- Email: [email protected]
NO conversion characteristics of coal chars prepared using different pyrolysis procedures under combustion conditions
Fuel 211 (2018) 484–491
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Full Length Article
Nitrogen/NO conversion characteristics of coal chars prepared using different pyrolysis procedures under combustion conditions ⁎
MARK
⁎
Jie Xua, Rui Suna, , Tamer M. Ismailb, , Shaozeng Suna, Zhuozhi Wanga, Lijin Maa a b
School of Energy Science and Engineering, Harbin Institute of Technology, Harbin 150001, China Department of Mechanical Engineering, Suez Canal University, Ismailia, Egypt
A R T I C L E I N F O
A B S T R A C T
Keywords: Secondary pyrolysis Reactivity Pore structure Char N/NO conversion Char combustion
In actual combustion facilities, coal chars are often generated using a variety of pyrolysis processes, such as secondary pyrolysis, which is characterized by a long residence time in high temperature zone. The effects of such processes on the conversion of char N to NO during combustion have seldom been explored. In this study, the releases of NO during the combustion of coal chars obtained from different pyrolysis processes in a drop tube and in a fixed bed reactor were investigated. In addition, the extent of char N/NO conversion was studied in relation to the char reactivity, pore surface structure and carbon conversion in a horizontal tube furnace. The results show that, compared with chars generated by a single pyrolysis, chars treated by a subsequent secondary pyrolysis process exhibit larger pore surface areas but less reactivity because of the thermal annealing resulting from a longer thermal history. Chars with higher intrinsic reactivity were also found to release a lower amount of NO. However, a weak correlation was identified between the apparent reactivity and char N/NO conversion, indicating that intrinsic reactivity is more important and directly determines the NO reduction process under combustion conditions. Moreover, char N/NO conversion was significantly affected by the coal rank, and a greater extent of conversion of char N to NO was observed in the case of high-rank coal chars. At a high combustion temperature (1373 K), variations in the bulk O2 concentration had little effect on the char N/NO conversion, and an apparent correlation was found between the extents of char N/NO conversion and the accessible pore surface area. These results indicate that at high temperatures, the char N/NO conversion is directly determined by the accessible pore surface area due to transportation limitations. The NO/(CO + CO2) ratio increased with increasing burn-off in the latter stages of char conversion, which can be attributed to decreases in both the BET surface area and accessible pore surface area available for NO reduction during combustion.
1. Introduction The subject of nitrogen oxide (NOx) emissions has captured the interest of both researchers and the general public because of the serious environmental impact of these compounds. In modern pulverized coal (PC) combustion systems, fuel-fixed nitrogen (fuel N) is typically the major source of total NOx emissions, and previous studies [1–3] have identified the processes by which nitrogen in either volatiles or char is converted to NOx. Volatile nitrogen has been shown to form HCN, NH3, and sootbound nitrogen as intermediate species during the pyrolysis process, all of which can be oxidized to NOx or react with NOx to produce N2 under a reducing atmosphere. In addition, a significant portion of char N can be directly oxidized to NOx. In modern low NOx combustion systems, such as PC systems or fluidized bed (FB) coal combustion systems, volatile nitrogen is transformed into N2 using a precisely controlled
⁎
reducing atmosphere; as a result, char nitrogen has been identified as the main contributor to the furnace exit NOx emissions [4]. The conversion process of nitrogen in chars is complex at PC combustion temperatures (which are higher than those used during FB combustion) because the concentrations of both reactants and products in the gas phase vary with the depth of their penetration into char pores, a depth that changes with char conversion. The transformation of char nitrogen to NOx is believed to proceed via the formation of NOx during char combustion, followed by the heterogeneous reduction of this NOx by carbon [5]. However, the simultaneous occurrence of oxidation and reduction reactions makes it difficult to distinguish the contribution of each route to the net conversion of char N to NOx. Several studies [3,6–8] have shown that coal char combustion conditions (such as reaction temperature, bulk oxygen concentration, and particle size) and factors related to the coal type (such as coal rank, char pore structure, char reactivity, nitrogen and ash level in the char) can
Corresponding authors. E-mail addresses: [email protected] (R. Sun), [email protected], [email protected] (T.M. Ismail).
http://dx.doi.org/10.1016/j.fuel.2017.08.078 Received 14 November 2016; Received in revised form 18 August 2017; Accepted 22 August 2017 Available online 10 October 2017 0016-2361/ © 2017 Elsevier Ltd. All rights reserved.
Fuel 211 (2018) 484–491
J. Xu et al.
The present study focused on the effects of char preparation under different pyrolysis conditions on the conversion of char N to NO. Ordinary pyrolysis, fast secondary pyrolysis, and slow secondary pyrolysis were all performed to prepare test batches of char, which were found to exhibit varying reactive characteristics and pore structures. The extent of char N/NO conversion was ascertained in each case to evaluate the effects of the char reactivity and pore texture on NO release during high-temperature oxidization.
greatly affect emission of NOx during char combustion. In these previous experimental works, chars were specially prepared by pyrolysis procedures; thus, some factors related to pyrolysis, such as the temperature, heating rate, residence time, atmosphere and reactor pressure, are likely to have affected the extent of NOx evolution [9–13]. In these studies, coal chars were typically prepared via a single hightemperature pyrolysis of coal, followed by cooling to ambient temperature. However, secondary pyrolysis chars undergo high-temperature pyrolysis more than once; such a process can significantly affect char reactivity through thermal annealing [14,15], although this effect has rarely been considered when assessing char N conversion. Note that heat treatment processes are known as reheating and/or re-cooling during original or secondary deep pyrolysis can greatly affect the pores and thus influence the apparent reactivity of the char [16]. At low combustion temperatures, micropores account for approximately 95% of the total reactive surface area and play a substantial role in the reactivity of char particles, while larger pores (mesopores and macro pores) provide channels for reactant gas transportation and directly affect the reaction rate at relatively high temperatures [17]. The char reactivity is also very sensitive to the heat treatment and preparation conditions; thus, the reactivity of chars can differ significantly, even when they originate from the same parent coal [18]. A study [19] by Arenillas investigated the effect of char structures obtained under various pyrolysis conditions on the reactivity and NO emissions. The results showed that an increase in the availability of active surface sites improved the apparent reactivity of the char, both for oxygen and for other reactive gases, in particular, NO. In term of the path of char-N/NO conversion especially at high temperature (approximately over FB reaction temperature), the heterogeneous mechanisms are significant to determine the NO formation, so the physical diffusion of O2 and NO in the pore may play an important role in NO formation because the reduction of NO mostly occurs on the pore surface. Previous studies mostly lay emphasis on the effect of the mass transfer limitations on combustion. However, few studies show the relationship between the diffusion of O2 and NO and the final NO formation. Arana [9] proposed the first order reaction model assumed with respect to NO to calculate the effectiveness factor associated to the pore diffusion using a simple reaction-transport model. But there were significant differences in values of the NO diffusion coefficient, so he reported that it is possible that there exist two different length scales of relevance correspond to NO transport in macropores or micropores. Recently, Xu [20] reported that the influence of O2 diffusion in the pore on NO reduction was similar to NO diffusion, so the depth of O2 diffusion directly affected the accessible surface area and reduction time when NO diffused from the internal surface area to the external. In that paper, the O2 diffusion was chartered by the effectiveness factor which was established by experimental data using first order reaction model and the experimental results of char-N/NO conversion were well explained.
