Scholarly article on topic 'Chemical-looping combustion of solid fuels in a 10 kWth pilot–batch tests with five fuels'

Chemical-looping combustion of solid fuels in a 10 kWth pilot–batch tests with five fuels Academic research paper on "Chemical engineering"

CC BY-NC-ND
0
0
Share paper
Academic journal
Energy Procedia
OECD Field of science
Keywords
{"Chemical-looping combustion" / "CO2 capture" / "Fluidized bed" / "Oxygen carrier" / "Solid fuels" / "Reaction rate constant" / "Oxygen demand"}

Abstract of research paper on Chemical engineering, author of scientific article — Carl Linderholm, Ana Cuadrat, Anders Lyngfelt

Abstract Chemical-looping combustion (CLC) is a combustion concept with inherent separation of CO2. The process uses a solid oxygen carrier, which consists of metal oxide, to transfer the oxygen from air to fuel. This paper presents findings from batch tests performed in a 10 kWth CLC pilot for solid fuels. The pilot, which is the world’s first chemical-looping combustor for solid fuels, is based on interconnected fluidized-bed technology and features a fuel reactor (FR) and an air reactor (AR) as the principal reaction chambers. In the FR, fuel is gasified with steam whereupon gasification products react with the oxygen carrier to form, ideally, CO2, H2O and SO2. Oxygen-carrier particles exit the FR through a weir and are led to the AR, where they are regenerated to their oxidized state. The pilot has been operated using a natural iron-titanium ore called ilmenite as oxygen carrier. Previous continuous tests have demonstrated a need for batch tests in order to obtain complementary information on system performance. In this study, five fuels were fed to the fuel reactor in batches of 20–25 g at four temperatures; 940 °C, 970 °C, 1000 °C and 1030 °C. By using devolatilized fuel, it was possible to determine (a) oxygen demand associated with syngas from char gasification as well as (b) kinetics of char conversion to gas. Rates of char conversion were found to be temperature dependent, as expected, whereas no temperature dependence was found for the oxygen demand. Activation energies for conversion of char to gas were calculated using Arrhenius plots. The minimum oxygen demand for char was found to be around 5%.

Academic research paper on topic "Chemical-looping combustion of solid fuels in a 10 kWth pilot–batch tests with five fuels"

Available online at www.sciencedirect.com

m\' ScienceDirect

Energy Procedía 4 (2011) 385-392

Energy Procedía

www.elsevier.com/locate/procedia

GHGT-10

Chemical-looping combustion of solid fuels in a 10 kWth pilot -

batch tests with five fuels

Carl Linderholma'*, Ana Cuadratb, Anders Lyngfelta

aDepartment of Energy and Environment, Chalmers University of Technology, S-41296 Göteborg, Sweden bInstituto de Carboquimica (CSIC), Department of Energy and Environment, Miguel Luesma Castan 4, Zaragoza 50018, Spain

Abstract

Chemical-looping combustion (CLC) is a combustion concept with inherent separation of CO2. The process uses a solid oxygen carrier, which consists of metal oxide, to transfer the oxygen from air to fuel. This paper presents findings from batch tests performed in a 10 kWth CLC pilot for solid fuels. The pilot, which is the world's first chemical-looping combustor for solid fuels, is based on interconnected fluidized-bed technology and features a fuel reactor (FR) and an air reactor (AR) as the principal reaction chambers. In the FR, fuel is gasified with steam whereupon gasification products react with the oxygen carrier to form, ideally, CO2, H2O and SO2. Oxygen-carrier particles exit the FR through a weir and are led to the AR, where they are regenerated to their oxidized state. The pilot has been operated using a natural iron-titanium ore called ilmenite as oxygen carrier. Previous continuous tests have demonstrated a need for batch tests in order to obtain complementary information on system performance. In this study, five fuels were fed to the fuel reactor in batches of 20-25 g at four temperatures; 940°C, 970°C, 1000°C and 1030°C. By using devolatilized fuel, it was possible to determine (a) oxygen demand associated with syngas from char gasification as well as (b) kinetics of char conversion to gas. Rates of char conversion were found to be temperature dependent, as expected, whereas no temperature dependence was found for the oxygen demand. Activation energies for conversion of char to gas were calculated using Arrhenius plots. The minimum oxygen demand for char was found to be around 5%. ©2010 Elsevier Ltd. All rights reserved

