Continuous Operation of Interconnected Fluidized Bed Reactor for

Article pubs.acs.org/EF

Continuous Operation of Interconnected Fluidized Bed Reactor for Chemical Looping Combustion of CH4 Using Hematite as Oxygen Carrier Jinchen Ma, Haibo Zhao,* Xin Tian, Yijie Wei, Yongliang Zhang, and Chuguang Zheng State Key Laboratory of Coal Combustion, Huazhong University of Science and Technology, Wuhan, P. R. China, 430074 ABSTRACT: Chemical looping combustion (CLC) is a novel technology for CO2 capture with low energy penalty. This technique involves the operation of the interconnected fluidized bed reactors, within which solid oxygen carrier particles circulate to provide active lattice oxygen for fuel combustion. An interconnected fluidized bed reactor was tested using natural hematite as oxygen carrier and CH4 as fuel. The thermal power was between 1.00 and 1.50 kWth in the continuous operation of 100 h. During the operation, the CLC unit showed a good performance, in which the gas leakage varied between 1.81% and 3.20%, and the CH4 conversion of 81.3% at maximum and the carbon capture efficiency of over 90% were attained. The effects of operational parameters on the performance of the new CLC reactor were investigated. The results indicated that the CH4 conversion and CO2 yield were influenced by the reactor temperature, bed inventory per unit thermal power, and the flow rate of the fluidizing gas agent in the fuel reactor. After the whole tests, no sintering and agglomeration of used particles were observed, and the thermogravimetric analyzer results of used samples showed the reactivity of oxygen carrier was well maintained. equal to the normal combustion heat released.4 The FR outlet gases consist mainly of CO2 and H2O. After condensation, concentrated CO2 can be attained for further disposal. The success of the CLC system mainly depends on suitable OCs with high reactivity, resistance to agglomeration, sintering, fragmentation, and abrasion. For an industrial plant, the OCs should also have characteristics of low cost and environmental friendliness.5 Many OCs have been investigated in thermogravimetric analyzer (TGA),6,7 batch fixed/fluidized bed reactors,8,9 and interconnected fluidized bed reactors.10−13 Some kinds of OC were synthetic oxides (like Ni-,14 Fe-,15,16 Cu-,17−19 Co-,20 Mn-based21 oxides), and some were natural ores such as hematite,22 ilmenite,23 manganese ore,24 copper ore,25 and the mixture of iron ore and copper ore.26 The experimental results in TGA or batch-operated reactors provided useful information about, e.g., reduction and oxidation kinetics, which are essential for the design and operation of a real CLC reactor.27 To further test the reactivity behavior of OCs during the continuously long-term operation as well as the feasibility of CLC technology, several groups have already built lab-scaled interconnected fluidized bed reactors, where the OCs continuously circulate in the units.12,14,28−31 CLC of gaseous fuels has been widely investigated using different OCs, such as NiO,32 Fe2O3,33 CuO materials,34 and natural minerals (ilmenite,35,36 manganese ore,24 and so on). Previous works of the CLC reactor were widely designed with an interconnected fluidized bed reactor as CLC combustor proposed by Lyngfelt et al.32 These FRs were generally operated in the bubbling regime to achieve a high residence time and the configuration of the OC recirculation was

1. INTRODUCTION It is becoming more and more important to implement carbon capture and storage (CCS) due to the increasing concentration of carbon dioxide in the atmosphere, which has been identified as the one of the biggest contributors to the ongoing climate change.1 The main CO2 emissions come from the combustion of fossil fuels.2 Chemical looping combustion (CLC) is a promising way of CO2 capture with high efficiency and low energy penalty.3 CLC is a process in which oxygen carriers (OC, usually metal oxides), rather than air, provide the active lattice oxygen for fuel combustion. Figure 1 shows the schematic map of CLC. The oxidized OCs are reduced by the fuel in the fuel reactor (FR) and transported to another reactor, namely, air reactor (AR), where the OCs are reoxidized by air. Generally, the reduction reaction is endothermic, the oxidation reaction is exothermic and the total amount of heat released in the two reactions is

Received: December 24, 2014 Revised: April 3, 2015

Figure 1. Schematic map of CLC. © XXXX American Chemical Society

A

DOI: 10.1021/ef502881x Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels Table 1. Configuration and Operational Experiences of the CLC Reactors location

size (kWth)

AR

FR

recirculation configuration

OC/fuels

Chalmers32 CSIC34 Vienna36 Xi’an34 Ohio38 Tsinghua12

10 10 120 3−10 25

bubbling fluidized bed bubbling fluidized bed fast fluidized bed pressure fluidized bed fluidized bed bubbling fluidized bed + riser

bubbling fluidized bed bubbling fluidized bed circulating fluidized bed pressure bubbling bed moving bed bubbling fluidized bed

overflow overflow lower loop seal overflow rotary solid feeder overflow

Ni-based OC/natural gas Cu-based OC/CH4 ilmenite/natural gas, et al. Fe-, Cu-based OC/coke oven gas Fe-based OC/syngas ilmentite (K+)/CO

Figure 2. Interconnected fluidized bed reactor for CLC.

overflow,12,34,37 except for the dual circulating fluidized beds (DCFB) at Vienna University of Technology,36 as shown in Table 1. However, for the initial experiments of the DCFB, the CH4 conversion was only between 30% and 70% in first run when using natural ilmenite as OC due to the limitation of the residence time at the circulating fluidized bed and the relatively low reactivity of the natural ilmenite. Recently, a moving bed was designed as FR for achieving the high syngas conversion at the Ohio State University.38 In most units, a circulated fluidized bed was usually configured for the AR to provide the driving force of solid circulation, and the nonmechanical loop seal was generally designed to prevent the gas leakage and transport the OC, and cyclone was designed for the separation of solid particles and gas. However, the OC circulation rate from the FR to the AR was generally restricted by the overflow in the FR, which was difficult to meet the criteria of solid circulation rate in the industrial plants. As reported, the solid circulation was around 1.8−2.3 t/h in the 1MWth CLC plant in Germany.39 Thus, in this work, the FR was designed as the configuration of “a bubbling fluidized bed plus a fast fluidized bed (riser) on the top” in order to attain enough residence time of fuel as well as to provide the driving force for the solid circulation from the FR to AR. The AR was also configured with a riser on the top and a turbulent fluidized bed at the bottom. In this unit, the OC particles were driven by the two risers instead of the overflow as the function of the solid recirculation, without the limitation of the solid circulation rate. The objective of the work was to investigate the effects of operational parameters on the performance (CH4 conversion, carbon capture efficiency, and CO2 yield) and the system behavior (the gas leakage between the two reactors). Moreover,