2. Experimental section 2.1. Char sample preparation Coal specimens were obtained from sources in China and had two different ranks and particles sizes in the range of 96–125 µm (a bituminous coal, denoted as type J, and a lignite coal, type Y). Various pyrolysis conditions were subsequently used to obtain chars with different reactivities and textural properties. First, standard pyrolysis char was obtained in a drop tube furnace under an inert atmosphere at 1173 K and with a residence time of approximately 2 s. A schematic diagram of this reactor has been previously published [21]. The resulting material is denoted as “original” char and was used to prepare two other types of char. After formation, this original char was quickly pushed into the center of a horizontal tube furnace (HTF) reactor (the apparatus discussed in Section 2.4) that had been preheated to 1373 K. After 30 min of residence time in this furnace, the char was promptly moved to the cooling section and cooled to ambient temperature using a water flow. This procedure is denoted as “fast secondary pyrolysis”, and the resulting material is termed fast secondary pyrolysis char. In an additional procedure, the original char was placed in the center of the HTF reactor at room temperature in advance and then heated to 1373 K at a constant rate of 10 K/min. A hold time of 30 min was applied when the reactor was at 1373 K, after which the char was pushed out to the cooling section and then cooled in preparation for collection in the same manner as that for the fast secondary pyrolysis char. This procedure is denoted as “slow secondary pyrolysis”, and the obtained char is termed “slow secondary pyrolysis char”. All of the above-described pyrolysis procedures were performed under an inert N2 atmosphere. Proximate and ultimate analysis of data for all of the prepared char samples are provided in Table 1, and the heat treatment details are summarized in Table 2. Measurements and tests were subsequently performed to obtain the pore structure parameters and reactivity of the prepared chars. 2.2. Pore structure characterization Coal char is a typical porous material and contains micropores (less than 2 nm), mesopores (2–50 nm) and macropores (more than 50 nm). The Brunauer–Emmett–Teller (BET) surface areas and porosities of the
Table 1 Proximate and Ultimate Analysis Data for the Two Original Parent Coals and their Chars. Samples
Coal J Char JO Char JF Char JS Coal Y Char YO Char YF Char YS a
Proximate analysis (wt% db)
Ultimate analysis (wt% d)
VM
FC
Ash
C
H
N
S
Ob
23.47 6.44 1.54 1.71 29.45 8.84 2.95 3.07
41.84 44.44 47.18 45.23 28.68 40.45 42.01 40.86
34.69 49.11 51.10 52.88 41.87 51.08 55.04 58.07
49.82 45.86 45.79 44.66 40.36 43.97 41.74 39.06
3.28 0.81 0.17 0.18 3.29 0.84 0.18 0.16
1.03 1.04 0.74 0.81 0.74 0.99 0.74 0.72
0.93 0.76 0.83 0.85 1.19 1.07 1.39 1.45
10.25 2.42 1.37 0.62 12.55 2.05 0.91 0.54
Subscripts O, F and S represent the original, fast secondary pyrolysis and slow secondary pyrolysis char samples, respectively. b Calculated by the difference. c Average diameter of char particles.
485
Diameter (µm)c
N/C atomic ratio
– 85.5 89.1 95.8 – 103.6 104.3 104.9
0.021 0.027 0.016 0.018 0.0183 0.023 0.017 0.037
Fuel 211 (2018) 484–491
J. Xu et al.
gas flow rate and reaction temperature were 10 mg, 2 L/min, and 1373 K, respectively, using 6%, 10%, and 20% O2/Ar as the reaction gas, respectively. The test temperature was set at 1373 K based on the heat endurance of the reactor material of construction. The primary gaseous products (CO, CO2 and NO) were analyzed online using Fourier transform infrared spectroscopy (accuracy of 0.01%). Other nitrogen compounds, such as HCN and N2O, were also detected at low temperatures, but their concentrations were negligible at 1373 K. Only the conversion of char N to NO was taken into consideration, that is, not including “thermal N” or “prompt N” conversion routines, because N2 in the air did not participate in the combustion trials. The char N/NO conversion αNO , was calculated using the equation
Table 2 Comparison of the Thermal Histories of Chars. Samples Char preparation
JO, YO
JF, YF
JS, JS
Pyrolysis temperature Secondary pyrolysis Heating rate
1173 K
1373 K
NO Fast, ∼104K/s 2s None
Yes Fast, ∼104 K/s
Heated from Room temperature to 1373 K Yes Slow, 10 K/s
Residence timea Thermal history
a
30 min Reheating/recooling
approximately 4 h Reheating/re-cooling
The time span that the sample remained in the reactor.
Q
αNO = chars were determined via the volumetric adsorption method with a Micromeritics ASAP 2010 instrument based on nitrogen adsorption at 77 K. The surface areas of the micropores, mesopores and macropores were obtained using the t-plot method from the N2 adsorption isotherms. In this study, the sum of mesopore and macro pore surface areas were compared with the microporous surface area. The average char particle sizes were determined using a Malvern laser granulometric particle size analyzer (Mastersizer 2000).
