2271

Proceedings of the Combustion Institute, Volume 28, 2000/pp. 2271–2278

EXPERIMENTAL INVESTIGATION OF NO FROM PULVERIZEDCHAR COMBUSTION

LARS SKAARUP JENSEN,1 HANS ERIK JANNERUP,1 PETER GLARBORG,2 ANKER JENSEN2

and KIM DAM-JOHANSEN2

F.L. Smidth & Co. A/S

Vigerslev Alle 77

DK-2500 Valby, Denmark2Department of Chemical Engineering

The Technical University of Denmark

Building 229

DK-2800 Lyngby, Denmark

NO formation and reduction during pulverized char combustion in the temperature range 850–1150 �Chave been investigated in fixed-bed combustion experiments. Chars from a high-volatile bituminous coaland an anthracite have been used. Under single-particle conditions the selectivity for NO formation fromcombustion of char from both fuel types lies in the range 65–100%. The NO formation selectivity undersingle-particle conditions was observed to be lowest at 850 �C, to have values close to 100% at 1050 and1150 �C, and to be independent of O2 concentration. When conditions deviate from single-particle con-ditions, net NO formation is significantly lower due to NO reduction taking place simultaneously with NOformation. Rate expressions for NO reduction on char both in the presence and in the absence of O2 havebeen determined. For bituminous coal char, these rates are 10–100 times more rapid than values previouslyreported in literature, but are consistent with reburn-type experiments employing char as fuel. This dis-crepancy is mainly attributed to rapid char deactivation prior to measuring of NO reduction rates inprevious determinations.

Shortly after pyrolysis, the effective NO-char reaction rate for pulverized bituminous coal char in thetemperature range 850–1150 �C has been found to be given by

3m6 14800/T(K)r � 6 •10 •e • [char C] • [NO]NO kg C • s

Introduction

During coal combustion at temperatures below1200 �C, the greatest fraction of fuel nitrogen is typ-ically in the form of char-bound nitrogen. Conse-quently, char-bound nitrogen is the greatest poten-tial source of NO because NOx formation bymechanisms other than fuel-N mechanisms typicallyare insignificant at these low temperatures.

Despite a large number of investigations of NOformation from char-bound N and NO reduction onchar, there are still significant uncertainties in theliterature, particularly regarding the effects of fueltype, combustion temperature, and O2 concentra-tion on NO formation selectivity. The main reasonfor uncertainty is that most experimental investiga-tions report the net amount of NO resulting fromsimultaneously proceeding NO formation and NOreduction reactions. Experimentally determined NOformation selectivities in a wide range of reactortypes vary between 0.1 and 1 [1–3,4 with reference

to 5,6]. A clear understanding of how net NO for-mation from char combustion depends on differentparameters requires a systematic approach in whichthe dependence of NO formation and reduction isanalyzed separately.

Fixed-bed and fluidized-bed char combustion ex-periments have shown that as the char mass burnedin these reactor types increases, net NO formationselectivity from char-N conversion decreases [7,8].The net NO formation selectivity is also observed todecrease as the particle size increases [7]. In bothcases, net NO formation decreases because NO re-duction reactions gain importance as the burnedchar mass and particle size increase. If intrinsic NOformation (i.e., in the absence of significant reduc-tion) is to be measured, small char sample masseswith a small particle size must therefore be em-ployed in experiments.

Thomas [9] provided an overview on how differentparameters affect NO formation from char combus-tion and concluded that NO is the primary char-N

2272 COMBUSTION OF SOLID FUELS

Fig. 1. Schematic representation of reactor section ofexperimental set-up (flows refer to 20 �C and 1 atm).

combustion product, but that it subsequently ispartly reduced on the char surface. However, NOformation and NO reduction have not been sepa-rately quantified. Therefore, it is not clear what theNO formation selectivity from char-N combustion isor how it is affected by char type and by combustionconditions.