Keywords: chemical-looping combustion; CO2 capture; fluidized bed; oxygen carrier; solid fuels; reaction rate constant; oxygen demand

1. Introduction

Chemical-looping combustion (CLC) is a combustion process where CO2 is inherently separated from other flue gases. The key component in the process is the oxygen carrier, consisting of metal oxide, which transfers oxygen from air to fuel. The CLC process is usually carried out in two interconnected fluidized beds, an air reactor and fuel reactor, shown in Fig. 1. The fuel is introduced into the fuel reactor where it reacts with the particulate oxygen carrier, e.g. ilmenite. The reduced oxygen carrier is transported to the air reactor where it is re-oxidized to its original state by air. The gas stream leaving the fuel reactor ideally consists of only CO2 and H2O.

* Corresponding author. Tel.: +46 31 772 1443 E-mail address: carl.linderholm@chalmers.se.

doi:10.1016/j.egypro.2011.01.066

In principle, all kinds of fuels can be oxidized with chemical-looping combustion. A process using gaseous fuels such as natural gas, refinery gas or synthesis gas would be easiest to realize, but there is considerable interest in CLC of solid fuels, such as coal, since it is a cheaper and much more abundant fuel.

In CLC with solid fuels, the fuel is primarily dried and devolatilized, whereupon gasification of the char can occur. Volatiles and gasification products react with the oxygen carrier to form CO2, H2O and SO2. In CLC, gasification takes place in an atmosphere with high concentrations of CO2 and H2O, which is an advantage compared to normal gasification. The main gasification reactions are

C + H2O ^CO + H2 (1)

C + CO2 CO (2)

The main reactions with the metal oxide, denoted MexOy, are

MexOy + H2 ^ MexOy-1 +H2O (3)

MexOy + CO ^ MexOy-1 + CO2 (4)

CLC of solid fuels has been demonstrated by Berguerand and Lyngfelt [1, 2, 3], and by Shen et al. [4]. Berguerand and Lyngfelt achieved approximately 90 h of continuous testing using the pilot described here with ilmenite as oxygen carrier. This pilot has also been used for batch tests [5]. The use of ilmenite as oxygen carrier in CLC of solid fuels has been investigated by Leion et al. [6]. Information about other potential oxygen carriers can be found in the work of Jerndal et al. [7], and reaction kinetics of various oxygen carriers has been examined by Abad et al. [8]. Several chemical-looping combustors for gaseous fuel ranging from 0.3-120 kWth have been designed and operated using a number of different oxygen carriers, e.g. [9, 10, 11], adding up to 4000 h of operation [12]. A 10 kW pilot for gaseous fuels has been operated for > 1000 h using a Ni-based oxygen carrier [13]. A number of literature studies have investigated the possibility to combine power cycles with the chemical-looping system. These process simulations have been performed using different types of oxygen carriers and natural gas or synthesis gas as fuel. Reviews of the literature in this area can be found in Wolf [14]. The process studies demonstrate that it is theoretically possible to achieve high thermal efficiencies using CLC integrated with CO2 capture.

2. Experimental

The reactor system consists of a fuel reactor (FR), where the fuel is gasified by steam and gasification products are oxidized by the oxygen carrier, an air reactor (AR), where the particles are oxidized, and a riser, which brings particles back to the fuel reactor via the cyclone and upper loop seal. The gas velocity in the AR and riser provides the driving force for particle circulation. However, the batch tests in this study were carried out without global circulation of solids. The upper and lower loop seals are present in the reactor system in order to eliminate gas leakages between the reactors.

The FR is divided into two main sections; a low-velocity section which is operated as a bubbling bed, and a carbon stripper which is operated between the terminal velocities of the oxygen carrier and the char, hence enabling separation of the two types of particles. The FR is also equipped with a recirculation loop, the function of which is to return elutriated char particles to the low-velocity section, but this loop was not really in operation during the present experiments. All reactions occur in the low-velocity section, which is fluidized with steam. The solids inventory in the low-velocity section is around 6 kg. A detailed description of the 10 kW pilot is given in [1]. Fig. 2 shows the whole reactor system along with a fuel-reactor close-up.

