Scholarly article on topic 'Natural minerals as oxygen carriers for chemical looping combustion in a dual circulating fluidized bed system'

Natural minerals as oxygen carriers for chemical looping combustion in a dual circulating fluidized bed system Academic research paper on "Chemical engineering"

CC BY-NC-ND
0
0
Share paper
Academic journal
Energy Procedia
OECD Field of science
Keywords
{"Carbon capture" / "Unmixed combustion" / "Chemical looping" / "Oxygen carriers" / Ilmenite / Olivine / "Dual fluidized bed"}

Abstract of research paper on Chemical engineering, author of scientific article — Tobias Pröll, Karl Mayer, Johannes Bolhàr-Nordenkampf, Philipp Kolbitsch, Tobias Mattisson, et al.

Abstract A first experimental campaign has been conducted at a 120 kW fuel power dual circulating fluidized bed installation for chemical looping combustion of gaseous fuels. In these test runs natural ilmenite (FeTiO3) has been used as oxygen carrier material. The plant consists of two interconnected circulating fluidized bed reactors (stainless steel construction, inner diameter: 0.15 m, height: air reactor 4.1 m, fuel reactor 3 m). Variations of fuel composition (natural gas, synthetic gas mixtures of H2 and CO), load, temperature and solids circulation rate have been performed for the bulk bed material. Further, natural olivine, (Fe,Mg)2SiO4, has been studied as an additive to increase hydrocarbon conversion. Despite the limited height of the risers, the results show reasonable fuel conversion for CO and H2 at 950 ∘C. The conversion of natural gas, i:e. CH4, on the other hand, is relatively low for the pure ilmenite material at about 30–40%. A certain dependency of fuel conversion on load is found especially for CH4. Addition of natural olivine results in a moderate increase of CH4 conversion.

Academic research paper on topic "Natural minerals as oxygen carriers for chemical looping combustion in a dual circulating fluidized bed system"

Available online at www.sciencedirect.com

ScienceDirect

Energy Procedia

ELSEVIER

Energy Procedía 1 (2009) 27-34

www.elsevier.com/locate/procedia

GHGT-9

Natural minerals as oxygen carriers for chemical looping combustion in a dual circulating fluidized bed system

Tobias Prolla*, Karl Mayera, Johannes Bolhar-Nordenkampf, Philipp Kolbitscha, Tobias Mattissonb, Anders Lyngfeltb, Hermann Hofbauera

aVienna University of Technology, Getreidemarkt 9/166, 1060 Wien, Austria bChalmers University of Technology, Hörsalvägen 7B, 412 96 Göteborg, Sweden

Abstract

A first experimental campaign has been conducted at a 120 kW fuel power dual circulating fluidized bed installation for chemical looping combustion of gaseous fuels. In these test runs natural ilmenite (FeTiO3) has been used as oxygen carrier material. The plant consists of two interconnected circulating fluidized bed reactors (stainless steel construction, inner diameter: 0.15 m, height: air reactor 4.1 m, fuel reactor 3 m). Variations of fuel composition (natural gas, synthetic gas mixtures of H2 and CO), load, temperature and solids circulation rate have been performed for the bulk bed material. Further, natural olivine, (Fe,Mg)2SiO4, has been studied as an additive to increase hydrocarbon conversion. Despite the limited height of the risers, the results show reasonable fuel conversion for CO and H2 at 950°C. The conversion of natural gas, i:e. CH4, on the other hand, is relatively low for the pure ilmenite material at about 30-40%. A certain dependency of fuel conversion on load is found especially for CH4. Addition of natural olivine results in a moderate increase of CH4 conversion. © 2009 Vienna University of Technology. Published by Elsevier Ltd. All rights reserved.