It is better understood how NO reduction on charis affected by reaction conditions, although availablerate data are inconsistent with reburn-type experi-ments [10,11], and large differences between therate constants for similar fuels are observed [3].Above 850 �C, the reaction order of the NO-charreaction is considered to be first order with respectto NO concentration [12]. Oxidizing gases such asO2, CO, and H2O affect the NO-char reaction rate[13]. Above 850 �C, the rate-enhancing effect of COseems insignificant [14]. In NO-containing atmo-spheres, virgin char has been observed to have atransient high reactivity toward NO, which rapidlydecreases, but can be partially or completely re-stored by gases such as CO or O2 [14]. This restor-able decrease in reactivity has been attributed to theformation of surface oxides [12,14]. In addition, ir-restorable decreases in reactivity (thermal anneal-ing) have been demonstrated to significantly affectNO-char reaction rates but to be different in naturefrom the effects of heat treatment on char-O2 reac-tion rates [13].

The main aim of this work has been to determinehow the NO formation selectivity from combustionof finely pulverized char is affected by char type andby combustion temperature. In addition, it has been

the goal to determine kinetic expressions describingNO reduction on char under conditions relevant tocombustion applications, and to determine the ef-fects of O2 concentration on both NO formation se-lectivity from char-N combustion and the NO re-duction rate kinetics on char.

Experimental

Varying masses of finely pulverized char mixedwith sand were burned in a laboratory fixed-bed re-actor at set-point temperatures between 850 and1150 �C. In addition, NO reduction rates on charwere measured under O2-deficient conditions.

The reactor set-up used is shown schematically inFig. 1. It is approximately 1 m long and consists ofa quartz glass reactor (with a diameter of about 30mm) in an electrically heated oven. Gas flows maybe admitted and withdrawn from the reactor contin-uously, while solids must be fed and removed batch-wise.

In the char combustion experiments, between 0.1and 20 mg of pulverized coal was mixed with about1 g of quartz sand and admitted to the preheatedreactor in an inert N2 atmosphere. On admission tothe reactor, the coal-sand mixture was heated,whereby pyrolysis took place. Pyrolysis gases werepurged, leaving char in the reactor. After 60 s, O2was admitted to the reactor through valve 5, afterwhich the O2 concentration in the reactor gas in-creased continuously (reaching 75% of the steady-state value in 5 s) while char combustion took place.The carbon and nitrogen contents of chars producedat different temperatures by the method describedabove were assessed in separate experiments, and sothe overall NO formation selectivity could subse-quently be determined by integrating the CO, CO2,and NO concentration-time profiles in the reactoreffluent.

NO reduction rates on char under O2-deficientconditions were measured by admitting coal-sandmixtures to the preheated reactor when passing aNO/N2 mixture through. The NO effluent concen-tration was continuously measured. As a final step inthe experiments, O2 was admitted whereby the charremaining in the char-sand bed was burned out. In-tegration of the CO and CO2 concentration-timeprofiles in the reactor effluent allowed the carbonmass remaining in the reactor to be known at alltimes during the experiments.

Two pulverized coal samples were used for theexperiments: a high volatile bituminous coal and ananthracite. The fuel analyses are shown in Table 1,and the mass-based mean particle size lies in therange 10–20 lm (see Ref. [15] for particle size dis-tribution).

NO FROM CHAR COMBUSTION 2273

TABLE 1Coal analyses

High-VolatileBituminous Coal(Colorado, USA)

Air-Dried

Anthracite(Vietnam)Air-Dried

Net calorific value (MJ/kg) 30.5 33.6Moisture (% w/w) 2.8 0.7Volatile matter (% w/w) 37.8 5.88Ash (% w/w) 5.5 3.8Fixed carbon (% w/w) 53.9 89.6

C (% w/w) 76.3 89.2H (% w/w) 5.15 3.04S (% w/w) 0.75 0.48N (% w/w) 1.54 1.05O (% w/w) 7.97 1.73Mineral matter (% w/w) 6.08 4.07

Fig. 2. Net NO formation selectivity from char com-bustion as a function of burned char mass. Experimentaldata are shown as points and the lines are two model fits(reactor set-point temperatures: 850, 1050 and 1150 �C;steady-state O2 concentration: 10%; gas flow: 1 L/min (at20 �C and 1 atm)).