The oxygen carrier used in the experiments was ilmenite, an iron-titanium oxide that has previously been used in the 10 kW pilot for approximately 90 h of operation [1, 2, 3, 5]. Ilmenite is a commonly occurring mineral which is

N2, (O2)

Mex°y ,

Air Fuel

Reactor Reactor

(AR) (FR)

, Mex°y-1

Figure 1. Schematic representation of CLC. MexOy and MexOy-1 denote oxidized and reduced forms of the oxygen carrier.

To Chimney

To Chimney

Figure 2. The 10 kWth pilot with peripherals (left) and fuel-reactor close-up (right). (a) air reactor (AR), (b) riser, (c) cyclone, and (d) fuel reactor (FR). LOVEL and CS denote low-velocity section and carbon stripper, both of which are parts of the fuel reactor.

produced in large quantities worldwide. Its chemical formula is FeTiO3, which is the reduced form. Advantages as oxygen carrier are low cost, no thermodynamical limitations, high melting point (1367°C), good mechanical strength and stability in fluidized beds, sufficient rates of oxidation and reduction, and no health or environmental issues. The ilmenite used in the tests described here has an average diameter of 170 p,m and the bulk density is approximately 2200 kg/m3. The minimum fluidization velocity in steam at 950°C is 0.016 m/s.

Five fuels were used in the experiments: three bituminous coals, a pet coke and a metallurgical coke (abbreviated met coke). The latter is used as reducing agent in blast furnaces and was produced from US and Australian bituminous coals by a Swedish steel company. Pet coke is a by-product from oil cracking. Four of the fuels were partially devolatilized in a coke oven before the tests, see table 1. The devolatilization process in the coke oven is not complete, as can be seen in the proximate analysis in table 1. Note especially the high amounts of volatiles for Colombian. The reason why volatiles are unwanted is explained below.

Table 1. Fuel specification with proximate analysis. M: moisture; A: ash; V: volatile matter; C: fixed carbon.

Fuel name Type Origin Devolatilized M (%) Proximate analysis A (%) V (%) C (%) V/(V+C)

RSA Bituminous coal South Africa Yes 4.3 22.4 2.8 70.6 3.8

Cerrejón Bituminous coal Colombia Yes 2.3 18.4 11.6 67.7 14.6

Colombian Bituminous coal Colombia Yes 2.7 6.2 29.1 62 31.9

Pet coke Petroleum coke Mexico Yes 2.3 6.4 2.5 88.9 2.7

Met coke Metallurgical coke SSAB, Sweden No 0.3 10.9 0.6 88.2 0.7

3. Data evaluation

In order to evaluate the performance of the 10 kW pilot certain parameters need be introduced:

• The oxygen demand, QOD, describes the fraction of oxygen n _ Q.S(CO) + 2{CHt) + O.S(H2) + l.S(H2^) lacking to achieve complete combustion of the gas produced in the FR, eq. (S), where @e is the molar ratio [nO2 required for combustion/kg fuel] / [nC /kg fuel].

• The instantaneous rate of fuel conversion, kinst, eq. (6), where mCFR (t)is the amount of non-reacted carbon present in the fuel reactor at time t, and mC(t) is the mass flow of gaseous carbon, i.e.

carbon converted to CO, CO2 and CH4.

• The fraction of carbon in the batch that is converted to gas, eq. (7), where te is the time when the batch is introduced and tend is the time when the experiment ends. The amount of carbon left in the fuel reactor at the end of each experiment, mCFR (tend), was zero for all

fuels except met coke.

• A relatively large fraction of the fuel escapes the reactor system unconverted. The total amount of fuel that is actually converted to gas, Xq^ci, is expressed in eq. (8), where mCfuel is the carbon introduced into the fuel reactor with the fuel.