Keywords: Carbon capture, Unmixed combustion, Chemical looping, Oxygen carriers, Ilmenite, Olivine, Dual fluidized bed

1. Introduction

Chemical looping combustion (CLC) is a novel combustion process that avoids direct mixing of fuel and combustion air. The CLC technology uses metal oxide particles for selective oxygen transport from the air reactor to the fuel reactor, thus pure CO2 is obtained in the fuel reactor exhaust stream after condensation of water without further gas separation needed. The process features 100% carbon capture rates, a highly concentrated stream of CO2 ready for sequestration, no NOx emissions [1], and no costs or energy penalties for gas separation. CLC uses well-established boiler technology very similar to circulating fluidized bed boilers, which also means that costs can be assessed with great accuracy. CLC is estimated to achieve significant CO2 capture cost reductions compared to today's best available technology, namely post combustion amine scrubbing [2].

The major challenge with CLC is the availability of adequate oxygen carriers. The requirements for a good oxygen carrier are high oxygen transport capacity, high reactivity, high mechanical strength and low production costs [3]. Typically, these requirements are best achieved if the active metal oxide is provided on a ceramic support

doi:10.1016/j.egypro.2009.01.006

structure with a high surface area. Such designed particles can be produced using different methods. A wide range of different oxide/support combinations have been selected and studied in recent years [4]. From thermodynamics point of view, the interesting redox systems for CLC are CuO/Cu, CoO/Co, Fe2O3/Fe3O, Mn3O4/MnO4 and NiO/Ni, as well as a number of perovskites. Cobalt and nickel based carriers show thermodynamic limitations with respect to full conversion of CO and H2, but they are very reactive towards hydrocarbons on the other hand. The other carriers thermodynamically allow full fuel conversion to CO2 and H2O, but their reactivity is lower on the other hand. Further, Fe, Mn, Co and Cu can take in more than two oxidation states. However, not all of these are relevant for CLC.

Ilmenite, a natural iron-containing mineral, has been proposed as cheap and readily available oxygen carrier for CLC of solid fuels [5]. Ilmenite is conventionally used as a raw material for TiO2 pigment production. Some processes for ilmenite upgrading and Fe/Ti separation for pigment production involve oxidation and subsequent reduction of ilmenite. The solid phase chemistry of ilmenite in different oxidation states has been studied by different authors [6,7]. In the raw mineral, most of the iron is present as Fen.in FeTiO3 (ilmenite phase). According to Nell [7], essentially two solid solution phases may occur if ilmenite is increasingly oxidized. The first region is represented by the ilmenite structure M2O3, where TiFeO3 (Fen) and and Fe2O3 (FeIH) are miscible within certain limits. Along with oxidation, rutile (TiO2) is formed according to:

2 FeTiO3 + 0.5 O2 ^ Fe2O3 + 2 TiO2 (1)

Rutile crystallites have been found outside of the M2O3 structure. At the same time, free cation sites are generated in the M2O3 structure due to the addition of anion (oxygen) sites. Electrical effects make the electron donating iron atoms move to the crystal surface [7].

If oxidation proceeds further, a second phase with a M3O5 structure appears. Again, there is a certain miscibility reported between FeTi2O5 (pseudobrookite = Fen) and Fe2TiO5 (ferric pseudobrookite = Feni). And again rutile is formed along with oxidation:

2 FeTi2O5 + 0.5 O2 ^ Fe2TiO5 + 3 TiO2 (2)

Despite the fact that rutile is formed and appears outside of the iron-containing phases the original ilmenite phase is re-formed during reduction of oxidized material. This is described for the Murso process [8] where the raw material is first oxidized to M3O5 phases (with TiO2 formed) and subsequently converted back to FeTiO3 in a reduction roasting. This implies that the rutile moves back in the original structure during reduction.

Leion et al. [9] have tested cyclic stability of ilmenite during repeated cycles of oxidation and reduction and found an increase in reactivity with increasing cycle number during the first 10-20 cycles. After that, the reactivity in terms of CO conversion has remained constant. It seems that ilmenite is a serious candidate for a cheap and environmentally sound oxygen carrier in chemical looping combustion. However, the reactivity towards methane is low for iron-based carriers, what turns into a problem for CH4 rich fuels like natural gas.

