1

Submitted, accepted and published by: Applied Energy 102 (2013) 860–867 Pollutant emissions in a bubbling fluidized bed combustor working in

oxy-fuel operating conditions. Effect of flue gas recirculation.

L.F. de Diego, M. de las Obras-Loscertales, A. Rufas, F. García-Labiano, P. Gayán, A.

Abad, J. Adánez.

Dept. Energy and Environment, Instituto de Carboquímica (ICB-CSIC)

Miguel Luesma Castán 4, 50018 Zaragoza. Spain

Corresponding author: ldediego@icb.csic.es

Abstract

In this work, pollutant emissions in a continuous bubbling fluidized bed combustor (~3

kWth) at oxy-firing conditions were measured. An anthracite coal was used as fuel and a

limestone was added for sulfur retention. Flue gas recirculation was simulated by

mixing different gases (CO2, SO2, steam, and NO) and the effect of varying the gas

composition of the recycled flow on the pollutant emissions was analyzed. It was

observed that the most important effect of CO2 recirculation was the increase of the

optimum temperature for SO2 retention from ~ 850 ºC (conventional air combustion) to

900-925 ºC. SO2 recirculation increased the Ca-based sorbent utilization and did not

affect the N2O emissions at any temperature. In addition, at 850 ºC the CO emission

increased and the NO emission decreased, however, at 925 ºC the CO variation was

negligible and the NO reduction was low. About 60-70% of the recycled NO was

reduced to N2 and N2O, being the NO converted to N2O lower than 5%. Steam

recirculation mainly led to a sharp decrease in NO emission. A synergetic effect among

the different recycled gases was not found in any case.

Keywords: Oxy-fuel combustion; fluidized bed; pollutant emission; gas recirculation.

1. Introduction.

2

Nowadays, there is an important public awareness in relation to the hazard of pollutant

gas emissions into the atmosphere from fossil fuel combustion to obtain energy. On the

one hand, CO2 gas has been recognized as one of the major contributors to the build-up

of greenhouse gases and, on the other hand, sulfur and nitrogen content in coal are

oxidized to SO2 and NOx respectively, which contribute to acid rain formation.

According to the IPCC 2005 [1], carbon dioxide capture and storage technologies would

be needed to mitigate CO2 emissions from large power plants into the atmosphere for

stabilization of atmospheric greenhouse gas concentrations. Oxy-fuel combustion is a

CO2 capture technology which is characterized by the use of a mixture of pure O2 and

CO2-rich recycled flue gas (instead of air which is used in conventional combustion) to

perform the combustion process. As a consequence, with this technology, the CO2

concentration in the flue gas may be enriched up to 95 vol% (dry), and therefore it is

possible to achieve an easy CO2 recovery.

There are mainly two types of boilers for coal combustion, pulverized coal (PC) boilers

and fluidized bed boilers. Although the majority of the researches in oxy-fuel

combustion have been developed for PC boilers [2,3], fluidized bed combustors (FBC),

and specially circulating fluidized bed (CFB) combustors, are also very appropriate for

this combustion system. This technology has the advantage that external solid heat

exchangers can be used to extract heat from the combustion process [4]. This will allow

the use of high oxygen concentrations in the combustor reducing the amount of recycled

flue gas and the area of the CFB combustor. Moreover, in situ desulfurization of

combustion gases, by feeding a low cost sorbent such as limestone [4,5], and relatively

low NOx emissions can be achieved [6,7].

Currently, the oxy-fuel combustion technology in CFB combustors is growing. Alstom

[8], VTT and Foster Wheeler [6], Metso [9], Czestochowa University of Technology

[7], and Canmet Energy [4,5,10,11] have performed oxy-fuel combustion experiments

with CFB combustors at scales up to 4 MWth. The Fundación Ciuden [12] in Spain has

built two plants able to operate at both conventional air and oxy-fuel combustion

conditions. The first plant is a 20 MWth PC boiler and the second plant is a CFB

combustor of 15 MWth operating in air-mode and 30 MWth operating in oxy-mode.

However, there are few works published which deal with the effect of the flue gas

recycle on the pollutant gas emission and thus there is still some lack of knowledge. The

research group of Canmet Energy has successfully operated two CFB combustors of

3

100 kWth [4,5,13] and 0.8 MWth [10,11] with flue gas recycle. At an operating

temperature of ~850 ºC, they found lower SO2 retention by calcium sorbents in oxy-fuel

combustion than in air combustion conditions, but the SO2 retention improved in oxy-

fuel combustion by increasing the operation temperature, that is, when the conditions

changed from direct to indirect sulfation. In addition, mainly due to flue gas

recirculation, lower NOx emissions in oxy-firing operation were observed compared to

air operation.

Our research group has recently carried out several tests in a continuous bubbling

fluidized bed (BFB) combustor (~3 kWth) where the effect of the temperature on the

sulfur retention by limestone addition was analyzed [14]. It was concluded that the

optimum operating temperature from the point of view of SO2 retention shifted from

850-870 ºC in combustion with enriched air to 900-925 °C in oxy-fuel combustion

mode.

