Fuel 139 (2015) 321–327

Contents lists available at ScienceDirect

Fuel

journal homepage: www.elsevier .com/locate / fuel

An experimental study of NO reduction by biomass reburning and thecharacterization of its pyrolysis gases

http://dx.doi.org/10.1016/j.fuel.2014.08.0710016-2361/� 2014 Published by Elsevier Ltd.

⇑ Corresponding author. Tel.: +86 10 84915188; fax: +86 10 84934516.E-mail address: zhangfan5188@vip.sina.com (F. Zhang).

Yun Shu, Fan Zhang ⇑, Hongchang Wang, Jinwei Zhu, Gang Tian, Chen Zhang, Yutao Cui, Jiayu HuangResearch Center of Air Pollution Control Technology, Chinese Research Academy of Environmental Sciences, Beijing 100012, China

h i g h l i g h t s

� Biomass type had a significant influence on NO reduction efficiency.� Difference in NO reduction was due to difference in yield of pyrolysis gas of biomass.� Hydrocarbons (mainly CH4) were mainly responsible for NO reduction.

a r t i c l e i n f o

Article history:Received 9 May 2014Received in revised form 14 July 2014Accepted 28 August 2014Available online 10 September 2014

Keywords:BiomassReburningNO reductionPyrolysis gas

a b s t r a c t

The reduction of NO by reburning using three biomass samples (rice husk, sawdust and corncob) wasinvestigated in a horizontal fixed-bed quartz reactor. The temperatures were ranging from 800 to1200 �C. The influence of the oxygen concentration entering the reburning zone, the particle size ofthe biomass and the initial NO concentration on the NO reduction efficiency were studied experimentally.In order to improve the understanding of the relative contribution of each of the pyrolysis gases on theNO reduction, the CO, H2, hydrocarbons (mainly CH4), HCN, and NH3 concentrations in the outlet gas fromthe reburning zone were measured at combustion temperatures of 800–1200 �C. The experimentalresults indicated that the biomass type had a significant influence on the NO reduction efficiency. Themaximum NO reduction efficiency of sawdust reburning (55 ± 2.4%) was much higher than those of ricehusk reburning (43 ± 1.8%) and corncob reburning (44 ± 2.1%). For reburning with sawdust, a highly effi-cient NO reduction was achieved at oxygen inlet concentrations of 0–1 vol%, particle sizes of 160–370 lmand initial NO concentration of 800 ppmv. For the present operating conditions, the difference in NOreduction for the three biomass samples could be ascribed to differences in the yield of pyrolysis gasesthrough the homogeneous reactions. Hydrocarbons (mainly CH4) were the key species for reducing theemissions of NO, whereas CO and H2 had little effect on NO reduction. The sum of the HCN and NH3 con-centrations could reflect the tendency for the conversion of the NO entering the reburning zone into N2.

� 2014 Published by Elsevier Ltd.

1. Introduction

Nitric oxides (mainly NO) are one of the main air pollutantsemitted from coal-fired boilers. They can cause a variety of envi-ronmentally harmful effects such as acid rain, ozone depletionand urban smog. In recent years, various technological approacheshave been used to reduce the nitric oxides emissions from combus-tion systems [1–3]. Reburning is recognized as one of the mostpromising combustion modification technologies for NO control.In the reburning process, a secondary fuel is injected downstreamfrom the main combustion region, to generate a fuel-rich reburnzone inside a furnace, where hydrocarbon radicals and

heterogeneous reactions promote the reduction of NO formed inthe main combustion region to N2. To complete the process,over-fire air is introduced prior to the furnace exit to oxidize car-bon monoxide, hydrogen, and any remaining combustible com-pounds exiting the reburning zone [4–6].

Many types of fuel have been investigated as reburning fuels,including natural gas [7], coal [8] and biomass [9,10]. Comparedwith other fuels, biomass as a reburning fuel has several advantages[11–14]: (1) biomass fuels contain lower levels of sulfur and nitro-gen, suggesting their lower SO2 and NOx emissions, (2) they arerenewable and nearly CO2-neutral fuels, and (3) they have a highercontent of volatile matter, leading to a greater reduction of NOx.

It is well known that NO reduction reactions with solid fuelsinclude homogeneous reactions between NO and volatiles and het-erogeneous reactions between NO and chars. Wendt [15] and

322 Y. Shu et al. / Fuel 139 (2015) 321–327

Mereb [16] hypothesized that the heterogeneous reactions contrib-uted to low levels of NO reduction during the reburning process.The reduction potential of a fuel depended on its ability to producevolatiles that react with NO. Assuming that only homogeneous gas-phase reactions participated in the reduction, the fuel with ahigher volatile matter content would be expected to achieve agreater NO reduction. Cancès et al. [17] also reported that homoge-neous reduction was the most efficient mechanism for reburningprocess. Furthermore, some researchers thought that the goodreburning performance of biofuels could be explained by their highvolatiles yields (mainly hydrocarbon), similarly to the mechanismof natural gas reburning [18]. In addition, the other pyrolysis gases,CO, H2 and some nitrogen-containing species (HCN and NH3), alsoplay an important role in NO reduction [19,20]. Indeed, the pyroly-sis gas from biomass has been shown to reduce NO effectively, anddifferent pyrolysis gases have different NO reduction potentials.Thus, it is thought that biofuels with different compositions underthe pyrolysis conditions in the reburning zone will be expected toproduce different pyrolysis gases that, in turn, can radically changethe NO reburning effectiveness. However, taking into accountexisting studies in the literature, a detailed characterization ofthe pyrolysis gases in the biomass reburning process is absent. Bet-ter knowledge of the relative contribution of each of the pyrolysisgas species to NO reduction is required.