although the synthetic OCs usually showed high reactivity with the gaseous fuel (actually, most units used synthetic OCs such as NiO/NiAl2O4, NiO/MgAl2O4, CuO/Al2O3, and attained very high fuel conversion of more than 90%), using natural ore as OC in larger CLC plants is preferred from the perspective of reducing the cost in investment and operation. Another objective of this work was to test the stability of a natural OC, namely, hematite in the interconnected fluidized bed reactor.

2. EXPERIMENTAL SECTION 2.1. Interconnected Fluidized Bed Reactor for CLC. In this CLC reactor system, there were four parts: the interconnected fluidized bed reactor; the preheating system of the gases and the electric furnace; the measurement and control system; off-gas analysis and treatment system. In the interconnected fluidized bed reactor system, the FR was operated in the bubbling fluidized bed regime, and the AR was operated in the turbulent fluidized bed regime. The riser was connected to each top of the reactors, which provided the driving force for solid circulation. The cyclone was used for the gas−solid separation, and the nonmechanical loop seal was located between the two reactors to prevent the gas leakage and transport the OC particles. Figure 2 shows the sketch map of this unit. Also as shown in Figure 2, the whole interconnected fluidized bed reactor system was located within an electrically heated furnace. Each kind of gas was preheated in the electrical heaters. There were 21 pressure sensors and 21 thermal couples evenly and axially distributed in the reactor. The temperature of the thermal couple located on the FR bottom (the main reaction zone) was selected as the reaction temperature (TFR). All the gases were controlled by the mass flow meters. The outlet gases from the AR and FR were cooled by the heat exchanger, and the fine particles entrained were collected by filters. At the start-up and shut-down stages, air was used as fluidizing gas agent in the reactors. During the continuous operation, air was also used to oxidize the reduced OC in the AR, and N2 and CO2 were used B


Article

Energy & Fuels as fluidizing agents when feeding CH4 for combustion, where N2 was used as the main fluidization agent and CO2 was used as a carbon gasification agent. All gases were heated to 800 °C in preheating heaters before pumping into reactors, except CH4. N2 was also used as a fluidization agent in the two loop seals. The operational conditions during the experiments were concluded in Table 2. PCH4 was the thermal power input of the reactor (based on

test number

air reactor

fuel reactor

operating temperature

thermal power

bed inventory

unit

FAR L/min

FFR, L/min

TFR °C

PCH4, kWth

mFR, kg

30

28

900

ca. 2.5

30

20

900

ca. 2.5

seals to identify the effect of the bed inventory on the performance of the unit. Furthermore, the OC samples were collected from the FR during the continuous tests. The off-gas samples were pumped into the online gas analyzers after the filters, respectively. O2 and CO2 from the AR outlet were measured by the online gas analyzer (Cubic, Gasbroad3000). CO2 was from the combustion of carbon deposited on the OC particles and the gas leakage through loop seals. The dry-basis gas concentrations from the FR outlet were measured by the online gas analyzer (Cubic, Gasbroad-3151), with CO2, CO, and CH4 being determined using nondispersive infrared analysis (NDIR) detector, H2 by thermal gas conductivity (TCD) detector, and O2 with electron capture detector (ECD). 2.2. Oxygen Carriers. In the tests, the natural hematite was selected as OC. The chemical analysis of hematite was analyzed by the X-ray fluorescence probe (XRF, EDAX EAGLE III) and X-ray diffraction (XRD, X’PertPRO) analysis, as shown in Table 3. The

30

25

900

1.00

ca. 2.5

Table 3. Chemical Analysis of Hematite

30

25

950

1.00

ca. 2.5

Table 2. Operational Parameters in the Tests

No. 1 (60 min) No. 2 (60 min) No. 3 (30 min) No. 4 (30 min) No. 5 (30 min) No. 6 (30 min) No. 7 (30 min) No. 8 (30 min) No. 9 (30 min) No. 10 (30 min) No. 11 (30 min) No. 12 (30 min) No. 13 (100 min)

30

25

1000

1.00

ca. 2.5

30

25

1000

1.25

ca. 2.5

30

25

1000

1.50

ca. 2.5

30

20

1000

1.00

ca. 2.5

30

20

950

1.00

ca. 2.5

30

28

900

1.00

ca. 2.5

30

28

900

1.25

ca. 2.5

30

25

950

1.25

ca. 2.5

30

25

900

1.00

ca. 2.5−2.0

Composition

wt %

Fe2O3 SiO2 Al2O3 others

65.9 25.2 5.4 3.5

particles were sieved in the size range between 150 and 250 μm, and the average particle diameter was 180 μm. The true density of the OC particles was 3472 kg/m3, measured by density analyzer (Micromeritics, AccuPyc1330). The crushing strength was 2.83 N by the digital dynamometer (Shimpo, FGP-100). The minimum fluidization velocity was 0.0187 m/s at 950 °C using nitrogen as fluidizing agent, and the terminal velocity was 1.12 m/s at 950 °C using air as fluidizing agent. In consideration of the using natural hematite rather than the calcined hematite, there were two aspects: first, it was 7.5 kg OC for the total bed inventory in the 1−1.5 kWth CLC reactor and about 300 kg OC of the bed inventory in the 50 kWth (in built). It was too much work and consuming time of calcining the hematite at the muffle furnace; then, in the preheating stage (0−1000 °C) and starting up stage (1000 °C), the OC in the system was fluidized by the air agent, which was similar to the calcined stage of the hematite OC.