In this work, pollutant gas emissions at different operating temperatures, mainly

working at oxy-fuel combustion mode, were measured in a continuous BFB combustor

(~3 kWth). However, the main objective of this work was to analyze the effect of the

flue gas recycled into the reactor on the pollutant emissions working in oxy-firing

mode. The gas inflow recycled was simulated by mixing different gases (CO2, SO2,

H2O, and NO), which allowed us to observe the influence of the different gases, both

separately and combined with each other.

2. Experimental section

2.1. Materials

A Spanish anthracite coal was selected as fuel for this study. The coal was crushed and

sieved, and the particle size in the range of 0.2-1.2 mm was used. Table 1 gives the

proximate and ultimate analyses of the coal. A high purity Spanish limestone

“Granicarb” (97.1 wt.% CaCO3) was used as calcium-based sorbent for sulfur retention.

Table 2 gives the analysis of the Granicarb limestone. The particle size of the limestone

was in the range of 0.3-0.5 mm. The porosities of the raw and after calcination sorbent

were 3.7 and 49 vol.%, respectively. Inert silica sand of size 0.2-0.6 mm was fed

together with the coal and the limestone during all the tests to control the residence time

of the sorbent in the fluidized bed reactor.

4

2.2. Experimental installation

The experimental installation consisted of a fluidized bed combustor (~3 kWth) and

different auxiliary systems for gas supply, solid feeding, solid recovering, and gas

analysis. Figure 1 shows a schematic diagram of the installation.

The combustor consisted of a stainless steel reactor of 9.5 cm i.d. and 60 cm height and

a freeboard of 15 cm i.d. and 50 cm height. The height of solid in the BFB was

maintained constant at 40 cm. A heat exchanger located inside the bed allowed the

perfect control of temperature. This heat exchanger could be moved vertically through

the reactor to modify the contact surface inside the bed, which allowed the extraction of

the needed heat from the combustor to reach the desired temperature.

The reactant gases, air, CO2, and O2, were supplied from cylinders by means of

electronic mass-flow controllers to simulate typical gas compositions entering into the

reactor in oxy-firing mode. The gases were fed into the reactor through a gas distributor

plate. An air pre-heater allowed the introduction of hot air inside the reactor during the

start-up of the plant. In the tests where the recycle of H2O, SO2, and/or NO was

simulated, the gas feeding system allowed the introduction of different gas mixtures in

order to analyze the influence of each gas in the process. H2O was supplied by a

peristaltic pump, subsequently evaporated and fed into the reactor through the gas

distributor plate together with the CO2 and O2. SO2 and NO were supplied from

cylinders by means of electronic mass-flow controllers just above the gas distribution

plate to avoid its corrosion.

The solids were fed to the reactor by means of water-cooled screw feeders located just

above the distributor plate. To ensure a constant Ca/S molar ratio, coal and sorbent were

fed together by means of two screw feeders in series: the first one controlled the

coal/sorbent feeding rate and the second one introduced the solid mixture as quick as

possible into the bed to avoid coal pyrolysis and plugging of the pipe. Other screw

feeder controlled the silica sand fed to the combustor. Table 3 shows the feeding rates of

solids and flow rate of gases used in the tests.

The flue gas stream leaving the combustor passed through a high efficiency cyclone to

recover the elutriated solids and it was afterwards sent to the stack. The gas composition

at the exit of the combustor was continuously analyzed after water condensation by on-

line gas analyzers. CO2, CO, and SO2 concentrations were measured in a Non-

5

Dispersive Infra-Red analyzer (NDIR, Siemens/Ultramat 23), O2 concentration in a

paramagnetic analyzer (Siemens/Oxymat 6), NO, N2O, and NO2 concentrations in an

infrared gas analyzer, via Fourier transform infrared (FTIR, Temet CX 4000).

The facility was equipped with pressure sensors and thermocouples to know the

pressure drops and temperatures in different locations of the installation. General

process data (temperatures, pressures, flow of gases, gas composition, etc.) were

continuously recorded by a computer.

2.3. Procedure

To start-up, the bed was filled with ~1.8 kg of silica sand and hot air was introduced

through the gas pre-heater to heat the bed up to the ignition temperature of coal. At this

temperature, the coal feeding started and the bed temperature rose rapidly due to the

coal combustion. When the desired temperature was reached, the preheating system was

turned off, the air was replaced by the desired gas mixture, sand and coal/limestone

mixture were fed into the bed, and the heat exchanger was introduced into the bed to

control the temperature. Once stable operating condition was reached, it was maintained

until steady state operation for SO2 retention was attained. An important feature of the

tests carried out in the continuous unit was the certainty that the results were obtained

under steady state conditions. This aspect was commented in detail in a previous paper

[14].

The SO2 retention (SR) was calculated by equation (1) as the molar fraction of sulfur

retained by the bed solids with respect to the sulfur contained in the coal feeding.

100.