τ
1000MN · 60 × 22.4 · ∫0 (CNO × 10−6) dτ m ·fN
(2)
where t is the experimental run time (s), Mn is the molar mass of elemental nitrogen (g/mol), Q is the volumetric flow (m3/s) of the reaction gas, CNO is the concentration of NO (ppm), m is the mass of char (mg) and fN is the percentage of nitrogen in the solid obtained from the ultimate analysis. 3. Results and discussion
2.3. Reactivity measurements
3.1. Effects of the intrinsic and apparent reactivity of char on NO emissions
Char intrinsic oxidation rate was measured with a thermogravimetric analyzer (TGA; SDTA851e, Mettler Toledo, Switzerland) at 20% O2/N2 atmosphere and a fixed reaction temperature of 773 K to ensure that there were no diffusion limitations. Kinetic measurements of chars over the temperature range of 773–873 K showed that each specimen reacted under conditions of chemical control, equivalent to regime I. In all of the experimental trials, approximately 5 mg of char sample was employed, which covered the bottom of the crucible. The gas flow rate was 100 mL/min and a linear heating rate of 40 °C/min was employed. When the reactor reached the test temperature, the inert atmosphere switched to20% O2/N2 atmosphere till the accomplishment of the isothermal combustion. The sample mass was continuously recorded by a data acquisition system until completion of the char conversion. All reported data for mass change are given on a dry, ash-free basis; each measurement was corrected for buoyancy using a blank run under the same heating condition. In this paper, the initial non-dimensional conversion rate (at X = 0.05) of the char was adopted to be the representative intrinsic reactivity determined from TGA data and the apparent reactivity measured in the HTF, as has been previously reported in the literature [22]. The apparent reactivity, R (s−1), is defined as
dX R=⎛ ⎞ ⎝ dt ⎠
Table 1 provides proximate and ultimate analysis data for chars obtained using different pyrolysis procedures. These data show that the average char particle diameter increases slightly between the original and secondary pyrolysis material; this behavior can be attributed to the release of a fraction of the remaining volatile content of char during the second pyrolysis. Nitrogen is preferentially retained in char during the first pyrolysis, whereas the N/C atomic ratio generally reduced during the secondary pyrolysis. The more release of volatile matters increases carbon contents and makes coal matrix ordered. This resulted in increased fixed carbon contents followed by decreased N/C ratio, that is, N release and carbon increase both effects to N/C ratio decrease. Previous studies [16,23–25] have reported that the thermal history of char, including the pyrolysis conditions, has a strong influence on the reactivity of the material. In this study, therefore, the reactivity of chars that had experienced different thermal histories was assessed to explore the subsequent effects on NO release. Figs. 2 and 3 show the evolution of the reaction rates of the bituminous and lignite chars during isothermal combustion at 773 K under 20 % oxygen in nitrogen, as determined from TGA data for the kinetic control regime I. In the literature, combustion rates at 773 K have been extensively used as a comparative measurement of the intrinsic reactivity of char [26]. These data indicate that following secondary pyrolysis, the reactivity of char is decreased dramatically. The maximum mass loss rates of the original pyrolysis chars are approximately 1.5 and 4.5 times higher than those of the secondary pyrolysis chars in the case of bituminous (type J) and lignite (type Y) coals. With regard to the bituminous chars JS and JF, the reactivity of the fast secondary pyrolysis char is higher than that of the slow secondary pyrolysis material, while the intrinsic reactivity of the lignite chars, YS and YF, are almost the same. These results suggest that secondary pyrolysis (either fast or slow) leads to a pronounced thermal annealing effect on the reactivity, possibly because of a more highly ordered char structure [16]. The relationship between the representative intrinsic reactivity and char N/NO conversion during combustion trials at high temperature (1373 K) and under O2 levels from 6% to 20% is summarized in Fig. 4. These data indicate that a lower intrinsic reactivity is associated with increases in char N/NO conversion. Similar results have been previously reported in the literature [25,26]. Radovic [27] found that chars that exhibit high reactivity under oxygen also show the highest
(1)
where X is the degree of char conversion, defined as X = 1−w / w0 , when ∫t (CO2 + CO) dt
using TGA data, and when X = 0 w ·C using HTF data. Here, the o carbon mass ratio is greater than 0.95, w0 is the combustible mass (daf) of the initial char, w is the combustible mass of the char at reaction time t, and C is the carbon fraction (daf) in the initial char. 2.4. Isothermal combustion and char N/NO conversion Experiments of char combustion were carried out in HTF reactor, a schematic of which is shown in Fig. 1. It consisted of three parts: a gas distribution system, a quartz reactor with an internal diameter of 34 mm and length of 1200 mm, and a gas analysis system. To maintain a completely inert atmosphere in the reactor during the fast pyrolysis process, deoxygenation equipment and a seal were used to remove trace oxygen and prevent air from leaking into the furnace, respectively. During the char combustion tests, the char sample mass, reaction 486
Fuel 211 (2018) 484–491
J. Xu et al.
Horizontal tube furnace reactor
Gas system
Analyzer system
To vent Mixed
To vent
Switch valve
Thermal couple Gas inlet Gas outlet
Mass flowmeter Deoxidation tube
Filter
Gas analyzer
Cooling water Ammeter and Controller Switch Voltmeter
O2
Laptop
Ar
Fig. 1. Schematic of horizontal tube furnace reactor.
0.45
1.2x10 -3
Char JO
20% O2
0.40
Char JF Char JS
10% O2 6% O2
0.35
dx/dt(s-1)
8.0x10 -4
20% O2 10% O2
ĮNO
0.30
4.0x10 -4
6% O2 0.25 0.20 0.15
0.0
0.0
0.2
0.4
0.6
0.8
0.10
1.0
Conversion X
4.0x10-4
8.0x10-4
1.2x10-3
1.6x10-3
2.0x10-3
-1
Intrinsic reactivity(s )
Fig. 2. Reaction rates of bituminous chars at 773 K, 20% O2/Ar.
Fig. 4. Char N to NO conversions under 20%, 10% and 6% O2/Ar at 1373 K as functions of the char representative intrinsic reactivity at 773 K. White symbols: bituminous chars (Char J); black symbols: lignite chars (Char Y).
Char YO
2.0x10-3
Char YF Char YS
NO reduction rates. This effect is believed to result from the highly reactive nature of these chars, leading to an increased extent of reduction of the primary product, NO, and a lower char N/NO conversion ratio. The results in this figure also suggest that when using a relatively high temperature that promotes diffusion control, the intrinsic reactivity of the char is still important for determining the extent of NO reduction on the carbon matrix in the chars. However, at 1373 K, there is a weak correlation between the apparent char reaction rates measured at different bulk O2 concentrations in the HTF and the NO evolution rates, as shown in Fig. 5. The data show what may be an inverse correlation, although a significant scattering of the points is observed. At high reaction temperatures, the char-O2 reaction is limited by O2 diffusion; thus, not all of the pore surfaces are involved in the nitrogen conversation reactions. The larger pores in the char (especially the mesopores and macropores) become even more important for O2 mass
-3
dx/dt(s-1)
1.5x10
1.0x10-3
5.0x10-4
0.0 0.0
0.2
0.4
0.6
0.8
1.0
conversion X Fig. 3. Reaction rates of lignite chars at 773 K, 20% O2/Ar.