Experimental Results

Combustion Experiments

Figure 2 shows how the net NO formation selec-tivity from char combustion depends on the amountof char burned, the temperature, and the char type.Both experimental data and fits for two models to be

presented below are shown. At single-particle con-ditions (i.e., the limit of zero char mass), the net charN to NO conversion lies between 65 and 100% forboth chars. Net NO formation (per char mass) de-creases as the amount of char in the reactor in-creases.

Model representationTwo different model representations were used in

the interpretation of the experimental data. Themodels facilitate data interpretation and allow therate of NO reduction on char during combustion tobe estimated. Both model representations assumethat char-N is liberated from each particle as eitherNO or N2, and that NO reduction on char is firstorder with respect to both NO concentration andchar concentration (mchar/Vf); this is given by theexpression:

mcharr � k • • [NO] (1)NO,red NO

Vf

where kNO is the NO reduction rate constant (in m3

kg�1 s�1), mchar is the char mass, and Vf is the freevolume in the char-sand bed. [NO] is the NO con-centration in the bulk gas.

Model expressions representing two extremes areemployed: (a) slow char conversion evenly throughthe char-sand bed, and (b) rapid char conversion ata reaction front. In the case of char conversionevenly through the char bed, net NO formation fromthe combustion of a batch sample of char burned ina fixed bed is given by

gNO,0g �NO

mchar,0k •NO

m0

�k m /m (1�X)1 NO char,0 01 � e• •dX (2)� � �0 1 � X

where mchar,0 is the initial char mass in the reactor;gNO,0 is the NO formation selectivity under single-particle conditions; X, over which integration takesplace, is the degree of char conversion; and m0 is thevolumetric gas flow rate through the char-sand bed.

If char conversion takes place at a reaction front,net NO formation is given by

�k m /mNO char,0 01 � eg � g • (3)NO NO,0

mchar,0k •NO

m0

The derivations of equations 2 and 3 can be foundelsewhere [15].

The dimensionless variable kNO • mchar,0/m0 ac-counts for effects of variation in the three variables:the kinetic rate constant (kNO), the initial char mass(mchar,0), and the volumetric flow rate (m0). As the

2274 COMBUSTION OF SOLID FUELS

Fig. 3. Effect of nominal (steady-state) O2 concentrationon net char-N to NO conversion from bituminous charcombustion at low char masses. A steady-state O2 concen-tration of 10% corresponds to an average O2 concentrationduring combustion of 4%, and a steady-state concentrationof 1% to 0.2%.

value of kNO • mchar,0/m0 increases, the net NO for-mation decreases for both of the above expressions.gNO,0 is the only other parameter.

From both equations 2 and 3, it may be shownthat increasing the initial char mass (mchar,0) reducesnet NO formation, and that in the limit of zero charmass, the NO formation selectivity under single-par-ticle conditions (gNO,0) is obtained.

TemperatureAt single-particle conditions, raising the combus-

tion temperature increases net NO formationslightly, whereas increasing temperature significantlydecreases net NO formation when the char contentin the reactor is greater (see Fig. 2).

At large reactor char contents (2–16 mg C), re-duced net NO formation with increased combustiontemperature can be accounted for by a higher rateof NO reduction on char. Increasing temperatureaffects two parameters in the dimensionless variablekNO • mchar,0/m0: the NO reduction-rate constant rap-idly increases as the temperature increases, and thevolumetric flow rate increases (as given by the idealgas law). Because the effect on the NO reduction-rate constant is the dominating effect on tempera-ture, increasing temperature results in an increasein the value of kNO • mchar,0/m0 and thereby a de-crease in net NO formation as the reactor tempera-ture is raised (according to the model equations).

RankThe data show that coal rank has a minor influence

on net NO formation selectivity at single-particle

conditions. At large char contents in the reactor (2–16 mg C), the bituminous coal char is observed toyield significantly lower net char-N to NO conver-sion than the anthracite char (at corresponding ex-perimental conditions). The reason is that the NOreduction-rate constant is much smaller for anthra-cite char than for bituminous coal char. Because therate of NO reduction on char typically decreases aschar rank increases, the value of kNO • mchar,0/m0 alsobecomes smaller, thereby increasing net NO for-mation as char rank increases.