/

/(%)

,,0

,2,,0

ScoalScoal

outSOoutScoalScoal

MxF

CQMxFSR

(1)

being F0,coal the coal feeding rate, xS,coal the coal sulfur content, MS the molecular weight

of S, CSO2,out the SO2 concentration in the flue gas at the exit of the reactor, and Qout the

gas flow rate at the reactor exit. Qout was calculated by means of a mass balance,

considering the coal and gas feeding flow rates and the flue gas composition. CSO2,out

was considered as an average value of the measurements taken during the whole test

duration in steady state conditions. The average concentrations of the other gases along

the test were also taken into account and were calculated in the same way as for SO2.

Steady state was maintained for at least 1 h for each experimental condition.

6

3. Results and discussion.

3.1. Influence of the combustion temperature.

First, combustion tests without feeding SO2, NO or steam were performed at different

temperatures between 800 and 950ºC. In all cases, the experimental conditions showed

in Table 3 with an inlet gas composition of O2/CO2 =35/65 were used. The coal feeding

rate was controlled to keep the O2 concentration at 3.75±0.75 vol% at the combustor

exit (dry basis).

Figure 2 shows the effect of the combustion temperature on the SO2, CO, NO, and N2O

emissions. NO2 emission was never detected (the lowest concentration reliably

detectable was 10 ppm). The SO2 emission decreased with increasing temperature up to

900-925 ºC and then, a further increase in temperature caused an increase in SO2

emission. In a previous work [14] it was found that calcium sorbent calcination had a

strong effect on SO2 retention under oxy-fuel combustion conditions. The SO2 retention

was higher working at calcining (indirect sulfation) than at non-calcining (direct

sulfation) operating conditions. So, the optimum combustor temperature from the point

of view of SO2 retention was 900-925 °C in oxy-fuel combustion mode.

The CO concentration decreased exponentially with rising temperature because the

overall coal combustion process improved as the temperature increased.

Figure 2 also shows a clear dependence of NO and N2O emissions with temperature.

The NO emission increased with temperature up to 900 ºC and then it was almost

constant until 950 ºC. On the contrary, the N2O emission decreased when the combustor

temperature increased.

A number of researchers have attempted to elucidate the effect of temperature on the

N2O emission in coal combustion. Some researchers [15,16] believed that

heterogeneous NO reduction by the char was the major source of N2O formation in the

fluidized bed combustion of coal. Thereby, the emission of N2O would decrease with

decreasing char concentration in the bed as the temperature increased. On the other

hand, other researchers [17,18] suggested that the homogeneous gas-phase reactions of

NCO played an important role in N2O formation. They concluded that the decrease in

the N2O emission at a higher temperature was the result of the NCO oxidation to NO

promoted by the higher temperature, decreasing the amount of NCO available for N2O

formation. In addition, homogeneous reduction of N2O by H or OH radicals was also

7

shown to be important [18,19]. The increasing NO emission with increasing bed

temperature may be due to the fact that a higher bed temperature reduced the char and

CO concentrations in the combustor, decreasing the heterogeneous reduction of NO on

the char surface. In addition, higher temperatures promoted the NCO oxidation to NO.

3.2. Effect of the reaction atmosphere.

To analyze the effect of reacting atmosphere, a series of experimental tests where N2

(O2/N2 =35/65) was fed in the system instead of CO2 was performed. The experimental

operating conditions showed in Table 3 and the temperature range (800-950 ºC) were

kept constants.

As can be seen in Figure 2, comparing the results obtained when working with O2/CO2

(oxy-fuel combustion) and O2/N2 (enriched air combustion), the most important effect

of the CO2 was to shift the optimum temperature for SO2 retention from ~860 ºC to 900-

925 ºC. The CO emission was slightly higher when working under oxy-fuel combustion

conditions than in enriched air combustion conditions. The NO and N2O emissions were

almost the same in both reacting atmospheres, which suggests that nitrogen oxides are

hardly produced from the N2 in air in the experimental conditions used (800-950ºC).

Hosoda and Hirama [20] burned a bituminous coal in a BFB and found the same

influence for N2O emission. However, they observed a slight decrease on NO emission

with increasing CO2 concentration.

3.3. Effect of the gases recycled with the CO2.

In the CO2-rich recycled flue gas, besides CO2 and O2, there are other gases in a lower

concentration that can affect to the pollutant emissions of the combustor. In this section,

a detailed study of the effect of the recycle of different gases, such as SO2, steam, and

NO, both separately and combined, was carried out at two different temperatures, 925

and 850 ºC. The first one was chosen because at this temperature the calcium sorbent

calcines (indirect sulfation) and, in addition, it was the optimum temperature for SO2

retention. The second one was chosen for comparison purposes because at 850 ºC the

calcium sorbent does not calcine (direct sulfation). All other operating conditions were

kept constants.

3.3.1. Effect of SO2 recirculation.

To simulate the flue gas recirculation of SO2, a flow of this gas was injected just above

the gas distribution plate. The pure SO2 flow rate introduced at each operating condition

8

was estimated by a mass balance of the system and it was modified until the feeding

flow had the same SO2 fraction as the recycled flow taking into account the SO2

concentration at the exit of the combustor. Thereby, a flow of SO2 of 2.0 ln/h in the test

carried out at 925 ºC and a flow of SO2 of 7.8 ln/h in the test carried out at 850 ºC were

injected. This corresponded to SO2 concentrations of 900 vppm and 3500 vppm,

respectively, in the inlet gas flow rate.