487
Fuel 211 (2018) 484–491
J. Xu et al. 0.45
then slightly decreases with increasing O2 concentration seen in Fig. 5. Previous reports [29–32] have sometimes been inconsistent, although there is a general agreement that if the reduction of NO occurs in the vicinity of the external surfaces of the char particles, then more NO will be released because the NO that formed in the pores will escape to the surface rapidly. In this scenario, the contribution of the NO reduction reaction on the pore surfaces will be minimal. Thus, increasing the oxygen concentration to provide a higher oxygen partial pressure will promote O2 penetration into the particles, and more of the NO that is generated will be reduced as a result of the improved access to more reactive pore surfaces, as shown in Fig. 6. Based on the above analysis, at high temperatures and in the O2 diffusion control regime, the rate-determining step will be the diffusion of oxygen and desorption of products from the external surface of the char or its mesopores and macro pores [20,33]. Reactions will primarily occur at the external surfaces of the particles; as a result, increasing the oxygen concentration may have little effect on the rate of O2 diffusion into the char pores and therefore have little impact on the NO reduction on the internal pore surfaces. These reasons are believed to account for the observation that the effect of oxygen on char N conversion to NO was weak.
0.40 0.35
0.25
20% O2
0.20
10% O2
0.15
20% O2
6% O2 10% O2 6% O2
0.10 -2
1.2x10
-2
1.6x10
-2
-2
2.0x10
2.4x10
2.8x10-2
-1
Apparent reactivity(s ) Fig. 5. Char N to NO conversions under 20%, 10% and 6% O2/Ar at 1373 K as functions of the char apparent reactivity at 1373 K. White symbols: bituminous chars (Char J); black symbols: lignite chars (Char Y).
transfer, and thus, a large difference may exist between the mass transfer rates of NO and O2 in the pores [28]. The effects of the coal rank and oxygen concentration on NO emissions were also examined. The properties of a particular coal have long been known to have some effects on the reactivity of char, which in turn influences NO emissions, and there is a general agreement that the nitrogen fractional conversion ratio under combustion conditions is lower in the case of chars prepared from low-rank coal [4]. In the present work, the coal J chars generated higher conversions of char N to NO than the coal Y chars. This result may be attributed to the presence of more catalytically active mineral constituents or to differences in the fundamental carbon matrix structure of the lower ranked coal Y. Lowrank coal chars have a structure with a higher degree of disorder, providing more active sites for NO reduction and O2 intrinsic reactivity, whereas high-rank coal chars have a more ordered carbon lattice structure and thus have fewer sites at which NO reduction can occur. With regard to the effect of the bulk oxygen concentration on the NO evolution, more experimental conditions at 973 K and 1173 K were carried out and the results were showed in Fig. 6. In the case of bituminous chars, the fractional conversion ratio gradually decreases with increasing O2 concentration and the NO conversion values are in the following order: JF, JS and JO. By contrast, the data obtained from lignite chars show that the char N/NO conversion initially increases and
3.2. Effects of the pore structure on NO emissions Previous studies have shown [26,34,35] that the development of certain porous surface structures on char plays an important role in its reactivity by exerting a significant effect on the overall heterogeneous reactions associated with the reductions of NO and N2O. Figs. 7 and 8 summarize the variations in the char pore structures and effects on char N/NO conversion under 6–20% O2/Ar at 1373 K for both the J and Y chars. These data show that the BET values, as well as the micropore and mesopore/macro pore surface areas, tend to increase following secondary pyrolysis. Slow secondary pyrolysis is associated with the largest BET value and the greatest micropore surface area. The pore structure of char Y changed to a greater extent than that of char J after secondary pyrolysis, showing that secondary pyrolysis, whether fast or slow, causes the elimination of volatiles, resulting in changes to the pore structure and an increased BET value. Char Y exhibits, even more, devolatilization and pore structure transformation following the deep pyrolysis treatment. However, attempts to establish a correlation between the pore structure variations and the char N to NO conversion ratio generate an unexpected result: the greater the surface area that is evident from increases in the BET values as well as the micropore and
0.7
20% O2
Char JO at 973 K Char YO at 973 K
0.6
Char JO at 1173 K
20
15 0.3
0.4
ĮNO
ĮNO
6% O2
0.4
Char YO at 1173 K
0.5
10% O2
BET surface area Micropore surface area Mesopore and macropore surface area
0.5
10
0.3
0.2
0.2
0.1
Surface area(m2/g)
ĮNO
0.30
5
0.1 2
6
10
0.0
20
0 JO
O2 (%)
JF
JS
Chars of J
Fig. 6. Char N to NO conversions as functions of the O2 concentration at two temperatures.
Fig. 7. Variations in the surface areas of bituminous char (type J) and the associated effects on NO emissions (1373 K, 6 % -20 % O2).
488
Fuel 211 (2018) 484–491
J. Xu et al.
20% O2
0.4
10% O2
6% O2
BET surface area Micropore surface area Mesopore and macropore surface area
Table 3 Results of calculated effectiveness factor for test chars at 1373 K at different oxygen concentrations.
100
80
0.2 40 0.1
Surface area(m2/g)
Į NO
60
Char Char Char Char Char Char
20
YO
YF
YS
Chars of Y Fig. 8. Variations in the surface areas of lignite char (char Y) and associated effects on NO emissions (1373 K, 6 % -20 % O2).