O2 concentrationFigure 3 shows a plot of the net NO formation

selectivity versus bituminous coal char mass fed tothe reactor at both low and high O2 concentrations.Both at 1050 �C and at 1150 �C, no significant dif-ferences between the data from experiments per-formed with average O2 concentrations of 0.2% and4% (i.e., a factor 20) are evident (see Ref. [15] foractual O2 profiles). Qualitatively, the same resultswere observed for the anthracite char. It can be con-cluded that the O2 concentration has no significanteffect on NO formation selectivity from coal charcombustion at single-particle conditions.

NO reduction under O2-free conditionsIn the experiments made to determine NO re-

duction rates on char under O2-free conditions, sev-eral processes occur simultaneously. Sample heatingaccompanied by pyrolysis takes place at the begin-ning of the experiments when pulverized coal is ad-mitted directly into the heated reactor throughwhich NO-containing gas passes. Char is also con-sumed by reaction with NO, and char deactivationmay take place. A result of the relatively slow sampleheating and rapid deactivation is that it is difficult todetermine the rate of the NO-char reaction imme-diately after pyrolysis (i.e., before significant deac-tivation has taken place).

Figure 4 shows an estimate of the time variationof the temperature in the center of the char-sandbed and the experimental results on which it isbased. Forty-five seconds after sample admission,the temperature at the coldest position in the char-sand bed (in the center) is within 40 �C of the steady-state temperature.

Figure 5 shows the rate constant for NO reductionon char calculated by the expression:

m [NO]0 0k � • ln (4)NO � �m [NO]char

where mchar and [NO] are the char-C mass and re-actor effluent NO concentration at a given time,

NO FROM CHAR COMBUSTION 2275

Fig. 4. Plot showing calculated estimates of the centretemperature in the char/sand bed vs. time after sampleadmission to the reactor during experiments at 850 �C and1050 �C. The experimentally measured temperature versustime profile in a corresponding sand bed without gas flow(forced convection), which has been used as a basis for theestimation, is included.

Fig. 5. Variation of reaction rate constant for NO re-duction on bituminous coal/char at 850 �C with time aftersample admission to reactor (the two curves indicate re-peatability of the experiments with 500 ppm NO in thereactor inlet gas).

Fig. 6. Arrhenius plot showing comparing rate constantsfor NO reduction on char determined by the 2 experimen-tal methods used in this work (I, II), with rate constantsdetermined from the literature (III–V). The rate 45 s aftersample admission is plotted representing O2 free condi-tions, I, while the range delimited by the values calculatedby equations 2 and 3 are plotted for O2 rich conditions, II.

m0 is the volumetric gas flow rate through the reactor,and [NO]0 is the inlet NO concentration.

The rate constant for NO reduction on char is seento vary significantly with time (Fig. 5). Within 5 min,deactivation of the char causes a drop in the NOreduction rate constant by about a factor 10. Theresults show that it is very important to consider chardeactivation during experiments. After 20 s, effectsof deactivation dominate over the effects of pyrolysisand sample heating on the NO reduction rate, andthe calculated rate constant decreases.

Because the NO reduction rate in the first fewseconds after pyrolysis typically is relevant for NOx

from pulverized coal combustion, determination ofthe NO-char reaction rate immediately after pyrol-ysis is of prime interest. It has been estimated by

calculating the value 45 s after coal admission to thereactor. This time has been chosen in order to ensurethat the transient effects of sample heating and py-rolysis essentially have been eliminated (while ac-knowledging that some deactivation will have takenplace). After 45 s, the deviation of the bed tempera-ture from the steady-state temperature is estimatedto be less than 40 �C (Fig. 4), which implies that thebulk of pyrolysis also is complete within 45 s.

The reactor effluent CO concentration versus timeprofiles from the experiments show that the bulk ofpyrolysis is complete within 10 s after sample ad-mission to the reactor. In the NO-char reaction ex-periments from which data are depicted in Fig. 5,the CO concentration was about 50 ppm 45 s afterthe sample addition to the reactor (30 ppm of thisCO is generated by the NO-char reaction) and 20ppm after 5 min (with negligible CO generated bythe NO-char reaction). Later experiments that wehave performed show that NO consumption by theNO-CO reaction is minor at these CO concentra-tions, which is in accordance with the literature re-ferred to in the Introduction [14].