Figure 3 shows the part of the tests just before and after SO2 injection. Obviously, at

both temperatures, the SO2 concentration in the outlet gas stream increased after its

injection. In a real oxy-fuel combustion system, SO2 is enriched in furnace owing to the

flue gas recirculation and consequently a good sulfation condition is provided. In this

work the SO2 recirculation was simulated by injecting SO2. The SO2 retention with

respect to the sulfur fed by the coal was calculated by means of the following

expression:

100.

/

/(%)

,,0

,2,2,,0

ScoalScoal

outSOoutinSOinScoalScoal

MxF

CQCQMxFSR

(2)

being F0,coal, xS,coal, MS, CSO2,out. and Qout the same as in equation (1), CSO2,in the SO2

concentration in the flue gas coming into the reactor, and Qin the gas flow rate at the

reactor inlet.

It was found that the SO2 retention increased from ~69% to ~81% at 925 ºC and from

~34% to ~39% at 850 ºC because of the higher SO2 concentration in the combustor. It

means that the Ca utilization was higher after SO2 injection. Therefore, although the

SO2 concentration at the outlet of the combustor increases due to the recirculation, the

total SO2 emission is much lower.

In Figure 3 it can be seen that the effect of the SO2 injection on the emissions of the

other gases depended on the temperature. On the one hand, at 850 ºC after SO2 injection

there was an important reduction in the NO concentration and an important increase in

the CO concentration at the outlet of the reactor. However, the NO reduction was lower

and the CO increase was negligible at 925 ºC. On the other hand, the SO2 recirculation

did not affect the N2O emission at both temperatures.

Amand et al. [21] worked at conventional air combustion in a FBC and found a

decrease in the NO emission and an increase in the N2O emission when SO2 was

injected into a sand bed (without limestone). The decrease in the NO emission was

9

explained by an increase in CO concentration, since the CO oxidation was moderately

inhibited by the addition of SO2 [22] and the increase in CO concentration improves the

reduction of NO on the char surfaces [23-25]. Some researchers [26-27] attributed the

inhibition of CO oxidation caused by the presence of SO2 to the decrease in the radical

inventory. Recently, Giménez-López et al. [27] observed that the presence of SO2

inhibited the oxidation of CO for both CO2 and N2 atmospheres and for all

stoichiometries.

3.3.2. Effect of NO recirculation

The effect of NO recirculation was investigated at two different temperatures, 850 and

925 ºC, by injecting different flow rates of NO (2 vol% NO in Ar), which were

calculated to be equivalent to 200, 350, 500 and 625 vppm with respect to the inlet gas

flow rate.

Figure 4 plots the gas distribution before and after the different injections of NO at 925

ºC, and Figure 5 shows the concentrations of NO measured at the combustor exit at both

temperatures. The amount of NO reduced (c in Fig. 5) was calculated by equation (3) as

the difference between the NO introduced in the inlet gas (a in Fig. 5) and the increase

in the NO concentration in the outlet gas due to NO injection (b in Fig. 5).

100.)(

(%),

,,,

recinNOin

outNOrecoutNOoutrecinNOinreduction CQ

CCQCQNO

(3)

being CNO,in rec the inlet NO concentration recirculated into the combustor, CNO,out rec the

the NO concentration in the flue gas at the exit of the reactor in the tests with NO

recirculation, and CNO,out the NO concentration obtained in the flue gas at the exit of the

reactor in the tests without NO recirculation. CNO,out rec and CNO,out were considered as

an average value of the measurements taken during the whole test duration in steady

state conditions.

It was observed that about 60-70% of the NO introduced was reduced to N2 and N2O,

and the reduction rate was almost independent of the initial concentration of NO and

temperature. The formation of N2O was very small, less than 5% of the NO introduced

(see Figures 6 and 7). In addition, neither SO2 nor CO emissions were affected by the

NO injection, as can be observed in Figure 4.

10

Guerrero et al. [28] found that if the amount of char is kept constant, an increase in the

reaction temperature leads to an increase in NO reduction. This behavior was explained

by the fact that thermal desorption is an activated phenomenon and increases with

increasing temperature. However, this phenomenon could not be observed in our

experimental tests because an increase in the temperature led to a decrease in char and

CO concentrations and as a consequence a decrease in the NO reduction. Therefore, it

could be that the opposite effects of the increase of the temperature and the decrease of

the char and CO concentrations compensate each other.

Hosoda and Hirama [20], in a similar test carried out in a BFB at 850 ºC using a

bituminous coal, found that about 80% of the added NO and N2O were reduced to N2

and 4-5% of the added NO was reduced to N2O. These results and the results found in

our tests seem to indicate that heterogeneous NO reduction on the char surface was not

the major source of N2O formation in the fluidized bed coal combustion.