the low heating value of CH4), ranging between 1.00 kWth and 1.50 kWth. FAR and FFR were the flow rates in the AR and FR. The CO2 flow rate in the FR was 5.0 L/min in the entire tests except tests No. 1 and 2. The N2 flow rates in LS1 and LS2 were varied from 0.90 to 1.32 L/ min in the transport chamber and from 1.50 to 2.20 L/min in the recycle chamber, respectively. The bed inventory in the FR (mFR) was approximately 2.5 kg, and the total inventory in the reactor system was around 7.5 kg. The gas leakage should be minimized in the interconnected fluidized bed reactor. During the initial hot tests No. 1 and No. 2 (TFR was 900 °C), the fresh hematite was circulated in the interconnected fluidized bed reactor without CH4 feeding in order to check the gas leakage between the AR and FR. The fluidizing agents to FR and two loop seals were high-purity nitrogen and the fluidizing agent to AR was air. The flow rate of N2 to FR was 28 L/min during test No. 1 and 20 L/min during test No. 2. The effects of operational parameters on the CLC performance were tested. Each test was continuously operated at least 30 min. Among tests No. 3, No. 4 and No. 5, three FR temperatures were investigated. The effect of CH4 feeding rate (or thermal power input) was demonstrated in tests No. 5, No. 6 and No. 7 (all operated at 1000 °C). For the investigation of the flow rate of FR fluidizing gas agent, test No. 8 was compared to test No. 5 at 1000 °C, as well as No. 9 vs No. 4 at 950 °C and No. 10 vs No. 3 at 900 °C. In addition, tests No. 11 and No. 12 were conducted to indentify the effect of CH4 feeding rates, in comparison with tests No. 10 (at 900 °C) and No. 4 (at 950 °C), respectively. During test No. 13 with the 100 min continuous operation, the solid circulation rates were regulated through the loop

3. DATA EVALUATION The gas leakage rate was calculated by the oxygen concentration from the FR outlet on the basis of oxygen balance. In the tests No. 1 and No. 2, the N2 was used as the fluidized gas agent in the FR and two LS, and just air (79% N2 and 21% O2) used in the AR as fluidized gas agent. Thus, the O2 can be viewed as tracer gas. On the basis of the definition of the gas leakage, eq 1 was used to calculate the gas leakage. FFR,outχO ,FR 2 η= FAR,inχO ,AR (1) 2

where FAR,in was the flow rate of AR inlet; χO2,AR was the O2 concentration of AR inlet; FFR,out was the flow rate of FR outlet and χO2,FR was the O2 concentration of FR outlet. The relevant performance parameter in CLC operation was the CH4 conversion, given by χCH ,FR 4 XCH4 = 1 − χCO,FR + χCH ,FR + χCO ,FR (2) 4

2

where χi,FR was the concentration of gas i from FR outlet, i = CO2, CO, or CH4. C


Article

Energy & Fuels

Figure 4 showed the pressure drop of the AR, FR (including the bottom part and the riser) and two LS during the continuous operation of tests No. 3, No. 4 and No. 8. The operating temperatures were also shown in Figure 4. During each test, the pressure drops were almost steady with some pressure fluctuation. In tests No. 3 and No. 4, the average pressure drop of the FR was 4500 Pa when the flow rate of the FR was 25 L/min. In comparison to test No. 8, in which the flow rate of the FR was 20 L/min, the average pressure drop of the FR was increased to 5000 Pa. At the same time, the pressure drop in the AR was slightly decreased. During the whole continuous operation, the pressure drops in the LS1 and LS2 were high enough to prevent the gas leakage between the AR and FR. The pressure fluctuations in the LS1 and LS2 were explained by the changes of the pressure at the outlet of the recycle pipes, which connected the reactors, and the growth and break of the bubbles in the recycle chamber, which was operated in the bubbling bed regime. The gas concentrations from FR and AR of six tests are shown in Figure 5. The effect of reactor temperature was first investigated. Figure 5 showed the CO 2 , CO and CH 4 concentrations from the FR outlet and the CO2 and O2 concentrations from the AR outlet during the hot tests No. 3, No. 4 and No. 5, which were operated at 900, 950 and 1000 °C, respectively. The main reactions associated with the fuel combustion in the fuel reactor, including OC reduction (eqs 5, 6 and 7), CH4 decomposition (eq 8), gasification of the deposit carbon (eq 9), et al., and the reactions occurred in the air reactor (eqs 10 and 11) were as follows.

The carbon capture efficiency was calculated as the ratio of carbon containing gas flow leaving the FR to the total carbon containing gas flow leaving the system,40 ηCC = FFR,out(χCO ,FR + χCO,FR + χCH ,FR ) 2

4

FAR,outχCO ,FR + FFR,out(χCO ,FR + χCO,FR + χCH ,FR ) 2

2

4

(3)

Here, FAR,out was the flow rate of AR outlet; χCO2,AR was the CO2 concentration of AR outlet. The fraction of CO2 of total gaseous carbon leaving the FR41 was defined by χCO ,FR 2 γCO = 2 χCO,FR + χCH ,FR + χCO ,FR (4) 4

2

4. RESULTS AND DISCUSSION 4.1. Gas Leakage. The gas leakage was tested at varying flow rates of FR fluidizing agent, which was 28 L/min during the first 60 min (test No. 1) and then decreased to 20 L/min during the second stage (60−120 min, test No. 2). As shown in Figure 3, the average gas leakage rate in the first stage was

CH4 + 12Fe2O3 → CO2 + 2H 2O + 8Fe3O4

(5)

CO + 3Fe2O3 → CO2 + 2Fe3O4

(6)

H 2 + 3Fe2O3 → H 2O + 2Fe3O4

(7)

CH4 → C + 2H 2

(8)

C + CO2 → 2CO

(9)

O2 + 4Fe3O4 → 6Fe2O3

(10)

C + O2 → CO2

(11)

