation to produce methanol (Li et al., 2011). 50% ННЛ’ ratio between natural gas and coal feeding rate, which allows both high utilization rate of coal and carbon efficiency improvement, was used in the process. Optimization analysis on the co-feed CTS process was conducted for identifying suitable Fe,03:C and H,0:C ratios to achieve autothermal operation and specific syngas compositions for downstream methanol synthesis.

The overall block diagram for the coal and natural gas co-feed CTS for a methanol production plant is shown in Figure 19, as based on a baseline case for a net plant output of 10,000 toimes of methanol per day. described in a DOE/NETL report (Summers, 2014). The co-feed CTS process is integrated into the methanol production plant with a second sub-model developed to balance steam generation against the steam loads in the system, which includes the steam for co-feed CTS process usage and power generation.

Table 7 listed the performance of the integrated process with coal and natural gas co-feed CTS process, as compared to the baseline cases with and without CO, capture as well as CTS process without natural gas co-feed. By co-feed natural gas with coal, the overall process was improved with reduced consumption of water, steam, and air, as well as reduced sulfur and carbon emissions. The co-feed case requires less parasitic energy for coal and ash handling, with coal consumption also reduced. An improvement in carbon efficiency with natural gas co-feed leads to a higher syngas quality from the syngas generation process resulting in a lower CO, concentration in the syngas and correspondingly

Overall block diagram for the co-feed CTS methanol production system

Figure 19. Overall block diagram for the co-feed CTS methanol production system.

Table 7. Overall performance of coal and natural gas co-feed CTS process in a 10,000 tonne, day methanol plant

(Summers, 2014).

Case

MBA

MB-B

CTS w/o co-feed

Co-feed CTS

Mass Flows (lb/hr)

As Received Coal

1,618,190

1,618,190

1,395,457

718,631

Oxygen from ASU containing 95% O,

1,010,968

1,010,968

NA

NA

Steam to Gasifier, Quench, Shift reactors, CTS

1,533,584

1.533,584

1,624,318

693,587

Air to Direct-fired boiler

121,518

121,518

181,009

606,106

Clean syngas for methanol production

1,183,080

1,183,080

1,025,106

1,039,864

Tail gas from Claus unit

61,476

61,476

50,089

25,589

Captured CO, (no capture for MB-A)

0

1,569,410

1,302,138

663,393

Electrical Load (kWe)

Total Gioss Power

320,680

390,170

20,830

31,491

Total Net Power* *

12,280

21,480

-323,504

-245,692

Notes ** Negative value indicates power purchase required.

a smaller acid gas removal (AGR) unit, CO, compressor, syngas compressors, and syngas cooling loads. Tlie results from the process efficiency standpoint alone show significant benefits associated with co-feeding natural gas to the reducer.

Table 8 listed the results of economic analysis for the cases of coal and natural gas co-feed CTS, baselines with and without CO, capture, and CTS without natural gas co-feed. The co-feed CTS case has a lower capital cost investment than that of the CTS without natural gas co-feed, owing to a higher carbon efficiency. However, the co-feed CTS case has a required selling price of methanol 5% higher than the CTS case, mainly due to the high natural gas price.

CO, and natural gas co-feed

Experimental results showed that co-feeding CO, with natural gas can effectively reduce or eliminate carbon deposition on the oxygen carriers (Rydeu et al., 2008). In addition, adding CO, in the feedstock

Table 8. Comparative summary of capital and operating costs of methanol production.

Case (2011 $)

MBA

MB-B

CTS w/o co-feed

Co-feed CTS

Total plant costs (Million $)

4,586

4,775

3,497

2,996

Total as-spent costs (Million $)

6,580

6,852

5,003

4,291

Capital costs ($/gal)

1.18

1.23

0,89

0.81

Fuel costs ($/gal)

0.24

0.26

0.18

0.39

O&M costs ($/gal)

0.23

0.23

0.16

0.14

CO, TS&M costs ($/gal)

0

0.06

0.05

0.03

Electricity' costs (S/gal)

0

0

0.14

0.11

Requued selling price* (S/gal)

1.64

1.78

1.41

1.48

*: calculated when the required return on equity equals the internal rate of return for 30 years of operation with assumed financial structure and escalations.

changes the syngas purity and the H::CO ratio in the gas product by affecting the reaction equilibrium in the CLR reducer. This co-feed scheme enables CLR to be CO, neutral, i.e., the CO, input to a process is equal to the CO, output from the process under a steady state condition, or even CO, negative, i.e., the CO, input to a process is greater than the CO, output from the process. A CO, neutral or negative process may involve a CO, recycle in the process system.