Olivine, (Fe,Mg)2SiO4, is applied as bed material in fluidized bed biomass gasification and some moderate catalytic activity towards hydrocarbon reforming has been reported [10]. In the present study, olivine is used as an additive to ilmenite in some experiments with the intention to improve methane conversion.

2. Bulk reactions and definitions

Generally, in the fuel reactor, the gaseous fuel species CO, H2 and CH4 may be oxidized in contact with the solid oxygen carrier. In the following expressions, the ilmenite chemistry is lumped into the formal expressions FeO (Fen) and Fe2O3 (FeIII), even though iron seems to always appear in combination with TiO2 if ilmenite is the raw material.

CO + Fe2O3 ^ CO2 + 2 FeO H2 + Fe2O3 ^ H2O + 2 FeO

CxHy + (2x + 0.5y) Fe2O3 ^ x CO2 + 0.5y H2O + (4x + y) FeO

In parallel to the oxidation Reactions (3)-(5), CO and H2 may be formed from hydrocarbons. This happens due to partially incomplete progress of Step (5) or via reforming of hydrocarbons. A detailed investigation of the complex chemical reaction mechanisms involved in hydrocarbon conversion in the fuel reactor would exceed the scope of the present effort. Instead, some basic definitions are made in the following to quantify experimental observations.

The conversion of a fuel species is generally defined as

v ! e FRout ' FRout

=1 —:-. (6)

e FRie Si,FRie

In chemical looping combustion, where the focus is on high fuel conversion to CO2, the CO2 yield from the fuel reactor is defined according to

Yco, =-.--\. (7)

FRie •Z^i • yi.FR.ie)

3. Experimental

A dual circulating fluidized bed test rig has been designed and built with hot commissioning in January 2008. The main design data are 120 kW fuel power natural gas and a global air ratio of 1.0.. .1.2. The design of the unit is discussed into detail elsewhere [11]. The air reactor (AR) of the CLC system is designed as fast fluidized bed determining the global solids circulation rate. The fuel reactor (FR) is designed as a second circulating fluidized bed reactor with the return loop of the entrained solids into itself. The FR may be optimized with respect to gas phase conversion. The lower loop seal connecting the two reactors represents a continuation of the reactor bodies and closes the global circulation loop. The main fluidization nozzles are arranged along the circumference of the cylindrical reactor shells. A sketch of the test rig with the most important geometric data is shown in Figure 1.

The fuel reactor may be supplied with either natural gas or with synthetic gas mixtures from gas cylinder stacks (H2, N2, CO). Fuel and air volume flows to the reactor system are continuously registered and controlled using rotary flow meters and control valves. Temperature and system pressure is taken at 30 positions at the test rig. In order to remove the heat released from combustion and to control the system temperature independently of the global air ratio, the air reactor shell is equipped with cooling jackets. These cooling jackets are operated either with boiling water or gaseous cooling media. The cyclone separators are designed according to Hugi [12]. The loop seals are fluidized with steam in nominal operation and may be switched to air fluidization during start-up and shut-down. The exhaust gas streams of the two reactors are cooled separately to about 573 K and analyzed on-line (Rosemount

Species wt.%

TiO2 44.08

Fe2+ 25.93

Fe3+ 9.14

O with Fe 11.36

P2O5 0.03

S 0.14

Cr2O3 0.08

SiO2 3.10

CaO 0.42

MgO 4.38

Al2O3 0.93

MnO 0.29

K2O 0.04

Na2O 0.10

Sum 100.00

Table 1: Chemical composition of the ilmenite used based on data supplied by Titania A/S.

Particle size [10-6 m]

Figure 2: Size distribution of the ilmenite used.

NGA 2000) to evaluate the conversion of the fuel as well as the leakages of the loop seals. The cooled exhaust gas streams pass valves. These allow for imposing defined backpressure on each reactor. The exhaust gas streams are then mixed and sent to a natural gas fired post combustion unit, cooled again, filtered and sent to the chimney. The whole laboratory plant is monitored and controlled using a commercial process control system.