In the present work, the NO reduction performances of differentbiomass samples as reburning fuels (rice husk, sawdust and corn-cob) were evaluated. The effects of crucial parameters, such as theoxygen concentration entering the reburning zone, the particlesizes of the biomass and the initial NO concentration, on thereburning process were analyzed. In the temperature range of800–1200 �C, much attention was paid to the detailed character-ization of the distribution of the key pyrolysis gaseous species(CO, H2, hydrocarbons (mainly CH4), HCN and NH3) for the threebiomass samples during the reburning process.

2. Experimental

2.1. Samples

Three different biomass samples, namely rice husk, sawdust,and corncob, were used in this study. All of the biomass sampleswere ground and sieved to a size range of 160–370 lm before anal-ysis and use in the experiments. The proximate and ultimate anal-yses of the biomass samples are shown in Table 1, according to thetest standard of coal. The composition of pyrolysis products col-lected by biomass is shown in a weight ratio (wt%) in Table 1. Asshown in the table, the biomass samples showed high volatile mat-ter and low N and S contents, and their heating value was in therange from 15.06 to 17.53 MJ/kg. These values indicated that the

Table 1Characteristics of biomass samples.

Samples Rice husk Sawdust Corncob

Proximate analysis (wt%, as air dried)Moisture 4.01 3.34 4.65Ash 15.48 2.98 5.26Volatile matter 64.43 74.38 72.09Fixed carbon 16.08 19.30 18.00

Ultimate analysis (wt%, as air dried)C 39.47 45.56 43.70H 4.85 5.57 5.27O 35.02 38.59 39.67N 0.48 1.28 0.54S 0.04 0.03 0.06LHV (MJ/kg) 15.06 17.53 16.38

samples were ideal renewable energy resources with minor poten-tial for environmental pollution. The ash analysis of the biomasssamples was performed using X-ray fluorescence spectroscopy(XRF) (results shown in Table S1). These results were reported onan oxide basis as a percentage of the ash mass. The main inorganicelements in the biomass samples were Na, K, Si, etc. The resultsshowed that sawdust had the highest amount of Na2O, whereasrice husk had the highest amount of SiO2, and corncob showedthe highest K2O.

2.2. Experimental system and procedures

The experimental setup was composed of a reaction system, agas feed system and a continuous analysis system, as illustratedin Fig. 1.

An annular electric furnace with a temperature control unit wasused to simulate the reaction conditions in the reburning zone. APt-Rh thermocouple measured the temperature in the electric fur-nace, and the precision of the temperature control was ±1 K. Thereburning experiments were performed in a horizontal fixed-bedquartz reactor with a length of 1200 mm and an inner diameterof 60 mm. The reactor was heated by the electric furnace. The reac-tor could withstand temperatures as high as 1250 �C, and theheated zone was 800 mm in length. The reaction temperature inthe reburning zone was controlled in the range of 800 to 1200 �C.An example of the temperature profile obtained for differentsystem temperatures is shown in Fig. 2. As shown, the shape ofthe temperature profiles was very similar for all of the system tem-peratures. There was a central zone (approximately 500 mm inlength) where the temperature could be considered fairly uniform.In the gas feed system, the high-purity N2, CO2, O2, and 2 vol% NOwith N2 gases from different gas cylinders first passed through themass flow controllers before entering a mixing chamber, where thedifferent gases were mixed prior to being fed into the horizontalquartz reactor as the simulated flue gas. Water vapor wasgenerated by passing N2 through a heated gas-wash bottle contain-ing deionized water. The volume fractions of CO2, O2, NO and H2Oin the simulated flue gas were 15 vol%, 0–4 vol% (when used),200–1200 ppmv (when used) and 6 vol%, respectively, and theseconcentrations were chosen according to measurements at the exitof the primary zone of a coal-fired boiler. The balance gas was N2.The flue gas residence time in the reburning zone was estimated tobe approximately 600 ms, corresponding to total flue flow ratesbetween 1500 and 2000 NL/h (corresponding to 101,325 Pa and0 �C on a dry basis) for reaction temperatures between 800 and1200 �C. The typical reburning zone parameters are shown in Table2. The reburning experiments were performed under atmosphericpressure. Before each experiment, a ceramic boat containing thebiomass sample (approximately 2 g) was placed in the cold sideof the reactor. Then, the reactant gases were flowed through thereactor, which was already preheated to the desired temperature,to measure the gas compositions. The biomass sample with aceramic boat was then rapidly introduced into the reactor by apush–pull rod to start the reburning experiment after the detectedconcentrations of the reactant gases were stable. The biomass sam-ples combusted, and the volatiles emitted from biomass couldreact with the flue gas to accomplish the reburning process.