Process simulation on a CLR process with CO, со-fed with natural gas, steam and recycled fuel gas from downstream processes to produce syngas for liquid fuels production, whose schematic diagram was shown in Figure 20, was conducted (Katlie et al., 2017). The CLR process used the moving bed CLR process with ITCMO particles, as described above. Like the beginning of section 2.2, the moving bed reducer of the CLR process was simulated using single stage RGibbs block fr om Aspen Plus®. The reducer was set to a temperature of 900 °C and a pressure of 1 atm. The equilibrium condition at the outlet of the reducer was simulated by the reducer model. Hie performance of the process was compared to the baseline case in section 2.1.3, which uses the conventional ATR for syngas production (Golleuer et al.,

2013).

CO, separated from the outlet stream of the Fisclier-Tropsch synthesis was recycled as CO, co-feed source to reduce CO, emission. The targets of syngas generation by the CLR process were to match the baseline performance with H,:CO ratio of 2.19 and S# greater than 1.58 at a FI, flow rate of at least 45,285 kmol/hr. 90% CO, recycling was assumed in the process simulation (Golleuer et al.. 2013).

The process simulation showed that an optimally designed CO, and natural gas co-feed CLR process can reduce the natural gas consumption by 22% and steam consumption by 27%, as compared to the ATR baseline case, as shown in Figure 21, with a H,:CO ratio matching the ATR baseline case and H, flow

Chemical looping CO, recycle scheme for liquid fuels production (Fan et al., 2017)

Figure 20. Chemical looping CO, recycle scheme for liquid fuels production (Fan et al., 2017).

rate and S# greater than the baseline case. At a natural gas price of S2 MMBtu, these benefits represented a $7,507.94/hr cost reduction for natural gas consumption, leading to an annual saving of $59.2 Million for a CLR plant, given an annual operation time of 90% of its 8,760 h/a designed operation time (U.S. Energy Information Administration, 2016; Gas-to-Liquids Conversion. 2012; Quality Guidelines for Energy System Studies, 2012).

When the CO, со-fed to the system, such as that shown in Figure 20, is higher than the available CO, recycled from downstream processes and, thus, external CO, sources are required, the process utilizes more CO, than it generates and is, thus, CO, negative (Fan et al., 2017). The simulated relationships among feedstock parameters of CO,:CH4 ratio and H,0:CH4 ratio, as well as syngas product parameters of H,:CO ratio, CO, concentration, and carbon recycle parameter (CRP) defined by the ratio of CO, entering the reducer to CO, exiting the reducer, are shown in Figure 22. From the figure, the expected syngas quality can be obtained if CO,, steam and methane co-feed ratios are given. In addition, the conclusion as to whether the process can be operated under CO, neutral or negative conditions can be obtained. The ability of the syngas production section to consume CO, is valuable to offset carbon emissions from other sections of the entire process or other processes and is able to transform the CO, market.

CO, and natural gas co-feed with reducer modularization

Further analysis on the above CO, co-feed scheme found that the amount of CO, со-fed to the CLR reducer posed a nonlinearity effect on the quality of syngas produced from the CLR process, as

ATR baseline case for liquid fuels production using syngas from autothermal reforming (Gollener et al., 2013)

Figure 21. ATR baseline case for liquid fuels production using syngas from autothermal reforming (Gollener et al., 2013).

Simulation of the effect of CO, input rates at various CRP and H,0:CH ratios (Fan et al., 2017)

Figure 22. Simulation of the effect of CO, input rates at various CRP and H,0:CH4 ratios (Fan et al., 2017).

exemplified in Figure 23, where 1 mole of CH, is со-fed with steam and CO, to a moving bed CLR reducer at a temperature of 900 °C, a pressure of 1 atm, and an effective Fe,0}:CH4 ratio of 0.40 (Fan et al., 2017). Such a nonlinearity relationship induced by CO, co-feed can be beneficial as it can be used to optimize the system configuration and enhance product yield using a reactor modularization approach, as exemplified in Figure 24, where a two reducer modularized system is shown (Fan et al., 2017).