3.3.3. Effect of steam recirculation.

Wet and dry gas recirculations are possible in oxy-fuel combustors. In this section the

effect of H2O(v) (steam) recirculation on the emission of gases was analyzed. The effect

of steam recirculation was simulated by feeding H2O instead of CO2 in the inlet gas

flow. The gas composition fed to the combustor in these tests was 35 vol% O2, 45-50

vol% CO2, and 15-20 vol% H2O.

In all the tests the gas composition was measured after steam condensation. So, to

compare the emissions, the measured gas concentrations for the tests with steam feeding

were corrected to the same outlet gas flow as in the tests without steam addition.

As it can be seen in Figure 8, when steam was introduced a sharp decrease in NO

emission was observed. Therefore, it could be inferred that wet recirculation would

produce a great decrease in the NO concentration in comparison with dry recirculation

in a fluidized bed furnace. Similar results have been observed by other researchers using

different coals [13,20]. It can be also observed in the figure that the change in the CO

concentration was negligible and the N2O concentration slightly increased. Hosoda and

Hirama [20] also observed a slight increase in the N2O concentration with H2O addition.

Regarding to SO2 concentration and thus the limestone sulfation conversion, Wang et al.

[29] found in TGA tests that the presence of steam improved the capacity of sulfation of

the limestones, in some cases up to a relative increase of 86.7%. Similar results were

11

found in a TGA and in a tube furnace by Stewart et al. [30], who added steam and

achieved limestone sulfation conversions up to 98% with sorbent particle sizes between

75 and 115 m and a residence time of 10 h. They suggested that the effect of H2O was

negligible in the kinetic-controlled stage but it was especially important in the diffusion-

controlled stage with an optimum of water concentration around 15-30 vol%. However,

these researchers [13] burnt petcoke in a CFB combustor and found an increase in the

sulfur capture in air-firing mode and a decrease in oxy-fuel firing mode when steam was

added. In our tests the effect of H2O addition (20 vol% at 850 ºC and 15 vol% at 925 ºC

+ 5 vol% generated by the coal combustion) on sulfation conversion was very small,

both at 850 ºC (direct sulfation) and 925 ºC (indirect sulfation). However, it must be

remarked that in this work the tests were carried out with a mean residence time of

solids of ≈ 2 hours, which could be not long enough to notice that effect.

3.3.4. Effect of combined NO and H2O recirculation.

As it was above showed, the H2O recirculation produced an important decrease in the

NO concentration in the outlet gas stream. Therefore, it would be expected a higher NO

reduction when H2O and NO are recycled together in comparison to the NO recycled

without H2O. Figure 6 shows the NO and N2O emissions and Figure 7 shows the

increase of NO and N2O emissions measured with and without H2O recirculation (wet

and dry gas recirculation) as a function of the recirculated NO concentration. It is

observed in these Figures that although the NO emission was lower with H2O

recirculation, the reduction of the NO recycled was almost the same with H2O and

without H2O recirculation. Therefore, it can be concluded that H2O recirculation has

only influence during the stage of NO formation but once NO is formed, the

contribution of the H2O to minimize the NO emission is very small.

3.3.5. Effect of combined H2O/SO2, SO2/NO, and H2O/SO2/NO recirculations.

In an oxy-fuel boiler, the recycled stream, besides CO2 and H2O, contains different

amount of minor gases, which could affect each other. The effect of the addition of

different gas combinations, H2O/SO2, SO2/NO and H2O/SO2/NO, in the inlet gas stream

was analyzed. The global effect on the pollutant emissions was approximately the same

to the tests with the addition of each gas separately and no a synergetic effect among

them was observed.

12

4. Conclusions

Combustion tests at oxy-firing conditions were carried out in a ~3 kWth continuous BFB

combustor with an anthracite coal as fuel and a limestone for sulfur retention. The effect

of the flue gas recirculation on pollutant emissions was analyzed.

A comparison of the results obtained feeding O2/CO2 (oxy-fuel combustion) and O2/N2

(enriched air combustion) as reactant gases showed that the most important effect of

CO2 was that the optimum temperature for SO2 retention shifted from ~860 ºC to 900-

925 ºC. The NO and N2O emissions were almost the same in both reacting atmospheres,

and the CO emission was slightly higher in oxy-fuel combustion than in enriched air

combustion conditions.

In oxy-firing operating conditions it was observed that:

- SO2 recirculation increased the Ca-based sorbent utilization and did not affect N2O

emission. At 850 ºC the CO emission increased and the NO emission decreased,

however, at 925 ºC the CO variation was negligible and the NO reduction was low.

- About 60-70% of the recycled NO was reduced to N2 and N2O. The reduction rate was

almost independent of the initial concentration of NO and temperature, and the

formation of N2O was very small. Less than 5% of the recycled NO was converted to

N2O. Neither SO2 nor CO emissions were affected by the NO recirculation.