The concentrations of CO, NO, CH4, HCN and NH3 in the outletgas from the reburning zone were continually measured by a Fou-rier transform infrared (FTIR) analyzer system (GASMET DX-4000),produced by Gasmet Technologies, Inc. in Finland. The O2 and NOconcentrations were measured using an electrochemical gasanalyser (Testo 340), and the concentrations of hydrocarbons(CH4, C2H2, C2H4 and C2H6) and hydrogen were measured by gaschromatography (Agilent Technologies, 3000 Micro GC) with aflame ionization detector (FID) and a thermal conductivity detector

gas feed system reaction system continuous analysis system

N2 O2 NO CO2

2

3

4

78

1

5 6

9

Fig. 1. Schematic diagram of the experimental system. (1) valve, (2) mass flow controller, (3) mixing chamber, (4) annular electric furnace, (5) horizontal fixed-bed quartzreactor, (6) ceramic boat, (7) push–pull rod, (8) filter, (9) gas analyzer.

0 200 400 600 800 1000 1200200

400

600

800

1000

1200

Tem

pera

ture

(o C)

Reactor length (mm)

800 oC 900 oC 1000 oC 1100 oC 1200 oC

Fig. 2. Temperature profiles inside the reburning reactor.

Table 2Typical reburning zone parameters.

Reburning zoneparameters

Values with the units

Pressure 101,325 PaResidence time 600 msInitial NO concentration 200, 300, 500, 800, 1200 ppmvO2 inlet concentration 0, 1, 2, 4 vol%Reburning temperature 800, 850, 900, 950, 1000, 1050, 1100, 1150,

1200 �C

Y. Shu et al. / Fuel 139 (2015) 321–327 323

(TCD). The measurement accuracies for CO, NH3 and HCN were ±2 -vol%, whereas those for NO and CH4 were ±1 vol%. The minimumdetectable concentrations for the hydrocarbon species (CH4, C2H2,C2H4 and C2H6) and hydrogen were 1 and 5 ppm, respectively.The FTIR analysis system was suitably configured for the NO con-centrations observed in this study, and the time interval of sam-pling was 1 s. A filter was connected to the sample line toremove fine particles from the flue gas. The temperature of thegas sample was maintained above 180 �C along the transport lineusing an internal heating element in the transport pipe of the FTIRanalysis system. Condensation and dissolution of the samples inwater were avoided. For the hydrogen and hydrocarbon speciesmeasurements, the gases were collected in an airtight bag andwere later analyzed using the gas chromatography. A good agree-ment between the GC and FTIR analyzers was found for the com-pounds measured by both techniques. The data of gasconcentration and NO reduction obtained from the reburningexperiments were a dynamic varying result (as shown in Fig. S1),since the delivery of reburning fuel was not continuous. Therefore,

the NO reduction efficiency and gas outlet concentration at eachreburning temperature were calculated by integrating the NOreduction efficiency and gas outlet concentration during the reac-tion time. Several parallel experiments were carried out to ensurethe reliability, and the standard deviation was ±5%. The calculationprocess was as follows:

gNO ¼uðNOÞin � 1

n

Pna¼0uðNOÞout;a

uðNOÞin� 100%

CT ¼1n

Xn

a¼0

Cout;a ða ¼ 0;1;2;3 . . . nÞ

where gNO is the NO reduction efficiency in %; u(NO)in is the NOinlet concentration in ppmv; u(NO)out,a is the NO outlet concentra-tion at each test point in ppmv; n is the total number of test points;CT is the gas outlet concentration in ppmv; and Cout,a is the gas out-let concentration at each test point in ppmv. For the present system,a burnout zone was not incorporated due to the very limited resi-dence time in the horizontal fixed-bed quartz reactor, and in thisrespect the reactor design was similar to the systems used by otherresearchers to investigated fundamental aspects of reburning[8,21]. Therefore, the NO reductions measured at the exit of thereburning zone were likely to be higher than if a burnout zonewas included (as in a practical system), since char nitrogen and vol-atile nitrogenous intermediates (e.g. HCN, NH3) not consumed inthe reburn zone might be oxidized to NO in the burnout zone [20].

3. Results and discussion

3.1. Effects of biomass type on NO reduction efficiency

The effects of the biomass type and reburning temperature onNO reduction are presented in Fig. 3. Error bars in this and all fig-ures represent the standard deviation of the three replicates. Thedata showed that the NO reduction efficiencies of all three biomasssamples followed a pattern of first increasing and then decreasingwith an increase in the reburning temperature. There was an opti-mum temperature window for the biomass reburning. At a rela-tively low temperature range, the herbaceous biomasses, such asrice husk and corncob, showed better performance compared tothat of sawdust. This different might be attributed to the lowerignition temperature and bigger volatiles combustion exothermiccapacity of the herbaceous biomass samples [22]. In contrast, ata relatively high temperature range, the woody sawdust biomassexhibited the best NO reduction. This finding suggested that thebiomass type had a significant influence on the NO reduction

800 900 1000 1100 12000

10

20

30

40

50

60

NO

redu

ctio

n ef

ficie

ncy

(%)

Temperature (oC)

inlet O2 =0% inlet O2 =1% inlet O2 =2% inlet O2 =4%

Fig. 4. Effect of the O2 inlet concentration on NO reduction efficiency. Experimentalconditions: residence time = 600 ms, initial NO = 800 ppmv, H2O = 6 vol%. Biomasssample: sawdust.