• Activation energies, Ea, are calculated using the Arrhenius equation, eq. (9), expressing the temperature dependence of the specific reaction rate, k. ke is the frequency factor, ta is the activation energy, R is the gas constant

(8.314 J mol-1 K-1), and T is the absolute temperature. 4. Background and purpose of study

This study deals with batch experiments of five fuels at four temperatures. The purpose of the tests is to evaluate kinetics of char conversion as a function of the temperature, and the oxygen demand associated with the syngas produced by the char. Most of the volatile matter in the fuels used was driven off by subjecting the fuel particles to high temperatures in a coke oven. This devolatilization was important since the purpose of the study was to investigate the reactivity of the char of different fuels.

Reactions between ilmenite and syngas are fairly quick at elevated temperatures [6]. It is thus reasonable to believe that the gasification of the fuel is the rate limiting step. Hence, diligent efforts have been made to find the rate of char gasification in this work. Prior experience from continuous and batch tests has shown that a rather high fraction of the reducing gases escapes the fuel reactor unconverted. These gases can either be volatiles or syngases, i.e. products from the char gasification. Using pet coke, Berguerand and Lyngfelt found that the oxygen demand, Qod, varied between 27 and 36% during stable, continuous operation [3]. An experiment aimed at quantifying the volatiles in the total gas outflow was performed by investigating the transitions in gas concentrations at the outlet of the fuel reactor during the first minute following start and stop of the fuel feed at TFR = 9S0 °C and a fuel feed corresponding to 6 kWth [3]. The experiment demonstrated that most of the CO and H2 and all the CH4 in the exiting gas from the fuel reactor actually come from the volatiles, which again is in line with expectations. The experiments indicated that as much as 80% of the oxygen demand was associated with volatiles.

The high amount of volatiles that escape the reactor system unconverted is due to the design of the fuel reactor and how the fuel is introduced. The time for devolatilization of a single, < 2S0 p,m, particle in an inert atmosphere at a temperature of 9S0°C, is in the order of 1 s or less. Since the height of the "hot" zone - i.e. the length of the fuel chute inside the oven - is 1 m and the maximum velocity of the particles attained in this zone is 1 m/s, there is enough time for devolatilization to occur. The volatiles then by-pass the fuel-reactor bed and never come into contact with the oxygen carrier. This suggests that the more volatile matter a fuel contains, the greater the oxygen demand in continuous testing. In-bed feeding would solve the problem by largely improving contact between oxygen carrier and fuel, but has not been considered for the 10 kW pilot.

0&COJ + (CO) + (CH4)] ( )

k,„„ (t) _-- m c (t) (6)

mc FR (t)

f mC (t)dt

Xc (t) _ ftw - (7)

I mc (t)dt + mc FR (tend ) Jt0

Y — 0

C, fuel

I" e"d rkc (t)dt

"C, fuel

k _ keEa'RT (9)

In summary, the rather high oxygen demand measured during previous tests is strongly associated with rapid devolatilization of the fuel in combination with the fuel feed design and the fuel reactor configuration itself, preventing the contact between volatiles and oxidized ilmenite. In order to reduce possible errors related to the presence of volatiles, the batch experiments conceived in this study were made with devolatilized fuel, which could provide vital information on (a) oxygen demand associated with syngas from char gasification as well as (b) kinetics of char conversion to gas. Efforts in this area are vital since char gasification is the rate limiting step in CLC with solid fuels.

5. Results

Table 2. Total fuel conversion, X,

C. fuel.

Five fuels were fed in batches of 20-25 g to the fuel reactor at four temperatures; 940°C, 970°C, 1000°C and 1030°C. Gas measurements of primarily CO2, CO, CH4 and H2 provide the basis for analysis in this study. Table 2 shows fuel conversion as the fraction of the carbon in the batch converted to gas. A lot of the incoming fuel is blown away to the chimney and never comes in contact with the oxygen carrier.