It was obvious that the CO2 concentration exiting from the FR in test No. 3 was lower, while the CO and CH 4 concentrations were considerably higher than in other two tests No. 4 and No. 5. These were ascribed to the reactivity of the OC being relativity lower at 900 °C. However, the CO2 concentrations from the FR were almost the same in tests No. 4 (950 °C) and No. 5 (1000 °C). The explanation for this was that more CO2 was reacted with carbon produced from the CH4 decomposition (CH4 → C + 2H2), which was also demonstrated by a slightly higher CO concentration from the FR in test No. 5 (CO2 + C → 2CO). Because of the limited residence time, there was still unconverted CO exiting from the FR. The CH4 concentration decreased with the increase of FR temperature. The CO2 was found in the AR outlet, which indicated that there was gas leakage from the FR to the AR or the carbon deposited was transported to the AR through OC entrainment. The CO2 concentration from the AR decreased with the increase of the reactor temperature. It was because of that that the carbon deposited on the surface of OC particles in the FR was more easily gasified at higher temperatures, reducing the amount of carbon entrained to the AR.

Figure 3. Gas leakage between the AR and the FR in two tests.

1.81% and then rapidly increased to 3.20% with the decrease of FFR to 20 L/min. Factually, as the flow rate of FR fluidizing agent decreased, the standard deviation of pressure drop in the FR was much larger due to the growth and break of more bubbles, which resulted in a larger pressure fluctuation in the FR. The pressure fluctuation lead to quite unsteady pressure differences between the FR and loop seal and then fluctuant bed height in the downcomer of the FR cyclone, which more easily caused gas leakage between two reactors. 4.2. Effects of Operational Parameters. The continuous operation of the unit was more than 300 h, including the coldmodel running, the heating running, the fuel-feeding running (nearly 100 h) and the cooling-down running. The natural hematite showed a stable and acceptable reactivity during the long-term operation, without obvious sintering or agglomeration. D


Article

Energy & Fuels

Figure 4. Pressure drops during the continuous tests.

Figure 5. Gas concentrations of the FR and AR during the continuous tests.

Tests No. 5, 6 and 7 were organized to investigate the effect of the thermal power input (or the CH4 feeding rates) on the performance of CLC reactor, as shown in Figure 5. Different CH4 feeding rates from 1.67 L/min (1.00 kWth) to 2.50 L/min (1.50 kWth) were tested at 1000 °C. As shown, the FR CO2 concentration decreased as the increase of thermal power. There was an obvious increase of CH4 concentration from the FR outlet with an increase of thermal power, which corresponds to the decrease of CH4 conversion. The CO concentration of the FR outlet was around 10%. On the other

hand, the CO2 concentration from the AR out increased with the thermal power, which indicated that the carbon transported to the AR increased. It seemed that the CH4 decomposition increased at a higher CH4 concentration in the FR. Factually, the CH4 decomposition was caused by the limitation of the bed inventory and the chemical reactivity of OC in the FR. The same trends were found at different operational temperatures (tests No. 10 and No. 11 at 900 °C, tests No. 4 and No. 12 at 950 °C). E


Article

Energy & Fuels A series of tests were designed to investigate the effects of the flow rates of FR fluidizing agent, which were varied from 20 L/ min to 28 L/min (as shown in Table 1). The CO 2 concentration from the FR outlet was 80% at a lower flow rate in test No. 8 (20 L/min N2), over 10% than that in test No. 5 of 25 L/min N2. Accordingly, the CH4 concentration from the FR outlet was also decreased to around 10% in test No. 8, in comparison with test No. 5. These were ascribed to the CH4 residence time increasing when the FR flow rate decreased. For the CO2 concentration of the AR, it decreased from 0.25% (in average) in test No. 5 to 0.15% in test No. 8. It was due to the longer carbon gasification time at a smaller FR flow rate to make less carbon entrained to the AR. The test No. 13 was continuously run for 100 min to investigate the effect of the FR bed inventory. The FR bed inventory was regulated by controlling the flow rates of the fluidizing agents in the loop seal. As shown in Figure 2, FLStran was the flow rate of the fluidizing agent in the transport chamber of loop seal (which connected the down comer of cyclone), and FLSrev was the flow rate of the fluidizing agent in the recycle chamber of loop seal (which connected the other reactor). On the basis of our previous cold-model experiments, the solid circulation rate increased with the flow rates of the fluidizing agent in the loop seal.42 Thus, an increase of the flow rates of the fluidizing agent in the loop seal (LS2) would lead to the solid circulation rate from the FR to the AR being higher than that from the AR to the FR, and then a gradual increase in the FR bed inventory. In test No. 13, the FLStran was 1.1 L/min in LS1 and 1.26 L/min in LS2; the FLSrec 1.5 L/min in LS1 and 1.7 L/min in LS2. The pressure drop of the FR, which was relevant to the bed inventory, was monitored during the hot tests. As shown in Figure 6, the average pressure drop was decreased from ca.

Figure 7. Gas concentrations of the FR and AR during the continuous test.

increased in the FR. At the same time, more carbon deposited on the OC particles was transported to the AR, as result of the increase of CO2 concentration from the AR outlet. Because of the same reason, the unconverted CO from the FR outlet increased gradually. 4.3. Performance of Interconnected Fluidized Bed Reactor. The CLC performance of the interconnected fluidized bed reactor was shown as functions of operational parameters (the reactor temperature, the thermal power input, the flow rates of the fluidizing agent, and the bed inventory). The CH4 conversion, the carbon capture efficiency and CO2 yield are discussed below, respectively. 4.3.1. CH4 Conversion. The CH4 conversion was affected by three factors: the reactor temperature, which is highly relevant to the OC reactivity; the flow rate of FR fluidizing agent (or the FR superficial gas velocity), which determines the CH4 residence time in the FR; the FR bed inventory, which determines the amount of lattice oxygen to oxidize the reducing gases (CH4, CO (the product of carbon gasification) and H2 (the product of CH4 decomposition)). A series of tests were well organized to investigate the effects of the three factors. The average CH4 conversion for 10 tests (each with 30 min continuous operation) is shown in Figure 8. The tests were

Figure 6. Pressure drop in the FR as a function of time.