324 Y. Shu et al. / Fuel 139 (2015) 321–327

efficiency. The maximum NO reduction efficiency of sawdustreburning (55 ± 2.4%) was much higher than those of rice huskreburning (43 ± 1.8%) and corncob reburning (44 ± 2.1%). Zhanget al. [23] thought that the pyrolysis products of CO and H2 frombiomass increased with an increasing in the temperature from800 to 900 �C, while the CH4 increased first and then reduced athigher temperatures. These effects would significantly influencethe NO reduction efficiency during the biomass reburning process.The variation tendency of the NO reduction efficiency with thereburning temperature might be ascribed to the different pyrolysisgases from the biomass sample. The contributions of pyrolysisgases to the NO reduction efficiency will be discussed in detail inthe following section.

3.2. Effect of the oxygen concentration entering the reburning zone onNO reduction efficiency

According to the literature, the oxygen entering the reburningzone is a very important parameter that influences NO reduction,and it is related to the excess air in the primary combustion zone[24]. Taking the sawdust sample as an example, the effect of theoxygen inlet concentration on NO reduction is shown in Fig. 4.Upon increasing the oxygen inlet concentrations from 0 to 1vol%, the sawdust showed a slight increase in its NO reduction effi-ciency under the different reburning temperatures. However, fur-ther increasing the oxygen inlet concentration clearly decreasedthe NO reduction, especially at a relatively high temperature range.The maximum NO reduction efficiency was obtained at an oxygeninlet concentration of 1 vol% during biomass reburning. Similarresults were also reported by Zhang et al. [25] and Badzioch andHawksley [26]. These researchers found that, when the oxygenconcentration was at a low level of approximately 1 vol%, the pyro-lysis reactions were similar to those induced in a reducing atmo-sphere, which was beneficial for the reduction of NO. Thus, ahighly efficient NO reduction could be achieved for sawdustreburning with an oxygen inlet concentration in range of 0 to1 vol%.

3.3. Effect of particle size on NO reduction efficiency

Fig. 5 shows the effect of the sawdust particle size on the NOreduction efficiency. It could be found that the NO reduction effi-ciency increased with a decrease in the sawdust particle size atthe same reburning temperature. When the particle size rangedecreased from 370–750 to 160–370 lm, the maximum reductionefficiency of sawdust increased from 45 ± 1.6% to 54 ± 2.7%. How-ever, with a further decrease in the particle size to less than

800 900 1000 1100 12000

10

20

30

40

50

60

NO

redu

ctio

n ef

ficie

ncy

(%)

Temperature (oC)

rice husk sawdust corncob

Fig. 3. Variation of NO reduction efficiencies with different biomass samples.Experimental conditions: residence time = 600 ms, initial NO = 800 ppmv, inletO2 = 0 vol%, H2O = 6 vol%.

160 lm, the maximum efficiency was not clearly increased, andthe corresponding value was 56 ± 2.1%. Finer biomass particlesmake it easier for volatiles to be released, which is beneficial forthe NO reduction. However, if the particle size of the biomasswas decreased past a certain threshold, this promoting effectwould fade [27]. Meanwhile, other factors, such as high energyconsumption for milling, would become the restrictions on reburn-ing process when the particle size of the biomass was decreased toa certain degree. Therefore, for the present study, the particle sizeof the biomass < 160 lm provided some minor benefits in terms ofNO reduction but at the cost of higher energy consumption formilling. The effect of initial NO concentration on NO reduction effi-ciency was also investigated (as shown in Fig. S2). At a reburningtemperature range of 800 to 1200 �C, increasing the initial NO con-centration from 200 to 800 ppmv enhanced the NO reduction effi-ciency of sawdust. A further increase in the initial NOconcentration from 800 to 1200 ppmv lowered the NO reductionefficiency. This suggested that there was an optimal initial NO con-centration for the sawdust reburning. When the initial NO concen-tration was 800 ppmv, the maximum reduction efficiency ofsawdust was 55 ± 2.3%. The effects of the oxygen inlet concentra-tion, particle size and initial NO concentration on the NO reductionefficiencies of rice husk and corncob were also investigated, andthe results were similar to those obtained with sawdust and arethus not shown here.

800 900 1000 1100 12000

10

20

30

40

50

60

NO

redu

ctio

n ef

ficie

ncy

(%)

Temperature (oC)

<160 µm 160-370 µm 370-750 µm

Fig. 5. Effect of the particle size on NO reduction efficiency. Experimentalconditions: residence time = 600 ms, initial NO = 800 ppmv, inlet O2 = 0 vol%,H2O = 6 vol%. Biomass sample: sawdust.