In a first experimental campaign directly after hot commissioning, ilmenite has been used as oxygen carrier material. The chemical composition and the size distribution of the ilmenite used are reported in Table 1 and Figure 2, respectively.

Due to defluidization problems when working with fresh ilmenite in chemical looping conditions, an oxidative pre-treatment of the ilmenite for 3-6 h at 850 °C has been performed at the installation before switching to chemical looping operation. This procedure solves the defluidization problem effectively. Each experimental run has lasted between 12 and 16 hours in total. Different operating points have been run during each day of experimentation. For natural gas operation, natural olivine has been tested as an additive to ilmenite. Table 2 reports the ranges of operating parameters during this first test campaign at the pilot rig with ilmenite as the bulk bed material.

Due to several difficulties at the beginning, only limited power ranges have been run for syngas mixtures. Fuel addition to the air reactor has been necessary in most cases to keep the system temperature. However, some basic results for natural minerals as oxygen carriers in CLC of gaseous fuels can be reported. The performance data can be compared to the data of designed NiO-based oxygen carriers tested in a later campaign and reported at this conference [13].

Unit Fuels

Parameter natural gas

H2 + n2 H2 + CO

Fuel power kW 24 - 43 40 - 92 20 - 130

Temperature air reactor °C 895 - 978 887 - 983 813 - 986

Temperature fuel reactor °C 894 - 966 878 - 976 839 - 969

Content of O2 in air reactor exhaust v-%(dry) 8.5 - 11.0 9.3 - 14.2 3.6 - 12.8

Total mass ilmenite kg 70 - 90 70 - 85 70 - 90

Total mass additive (olivine) kg 0 0 0 - 15

Table 2: Ranges of operating parameters during ilmenite runs.

4. Results and discussion

A key towards the interpretation of results is a general picture of the fluid dynamics of the system. Figure 3 shows the pressure profile of the system measured at 120 kW operation with ilmenite. An exponential decay of pressure is found in the air reactor. This is typical for circulating fluidized bed risers and indicates the decrease of solids concentration with height. The FR on the other hand, shows a concentration of solids in the bottom region, indicated by the steep pressure change between positions FR 1 and FR 2 (Figure 3). The upper regions of the fuel reactor are, therefore, very lean in solids and no real circulating regime is reached in this case for the fuel reactor. The reasons for the limited development of the circulating regime are lower gas velocities due to limited fuel conversion and increased mean particle size of the solids. In comparison, the system has been designed for a mean particle diameter of 0.120 mm and full fuel conversion has been assumed for the design case [11].

The fuel conversion for CO and H2 is shown versus fuel power in Figure 4 for a 1:1 mixture of CO and H2 as fuel. The results mirror the apparent reactivity of the system. A slight decrease of fuel conversion is found with increasing load. This indicates that the gas-solids contact time, which decreases with increasing load, governs the fuel reactor performance in this case. Taking the non-optimal fluid dynamic situation in the fuel reactor into account, it can be assumed that there is room for improvement of syngas conversion rates in the future. Combustion of CO and H2 plays an essential role in the context of CLC for solid fuels.

Figure 5 shows the CH4 conversion and the CO2 yield versus fuel power with natural gas as fuel in the fuel reactor. The CH4 conversion is generally relatively low in these runs. A high dependency of fuel reactor performance on load is witnessed. A change of the global solids inventory from 70 kg to 85 kg shows little effect.

ra <u CE

5.0 4.0 3.0 2 0 1.0 0.0

cyclone j exit\ I

FR 4 FR 3

AR cyclone exit —AR 4 (top)

■ AR downcomer AR 3

FR FR 2 downcomer

internal loopseal

AR 1 (bottom) / lower loop seal

6 8 10 12 Pressure [kPa]

14 16 18

— 0.8

o E 15

ê 0.4

lu £ 0.2

♦ X_H2

■ 1 ■ ■ X_CO

T M 960°C ± 10°C

System solids inventory: 70 kg ilmenite

0 20 40 60 80 100 120 140 Fuel power to fuel reactor [kW]

Figure 3: Pressure profile in the reactor system with ilmenite at 120 kW operation with natural gas.