- Steam recirculation produced a sharp decrease in NO emission, a slight increase in

N2O emission, and did not affect the CO emission. Steam affected the formation of NO

but once formed it did not have any influence in the NO reduction. In our tests, with a

mean residence time of solids in the combustor of ~2 hours, the effect of steam on

sulfation conversion was very small, at both 850 ºC (direct sulfation) and 925 ºC

(indirect sulfation).

- The combined effect of the recirculation of different gases was almost the same as the

sum of the effects of the recirculation of each gas separately and a synergetic effect

among them was not observed.

Acknowledgements

This work has been supported by Spanish Ministry of Science and Innovation

(MICINN, Project: CTQ2008-05399/PPQ) and by FEDER. M. de las Obras thanks

13

MICINN for a FPI fellowship grant and A. Rufas thanks CSIC for a JAE fellowship

(grant co-funded by the European Social Fund).

References

[1] IPCC, 2005: IPCC Special Report on Carbon Dioxide Capture and Storage. [Metz

B, Davidson O, de Coninck HC, Loos M, and Meyer LA, editors]. Cambridge

University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 442.

[2] Toftegaard MB, Brix J, Jensen PA, Glarborg P, Jensen AD. Oxy-fuel combustion of

solid fuels. Prog Energ Combust Sci 2010;36:581-625.

[3] Scheffknecht G, Al-Makhadmeh L, Schnell U, Maier J. Oxy-fuel coal combustion-A

review of current state-of-the-art. Int J Greenhouse Gas Control 2011;5S:S16-S35.

[4] Jia L, Tan Y, Anthony EJ. Emissions of SO2 and NOx during oxy-fuel CFB

combustion tests in a mini-circulating fluidized bed combustion reactor. Energy

Fuels 2010;24:910-5.

[5] Jia L, Tan Y, Wang C, Anthony EJ. Experimental study of oxy-fuel combustion and

sulfur capture in a mini-CFBC. Energy Fuels 2007;21:3160-4.

[6] Myöhänen K, Hyppänen T, Pikkarainen T, Eriksson T, Hotta A. Near zero CO2

emissions in coal firing with oxy-fuel circulating fluidized bed boiler. Chem Eng

Technol 2009;32:355-63.

[7] Czakiert T, Sztekler K, Karski S, Markiewicz D, Nowak W. Oxy-fuel circulating

fluidized bed combustion in a small pilot-scale test rig. Fuel Process Technol

2010;91:1617-23.

[8] Liljedahl GN, Turek DG, Nsakala NY, Mohn NC, Fout TE. Alstom´s oxygen-fired

CFB technology development status for CO2 mitigation. In: 31st International

Technical Conference on Coal Utilization and Fuel Systems, Clearwater, Florida,

USA, May 21-25, 2006.

[9] Varonen M. 4MWth Oxy-CFB Test Runs. 63rd IEA FBC Metting, Ponferrada,

Spain, November 29-30, 2011

[10] Tan Y, Jia L, Wu Y, Anthony EJ. Experiences and results on a 0.8 MWth oxy-fuel

operation pilot-scale circulating fluidized bed. Appl Eng 2012;92:343-7.

14

[11] Jia L, Tan Y, McCalden D, Wu Y, He I, Symonds R, Anthony EJ. Commissioning

of a 0.8 MWth CFBC for oxy-fuel combustion. Int J Greenhouse Gas Control

2011;doi:10.1016/j.ijggc.2011.10.009.

[12] Lupion M, Navarrete B, Otero P, Cortés VJ. Experimental programme in

CIUDEN´s CO2 capture technology development plant for power generation. Chem

Eng Res Des 2011;89:1494-500.

[13] Stewart MC, Symonds T, Manovic V, Macchi A, Anthony EJ. Effects of steam on

the sulfation of limestone and NOx formation in an air- and oxy-fired pilot-scale

circulating fluidized bed combustor. Fuel 2012;92:107-15

[14] de Diego LF, Rufas A, García-Labiano F, de las Obras-Loscertales M, Abad A,

Gayán P, Adánez J. Optimum temperature for sulphur retention in fluidised beds

working under oxy-fuel combustion conditions. Fuel 2012;

doi:10.1016/j.fuel.2012.02.064.

[15] Amand L-E, Leckner B, Andersson S. Formation of N2O in circulating fluidized

bed boilers. Energy Fuels 1991;5:815-23

[16] Goel S, Zhang B, Sarofim AF. NO and N2O formation during char combustion: is

it HCN or surface attached nitrogen?. Combust Flame 1996;104:213-7

[17] Oude Lohuis JA, Tromp PJJ, Moulijn JA. Parametric study of N2O formation in

coal combustion. Fuel 1992;71:9-14

[18] Kilpinen P, Hupa M. Homogeneous N2O chemistry at fluidized bed conditions: a

kinetic modeling study. Combust Flame 1991;85:94-104

[19] Kramlich JC, Cole JA, McCarthy JM, Lanier WS, McSorley JA. Mechanisms of

nitrous oxide formation in coal flames. Combust Flame 1989;77:375-84

[20] Hosoda H, Toshimasa H. NOx and N2O emission in bubbling fluidized-bed coal

combustion with oxygen and recycled flue gas: Macroscopic characteristics of their

formation and reduction. Energy Fuels 1998;12:102-8.