Figure 6 shows Arrhenius plots comparing rate ex-pressions for the NO-char reaction determined inthis work and based on literature data. The rate ex-pressions based on this work include (for both bi-tuminous and anthracite char plotted as a functionof steady-state temperature) the following:

2276 COMBUSTION OF SOLID FUELS

I. The values 45 s after sample admission in exper-iments with O2-free conditions

II. The calculated range under O2-rich conditions(with simultaneous combustion) calculated byfitting equations 2 and 3 to the NO data fromthe char combustion experiments

Following are the three groups of rate expressionsbased on literature data:

III. For bituminous coal chars under O2-free con-ditions in fixed-bed reactors at temperaturesrelevant to fluidized-bed combustion [16–18]

IV. For lignite chars under O2-free conditions in en-trained-flow reactors at temperatures relevantto pulverized coal firing [19,20] (correspondingdata for bituminous coal char have not beenavailable)

V. Calculated in the present work based on reburn-type experiments in an entrained-flow reactorusing bituminous coal char [21]; the uncertaintylimits reflect the range of values encompassedby calculations assuming NO reduction on charto be first order with respect to both char con-centration and NO concentration, and with ei-ther instantaneous combustion of char corre-sponding to the amount of O2 present (formingan amount of NO corresponding to the N con-tent in the burned char) or no conversion ofchar (i.e., without any NO formation during theexperiments)

For bituminous coal char, the data from this workshow that the NO reduction rate under O2-free con-ditions (Ia in Fig. 6) is congruent with the lowerrange limit for NO reduction rate under O2-rich con-ditions (IIa). The upper limit is greater by a factorof 2–4 (for O2-rich conditions). Because some de-activation has taken place before the NO-char re-action rates under O2-free conditions were mea-sured, the results indicate that simultaneouscombustion (or presence of O2) does not have a sig-nificant effect on the NO reduction rate for bitu-minous coal char (within a factor of 2). Fig. 6 showsthat the range of the NO-char reaction rates deter-mined for bituminous coal char in this work (Ia, IIa)lies within the range of values calculated based onreburn-type experiments (V), but is a factor of 10–100 greater than the range of the literature data fromexperiments under O2-free conditions (III, IV).

The NO-char reaction rate 45 s after admission ofthe pulverized bituminous coal sample (i.e., shortlyafter pyrolysis) under O2-free conditions (IIa) is wellrepresented by the expression

3m6 �14800/T(K)r � 6 •10 •eNO kg C • s

• [char C]• [NO] (5)

The reduction rate will be in mol/m3/s if units of

kg C/m3 are used for char-C and mol/m3 are usedfor NO.

Despite the small amount of data representingNO reduction on anthracite char under O2-free con-ditions, it is evident that the NO reduction rate mea-sured under O2-free conditions is significantly lowerthan the range calculated for combustion conditions.The reduction rates measured under O2-free con-ditions are about a factor of 5 smaller than with si-multaneous combustion. This can possibly be ex-plained by the NO-char reaction rate increasingsignificantly as anthracite char conversion proceeds(the reactive surface area of chars has been found toincrease with increasing conversion [22]).

Discussion

The two most remarkable results of the presentwork are that NO formation selectivities close to100% for char-N oxidation have been determined,and that significantly greater NO-char reaction ratesthan generally reported in the combustion literaturehave been found. The first can be explained by NOreduction reactions being essentially eliminated un-der the single-particle conditions in the presentwork, whereas this has not commonly been the casefor the experimental results reported in the litera-ture. The second is discussed below.

Because the employed experimental proceduresvary in the reported experiments from the literature,there may be a large number of factors responsiblefor the observed discrepancies. We attribute the dif-ferences to two main factors:

• Significant deactivation may have taken place priorto measurement of the NO-char reaction rates inthe reported experiments from the literature un-der O2-free conditions (as is evident from Fig. 5).

• Temperature increases during combustion in theexperiments with presence of O2 may result in ac-tual char temperatures during experiments beingsignificantly greater than reactor set-point tem-peratures.