4500 Pa in the initial stage to ca. 3500 Pa. It was estimated that the bed inventory in the FR was decreased by 0.5 kg according to the reduced FR bed height when the test was stopped. Figure 7 showed the gas concentrations from the FR and AR outlets during the continuous test No. 13. The CO 2 concentration from the FR continuously decreased from the peak value of 70% to almost 50%. It was attributed to insufficient bed inventory that cannot provide enough lattice oxygen for CH4 combustion. Therefore, the unconverted CH4

Figure 8. CH4 conversion at different operational parameters. F


Article

Energy & Fuels operated at three typical reactor temperatures, 900, 950, and 1000 °C. At the FR flow rate of 25 L/min and the thermal power of 1.00 kWth, the CH4 conversion increased from 71.38% to 78.58% when the reactor temperature increased from 900 to 1000 °C. The same trend also held at the thermal power of 1.25 kWth. Furthermore, the flow rates of FR fluidizing agent had an impact on the CH4 conversion. This operational parameter was varied from 20 L/min, 25 L/min, to 28 L/min at the thermal power of 1.00 kWth. The highest CH4 conversion of over 80% was achieved at the lowest FR flow rate (20 L/min) at 950 and 1000 °C. When the FR flow rate was higher, the CH4 conversion was decreased to below 80% and even lower than 40% at 900 °C. It was explained by the insufficient CH4 residence time. In addition, there was more unconverted CH4 at the FR outlet in test No. 5, compared with test No. 8, as shown in Figure 4. However, the almost same CH4 conversions were obtained at tests No. 8 and No. 9, only with different reactor temperatures (950 and 1000 °C). The experimental result indicated that the temperature variation (in the range of 950−1000 °C) was not crucial when the time for CH4−oxygen carrier contact was sufficient at the flow rate of 20 L/min in the FR. According to test No. 13 (continuous 100 min at 1.00 kWth and 900 °C), there was an obvious decrease of the CH4 conversion with a decrease of the pressure drop in the FR (see Figure 6), which is relevant to the bed inventory. The CH4 conversion was reduced by 14% when the bed inventory was reduced by 0.5 kg as mentioned above. Similar trend was observed when the thermal power increased from 1.00 kWth to 1.50 kWth at 1000 °C. No matter when the bed inventory decreased or the thermal power increased, both of them resulted in the lower bed inventory per unit of thermal power. 4.3.2. Carbon Capture Efficiency. The carbon capture efficiency, which is highly relevant to the CO2 concentration from the AR outlet, was influenced by a few factors, such as methane conversion and decomposition, which resulted in the amount of carbon entrained to the AR, and the fluidizing velocity, which had the impact on the gas leakage as discussed in section 4.1. Considering the methane conversion and decomposition, it was better that the FR was operated at the higher temperatures, because the lower amount of carbon generated in the FR resulted in less carbon entrained to the AR. For another factor, the gas leakage should be minimized. In Figure 9, when the flow rate of the FR fluidizing agent was 20 L/min at 950 and 1000 °C and with the thermal power of 1 kWth, 96% of carbon capture efficiencies were attained, indicating that more carbon was gasified by CO2 (C + CO2 → 2CO) due to more carbon residence time compared with tests under 25 L/min. However, the highest carbon capture efficiencies were achieved in the tests with the highest flow rate of FR fluidizing agent (28 L/min) and the lowest reactor temperature (900 °C). The explanations for this were (1) the smallest gas leakage at 28 L/min, as discussed in section 4.1; (2) less methane decomposition at 900 °C. The bed inventory per unit thermal power was defined as the ratio of the bed inventory and the thermal power. The results showed that the increase of the thermal power (from the 1.00 kWth to 1.5 kWth) and the decrease of the bed inventory in the FR (test No. 13) will cause the decrease of the bed inventory per unit thermal power, resulting in the decrease of the carbon capture efficiency, as shown in Figure 9.

Figure 9. Carbon capture efficiency at different operational parameters.

4.3.3. CO2 Yield. The influencing factors to the CO2 yield were almost the same to these of CH4 conversion. However, some differences were observed. When the thermal power was 1.00 kWth and the flow rates of FR fluidizing agent were 25 L/min, the CO2 yield at 950 °C was the highest (70.85%). This was different with the varying trend of the CH4 conversion. It can be explained by the competition between the CH 4 reduction and the CH 4 decomposition. As known, the higher the reactor temperature, the higher the reactivity between CH4 and OC (resulting in more CO 2 and less CH 4 ), however the faster CH 4 decomposition (resulting in less CH4 and more carbon formation) and the higher reactivity between carbon and CO2 (resulting in less CO2 and more CO). Factually, as a result, there was less CO from the FR in test No. 4 at 950 °C, compared to other two cases. Similar results were found in tests No. 8 (1000 °C) and No. 9 (950 °C) both with flow rate of 20 L/min. When the flow rate in the FR was 28 L/min, the CO2 yields were both below 35% in tests No. 11 and 12. The results showed that a higher gas superficial velocity in the FR resulted in a lower CH4 residence time. Thus, the CO2 yield significantly increased with the decrease of the FR flow rates. During the continuous operation in test No. 13, the CO2 yield was reduced from 70% to 51% as the decrease of the bed inventory. It was also because more bed inventory could provide more lattice oxygen to CH4 reduction in FR. 4.4. OC Characterization. Characterization analysis of the OC particles was performed to investigate the structure and composition changes between the fresh and used OC particles. The oxidized sample and reduced sample were collected from the AR and FR at the end of the experiments, respectively. When all tests were finished, the used particles were collected to analyze the particle size distributions of the fresh and used samples by laser diffraction particle size analyzer (Master Min, Malvern), as shown in Figure 11. It was obvious that the percentage of the small used particles (below the 200 μm) increased compared with the fresh sample, while that between 200 and 250 μm was decreased. An explanation is that the OC collided and attrited in the reactor, especially in the cyclone due to the high gas velocity. However, the natural hematite showed high strength during circulation in the real CLC reactor. G


Article

Energy & Fuels

Figure 10. CO2 yield at different operational parameters.