[21] Amand L-E, Leckner B, Dam-Johansen K. Influence of SO2 on the NO/N2O

chemistry in fluidized bed combustion: 1. Full-scale experiments. Fuel 1993;72:557-

64

15

[22] Dam-Johansen K, Amand L-E, Leckner B. Influence of SO2 on the NO/N2O

chemistry in fluidized bed combustion: 2. Interpretation of full-scale observations

based on laboratory experiments. Fuel 1993;72:565-71

[23] Chan LK, Sarofim AF, Beér JM. Kinetics of the NO-carbon reaction at fluidized

bed combustor conditions. Combust Flame 1983;52:3745

[24] Johnsson JE, and Dam-Johansen K. Formation and reduction of NOx in a fluidized

bed combustor. In: Anthony EJ, editor. Proceedings of the 11th International

Conference on Fluidized Bed Combustion, ASME, New York, 1991, pp. 1389-96

[25] Furusawa T, Tsunoda M, Tsujimura M, Adschiri T. Nitric oxide reduction by char

and carbon monoxide: Fundamental kinetics of nitric oxide reduction in fluidized

bed combustion of coal. Fuel 1985;64:1306-9

[26] Wójtowicz MA, Pels JR, Moulijn JA. N2O emission control in coal combustion.

Fuel 1994;73:1416-22

[27] Giménez-López J, Martínez M, Millera A, Bilbao R, Alzueta MU. SO2 effects on

CO oxidation in a CO2 atmosphere, characteristic of oxy-fuel conditions. Combust

Flame 2011;158:48-56.

[28] Guerrero M, Millera A, Alzueta MU, Bilbao R. Experimental and kinetic study at

high temperatures of the NO reduction over eucalyptus char produced at different

heating rates. Energy Fuels 2011;25:1024-33.

[29] Wang C, Jia L, Tan Y, Anthony EJ. The effect of water on the sulphation of

limestone. Fuel 2010;89:2628-32

[30] Stewart MC, Manovic V, Anthony EJ, Macchi A. Enhancement of indired

sulphation of limestone by steam addition. Environ Sci Technol 2010;44:8781-6

16

Captions for Tables and Figures

Table 1. Analysis of anthracite coal.

Table 2. Granicarb limestone characteristics.

Table 3. Feeding rates of solids and flow rate of gases used in the tests.

Figure 1. Oxy-fuel Bubbling Fluidized Bed Combustor at ICB-CSIC. Measurements of

temperature (T) and pressure (P).

Figure 2. Effect of temperature on SO2, CO, NO, and N2O emissions. Empty symbols:

enriched air conditions (O2/N2= 35/65). Filled symbols: oxy-fuel conditions (O2/CO2=

35/65)

Figure 3. Effect of SO2 recirculation on SO2, CO, NO, and N2O emissions.

Figure 4. Typical test showing the effect of NO recirculation on the SO2, CO, NO, and

N2O emissions. T=925ºC.

Figure 5. NO emission and reduction as a function of the recirculated NO

concentration. a= recirculated NO (vppm), b= increase in NO concentration at the outlet

due to NO recirculation (vppm), c= decrease of the recirculated NO (vppm)

Figure 6. NO and N2O emissions as a function of the recirculated NO concentration for

different inlet gas composition. Empty symbols: 850 ºC. Filled symbols: 925 ºC.

Figure 7. Increase of NO and N2O emissions as a function of the recirculated NO

concentration for different inlet gas compositions. Empty symbols: 850 ºC. Filled

symbols: 925 ºC.

Figure 8. Effect of steam recirculation on SO2, CO, NO, and N2O emissions.

Continuous lines = measured values (dry basis). Dotted lines = corrected values (wet

basis), Ci,out(wet)=Ci,out(dry)/(1-CH2O,in)

17

Table 1. Analysis of anthracite coal.

Proximate analysis (wt %)Moisture 2.3 Ash 31.7 Volatiles 5.6 Fixed C 60.4 Ultimate analysis (wt %, daf)C 90.36 H 2.53 N 1.41 S 2.30 LHVa (kJ/kg) 21807

a Lower Heating Value.

Table 2. Granicarb limestone characteristics.

Composition (wt %) CaCO3 97.1 MgCO3 0.2 Na2O 1.1 SiO2 <0.1 Al2O3 <0.1 Fe2O3 <0.1 Porosity (%)

Raw 3.7 Calcinedb 49

b Calcined in N2 atmosphere at 900 ºC for 10 minutes.

Table 3. Feeding rates of solids and flow rate of gases used in the tests.