Figs. 3 and 4 show dry gas concentrations at the outlet of the fuel reactor. Note that steam is the fluidizing agent in the LOVEL, whereas N2 is used (a) for fluidization of the carbon stripper and the loop seals, and (b) as sweep gas for the fuel. Thus, large amounts of nitrogen dilute the gas originating from the fuel, which explains the rather low concentrations in Figs. 3 and 4. Fig. 3 shows typical concentrations versus time for Cerrejón at 1000°C. Fig. 4 shows CO2 concentrations versus time for the same fuel at different temperatures. The general appearance of these curves is the same for all five fuels.

Fuel TFR (°C)

940 970 1000 1030

RSA Cerrejón Colombian Pet coke Met coke 0.39 0.45 0.46 0.52 0.47 0.53 0.64 0.64 0.24 0.48 0.51 0.64 0.63 0.33 0.49 0.41 0.60 0.61 0.41

....... 940°C

----- 970°C

---1000°C

-1030°C

3 4 5 6 Time (min)

Figure 3. Gas concentrations for Cerrejón at 1000°C.

3 4 5 6 Time (min)

Figure 4. CO2 concentrations at different temperatures with Cerrejón.

Figs. 5 and 6 show reaction rate constants versus fuel conversion. Note that in Fig. 5, the line for met coke ends abruptly at XC = 0.375. This is because the met coke reacted much slower than the other fuels and there was still a lot of fuel left in the fuel reactor when the experiment was terminated after 25 min. The reason the tests had to be terminated after this time was to avoid excessive thermal stress in the reactor system, since the temperature in the air reactor drops rapidly in batch tests, thus creating a substantial temperature difference between AR and FR which increases over time. Note also the initial "hump" for Colombian, which is due to volatiles.

Fig. 7 shows the Arrhenius plot of ln kimtmax versus Tfor five fuels. Values for kimtmiax were obtained by taking the average of kinst between XC = 0.33 and XC = 0.67, except for met coke, for which the corresponding "mid-third" was used, e.g. for met coke at 970°C, kimt_max was obtained as the average of kinst between XC = 0.125 andXC = 0.25, cf. Fig 5. The reason kimmax is used, is that it is believed to best represent the actual reaction rate constant for char conversion to gas. At the start of a test, the effect of back-mixing in the gas measurements is considerable, due to the presence of dead volumes, such as cooling systems. Studying Fig. 5, it is also clear that the initial volatile-induced peak for Colombian is not representative for char conversion and should thus be excluded when determining kimt_ max, which is accomplished by choosing the mid-third with respect to XC. The reaction rate constant decreases towards

- 0-8 'c

1. 0.6

the end of the experiments, notably for those carried out at higher temperatures. Here, concentrations are lower, and there is therefore more uncertainty in the measurements. More importantly, it is suspected that the long "tail" of synthesis gases seen in all experiments can be explained by gasification occurring outside LOVEL. This would happen if (a) a fraction of the volatiles released initially formed carbon on the walls downstream of LOVEL, or (b) fuel that did not enter the LOVEL-bed get stuck on the walls downstream of LOVEL or in the recirculation loop.

Table 3 shows activation energies and ratios for ki„slrT1/kinslrT2 for the fuels used in the experiments. It can be noted that the more volatiles a fuel contains, the higher the calculated activation energy. Similarly, the ratio kimt,le3e°C/ kimt,97e°C increases with amount of volatile matter present in the fuel.

Table 3. Activation energies and ratios for

k¡mt. Ti/k¡„st . T2 at different temperatures_

Fuel_Ea (kJ/mol) kimt.,ese°c/ kimt.g?e°c

RSA 153 2

Cerrejón 189 2.4

Colombian 222 3.2

Pet coke 119 1.8

Met coke 99 1.6

Figs. 8-10 show oxygen demand, QOD, versus fuel conversion. It is assumed that minimum values of QOD best represent the QOD of the char. Initially there is the presence of volatiles creating a peak and the effect of back-mixing in the measurement system, discussed above, that makes the peak wider and less tall. Towards the end of the experiment, the fraction of CO and H2 - relative to CO2 - increases. The presence of syngases may, as briefly discussed above, be explained by gasification outside the LOVEL section of the fuel reactor. Since there is only little fuel left in the fuel reactor towards the end of the experiment, such lingering gasification would yield an increase in oxygen demand, as seen to the right in Figs. 8-10. Fig. 11 shows the minimum oxygen demand as function of temperature for five fuels. The three points in parentheses along the x-axis represent measured minimum oxygen demand for the met coke. However, concentrations of CO and H2 were very low, which makes the measurement uncertainty quite large. There is one measurement point for met coke at 1000°C in Fig. 11 that is not in parenthesis. This test was made with a S0 g batch and much reduced sweep gas (N2) flow, which reduced the dilution of carbonaceous gases, i.e. increased the concentrations of CO and H2. This test with met coke can also be seen in Fig. 10.