mesopore/macro pore surface areas leads to a greater degree of NO conversion, a result that contradicts several previous reports [28,36,37] that showed that smaller surface areas were associated with higher conversion.Visona and Stanmore [34] conducted an analysis of the sensitivity of char N to NO conversion to variations in the pore internal surface area and found that the conversion remained unchanged upon increasing the internal area by an order of magnitude at 1750 K. However, another study by Spinti [35] found the same result as in the present study, showing that the char with the highest char N to NOx conversion ratio under FB conditions had the highest surface area. In the present study, the high temperature (1373 K) combustion conditions resulted in transportation limitations that affected O2 and NO diffusion into the pores, such that the char-O2 or char-NO reactions occur less frequently in micro pores than in mesopores/macropores. Therefore, under such conditions, mesopores/macropores not only transport gaseous reactants (O2 and NO) into the pores but also provide reaction surfaces. The extent to which O2 (or NO) diffuses into the pores determines the O2-char and NO-char rates. The accessible pore surface area Sa can be defined as the surface area that is most effective for the reaction of O2 as follows:
Sa = SBET ·ηp
ηp =
(3)
r
rm
ηp
rm
ηp
rm
ηp
2.02 8.88 3.17 2.01 3.97 4.15
0.02277 0.02109 0.02042 0.01995 0.02201 0.02018
0.0113 0.0024 0.0064 0.0099 0.0055 0.0049
0.01495 0.02038 0.01528 0.01563 0.01636 0.01296
0.0074 0.0023 0.0048 0.0078 0.0041 0.0031
0.01165 0.01192 0.01184 0.01288 0.01561 0.01246
0.0058 0.0013 0.0037 0.0064 0.0039 0.0030
rm r
(6)
0.5
20% O2 10% O2 0.4
6% O2 Fitting
R=0.9 0.3
ĮNO
(4)
where R is the gas constant and PO2 is the partial pressure of oxygen. The activation energy E and pre-exponential factor A were calculated by multiple linear regressions using TGA data over the temperature range of 773–873 K. To establish the conditions for diffusion of oxygen in the pores, the effectiveness factor, ηp, was used to account for diffusion resistance in the pores during the overall char combustion process. The overall reaction rate can be expressed as follows:
rm = ηp Aint e−Eint /RT (1−X) PO2
6% O2/Ar
The results of the effectiveness factor under different O2 concentrations were presented in Table 3, and the accessible surface area was calculated by Eq. (3) seen in Fig. 9. Fig. 9 summarizes the relationship between the conversion of char N to NO and the accessible pore surface area of the char. Experimental data of test chars were obtained at 1373 K of the O2 concentrations 6%, 10% and 20%. The fitting was well done (R = 0.9) when just char-N conversion vs the accessible surface area was in consideration. An apparent correlation is found: increasing the accessible surface area results in less NO being released. This shows that the accessible surface area may profoundly influence the conversion of char-N to NO during the char combustion. This occurs because there is a reduced possibility of NO reduction on the micropore surfaces because of the diffusion limitation compared with the reaction at a lower temperature. However, note that the value
where SBET is the BET surface area respectively measured by the N2 adsorption t-plot method and ɳp is an effectiveness factor that accounts for the diffusion resistance in the pores during the overall char combustion process [38]. In this study, the first-order oxidation model was used to fit the intrinsic kinetics data measured in TGA, because this model was widely used in investigations of char intrinsic kinetics. The reaction rate is expressed as:
r= Ae−E / RT (1−X)PO2
JO JF JS YO YF YS
10% O2/Ar
effects of inter-particles and the external mass transfer coefficient. The overall reaction rate can be considered as the rate influenced by pore diffusion, when the external mass transfer resistance was calculated to be less than 0.01% of the total mass resistance according to a procedure found in the literature [39] using a correlation that was valid for a fixed bed with inert and active particles. The apparent reaction rate in the HTF, ra, was calculated based on the time integrals of the measured CO2 and CO molar quantities. The initial gasification rate (x = 0.05) was employed as a representative rate because the char structures are believed to be the same at the initial gasification stage. And then the effectiveness factor was calculated as the ratio of the observed rate to the intrinsic rate at the test temperature as [40].
0
0.0
20% O2/Ar
Samples
0.3
0.2
0.1
0.0 0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Sa(m2/g)
(5)
where the subscripts m and int represent experiments performed with and without diffusional (intrinsic) effects, respectively. In this study, the char particles were mixed with 1 g quartz sand to eliminate the
Fig. 9. Char N to NO conversions at 1373 K, 6 % -20 % O2/Ar as functions of the accessible pore surface area. White symbols: bituminous chars (Char J); black symbols: lignite chars (Char Y).
489
Fuel 211 (2018) 484–491
J. Xu et al.
1.6
JO(6% O2)
Char JF
JO(20% O2)
Char JS 1.2
White symbol
80
16
JO(20% O2)
Char YO
14
SBET(m2/g)
Char YF
1.0
18
JO(6% O2)
Char YS
0.8
12
60
10 8
40
0.6
6
Black symbol
4
20
0.4
Sa*10(m2/g)
1.4
NO/(CO+CO2)*100
20
100
Char JO
2
0.2
0
0
0.2
0.4
0.6
0.8
0.0
1.0
0.2
0.4
0.6
0.8
1.0
Conversion X
Carbon conversion X
Fig. 11. Variations in surface area with the conversion of bituminous char JO at different oxygen concentrations at 1373 K. White symbols: the BET surface area; black symbols: the accessible pore surface area.
Fig. 10. Isothermal combustion gas profiles from char conversion at 1373 K in 20% O2/ Ar. White symbols: bituminous chars; black symbols: lignite chars.
of Sa calculated from the char-O2 reaction may not accurately reflect the reaction surface area of the char-NO and char-O2 reactions because of the higher effective diffusion coefficient of NO and the lower reaction rate (the char-NO reduction rate may less than one tenth of the char-O2 reaction rate). However, it is reasonable to expect that a greater penetration of O2 into the pores and the subsequent reaction with a larger pore surface area indicates that NO will be simultaneously reduced on the larger pore surface area while NO diffuses into the bulk gas phase.
250
20 YO(6% O2)
18
YO(20% O2) YO(6% O2)
200
16
YO(20% O2)
White symbol
SBET(m2/g)
14
3.3. Conversion of char N to NO with char conversion
150
12 10
100
8
Sa*10(m2/g)
0.0
Black symbol 6
Fig. 10 shows the isothermal combustion profiles in the HTF for the various chars at 1373 K in 20% O2/Ar as a function of carbon conversion. These data demonstrate that during the first stage of the burn-off, the NO/(CO + CO2) ratio remains essentially constant, whereas, at the end of the combustion, there is an apparent sharp increase in the ratio. Tullin et al. [41] proposed that this phenomenon can be attributed to the formation of NO during char oxidation, a portion of which is reduced as it diffuses out of the pores. Because the diameter of the char particles decreases with increasing carbon conversion, there will be a greater or lesser possibility of a NO reaction based on the extent of char conversion. However, Harding [26] explained that at low temperatures, it is more likely that changes in the surface area rather than in particle diameter will affect the availability of surface sites for NO reduction. Harding suggested that the reason for the rapid increase in the NO/(CO + CO2) ratio results from nitrogen retention in the char and a lower extent of reduction in the pores or from decreasing char surface area at the end of the burn-off. An apparent change in the pore structure of the char with increasing carbon conversion was identified in the present study, and Figs. 11 and 12 show the changes in the BET surface area and accessible surface area with burn-off of the original chars. Here, given the effectiveness factor remains constant during char oxidation. From these data, it is evident that both the BET surface area and accessible surface area initially increased because of the opening of previously closed pores and widening of existing pores up to a maximum value during the early burn-off stage, followed by a decrease as a result of the formation of mesopores and macropores. Hence, the observed dramatic increase in the NO/(CO + CO2) ratio in the latter stages of char conversion can be attributed to the much lower reaction surface area available for NO reduction. Moreover, during the initial oxidation stage (at less than 45% conversion), the NO/(CO + CO2) ratio does not change significantly. This may be attributed to the increase in the BET and accessible surface area. Different O2 concentrations were used to examine the manner in which the surface structure developed, and almost identical trends in
50
4 2
0
0 0.0
0.2
0.4
0.6
0.8
1.0
Conversion X Fig. 12. Variations in surface area with the conversion of lignite char YO at different oxygen concentrations at 1373 K. White symbols: the BET surface area; black symbols: the accessible pore surface area.