Figure 4: Fuel conversion in the fuel reactor for syngas operation (CO:H2 = 1:1).

o E ö £ 0.6

TFR = 950°C ± 5°C

♦ * ♦

System solids inventory: ■ 70kg Ilmenite + 15kg Olivine

♦ 85kg Ilmenite

* 70kg Ilmenite

20 40 60 80 100 120 140 Fuel power to fuel reactor [kW]

TFR = 950°C ± 5°C

♦ * ♦

Î." -

System solids inventory: ■ 70kg Ilmenite + 15kg Olivine

♦ 85kg Ilmenite

* 70kg Ilmenite

0 20 40 60 80 100 120 140 Fuel power to fuel reactor [kW]

Figure 5: Fuel conversion (a) and CO2 yield (b) for 100% natural gas operation with different solids inventories.

=3 0.6

An interesting aspect is that the 85 kg run where about 18% of the bed material have been replaced by natural olivine shows some improvement in CH4 conversion. Again, the results indicate that the reactivity in the fuel reactor limits the system performance. The amount of CO and H2 leaving the fuel reactor can be derived from the CO/CO2 ratio and from the H2/CO ratio shown in Figure 6. It turns out that the slip of the uncombusted species CO and H2 through the fuel reactor is slightly increasing with increasing load and, therefore, with increasing hydrocarbon slip. Olivine addition shows some effect on the fuel reactor exhaust stream and slightly reduces CO as well as H2. All in

all, also these results for natural gas conversion own some potential for future improvement. It is expected that with fluidized bed riser reactors of adequate height (12-18 m) and reasonable fluidizing velocities for solids distribution over the whole height, significantly better fuel conversion rates can be achieved.

0.18 2.0 :

0.16 ♦ ♦ A 1.8 A 4 i 4

0.14 ♦ A ♦ ♦ A 1.6

0.12 A** . " 1.4 ' . ■

♦♦ . 1 o E 1.2

0.10 ■ lo £ 1.0 : TFR = 950°C ± 5°C

0.08 : TFR = 950°C ± 5°C O

0.06 - System solids inventory: X 0.6 : System solids inventory:

0.04 ♦ 85kg Ilmenite 0.4 ♦ 85kg Ilmenite

0.02 a 70kg Ilmenite a 70kg Ilmenite

■ 70kg Ilmenite + 15kg Olivine 0.2 : 70kg Ilmenite + 15kg Olivine

0.00 ........................... 0.0 :..........................

0 20 40 60 80 100 120 140 Fuel power to fuel reactor [kW]

0 20 40 60 80 100 120 140 Fuel power to fuel reactor [kW]

Figure 6: Ratio CO/CO2 (a) and ratio H2/CO in fuel reactor exhaust gas for natural gas operation with different solids inventories.

5. Conclusions

A first experimental campaign has been performed on a novel dual circulating fluidized bed system for chemical looping combustion using natural ilmenite as bed material. The system has been designed for 120 kW fuel power natural gas operation with NiO-based oxygen carriers. Several conditions have still not been optimized during these first runs using ilmenite. Coarser particles and lower gas flow rates compared to the design case caused a nonoptimal solids distribution in the fuel reactor. Nevertheless, reasonable fuel conversion between 60 and 90 % has been found for syngas (CO and H2), which is the main product of coal gasification. The operation with natural gas shows moderate conversion of CH4 and a strong dependency on load. Partial replacement of ilmenite by natural olivine improves the hydrocarbon conversion moderately. In conclusion, combinations of natural minerals offer some potential as cheap and environmentally sound oxygen carriers, especially for fuels with low gaseous hydrocarbon contents like gasification product gas or solid fuels. It is important to mention that the results have been obtained from a pilot plant and the performance of the oxygen carriers reported needs to be regarded as the performance of the carrier in this specific system at the parameters operated. The chemistry of ilmenite during repeated cycles of oxidation and reduction will certainly require more attention in the future.