Y. Shu et al. / Fuel 139 (2015) 321–327 325

3.4. Influence of biomass type on the CO, H2, hydrocarbons (mainlyCH4), HCN and NH3 emissions from reburning zone

It is well known that the pyrolysis gases in the volatiles have amajor contribution to biomass reburning through heterogeneousreactions, and different pyrolysis gases have different NO reductionpotentials. The composition of the pyrolysis gas in the volatiles isgreatly influenced by the choice of reburning fuels and their pyro-lysis characteristics. Additionally, the present study found that thebiomass type had a significant influence on the NO reduction effi-ciency (as shown in Fig. 3). In the following sections, the variationsin the CO, H2, hydrocarbons (mainly CH4), HCN and NH3 emissionsas function of temperature for the three biomass samples wereevaluated to study the relative contribution of pyrolysis gases toNO reduction. As shown in Fig. 6, the CO concentrations in the out-let gas from the reburning zone for the three biomass samplesincreased initially as the temperature was increased from 800 to1000 �C. According to the literature [24,25], this increase mightbe ascribed to the predominance of biomass pyrolysis with anincrease in the reburning temperature, which resulted in the pro-duction of hydrogen and carbon that would react with CO2 to yieldCO. A further increase in the temperature from 1000 to 1200 �Cresulted in nearly constant CO concentrations. In the reburningheterogeneous reaction, CO reacts with NO on the char surfacevia [28,29]

2NOþ 2CO !char2CO2 þ N2 ð1Þ

Among the three biomass samples, the lowest amount of COwas observed for the sawdust reburning throughout the tempera-ture range from 800 to 1200 �C, but the sawdust exhibited the bestperformance for NO reduction at relatively high temperatures (asshown in Fig. 3). The results indicated that the heterogeneous reac-tion between CO, NO and the char had a minor contribution towardNO reduction under present experiment conditions.

As shown in Fig. 7, the H2 emissions for the three biomass sam-ples monotonously increased in the temperature range from 800 to1200 �C. This result was similar to the experimental results fromZanzi et al. [30] where H2 increased over the pyrolysis temperaturerange from 600 to 1400 �C. They thought that it might beprincipally attributed to dehydrogenation through the cleavage ofa C–H or O–H bond in the biomass and gasification reactionsbetween H2O and the pyrolyzed products (chars, tars, and hydro-carbon gases). Of the three biomass samples, rice husk producedthe highest amount of H2 from 800 to 1200 �C but showed the low-est NO reduction efficiency at relatively high temperature range (asshown in Fig. 3). This indicated that the H2 released from the three

800 900 1000 1100 12002000

3000

4000

5000

6000

7000

8000

CO

con

cent

ratio

n (p

pmv)

Temperature (oC)

rice husk sawdust corncob

Fig. 6. CO concentrations in the outlet gas from the reburning zone versusreburning temperature. Experimental conditions: residence time = 600 ms, initialNO = 800 ppmv, inlet O2 = 0 vol%, H2O = 6 vol%.

biomass samples contributed little to the NO reduction under thepresent experimental conditions. This was consistent with theresults reported by Glarborg et al. [31] and Dagaut and Lecomte[32], who found that the ability of non-hydrocarbon fuels, suchas hydrogen and carbon monoxide, to reduce NO to N2 was verylimited compared to that achieved with the hydrocarbon mixtureunder typical reburning conditions.

The detectable hydrocarbons in these experiments were CH4,C2H2, C2H4, and C2H6, and the concentrations of C2H2, C2H4 andC2H6 were at quite low levels (at most 25 ppm), thus the primarycomponent of the hydrocarbon emissions was CH4. The sum ofthe concentrations of all detectable hydrocarbons, CxHy, is shownin Fig. 8. Interestingly, the profiles of the CxHy concentration forthe three biomass samples reburning were similar to those of theNO reduction efficiencies, as indicated in Fig. 3, i.e., CxHy increasedinitially and then decreased as the reburning temperature wasincreased. After the biomass fuel was introduced into the reburn-ing zone, thermal pyrolysis reactions occurred, and the volatilematter released contained a high proportion of hydrocarbon spe-cies (CH4, C2H2, C2H4, and C2H6). These species further decomposedinto C-containing radicals (HCCO and CHi) under the reducingconditions, and the CHi and HCCO radicals reacted with the NOentering the reburning zone through the following reactions[10,14,33–35]:

CHi þ NO! HCNþHi�1O ð2Þ

HCCOþ NO! HCNþ CO2 ð3Þ

HCCOþ NO! HCNOþ CO ð4Þ

Reaction 4 was followed by reaction 5 [35–37]:

HCNOþH! HCNþ OH ð5Þ

HCNþ O and OH! CN;NCO and NH ð6Þ

CNþ O! NCO ð7Þ

NCOþH! NHþ CO ð8Þ

NHþ NO! N2 þ OH ð9Þ

According to the literature, the sequence of reactions 2 to 5 wasmainly responsible for the reduction of NO by hydrocarbons. It wasfound that this process proceeded mainly via reactions 3 and 4[35,37]. Combined with our results, it could be concluded thatthe hydrocarbons (mainly CH4) released from the biomass werelikely responsible for causing a reduction in the emission of NO.