pore surface development were found. It may, therefore, be reasonable to assume that the surface structure of the secondary pyrolysis chars develops in the same manner, although data supporting this contention are not presented here. Note that although the accessible surface areas of the chars during burn-off are presented and these values tend to correlate with the overall variations in char N/NO conversion, these values should not be considered as the exact values of the reactive surface areas for the char to NO reaction since the effectiveness factor changes with the burn-off in actual combustion process.
4. Conclusions The release of nitrogen during the combustion of coal chars generated using different pyrolysis processes was studied. Following secondary pyrolysis of the char, variations in the reactivity and pore structure were found to influence the char N to NO conversion. There is a better correlation of the char N/NO conversion with intrinsic reactivity at 773 K than with apparent reactivity at 1373 K, indicating that increases in the intrinsic reactivity may allow NO to react with char at a higher reduction rate to form other nitrogen species (such as N2), such that the conversion of char N to NO decreases. Moreover, highrank bituminous coal char yielded more NO because it had a less accessible pore surface area after pyrolysis. The extent of char N/NO 490
Fuel 211 (2018) 484–491
J. Xu et al.
[15] Senneca O, Salatino P, Masi S. Heat treatment-induced loss of combustion reactivity of a coal char: the effect of exposure to oxygen. Exp Thermal Fluid Sci 2004;28:735–41. [16] Lu L, Kong C, Sahajwalla V, Harris D. Char structural ordering during pyrolysis and combustion and its influence on char reactivity. Fuel 2002;81:1215–25. [17] Simons GA. The role of pore structure in coal pyrolysis and gasification. Prog Energy Combust Sci 1983;9:269–90. [18] Gale TK, Bartholomew CH, Fletcher TH. Effects of pyrolysis heating rate on intrinsic reactivities of coal chars. Energy Fuels 1996;10:766–75. [19] Arenillas A, Rubiera F, Parra JB, Pis JJ. Influence of char structure on reactivity and nitric oxide emissions. Fuel Process Technol 2002;77:103–9. [20] Xu J, Sun R, Ismail TM, Sun S, Wang Z. Effect of oxygen concentration on NO formation during coal char combustion. Energy Fuels 2017. [21] Fei J, Sun R, Yu L, Liao J, Sun S, Kelebopile L, et al. NO emission characteristics of low-rank pulverized bituminous coal in the primary combustion zone of a drop-tube furnace. Energy Fuels 2010;24:3471–8. [22] Cai H-Y, Güell A, Chatzakis IN, Lim J-Y, Dugwell D, Kandiyoti R. Combustion reactivity and morphological change in coal chars: effect of pyrolysis temperature, heating rate and pressure. Fuel 1996;75:15–24. [23] Zygourakis K. Effect of pyrolysis conditions on the macropore structure of coalderived chars. Energy Fuels 1993;7:33–41. [24] Cetin E, Gupta R, Moghtaderi B. Effect of pyrolysis pressure and heating rate on radiat a pine char structure and apparent gasification reactivity. Fuel 2005;84:1328–34. [25] Wang WX, Thomas KM, Cai HY, Dugwell DR, Kandiyoti R. NO release and reactivity of chars during combustion: the effect of devolatilization temperature and heating rate. Energy Fuels 1996;10:409–16. [26] Harding A, Brown S, Thomas K. Release of NO from the combustion of coal chars. Combust Flame 1996;107:336–50. [27] Radovic L, Illan-Gomez M, Salinas-Martinez de Lecea C, Linares-Solano A. Reduction of nitric oxide by low-rank coal chars, in. Washington, DC (United States): American Chemical Society; 1996. [28] Calo J, Suuberg E, Aarna I, Linares-Solano A, Salinas-Martínez de Lecea C, IllanGomez M. The role of surface area in the NO-carbon reaction. Energy Fuels 1999;13:761–2. [29] Winter F, Löffler G, Wartha C, Hofbauer H, Preto F, Anthony EJ. The NO and N2O formation mechanism under circulating fluidized bed combustor conditions: From the single particle to the pilot-scale. Can J Chem Eng 1999;77:275–83. [30] Hayhurst A, Lawrence A. The amounts of NOx and N2O formed in a fluidized bed combustor during the burning of coal volatiles and also of char. Combust Flame 1996;105:341–57. [31] Chambrion P, Kyotani T, Tomita A. C−NO reaction in the presence of O2. Symposium (International) on Combustion. Elsevier; 1998. p. 3053–9. [32] Ashman PJ, Haynes BS, Nicholls PM, Nelson PF. Interactions of gaseous NO with char during the low-temperature oxidation of coal chars. Proc Combust Inst 2000;28:2171–9. [33] Lorenz H, Carrea E, Tamura M, Haas J. The role of char surface structure development in pulverized fuel combustion. Fuel 2000;79:1161–72. [34] Visona S, Stanmore B. Modeling NOx release from a single coal particle II. Formation of NO from char-nitrogen. Combust Flame 1996;106:207–18. [35] Spinti JP, Pershing DW. The fate of char-N at pulverized coal conditions. Combust Flame 2003;135:299–313. [36] Johnsson JE. Formation and reduction of nitrogen oxides in fluidized-bed combustion. Fuel 1994;73:1398–415. [37] Li Y, Radovic L, Lu G, Rudolph V. A new kinetic model for the NO–carbon reaction. Chem Eng Sci 1999;54:4125–36. [38] Satterfield CN. Mass transfer in heterogeneous catalysis. RE Krieger Publishing Company; 1981. [39] Johnsson JE, Jensen A. Effective diffusion coefficients in coal chars. Proc Combust Inst 2000;28:2353–9. [40] Huo W, Zhou Z, Wang F, Wang Y, Yu G. Experimental study of pore diffusion effect on char gasification with CO2 and steam. Fuel 2014;131:59–65. [41] Tullin CJ, Goel S, Morihara A, Sarofim AF, Beer JM. Nitrogen oxide (NO and N2O) formation for coal combustion in a fluidized bed: effect of carbon conversion and bed temperature. Energy Fuels 1993;7:796–802.