6. Acknowledgement

This work was part of the EU financed project Chemical Looping Combustion CO2 Ready Gas Power (CLC GAS POWER - FP6 Contract No. 019800), coordinated by Chalmers University of Technology. The project is also part of phase II of CCP (CO2 Capture Project) via Shell.

The ilmenite used in the experiments has been supplied by Titania A/S, Norway.

7. Notation

temperature

conversion of species i

gas phase mole fraction

CO2 yield based on carbon supplied

number of carbon atoms in molecule of species i

molmol" molmol" molmol" molmol"

Super-/Subscripts AR air reactor

FR fuel reactor

i reference to gas species

in gas stream into reactor

out gas stream from reactor

8. References

[1] H.J. Ryu, G.T. Jin and C.K. Yi, Demonstration of inherent CO2 separation and no NOx emission in a 50kWth chemical-looping combustor: continuous reduction and oxidation experiment, in: Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies (GHGT-7), 2004, pp. 1907-1910.

[2] D.C. Thomas, 2005, Carbon Dioxide Capture for Storage in Deep Geologic Formations - Results from the CO2 Capture Project, Elsevier, Oxford, UK.

[3] A. Lyngfelt, B. Leckner and T. Mattisson, A fluidized bed combustion process with inherent CO2 separation, application of chemical looping combustion, Chemical Engineering Science 56, 2001, 3101-3113.

[4] M. Johansson, T. Mattisson, M. Ryden, A. Lyngfelt, Carbon Capture via Chemical-Looping Combustion and Reforming, in: International seminar on carbon sequestration and climate change, 2006, Rio de Janeiro, Brazil.

[5] H. Leion, T. Mattisson and A. Lyngfelt, Solid fuels in chemical-looping combustion, International Journal of Greenhouse Gas Control, 2, 2008, 180-193.

[6] K. Borowiec and T. Rosenqvist, Phase relations and oxidation studies in the system Fe-Fe2O3-TiO2 at 7001000 degree C, Scandinavian Journal of Metallurgy, 10, 1981, 217-224.r

[7] J. Nell, An overview of the phase-chemistry involved in the production of high-titanium slag from ilmenite feedstock, Journal of The South African Institute of Mining and Metallurgy, 100, 2000, 35-44.

[8] T.P. Battle, D. Nguyen and J.W. Reeves, The processing of titanium-containing ores, in: R.G. Reddy and R.N. Weizenbach (Eds.), Extractive Metallurgy of Copper, Nickel and Cobalt, Vol. 1: Fundamental Aspects, 1, The Minerals, Metals and Materials Society, 1993, pp. 925-943.

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

[10] R. Rauch, C. Pfeifer, K. Bosch, H. Hofbauer, D. Swierczynski, C. Courson and A. Kiennemann, Comparison of different olivines for biomass steam gasification, in: A.V. Bridgwater and D.G.B. Boocock (Eds.), Science in Thermal and Chemical Biomass Conversion, CPL Press, April 2006.

[11] P. Kolbitsch, J. Bolhar-Nordenkampf, T. Pröll and H. Hofbauer, Design of a chemical looping combustor using a dual circulating fluidized bed (DCFB) reactor system, in: J. Werther et al. (Eds.), Circulating Fluidized Bed Technology IX, TuTech, Hamburg, 2008, pp. 795-800.

[12] E. Hugi, Auslegung hochbeladener Zyklonabscheider für zirkulierende Gas/Feststoff-Wirbelschicht-Reaktorsysteme, Fortschr.-Ber. VDI 3 (502), VDI, Düsseldorf, Germany, 1997 (written in German).

[13] J. Bolhar-Nordenkampf, P. Kolbitsch, T. Pröll and H. Hofbauer, Performance of a NiO-based oxygen carrier for chemical looping combustion and reforming in a 120kW unit, this conference (GHGT-9), 2008.