Figure 12. SEM images: (a) fresh sample 1500×, (b) fresh sample 3000×, (c) used sample 1000×, (d) used sample 2000×.

collected from FR during the continuous tests were determined by X-ray diffraction (XRD, X’Pert Pro) using Cu Kα radiation (λ = 1.542 nm) in a 2θ range of 10−90° with an accelerating voltage of 40 kV and a tube current of 40 mA. It can be seen that the fresh ore mainly contains Fe2O3, although a small amount of Fe3O4 still exists, which may be explained by incomplete oxidation in the OC preparation step. For the used iron ore sampled from AR, the XRD results indicated that the reduced ore from FR can mostly be oxidized into Fe2O3 in AR by air in the continuous operation. Meanwhile, as shown above, the OC was mostly reduced from Fe2O3 to Fe3O4 by CH4 in FR. Figure 14 showed the redox reactivity of both the fresh and used iron ore OC sample in a thermogravimetric analyzer (TGA, BOIF WCT-1D/2D). ω is the mass-based degree of conversion,43 eqs 12 and 13; dXr/dt and dX0/dt are the conversion rates in the reduction stage and the oxidization

Figure 11. Particle size distributions of the fresh and used samples.

The microstructure of fresh and used samples after the continuous CLC tests were investigated by scanning electron microscopy (SEM, FEI Quanta 200) with 20.0 kV accelerating voltage. The fresh OC sample contained amounts of fines although it had been screened thoroughly before the test. There were many small particles to be adhered on the surface of the fresh OC, and a compact structure existed. Under the influence of thermal effect and mechanical abrasion, the surface of the OC became porous and the irregular protrusions structure became smooth to some extent. In the SEM image of the used OC sample, small particles were not found, which indicated a high separation efficiency of the cyclone. It was worth noting that no sintering or agglomeration was observed for these used OCs. The Brunauer−Emmett−Teller (BET) surface area analysis (Micromeritics, ASAP-2020) was consistent with the SEM characterization. The specific surface areas were 3.390 m2/g for the fresh particles but 1.768 m2/g for the used one. Actually, the fresh sample was surrounded by many small particles, which led to a higher BET surface area. Note that, even enough the surface area and pore diameter were decreased, the used OC showed an acceptable reactivity with CH4. The crystalline phase compositions of the fresh iron ore samples, used samples collected from AR and used samples

Figure 13. XRD results: (a) fresh sample, (b) used sample from AR, (c) used sample from FR. H


Article

Energy & Fuels

5. CONCLUSION An interconnected fluidized bed reactor for chemical looping combustion has been operated with the hematite as oxygen carrier and CH4 as fuel. A series of tests fed with CH4 have been performed at different operational parameters, such as the FR temperature, the thermal power input, the flow rate of the gas agent in the FR, and the bed inventory, to investigate the influence of different operational parameters on the performance of the interconnected fluidized bed reactor for CLC. The almost same batch of hematite had been used for the whole operation without sintering and agglomeration. The operation continued for 30 days, with typically 12 h per day. The following conclusions were drawn: • The highest CH4 conversion of 81.4% was achieved at 1000 °C with sufficient residence time and bed inventory. • The carbon capture efficiencies were dependent on the CH4 decomposition, the bed inventory in the FR, and the gas leakage in the unit. • The flow rate of the fluidizing gas agent and the bed inventory were the important factors in the CO2 yield. However, the operational temperature should be optimized to control the CH4 conversion and CO2 yield when the hematite was used as OC and CH4 as gas fuel, because a higher temperature is beneficial to CH4 conversion as well as CH4 decomposition. This work showed the successfully continuous operation and the different operational parameters during the tests, which made contributions to the scale-up of the CLC reactor. In further research, the steam as a gasification agent could be favorable to achieve high carbon capture efficiency, which can limit the methane decomposition, and a carbon stripper should be adopted to reduce the carbon entrained by the reduced OC from the FR to the AR.

Figure 14. TGA results: (a) reduction stage in the fifth cycle, (b) oxidation stage in the fifth cycle.

stage, respectively, eq 14. The OC was alternately exposed in the mixture gases of 25% H2 + 35% CO + 40% CO2 for reduction and the air for oxidation, both at the temperature of 900 °C. It can be clearly seen that the fresh and used hematite showed acceptable reactivity during both the oxidation and reduction stages. The conversion rate of the used sample was slightly higher than that of fresh sample in both the reduction stage and the oxidization stage. However, the oxygen transport capacity of the used sample was decreased in both the reduction stage and the oxidization stage due to a decrease of the surface area, as shown in Figure 15. Overall, the reactivity of

■

AUTHOR INFORMATION

Corresponding Author

*Tel: +86-027-87545526-8208. Fax: +86-027-8754-5526. Email: [email protected], [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by “National Natural Science Foundation of China (51390494)”, and “National Key Basic Research and Development Program (2011CB707300)”. The authors acknowledge the support from the Australian Academy of Technological Sciences and Engineering (ATSE) for funding through the Australia−China Joint Coordination Group on Clean Coal Technology (JCG) for this work.