conversion generally decreased as the oxygen concentration was increased, an effect that was attributed to the significant depth of oxygen diffusion and the subsequent greater degree of NO reduction on the surface. As a result of the diffusion limitation of this process, the conversion of char N to NO was well correlated with the accessible pore surface area, demonstrating that at higher temperatures, due to transportation limitations of O2 and NO in the pores, the accessible pore surface area make the primary contribution to the char N reaction surface area. Char N/NO conversion was also found to thus, a greater fraction of the already formed NO was evidently reduced on the increased accessible pore surface area. In addition, the NO/(CO + CO2) ratio was observed to sharply increase with final char burn-off in combustion trials. This may also be attributed to a decrease in both the BET surface area and accessible pore surface area, which reduces the amount of NO formed during the combustion process. Acknowledgments The financial support of the National Natural Science Foundation of China (No. 51476046) and the Innovative Research Groups of the National Natural Science Foundation of China (Grant No. 51421063) is gratefully acknowledged. References [1] Hill S, Glarborg P, Dam-Johansen K. Influence of process parameters on nitrogen oxide formation in pulverized coal burners. Prog Energy Combust Sci 1997;23:349–77. [2] Hill S, Smoot LD. Modeling of nitrogen oxides formation and destruction in combustion systems. Prog Energy Combust Sci 2000;26:417–58. [3] Thomas KM. The release of nitrogen oxides during char combustion. Fuel 1997;76:457–73. [4] Wang W, Brown SD, Hindmarsh CJ, Thomas KM. NOx release and reactivity of chars from a wide range of coals during combustion. Fuel 1994;73:1381–8. [5] Molina A, Eddings EG, Pershing DW, Sarofim AF. Char nitrogen conversion: implications to emissions from coal-fired utility boilers. Prog Energy Combust Sci 2000;26:507–31. [6] Glarborg P, Jensen A, Johnsson JE. Fuel nitrogen conversion in solid fuel fired systems. Prog Energy Combust Sci 2003;29:89–113. [7] Aarna I, Suuberg EM. A review of the kinetics of the nitric oxide-carbon reaction. Fuel 1997;76:475–91. [8] Sung Y, Moon C, Eom S, Choi G, Kim D. Coal-particle size effects on NO reduction and burnout characteristics with air-staged combustion in a pulverized coal-fired furnace. Fuel 2016;182:558–67. [9] Hayashi Ji, Hirama T, Okawa R, Taniguchi M, Hosoda H, Morishita K, et al. Kinetic relationship between NO/N2O reduction and O2 consumption during flue-gas recycling coal combustion in a bubbling fluidized-bed. Fuel 2002;81:1179–88. [10] Yan X, Che D, Xu T. Effect of rank, temperatures and inherent minerals on nitrogen emissions during coal pyrolysis in a fixed bed reactor. Fuel Process Technol 2005;86:739–56. [11] Arenillas A, Rubiera F, Pis JJ, Jones JM, Williams A. The effect of the textural properties of bituminous coal chars on NO emissions. Fuel 1999;78:1779–85. [12] Tsubouchi N, Abe M, Xu C, Ohtsuka Y. Nitrogen release from low rank coals during rapid pyrolysis with a drop tube reactor. Energy Fuels 2003;17:940–5. [13] Park D-C, Day SJ, Nelson PF. Nitrogen release during reaction of coal char with O2, CO2, and H2O. Proc Combust Inst 2005;30:2169–75. [14] Suuberg E. Thermally induced changes in reactivity of carbons. Fundamental issues in control of carbon gasification reactivity. Springer; 1991. p. 269–305.
491
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Full Length Article
Nitrogen/NO conversion characteristics of coal chars prepared using different pyrolysis procedures under combustion conditions ⁎
MARK
⁎
Jie Xua, Rui Suna, , Tamer M. Ismailb, , Shaozeng Suna, Zhuozhi Wanga, Lijin Maa a b
School of Energy Science and Engineering, Harbin Institute of Technology, Harbin 150001, China Department of Mechanical Engineering, Suez Canal University, Ismailia, Egypt
A R T I C L E I N F O
A B S T R A C T
Keywords: Secondary pyrolysis Reactivity Pore structure Char N/NO conversion Char combustion
In actual combustion facilities, coal chars are often generated using a variety of pyrolysis processes, such as secondary pyrolysis, which is characterized by a long residence time in high temperature zone. The effects of such processes on the conversion of char N to NO during combustion have seldom been explored. In this study, the releases of NO during the combustion of coal chars obtained from different pyrolysis processes in a drop tube and in a fixed bed reactor were investigated. In addition, the extent of char N/NO conversion was studied in relation to the char reactivity, pore surface structure and carbon conversion in a horizontal tube furnace. The results show that, compared with chars generated by a single pyrolysis, chars treated by a subsequent secondary pyrolysis process exhibit larger pore surface areas but less reactivity because of the thermal annealing resulting from a longer thermal history. Chars with higher intrinsic reactivity were also found to release a lower amount of NO. However, a weak correlation was identified between the apparent reactivity and char N/NO conversion, indicating that intrinsic reactivity is more important and directly determines the NO reduction process under combustion conditions. Moreover, char N/NO conversion was significantly affected by the coal rank, and a greater extent of conversion of char N to NO was observed in the case of high-rank coal chars. At a high combustion temperature (1373 K), variations in the bulk O2 concentration had little effect on the char N/NO conversion, and an apparent correlation was found between the extents of char N/NO conversion and the accessible pore surface area. These results indicate that at high temperatures, the char N/NO conversion is directly determined by the accessible pore surface area due to transportation limitations. The NO/(CO + CO2) ratio increased with increasing burn-off in the latter stages of char conversion, which can be attributed to decreases in both the BET surface area and accessible pore surface area available for NO reduction during combustion.