---Cerrejón

- — Colombian

..... Pet coke

-Met coke

Figure 5. kinst versus fuel conversion at 970°C for five different fuels.

Figure 6. kimt versus fuel conversion for different temperatures using Cerrejón as fuel.

-0.5 -1

I -1.5 c

~ -2.5 -3 -3.5

7.6 7.7 7.8 7.9

"1(K"1)

- Cerrejón

- Colombian

- Pet coke

- Met coke

Figure 7. ln k„st versus 1/TFR for five fuels.

1 0.6 E

с „ ,

О) 0.4

....... 940°С

----- 970°С

---1000°С

-1030°С

Figure 8. Oxygen demand versus fuel conversion for different temperatures using Cerrejón as fuel.

---Cerrejón

-----Colombian

....... Pet coke

-Met coke

Figure 9. Oxygen demand versus fuel conversion at 970°C for five fuels.

0.12 0.1

n 0.06

с 0.04

О 0.02

0.4 0.6

Figure 10. Oxygen demand versus fuel conversion for pet coke and met coke.

—Д- -RSA

—e- Cerrejón

—it— Colombian

-B— - Pet coke

* Met coke

( Ht )

( Ж )

980 Temp fC)

Figure 11. Minimum oxygen demand as function of TFR for five fuels.

6. Discussion

Values obtained for activation energy of the char seem reasonable. Wen and Lee compared literature data on activation energies for steam gasification of different chars and found an average of 146 kJ/mole [15].

The oxygen demand, QOD, was thoroughly investigated for all fuels. The minimum QOD was assumed to best represent the QOD of the char, since (a) it was clear that unconverted volatiles are responsible for the initial peak in QOD, and (b) it was suggested that gasification of fuel outside the LOVEL created a long tail of syngases, as discussed above, which would explain the increase in QOD once the minimum had been reached. This would also imply that minimum values for QOD are overestimations of the actual QOD of the char. Modelling work by Berguerand et al. [16] suggests that, using ilmenite in the 10 kW pilot, the oxygen demand of the char should be around 1.5% at 1000°C. Another possible explanation of the increase in QOD with increasing XC would be decreased reactivity of ilmenite as the amount of available oxygen decreases.

Fig. 11 suggests that the minimum QOD is lower for fuels with little volatile matter. This result is also in line with findings by Berguerand and Lyngfelt [5] who found the QOD for non-devolatilized pet coke to be 8.4-11%. QOD for the char is generally expected to be similar for all fuels, since there is no difference in the syngas generated. However, the difference in QOD for the different fuels could be explained by (a) the previously discussed carbon gasification downstream of LOVEL, which should be higher for fuels with more volatiles; (b) back-mixing, which is more likely to affect rapidly reacting fuels; note how back-mixing and high volatile content move the minimum Qod to high XC for 'Colombian' in Fig. 9; and (c) mixing of fuel into the bed; fuels with high rates of conversion may not have sufficient time to mix, i.e. more fuel is gasified in the top part of the bed, giving less contact with the oxygen

carrier. In conclusion, the actual oxygen demand of the char is best represented by pet coke and met coke, which gasify slowly and contain little volatile matter.

7. Conclusions

Batch experiments with five different fuels, four of which were devolatilized, were carried out in a 10 kW CLC pilot. Batches of 20-2S g were fed to the fuel reactor at TFR = 940°C, 970°C, 1000°C, and 1030°C. The aim of the tests was to determine vital parameters for char conversion. It was possible to determine (a) oxygen demand associated with syngas from char gasification as well as (b) kinetics of char conversion to gas. Instantaneous rates of char conversion to gas was found to be temperature dependent, as expected, whereas no temperature dependence was found for the oxygen demand. Activation energies for conversion of char to gas were calculated using Arrhenius plots.