800 900 1000 1100 12000

500

1000

1500

2000

2500

H2 c

once

ntra

tion

(ppm

v)

Temperature (oC)

rice husk sawdust corncob

Fig. 7. H2 concentrations in the outlet gas from the reburning zone versusreburning temperature. Experimental conditions: residence time = 600 ms, initialNO = 800 ppmv, inlet O2 = 0 vol%, H2O = 6 vol%.

800 900 1000 1100 1200

400

500

600

700

800

900

1000

1100

1200C

xH

y con

cent

ratio

n (p

pmv)

Temperature (oC)

rice husk sawdust corncob

Fig. 8. CxHy concentrations in the outlet gas from the reburning zone versusreburning temperature. Experimental conditions: residence time = 600 ms, initialNO = 800 ppmv, inlet O2 = 0 vol%, H2O = 6 vol%.

800 900 1000 1100 1200

60

80

100

120

140

160

180

200

220

HC

N c

once

ntra

tion

(ppm

v)

Temperature (oC)

rice husk sawdust corncob

Fig. 9. HCN concentrations in the outlet gas from the reburning zone versusreburning temperature. Experimental conditions: residence time = 600 ms, initialNO = 800 ppmv, inlet O2 = 0 vol%, H2O = 6 vol%.

800 900 1000 1100 12000

10

20

30

40

NH

3 con

cent

ratio

n (p

pmv)

Temperature (oC)

rice husk sawdust corncob

Fig. 10. NH3 concentrations in the outlet gas from the reburning zone versusreburning temperature. Experimental conditions: residence time = 600 ms, initialNO = 800 ppmv, inlet O2 = 0 vol%, H2O = 6 vol%.

326 Y. Shu et al. / Fuel 139 (2015) 321–327

Therefore, the high temperatures used in this study reduced theeffective hydrocarbon species (especially CH4), and thereby thecorresponding C-containing radicals, which resulted in a decreasein the NO reduction efficiency of the three biomass samples. Inaddition, the comparison of the results shown in Fig. 3 and Fig. 8revealed that the peaks of the NO reduction efficiency for the threebiomass samples reburning occurred at higher temperatures thanthe CxHy concentration peaks. According to the Arrhenius law,the reaction rate between NO and hydrocarbons should speed upwith an increase in the reburning temperature. Thus, the NOreduction efficiency could be maintained at a high level eventhough the CxHy concentration showed a decrease with an increasein the reburning temperature.

During the reburning process, HCN and NH3, which are twomajor nitrogen-containing intermediates in flue gas, are importantspecies for NO reduction. Fig. 9 represents the variations in theHCN emissions as a function of the reburning temperature forthe three biomass samples. As shown, the profiles of the HCN con-centration were in agreement with the patterns of NO reductionand CxHy concentration (as shown in Fig. 3 and Fig. 8) for the threebiomass samples. Some researchers [14,38] reported that volatilenitrogenous intermediates (e.g. HCN, NH3) in the reburning processcould be generated by the reaction between NO and hydrocarbonsas well as the direct pyrolysis of reburning fuel. Combined with ourexperimental results, it was thought that HCN in reburning processwas mainly produced by hydrocarbons reacting with NO, while theHCN generated directly as a pyrolysis product contributed little toHCN production. Therefore, according to the above experimentalresults, HCN was an important nitrogen-containing intermediatein the NO reduction process.

As shown in Fig. 10, NH3 had relatively lower concentrationsthan HCN during the three biomass samples reburning process.Sawdust showed the highest NH3 production, whereas that ofcorncob was moderate, and rice husk presented the lowest NH3

production. This trend was closely related with the volatile mattercontent of the biomass. As listed in Table 1, the volatile content insawdust was 74.38 wt%, whereas the volatile contents in corncoband rice husk were 72.09 wt% and 64.43 wt%, respectively. Thus,it was assumed that NH3 was mainly generated from the volatilenitrogen in the biomass during the reburning process, which wasconsistent with the experimental results reported by otherresearchers [39–41], who thought that nitrogen bound to proteinsand amino acids was the main source of NH3 and that volatilenitrogen in biomass was normally considered to form NH3 duringthe pyrolysis process. Moreover, rice husk produced the highesttotal concentration of HCN and NH3 among the three biomass sam-ples at relatively low temperatures, whereas sawdust produced the

most HCN and NH3 at relatively high temperatures. These observa-tions were in agreement with the NO reduction levels observed inFig. 3. For the present operating conditions, a high total concentra-tion of HCN and NH3 indicated that the homogeneous reactionsbetween volatiles (CHi and HCCO) and NO were enhanced, andthus, the reburning mechanism should be more effective. Aboveall, the sum of the HCN and NH3 concentrations could reflect thetendency for the NO entering the reburning zone to be convertedinto N2. In addition, the levels of HCN and NH3 concentrations inthe present study were similar to those presented in Ballester’swork [14] where the oak sawdust was used as reburning fuel.While for the actual situation, due to lack of the burnout zonethe exhaust concentrations of HCN and NH3 for the present studywould be relative lower than those for the Ballester’s work. Andthe level of NO reduction in the present work would be reducedaccordingly due to the oxidation of HCN/NH3 to NO in the burnoutzone.