Figure 15. Oxygen transport capacity of the fresh and used sample in TGA: (a) reduction stage in the five cycles, (b) oxidation stage in the five cycles.

hematite was well maintained after continuous operation of over 100 h. It could be a suitable choice as OC after improving the reactivity by decorating copper44 or nickel.45 X r = (m − mr )/m initial × 100%

(12)

Xo = (mo − m)/m initial × 100%

(13)

dX /dt = dm /(m initial dt ) × 100%

(14)

■

REFERENCES

(1) Goldthau, A.; Witte, J. Back to the future or forward to the past? Strengthening markets and rules for effective global energy governance. Int. Affairs 2009, 85 (2), 373−390. (2) Parry, M. L. Climate Change 2007: impacts, adaptation and vulnerability: contribution of working group II to the fourth assessment report of the intergovernmental panel on climate change. Cambridge University Press: Cambridge, U.K., 2007; Vol. 4. (3) Xu, M.; Ellis, N.; Jim Lim, C.; Ryu, H. J. Mapping of the operating conditions for an interconnected fluidized bed reactor for

where m0 and mr represent the weight of OC at the oxidized state and reduced state; m is the weight of the OC at time t; minitial is the weight of the OC at the beginning. I


Article

Energy & Fuels CO2 separation by chemical looping combustion. Chem. Eng. Technol. 2009, 32 (3), 404−409. (4) Peltola, P.; Ritvanen, J.; Tynjälä, T.; Pröll, T.; Hyppänen, T. Onedimensional modelling of chemical looping combustion in dual fluidized bed reactor system. Int. J. Greenhouse Gas Control 2013, 16, 72−82. (5) Chandel, M.; Hoteit, A.; Delebarre, A. Experimental investigation of some metal oxides for chemical looping combustion in a fluidized bed reactor. Fuel 2009, 88 (5), 898−908. (6) Nasr, S.; Plucknett, K. P. Kinetics of iron ore reduction by methane for chemical looping combustion. Energy Fuels 2014, 28 (2), 1387−1395. (7) Wang, B.; Gao, C.; Wang, W.; Kong, F.; Zheng, C. TGA-FTIR investigation of chemical looping combustion by coal with CoFe2O4 combined oxygen carrier. J. Anal. Appl. Pyrol. 2014, 105 (0), 369−378. (8) Mei, D.; Abad, A.; Zhao, H.; Adánez, J.; Zheng, C. On a highly reactive Fe2O3/Al2O3 oxygen carrier for in situ gasification chemical looping combustion. Energy Fuels 2014, 28 (11), 7043−7052. (9) Guo, L.; Zhao, H.; Ma, J.; Mei, D.; Zheng, C. Comparison of large-scale production methods of Fe2O3/Al2O3 oxygen carriers for chemical-looping combustion. Chem. Eng. Technol. 2014, 37 (7), 1211−1219. (10) Linderholm, C.; Mattisson, T.; Lyngfelt, A. Long-term integrity testing of spray-dried particles in a 10-kW chemical-looping combustor using natural gas as fuel. Fuel 2009, 88 (11), 2083−2096. (11) Kolbitsch, P.; Bolhar-Nordenkampf, J.; Pröll, T.; Hofbauer, H. Comparison of two Ni-based oxygen carriers for chemical looping combustion of natural gas in 140 kW continuous looping operation. Ind. Eng. Chem. Res. 2009, 48 (11), 5542−5547. (12) Bao, J.; Li, Z.; Sun, H.; Cai, N. Continuous test of ilmenitebased oxygen carriers for chemical looping combustion in a dual fluidized bed reactor system. Ind. Eng. Chem. Res. 2013, 52 (42), 14817−14827. (13) Gu, H.; Shen, L.; Zhong, Z.; Niu, X.; Ge, H.; Zhou, Y.; Xiao, S. Potassium-modified iron ore as oxygen carrier for coal chemical looping combustion: continuous test in 1 kW reactor. Ind. Eng. Chem. Res. 2014, 53 (33), 13006−13015. (14) Adánez, J.; Dueso, C.; de Diego, L. F.; García-Labiano, F.; Gayán, P.; Abad, A. Methane combustion in a 500 Wth chemicallooping combustion system using an impregnated Ni-based oxygen carrier. Energy Fuels 2008, 23 (1), 130−142. (15) Bao, J.; Liu, W.; Cleeton, J.; Scott, S.; Dennis, J.; Li, Z.; Cai, N. Interaction between Fe-based oxygen carriers and n-heptane during chemical looping combustion. Proc. Combust. Inst. 2013, 34 (2), 2839−2846. (16) Ge, H.; Shen, L.; Gu, H.; Jiang, S. Effect of co-precipitation and impregnation on K-decorated Fe2O3/Al2O3 oxygen carrier in chemical looping combustion of bituminous coal. Chem. Eng. J. 2014, 262, 1065−1076. (17) Chang, F.; Liao, P.; Tsai, C.; Hsiao, M.; Paul Wang, H. Chemical-looping combustion of syngas with nano CuO−NiO on chabazite. Appl. Energy 2014, 113, 1731−1736. (18) Wang, B.; Yan, R.; Zhao, H.; Zheng, Y.; Liu, Z.; Zheng, C. Investigation of chemical looping combustion of coal with CuFe2O4 oxygen carrier. Energy Fuels 2011, 25 (7), 3344−3354. (19) Mei, D.; Zhao, H.; Ma, Z.; Zheng, C. Using the sol−gel-derived CuO/CuAl2O4 oxygen carrier in chemical looping with oxygen uncoupling for three typical coals. Energy Fuels 2013, 27 (5), 2723− 2731. (20) Mei, D.; Zhao, H.; Yang, W.; Fang, Y.; Zheng, C. Oxygen release kinetics and mechanism study on Cu-, Co-, Mn-based oxygen carrier. J. Fuel Chem. Technol. 2013, 41 (2), 235−242. (21) Azimi, G.; Leion, H.; Mattisson, T.; Lyngfelt, A. Chemicallooping with oxygen uncoupling using combined Mn-Fe oxides, testing in batch fluidized bed. Energy Proc. 2011, 4, 370−377. (22) Linderholm, C.; Schmitz, M.; Knutsson, P.; Källén, M.; Lyngfelt, A. Use of low-volatile solid fuels in a 100 kW chemical-looping combustor. Energy Fuels 2014, 28 (9), 5942−5952.