The first of the above is expected to be the mostimportant. This conclusion can be drawn based onthe experiments at O2-deficient conditions (II) be-cause there is no significant increase in temperaturein this case. Furthermore, Fig. 5 clearly shows thatthe activity of char for NO reduction decreases rap-idly with the holding time at high temperature in aNO/N2 environment. The chars used for the exper-iments, from which the rate expressions III (a–d)and IV (a and b) have been calculated, were derivedby heating coals to high temperatures with longholding times, making it probable that their activitiesmay have been reduced severely before NO reduc-tion rates were measured. Reductions in NO-charreactivity over time have been observed before and

NO FROM CHAR COMBUSTION 2277

were attributed to formation of surface oxides[12,14] and to thermal annealing [13].

Besides the above factors, the calculated tempera-ture increases caused by char combustion (as dis-cussed below) can, at most, result in an increase ofthe rate constant by about a factor of 5 (i.e., only afraction of the factor 10–100 discrepancy). Calcula-tion of the particle temperature rise (overshoot) dur-ing char combustion indicates that for 75% of thechar particles, the temperature during combustionis at most 15 �C greater than the char-sand-bed bulktemperature. In the char combustion experimentsperformed, the bulk temperature of the char-sandbed may increase by up to 30 �C per mg of charburned (adiabatic temperature increase), and in thereburn-type experiments [21], calculations show thatan adiabatic temperature increase of up to 170 �Cmay be experienced.

Conclusions

The experimental data show that the intrinsic NOformation selectivity from char-N during pulverizedcoal char combustion at temperatures in the range1050–1150 �C is close to 100% and about 65% at850 �C. Temperature only has a weak effect on theintrinsic NO formation selectivity, and no discernibleeffect of O2 concentration has been found on theNO formation selectivity. Because the bulk of fuel-N is contained in the char, NO formation from char-N combustion is the dominating NO-forming reac-tion during coal combustion at temperatures up to1200 �C.

Reduction of NO by reaction with char is the mainreaction accounting for net NO formation from charcombustion being significantly lower than 100%.The reaction rates determined in this work and therates calculated based on results from reburn-typeexperiments are a factor of 10–100 greater than typ-ical values reported in the literature. The NO re-duction rates on bituminous coal char under O2-de-ficient conditions shortly after pyrolysis seem todiffer by less than a factor of 2 from the NO reduc-tion rates simultaneously with combustion. For an-thracite char, the NO reduction rate shortly after py-rolysis and under O2-deficient conditions is about afactor of 5 slower than the average NO reductionrate simultaneously with combustion.

The results show that when working with NO-charreactions during combustion, it is important to rec-ognize that the NO-char reaction is dynamic in na-ture. Therefore, kinetic parameters determined atquasi-steady-state conditions are not directly appli-cable to most combustion processes. Both the for-mation of reaction inhibiting surface oxides and ther-mal annealing are expected to be important inaccounting for the time dependence of the NO re-duction rate on char.

Acknowledgments

The authors would like to acknowledge funding for thiswork from F. L. Smidth & Co. A/S, the financial sponsorsof the Combustion and Harmful Emission Control re-search program, the Danish Energy Research Program,and the Danish Agency for Trade and Industry under theMinistry of Business and Industry.

REFERENCES

1. Pohl, J., and Sarofim, A., Proc. Combust. Inst. 16:491–501 (1976).

2. Song, Y. H., Pohl, J. H., Beer, J. M., and Sarofim, A.,Combust. Sci. Technol. 28:31–39 (1982).

3. Johnsson, J. E., Fuel 73:1398–1415 (1994).4. Juntgen, H., Erdol und Kohle–Erdgas–Petrochemie

vereinigt mit Brennstoff-Chemie 40:204–208 (1987).5. Schuler, J., “Internal Report: Fundamental Research

BF/BMFT 1986,” thesis, University of Essen, Essen,Germany, 1987.

6. Wang, W., Brown, S. D., Hindmarsh, C. J., andThomas, K. M., Fuel 73:1381–1388 (1994).

7. Ninomiya, Y., Yokoi, K., Arai, N., and Hasatani, M.,Int. Chem. Eng. 29:512–516 (1989).

8. Furusawa, T., Shimizu, T., Kawaguchi, H., Kojima, T.,and Chihara, Y., in International Conference on Coal

Science, Elsevier, Amsterdam, 1987, pp. 853–856.9. Thomas, K. M., Fuel 76:457–473 (1997).