4. Conclusions

Experimental studies were performed on a horizontal fixed-bedquartz reactor to evaluate the NO reduction performances of ricehusk, sawdust and corncob. To study the relative contribution ofeach pyrolysis gas in the biomass reburning process, much atten-tion was paid to the characterization of the distribution of thekey pyrolysis gaseous species (CO, H2, hydrocarbons (mainlyCH4), HCN and NH3) at different reburning temperatures for the

Y. Shu et al. / Fuel 139 (2015) 321–327 327

three biomass samples. According to the experimental results, thefollowing conclusions can be drawn: (1) The biomass type had animportant influence on the NO reduction efficiency. The maximumNO reduction efficiency of sawdust reburning (55 ± 2.4%) wasmuch higher than those of rice husk reburning (43 ± 1.8%) andcorncob reburning (44 ± 2.1%). (2) For sawdust reburning, therewas an optimal initial NO concentration for NO reduction, andhighly efficient NO reduction could be achieved at oxygen inletconcentrations of 0–1 vol% and particle sizes of 160–370 lm. (3)The difference in the NO reduction efficiencies of the three biomasssamples could be ascribed to differences in the yield of pyrolysisgases from the biomasses. Hydrocarbons (mainly CH4) were mainlyresponsible for causing a reduction in the emissions of NO,whereas CO and H2 had little effect on NO reduction. The sum ofthe HCN and NH3 concentrations could reflect the tendency forthe conversion of the NO entering the reburning zone into N2. Itis expected that this study provides insight into the relative contri-bution of pyrolysis gases to the biomass reburning process, whichmay help researchers choose a feedstock of biomass fuels thatwould increase the yield of the desired species.

Acknowledgements

This work was supported by the National High TechnologyResearch and Development Program (‘‘863’’ Program) of China(Nos. 2012AA062505, 2012AA06A113) and the Central ResearchInstitutes of Basic Research and Public Service Special Operations(No. 2013-YSKY-01).

Appendix A. Supplementary material

Ash chemical composition of experimental fuels, time evolutionof measured CH4 concentration during the sawdust reburning pro-cess and effect of the initial NO concentration on NO reduction effi-ciency are shown. This material is available free of charge via theInternet at http://www.sciencedirect.com. Supplementary dataassociated with this article can be found, in the online version, athttp://dx.doi.org/10.1016/j.fuel.2014.08.071.

References

[1] Roger S. Fuel staging for NOx reduction in biomass combustion: experimentsand modeling. Energy Fuels 2001;15:575–82.

[2] Ehsan H, Roger AK, Terese L, Øyvind S. Enhanced NOx reduction by combinedstaged air and flue gas recirculation in biomass grate combustion. Energy Fuels2012;26:3003–11.

[3] Bowman CT. Control of combustion-generated nitrogen oxide emissions:technology driven by regulation. Proc Combust Inst 1992;24:859–78.

[4] Carlin NT, Annamalai K, Harman WL, Sweeten JM. The economics of reburningwith cattle manure-based biomass in existing coal-fired power plants for NOx

and CO2 emissions control. Biomass Bioenergy 2009;33(9):1139–57.[5] Smoot LD, Hill SC, Xu H. NOx control through reburning. Prog Energy Combust

Sci 1998;24(5):385–408.[6] Burch TE, Tillman FR, Chen WY, Lester TW, Conway RB, et al. Partitioning of

nitrogenous species in the fuel rich state of reburning. Energy Fuels1991;5(2):231–7.

[7] Bilbao R, Millera A, Alzueta MU. Evaluation of the use of different hydrocarbonfuels for gas reburning. Fuel 1997;76:1401–7.

[8] Liu H, Hampartoumian E, Gibbs BM. Evaluation of the optimal fuelcharacteristics for efficient NO reduction by coal reburning. Fuel1997;76(11):985–93.

[9] Spliethoff H, Greul U, Rudiger H, Hein KRG. Basic effects on NOx emissions inair staging and reburning at a bench-scale test facility. Fuel 1996;75(5):560–4.

[10] Adams BR, Harding NS. Reburning using biomass for NOx control. Fuel ProcessTechnol 1998;54:249–63.

[11] Harding NS, Adams BR. Biomass as a reburning fuel: a specialized cofiringapplication. Biomass Bioenergy 2000;19:429–45.

[12] Maly PM, Zamansky VM, Zamansky LH. Alternative fuel reburning. Fuel1999;78:327–34.

[13] Rudiger H, Kicherer A, Greul U. Investigations in combined combustion ofbiomass and coal in power plant technology. Energy Fuels 1996;10:789–96.

[14] Ballester J, Ichaso R, Pina A, González MA, Jiménez S. Experimental evaluationand detailed characterization of biomass reburning. Biomass Bioenergy2008;32(10):959–70.