(23) Moldenhauer, P.; Rydén, M.; Mattisson, T.; Younes, M.; Lyngfelt, A. The use of ilmenite as oxygen carrier with kerosene in a 300W CLC laboratory reactor with continuous circulation. Appl. Energy 2014, 113, 1846−1854. (24) Linderholm, C.; Lyngfelt, A.; Cuadrat, A.; Jerndal, E. Chemicallooping combustion of solid fuels - operation in a 10kW unit with two fuels, above-bed and in-bed fuel feed and two oxygen carriers, manganese ore and ilmenite. Fuel 2012, 102, 808−822. (25) Zhao, H.; Wang, K.; Fang, Y.; Ma, J.; Mei, D.; Zheng, C. Characterization of natural copper ore as oxygen carrier in chemicallooping with oxygen uncoupling of anthracite. Int. J. Greenhouse Gas Control 2014, 22, 154−164. (26) Yang, W.; Zhao, H.; Wang, K.; Zheng, C. Synergistic effects of a mixture of iron ore and copper ore as oxygen carriers in chemicallooping combustion. Proc. Combust. Inst. 2014, 35 (3), 2811−2818. (27) Zhao, H.; Liu, L.; Wang, B.; Xu, D.; Jiang, L.; Zheng, C. Sol-gelderived NiO/NiAl2O4 oxygen carriers for chemical-looping combustion by coal char. Energy Fuels 2008, 22 (2), 898−905. (28) Lyngfelt, A. Oxygen carriers for chemical looping combustion 4000 h of operational experience. Oil Gas Sci. Technol. 2011, 66 (2), 161−172. (29) Linderholm, C.; Lyngfelt, A.; Dueso, C. Chemical-looping combustion of solid fuels in a 10kW reactor system using natural minerals as oxygen carrier. Energy Proc. 2013, 37, 598−607. (30) Abad, A.; Adánez-Rubio, I.; Gayán, P.; García-Labiano, F.; de Diego, L. F.; Adánez, J. Demonstration of chemical-looping with oxygen uncoupling (CLOU) process in a 1.5 kWth continuously operating unit using a Cu-based oxygen-carrier. Int. J. Greenhouse Gas Control 2012, 6, 189−200. (31) Ma, J.; Zhao, H.; Tian, X.; Wei, Y.; Rajendran, S.; Zhang, Y.; Bhattacharya, S.; Zheng, C. Chemical looping combustion of coal in a 5 kwth interconnected fluidized bed reactor using hematite as oxygen carrier. Appl. Energy 2015, DOI: 10.1016/j.apenergy.2015.03.124. (32) Linderholm, C.; Abad, A.; Mattisson, T.; Lyngfelt, A. 160h of chemical-looping combustion in a 10kW reactor system with a NiObased oxygen carrier. Int. J. Greenhouse Gas Control 2008, 2 (4), 520− 530. (33) Johansson, M.; Mattisson, T.; Lyngfelt, A. Investigation of Fe2O3 with MgAl2O4 for chemical-looping combustion. Ind. Eng. Chem. Res. 2004, 43 (22), 6978−6987. (34) Adánez, J.; Gayán, P.; Celaya, J.; de Diego, L. F.; GarcíaLabiano, F.; Abad, A. Chemical looping combustion in a 10 kWth prototype using a CuO/Al2O3 oxygen carrier: Effect of operating conditions on methane combustion. Ind. Eng. Chem. Res. 2006, 45 (17), 6075−6080. (35) Kolbitsch, P.; Bolhàr-Nordenkampf, J.; Pröll, T.; Hofbauer, H. Operating experience with chemical looping combustion in a 120kW dual circulating fluidized bed (DCFB) unit. Int. J. Greenhouse Gas Control 2010, 4 (2), 180−185. (36) Pröll, T.; Mayer, K.; Bolhàr-Nordenkampf, J.; Kolbitsch, P.; Mattisson, T.; Lyngfelt, A.; Hofbauer, H. Natural minerals as oxygen carriers for chemical looping combustion in a dual circulating fluidized bed system. Energy Proc. 2009, 1 (1), 27−34. (37) Wang, S.; Wang, G.; Jiang, F.; Luo, M.; Li, H. Chemical looping combustion of coke oven gas by using Fe2O3/CuO with MgAl2O4 as oxygen carrier. Energy Environ. Sci. 2010, 3 (9), 1353−1360. (38) Sridhar, D.; Tong, A.; Kim, H.; Zeng, L.; Li, F.; Fan, L.-S. Syngas chemical looping process: design and construction of a 25 kWth subpilot unit. Energy Fuels 2012, 26 (4), 2292−2302. (39) Ströhle, J.; Orth, M.; Epple, B. Design and operation of a 1MWth chemical looping plant. Appl. Energy 2014, 113, 1490−1495. (40) Berguerand, N.; Lyngfelt, A. Chemical-looping combustion of petroleum coke using ilmenite in a 10 kWth unit− high-temperature operation. Energy Fuels 2009, 23 (10), 5257−5268. (41) Berguerand, N.; Lyngfelt, A. Design and operation of a 10 kWth chemical-looping combustor for solid fuels-testing with South African coal. Fuel 2008, 87 (12), 2713−2726. J


Article

Energy & Fuels (42) Ma, J.; Zhao, H.; Mei, D.; Guo, L.; Zheng, C. Cold-model experiment of the interconnected fluidized bed for chemical looping combustion. J. Chin. Soc. Power Eng. 2013, 32 (12), 909−915. (43) Johansson, M.; Mattisson, T.; Lyngfelt, A. Creating a synergy effect by using mixed oxides of iron-and nickel oxides in the combustion of methane in a chemical-looping combustion reactor. Energy Fuels 2006, 20 (6), 2399−2407. (44) Yang, W.; Zhao, H.; Ma, J.; Mei, D.; Zheng, C. Copperdecorated hematite as an oxygen carrier for in situ gasification chemical looping combustion of coal. Energy Fuels 2014, 28 (6), 3970−3981. (45) Chen, D.; Shen, L.; Jun, X.; Tao, S.; Gu, H.; Zhang, S. Experimental investigation of hematite oxygen carrier decorated with NiO for chemical-looping combustion of coal. J. Fuel Chem. Technol. 2012, 40 (3), 267−272.

K


Continuous Operation of Interconnected Fluidized Bed Reactor for

Recommend Documents