10. Smoot, L. D., Hill, S. C., and Xu, H., Prog. Energy

Combust. Sci. 24:385–408 (1998).11. Østberg, M., Glarborg, P., Jensen, A., Johnsson, J. E.,

Pedersen, L. S., and DamJohansen, K., Proc. Combust.

Inst. 27:3027–3035 (1998).12. Aarna, I., and Suuberg, E. M., Proc. Combust. Inst.

27:3061–3068 (1998).13. Aarna, I., and Suuberg, E. M., Fuel 76:475–491 (1997).14. Aarna, I., and Suuberg, E. M., Energy Fuels 13:1145–

1153 (1999).15. Jensen, L. S., “NOx From Cement Production: Reduc-

tion by Primary Measures,” Ph.D. thesis, TechnicalUniversity of Denmark, Lyndty, Denmark, 1999.

16. Suzuki, Y., Moritomi, H., and Kido, N., “On the For-mation Mechanism of N2O During Circulating Flu-idized Bed Combustion,” Fourth SCEJ Symposium onCirculating Fluidized Beds, 1991.

17. Johnsson, J. E., and Dam-Johansen, K., in Proceedings

of the Eleventh International Conference on Fluidized

Bed Combustion (E. J. Anthony, ed.), ASME, 1991,pp. 1389–1396.

18. De Soete, G. G., Proc. Combust. Inst. 23:1257–1264(1990).

19. Levy, J. M., Chan, L. K., Sarofim, A. F., and Beer,J. M., Proc. Combust. Inst. 18:111–120 (1981).

20. Song, Y. H., Beer, J. M., and Sarofim, A. F., Combust.

Sci. Technol. 25:237–240 (1981).21. Chen, W-Y., and Ma, L., AIChE J. 42:1968–1976

(1996).22. Aarna, I., and Suuberg, E. M., Proc. Combust. Inst.

27:2933–2939 (1998).

2278 COMBUSTION OF SOLID FUELS

COMMENTS

Adel F. Sarofim, University of Utah, USA. Given the highreactivity of oxygen with char relative to that of NO, oneneeds to be concerned with O2 depletion, particularly athigh char loadings. Is this a problem that you have ad-dressed? The apparatus you have developed has flexibilityto vary parameters not possible with entrained flow reac-tors. It would be interesting to increase the particle di-ameter in order to enhance the potential for NO reductionduring diffusion out of a reacting particle.

Author’s Reply. We have observed O2 depletion at highchar loadings in the bed. This was seen when comparingdeconvoluted O2 concentration versus time profiles in thereactor effluent for varying char mass and temperature.During the experiments performed with the smallest charmasses, no O2 depletion was evident, even at 1150 �C. TheO2 depletion was not accurately quantified as a function ofexperimental conditions. The range of rate constants de-termined by the two model representations used to inter-pret the experimental data take O2 depletion into account.

Yes, it would be simple to perform experiments withvarying particle size in the employed experimental appa-ratus in order analyze NO reduction during diffusion outof a reacting particle. However, as the char particle sizeincreases, the temperature rise in the particle during com-bustion will become increasingly significant.

●

Robert Hurt, Brown University, USA. I am aware thatyour laboratory is also working on annealing effects in thecarbon/oxygen reaction. Do you have any information onwhether annealing has a similar effect on the NO/C andthe O2/C reactions? Could perhaps the models currentlyused for annealing in O2/C be applied to NO/C? Also, werethe temperatures chosen representative of those present incommercial cement kilns?

Author’s Reply. We have not directly compared anneal-ing effects on the NO/C reaction with the O2/C reaction.Also, we do not believe that the models currently used forannealing in O2/C to be directly applicable to describingthe loss of apparent char reactivity in NO/C observed inour experiments. As described in the paper, the loss in ap-parent char reactivity in NO/C is the result of both anirrestorable decrease (thermal annealing) and a restorableincrease (possibly due to the formation of surface oxides).Additionally, the temperatures chosen are representativeof those present in calciners of modern commercial kilnsystems for cement production. In rotary kilns, tempera-tures are significantly higher.