Qin (lN/h) 2230 F0,coal (g/h) 576 ± 14 F0,limestone (g/h) 82 ± 2 F0,sand (g/h) 900 ± 20

18

Coal +

LimestoneSilicasand

Bubbling fluidized bed

Freeboard

Drainage deposit

Cyclone

Heat exchanger

Water-cooled screw-feeders

O2

T3

T2

T4

T5P3

AnalyzerCO2, CO, SO2, O2

Stack

Cool water

Hot water

T1

P1

Preheater

Air

CO2

NO

H2O

Analyzer (FTIR) NO, N2O, NO2

Evaporator

SO2

P2

Coal +

LimestoneSilicasand

Bubbling fluidized bed

Freeboard

Drainage deposit

Cyclone

Heat exchanger

Water-cooled screw-feeders

O2

T3

T2

T4

T5P3

AnalyzerCO2, CO, SO2, O2

Stack

Cool water

Hot water

T1

P1

Preheater

Air

CO2

NO

H2O

Analyzer (FTIR) NO, N2O, NO2

Evaporator

SO2

P2

Figure 1. Oxy-fuel Bubbling Fluidized Bed Combustor at ICB-CSIC. Measurements of

temperature (T) and pressure (P).

19

Temperature (ºC)

800 825 850 875 900 925 950 975 1000

SO

2 (v

pp

m,

dry

bas

is)

0

500

1000

1500

2000

2500

Temperature (ºC)

800 825 850 875 900 925 950 975 1000

CO

(vp

pm

, d

ry b

as

is)

0

200

400

600

800

1000

Temperature (ºC)

800 825 850 875 900 925 950 975 1000

N2O

(vp

pm

, d

ry b

asis

)

0

50

100

150

200

Temperature (ºC)

800 825 850 875 900 925 950 975 1000

NO

(vp

pm

, d

ry b

as

is)

0

200

400

600

800

Figure 2. Effect of temperature on SO2, CO, NO, and N2O emissions. Empty symbols:

enriched air conditions (O2/N2= 35/65). Filled symbols: oxy-fuel conditions (O2/CO2=

35/65)

20

Time (min)

0 20 40 60 80 100 120 140 160

SO

2 (

vpp

m,

dry

bas

is)

0

1000

2000

3000

4000

5000

6000

CO

, N

O,

N2O

(vp

pm

, d

ry b

asis

)

0

200

400

600

800

1000

NO

N2O

SO2

CO

NO

SO2

CO

SO2 recirculation850 ºC

Time (min)

0 20 40 60 80 100 120 140 160 180 200

CO

, N

O,

N2O

(vp

pm

, d

ry b

asis

)

0

100

200

300

400

500

600

700

800

SO

2 (

vpp

m,

dry

ba

sis

)

0

300

600

900

1200

1500

1800

SO2

NO

CO

N2O

SO2 recirculation 925 ºC

Figure 3. Effect of SO2 recirculation on SO2, CO, NO, and N2O emissions.

21

Time (min)

0 20 40 60 80 100

Ga

s co

nc

entr

ati

on

(vp

pm

, d

ry b

asis

)

0

200

400

600

800

1000

1200

SO2

NO

CO

N2O

recirculation 0 vppm

200350

500625

Figure 4. Typical test showing the effect of NO recirculation on the SO2, CO, NO, and

N2O emissions. T=925ºC.

22

NO

co

nce

ntr

atio

n (

vpp

m)

0

200

400

600

800

1000

1200

1400850 ºC

6255003502000

NO recirculated (vppm)

NO decreaseNO measured

925 ºC

6255003502000

NO recirculated (vppm)

a

b

c

Figure 5. NO emission and reduction as a function of the recirculated NO

concentration. a= recirculated NO (vppm), b= increase in NO concentration at the outlet

due to NO recirculation (vppm), c= decrease of the recirculated NO (vppm)

23

NO recirculated (vppm)

0 100 200 300 400 500 600 700

NO

mea

sure

d (

vpp

m)

0

200

400

600

800

1000

1200

NO

N2O

Non NO recirculation NO + SO2

NO + H2O

NO + SO2 + H2O

Figure 6. NO and N2O emissions as a function of the recirculated NO concentration for

different inlet gas compositions. Empty symbols: 850 ºC. Filled symbols: 925 ºC.

24

NO recirculated (vppm)

0 100 200 300 400 500 600 7000

50

100

150

200

250

300

NO

N2O

Non NO recirculationNO + SO2

NO + H2O

NO + SO2 + H2O

N

O a

nd

N

2O(v

pp

m)

Figure 7. Increase of NO and N2O emissions as a function of the recirculated NO

concentration for different inlet gas compositions. Empty symbols: 850 ºC. Filled

symbols: 925 ºC.

25

Time (min)

0 20 40 60 80 100 120

Ga

s c

on

cen

tra

tio

n (

pp

mv)

0

500

1000

1500

2000

2500

SO2

CO

NON2O

H2O recirculation 850 ºC

Time (min)

0 20 40 60 80 100 120

Ga

s c

on

cen

tra

tio

n (

vpp

m)

0

200

400

600

800

1000

1200

SO2

NO

CO

N2O

H2O recirculation 925 ºC

Figure 8. Effect of steam recirculation on SO2, CO, NO, and N2O emissions.

Continuous lines = measured values (dry basis). Dotted lines = corrected values (wet

basis), Ci,out(wet)=Ci,out(dry)/(1-CH2O,in)