When the system uses two reducers, one producing syngas product with a H,:CO ratio of 2.50 and the other with a ratio of 1.27, the system can provide the same syngas with a combined methane consumption of only 0.83 mol/s, a 17% reduction. The details of this operating condition are shown in Table 9. The decreased carbon and hydrogen supply from methane are offset by the increase of steam and CO, co-feed rates. The economic benefits of the reduction in natural gas flow with a two-reducer system will outweigh the increased cost of higher steam and CO, input.

When the two reducer moving bed CLR system is integrated into the 50,000 bpd GTL plant, as shown in Figure 7, with one reducer processing 8850 kmol/hr of natural gas and the other processing 6000 kmol/hr, a liquid fuel production equivalent to that shown in Table 6 can be obtained. Compared to the single reducer system with CO, co-feed which consumes 15,500 kmol/hr of natural gas, the two reducer system reduces the natural gas consumption by 650 kmol/hr when 50,000 bpd of liquid fuel is produced. Considering this reduction is in addition to the reduction of 4349 kmol/hr (or 22%) over the

Simulated syngas quality as a function of H,OCH and CO,:CH ratios at 900 °C and 1 atm for a cocurrent moving bed reactor and an effective Fe,0,:CH ratio of 0.8 (Fan et al., 2017)

Figure 23. Simulated syngas quality as a function of H,OCH4 and CO,:CH4 ratios at 900 °C and 1 atm for a cocurrent moving bed reactor and an effective Fe,0,:CH4 ratio of 0.8 (Fan et al., 2017).

Chemical looping system operated with two reducers in parallel and a smgle combustor reactor with CO, input

Figure 24. Chemical looping system operated with two reducers in parallel and a smgle combustor reactor with CO, input

(Fan etal., 2017).

Table 9. Comparison of the syngas yields for the 2-reducer and the 1-reducer system.

2-Reducer modularized system

1-Reducer system

Reducer 1

Reducer 2

combined

Input conditions

CH4 m (mol/s)

0.68

0.15

0.83

1

H.O m (mol/s)

0.54

0.03

0.57

0.23

CO,m (mol/s)

0.07

0.10

0,17

0

Output conditions

H2 ou, (m0l/S)

1.74

0.30

2.04

2.04

СОш (mol/s)

0.70

0.23

0,93

0,93

H,/CO Ratio

2.50

1.27

2.19

2.19

ATR baseline case by using the CO, co-feed scheme, the overall benefit of CO, co-feed with reducer modularization configuration is, therefore, significant.

It should be noted that reducer modularization is not only limited to two reducers, the concept can be expanded to ‘n’ reducers (Fan et al., 2017). While each reducer may perform differently, their configurations can be optimized such that the combined performance can be optimal for product yield, considering temperature, pressure, reactor reaction time, H,:CO ratio, and Fe,03:CH4 ratio. The optimization of these reducers can target to reduce overall cost and maximize the efficiency under given operating conditions (Fan et ah, 2017).

Summarizing the moving bed CLR of methane process, it can efficiently convert feedstock into a high-quality syngas using an iron-based oxygen carrier. Methane can also be used as a со-fed feedstock with solid feedstock, such as coal or biomass, to effectively adjust the H,:CO ratio of the syngas and improve the quality of the syngas product. Owing to the uniform and controllable residence time of oxygen carrier particles and avoided solids backmixing typically presented in fluidized bed reactors, the oxidation state of the oxygen carrier at the reducer outlet is controlled to those thermodynamically favorable for syngas generation. The cocurrent flow pattern in the moving bed reducer ensures that the syngas is only in contact with the desired oxidation state of the oxygen carrier, to ensure a favorable composition of the syngas. The oxygen earner particles can donate more oxygen than those used in fluidized bed reactors as the oxygen earner conversion from a moving bed reducer is higher. As a result, the moving bed CLR process requires a lower solids circulating rate than the fluidized bed CLR process operated at the same feedstock processing capacity. In addition, by limiting the reduction degree of oxygen carrier in order to avoid the formation of metallic iron, carbon deposition is effectively inhibited. With steam or CO, co-feed options, the H,:CO ratio in the syngas can be adjusted to meet the downstream processing requirement, for example, a H,:CO ratio of 2:1 for Fisclier-Tropsch synthesis or methanol production. With the source of со-fed CO, coming from the recycled CO, of downstream processes, the moving bed CLR process can be CO, neutral or negative. As the amount of CO, со-fed to the CLR reducer poses a nonlinearity effect on the quality of syngas, CLR reducer modularization can further reduce the overall cost and maximize the efficiency under given operating conditions.

 
<<   CONTENTS   >>

Related topics