[15] Wendt JOL. Mechanisms governing the formation and destruction of NOx andother nitrogenous species in low NOx coal combustion systems. Combust SciTechnol 1995;108:323–44.

[16] Mereb JB, Wendt JOL. Air staging and reburning mechanisms for NOx

abatement in a laboratory coal combustor. Fuel 1994;73(7):1020–7.[17] Cancès J, Commandré JM, Salvador S, Dagaut P. NO reduction capacity of four

major solid fuels in reburning conditions -experiments and modeling. Fuel2008;87:274–89.

[18] Sarma V, Pisupati SB. Numerical modeling of NOx reduction using pyrolysisproducts from biomass-based materials. Biomass Bioenergy2008;32(2):146–54.

[19] Rudiger H. Pyrolysis gas from biomass and pulverized biomass as reburn fuelsin staged coal combustion. Energy Fuels 1998;39(5):354–61.

[20] Burch TE, Tillman FR, Chen WY, et al. Partitioning of nitrogenous species in thefuel-rich state of reburning. Energy Fuels 1991;5(2):231–7.

[21] Burch TE, Chen WY, Lester TW, et al. Interaction of fuel nitrogen with nitricoxide during reburning with coal. Combust Flame 1994;98:391–401.

[22] Han KH, Liu ZC, Gao P, Lu CM, Cheng ZJ, et al. Experimental study oncharacteristics of nitrogen oxides reduction by biomass reburning. J China CoalSoc 2008;33(5):570–4.

[23] Zhang J, Fan ZL, Lin XF, Xu YQ. Online measurement of products during fastpyrolysis of biomass. J Southeast Univ 2005;35(1):16–9.

[24] Bilbao R, Millera A, Alzueta MU. Influence of the temperature and oxygenconcentration on NOx reduction in the natural gas reburning process. Ind EngChem Res 1994;33:2846–52.

[25] Zhang Y, Kajitani S, Ashizawa M, Miura K. Peculiarities of rapid pyrolysis ofbiomass covering medium- and high-temperature ranges. Energy Fuels2006;20:2705–12.

[26] Badzioch S, Hawksley P. Kinetics of thermal decomposition of pulverized coalparticles. Ind Eng Chem Res 1970;9(2):521–7.

[27] Han KH, Niu SL, Lu CM. Experimental study on biomass advanced reburning fornitrogen oxides reduction. Process Saf Environ 2010;88:425–30.

[28] Lu P, Xu S, Zhu X. Study on NO heterogeneous reduction with coal in anentrained flow reactor. Fuel 2009;88(1):110–5.

[29] Miller JA, Bowman CT. Mechanism and modeling of nitrogen chemistry incombustion. Prog Energy Combust Sci 1989;15(4):287–337.

[30] Zanzi R, Sjöström K, Björnbom E. Rapid high-temperature pyrolysis of biomassin a free-fall reactor. Fuel 1996;75(5):545–50.

[31] Glarborg P, Kristensen PG, Dam-Johansen K. Nitric oxide reduction by non-hydrocarbon fuels implications for reburning with gasification gases. EnergyFuels 2000;14:828–38.

[32] Dagaut P, Lecomte F. Experiments and kinetic modeling study of NO-reburningby gases from biomass pyrolysis in a JSR. Energy Fuels 2003;17:608–13.

[33] Casaca C, Costa M. NOx control through reburning using biomass in alaboratory furnace: effect of particle size. Proc Combust Inst2009;32(2):2641–8.

[34] Zhang RA, Liu CY, Yin RH, Duan J, Luo YH. Experimental and kinetic study ofthe NO-reduction by tar formed from biomass gasification using benzene as atar model component. Fuel Process Technol 2011;92(1):132–8.

[35] Dagaut P, Lecomte F. Experimental and kinetic modeling study of thereduction of NO by hydrocarbons and interactions with SO2 in a JSR at 1atm. Fuel 2003;82(9):1033–40.

[36] Dagaut P, Lecomte F, Chevailler S. Experimental and kinetic modeling of nitricoxide reduction by acetylene in an atmospheric pressure jet-stirred reactor.Fuel 1999;78:1245–52.

[37] Dagaut P, Lecomte F, Chevailler S, Cathonnet M. Experimental and detailedkinetic modeling of nitric oxide reduction by a natural gas blend in simulatedreburning conditions. Combust Sci Technol 1998;139:329–63.

[38] Luan JY, Sun R, Wu SH, Lu JF, Yao N. Experimental studies on reburning ofbiomasses for reducing NOx in a drop tube furnace. Energy Fuels2009;23:1412–21.

[39] Chen Z, Yuan S, Liang Q, Wang F, Yu Z. Distribution of HCN, NH3, NO and N2 inan entrained flow gasifier. Chem Eng J 2009;148:312–8.

[40] Yuan S, Zhou ZJ, Li J, Chen XL, Wang FC. HCN and NH3 released from biomassand soybean cake under rapid pyrolysis. Energy Fuels 2010;24:6166–71.

[41] Hansson KM, ÅA0

mand LE, Habermann A, Winter F. Pyrolysis of poly-L-leucineunder combustion-like conditions. Fuel 2003;82(6):653–60.