• The activation energies were found to vary from 99 kJ/mole for met coke to 222 kJ/mole for Colombian.

• The reaction rate constant was found to increase with a factor 1.6 - 3.2 between 970°C and 1030°C.

• The minimum oxygen demand, suggested to be the same as the oxygen demand for char, was found to be 4-S% for pet coke and met coke, and 6-11% for the bituminous fuels.

Acknowledgement

This work has been made with financial grant from the Research Fund for Coal and Steel of the European Community (Contract No: RFCP-CT-2008-00008). The authors wish to thank SSAB for supplying the met coke.

References

[1] Berguerand N, Lyngfelt A. Design and operation of a 10 kWth chemical-looping combustor for solid fuels - Testing with South African coal. Fuel 2008; 87:2713-2726.

[2] Berguerand N, Lyngfelt A. The use of Petroleum coke in a 10 kWth Chemical-Looping Combustor. International Journal of Greenhouse Gas Control 2008; 2(2):169-179.

[3] 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): S2S7-S268.

[4] Shen L, Wu J, Xiao J. Experiments on Chemical-looping Combustion of Coal with a NiO based Oxygen Carrier. Combustion and Flame 2009; 1S6(3): 721-728.

[5] Berguerand N, Lyngfelt A. Batch Testing of Solid Fuels with Ilmenite in a 10 kWth Chemical-Looping Combustor. Fuel 2010; 89:1749-1762.

[6] Leion H, Lyngfelt A, Johansson M, Jerndal E, Mattisson T. The use of ilmenite as an oxygen carrier in chemical-looping combustion. Chemical Engineering Research and Design 2008; 86:1017-1026.

[7] Jerndal E, Mattisson T, Lyngfelt A. Thermal Analysis of Chemical-Looping Combustion. Chemical Engineering Research and Design 2006; 84:79S-806.

[8] Abad A, Garcia-Labiano F, de Diego L, Gayan P, Adanez J. Reduction kinetics of Cu-, Ni, and Fe-based oxygen carriers using syngas (CO + H2) for chemical-looping combustion. Energy&Fuels 2007; 21:1843-18S3.

[9] Ryu HJ, Bae DH, Jin GT. Effect of temperature on reduction reactivity of oxygen carrier particles in a fixed bed chemical-looping combustor. Kor J Chem Eng 2003; 20:960-966.

[10] de Diego LF, Garcia-Labiano F, Gayan P, Celaya J, Palacios JM, Adanez J. Operation of a 10 kWth chemical-looping combustor during 200 h with a CuO-Al2O3 oxygen carrier. Fuel 2007; 86:1036-104S.

[11] Kolbitsch P, Proll T, Bolhar-Nordenkampf J, Hofbauer H. Operating experience with chemical looping combustion in a 120 kW dual circulating fluidized bed (DCFB) unit. GHGT-9. Washington DC, 16-20 Nov. 2008.

[12] Lyngfelt A. Oxygen carriers for chemical-looping combustion - 4000 h of operational experience. Submitted to Oil and Gas Science and Technology - La Revue de l'IFP. (/1st Int. Conf. on Chemical Looping. Lyon, 2010.)

[13] 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; 88:2083-2096.

[14] Wolf J. CO2 mitigation in advanced power cycles. PhD Thesis, Department of Chemical Engineering and Technology. Royal Institute of Technology, Stockholm, Sweden. 2004.

[15] Wen CY, Lee ES. Coal Conversion Technology. Reading, Massachusetts: Addison-Wesley Publishing Company; 1979.

[16] Berguerand N, Lyngfelt A, Mattisson T, Markstrom P. Chemical-Looping Combustion of Solid Fuels in a 10 kWth Unit. Submitted to Oil and Gas Science and Technology - La Revue de l'IFP. (/1st Int. Conf. on Chemical Looping. Lyon, 2010.)