1. Introduction The subject of nitrogen oxide (NOx) emissions has captured the interest of both researchers and the general public because of the serious environmental impact of these compounds. In modern pulverized coal (PC) combustion systems, fuel-fixed nitrogen (fuel N) is typically the major source of total NOx emissions, and previous studies [1–3] have identified the processes by which nitrogen in either volatiles or char is converted to NOx. Volatile nitrogen has been shown to form HCN, NH3, and sootbound nitrogen as intermediate species during the pyrolysis process, all of which can be oxidized to NOx or react with NOx to produce N2 under a reducing atmosphere. In addition, a significant portion of char N can be directly oxidized to NOx. In modern low NOx combustion systems, such as PC systems or fluidized bed (FB) coal combustion systems, volatile nitrogen is transformed into N2 using a precisely controlled
⁎
reducing atmosphere; as a result, char nitrogen has been identified as the main contributor to the furnace exit NOx emissions [4]. The conversion process of nitrogen in chars is complex at PC combustion temperatures (which are higher than those used during FB combustion) because the concentrations of both reactants and products in the gas phase vary with the depth of their penetration into char pores, a depth that changes with char conversion. The transformation of char nitrogen to NOx is believed to proceed via the formation of NOx during char combustion, followed by the heterogeneous reduction of this NOx by carbon [5]. However, the simultaneous occurrence of oxidation and reduction reactions makes it difficult to distinguish the contribution of each route to the net conversion of char N to NOx. Several studies [3,6–8] have shown that coal char combustion conditions (such as reaction temperature, bulk oxygen concentration, and particle size) and factors related to the coal type (such as coal rank, char pore structure, char reactivity, nitrogen and ash level in the char) can
Corresponding authors. E-mail addresses: [email protected] (R. Sun), [email protected], [email protected] (T.M. Ismail).
http://dx.doi.org/10.1016/j.fuel.2017.08.078 Received 14 November 2016; Received in revised form 18 August 2017; Accepted 22 August 2017 Available online 10 October 2017 0016-2361/ © 2017 Elsevier Ltd. All rights reserved.
Fuel 211 (2018) 484–491
J. Xu et al.
The present study focused on the effects of char preparation under different pyrolysis conditions on the conversion of char N to NO. Ordinary pyrolysis, fast secondary pyrolysis, and slow secondary pyrolysis were all performed to prepare test batches of char, which were found to exhibit varying reactive characteristics and pore structures. The extent of char N/NO conversion was ascertained in each case to evaluate the effects of the char reactivity and pore texture on NO release during high-temperature oxidization.
greatly affect emission of NOx during char combustion. In these previous experimental works, chars were specially prepared by pyrolysis procedures; thus, some factors related to pyrolysis, such as the temperature, heating rate, residence time, atmosphere and reactor pressure, are likely to have affected the extent of NOx evolution [9–13]. In these studies, coal chars were typically prepared via a single hightemperature pyrolysis of coal, followed by cooling to ambient temperature. However, secondary pyrolysis chars undergo high-temperature pyrolysis more than once; such a process can significantly affect char reactivity through thermal annealing [14,15], although this effect has rarely been considered when assessing char N conversion. Note that heat treatment processes are known as reheating and/or re-cooling during original or secondary deep pyrolysis can greatly affect the pores and thus influence the apparent reactivity of the char [16]. At low combustion temperatures, micropores account for approximately 95% of the total reactive surface area and play a substantial role in the reactivity of char particles, while larger pores (mesopores and macro pores) provide channels for reactant gas transportation and directly affect the reaction rate at relatively high temperatures [17]. The char reactivity is also very sensitive to the heat treatment and preparation conditions; thus, the reactivity of chars can differ significantly, even when they originate from the same parent coal [18]. A study [19] by Arenillas investigated the effect of char structures obtained under various pyrolysis conditions on the reactivity and NO emissions. The results showed that an increase in the availability of active surface sites improved the apparent reactivity of the char, both for oxygen and for other reactive gases, in particular, NO. In term of the path of char-N/NO conversion especially at high temperature (approximately over FB reaction temperature), the heterogeneous mechanisms are significant to determine the NO formation, so the physical diffusion of O2 and NO in the pore may play an important role in NO formation because the reduction of NO mostly occurs on the pore surface. Previous studies mostly lay emphasis on the effect of the mass transfer limitations on combustion. However, few studies show the relationship between the diffusion of O2 and NO and the final NO formation. Arana [9] proposed the first order reaction model assumed with respect to NO to calculate the effectiveness factor associated to the pore diffusion using a simple reaction-transport model. But there were significant differences in values of the NO diffusion coefficient, so he reported that it is possible that there exist two different length scales of relevance correspond to NO transport in macropores or micropores. Recently, Xu [20] reported that the influence of O2 diffusion in the pore on NO reduction was similar to NO diffusion, so the depth of O2 diffusion directly affected the accessible surface area and reduction time when NO diffused from the internal surface area to the external. In that paper, the O2 diffusion was chartered by the effectiveness factor which was established by experimental data using first order reaction model and the experimental results of char-N/NO conversion were well explained.
2. Experimental section 2.1. Char sample preparation Coal specimens were obtained from sources in China and had two different ranks and particles sizes in the range of 96–125 µm (a bituminous coal, denoted as type J, and a lignite coal, type Y). Various pyrolysis conditions were subsequently used to obtain chars with different reactivities and textural properties. First, standard pyrolysis char was obtained in a drop tube furnace under an inert atmosphere at 1173 K and with a residence time of approximately 2 s. A schematic diagram of this reactor has been previously published [21]. The resulting material is denoted as “original” char and was used to prepare two other types of char. After formation, this original char was quickly pushed into the center of a horizontal tube furnace (HTF) reactor (the apparatus discussed in Section 2.4) that had been preheated to 1373 K. After 30 min of residence time in this furnace, the char was promptly moved to the cooling section and cooled to ambient temperature using a water flow. This procedure is denoted as “fast secondary pyrolysis”, and the resulting material is termed fast secondary pyrolysis char. In an additional procedure, the original char was placed in the center of the HTF reactor at room temperature in advance and then heated to 1373 K at a constant rate of 10 K/min. A hold time of 30 min was applied when the reactor was at 1373 K, after which the char was pushed out to the cooling section and then cooled in preparation for collection in the same manner as that for the fast secondary pyrolysis char. This procedure is denoted as “slow secondary pyrolysis”, and the obtained char is termed “slow secondary pyrolysis char”. All of the above-described pyrolysis procedures were performed under an inert N2 atmosphere. Proximate and ultimate analysis of data for all of the prepared char samples are provided in Table 1, and the heat treatment details are summarized in Table 2. Measurements and tests were subsequently performed to obtain the pore structure parameters and reactivity of the prepared chars. 2.2. Pore structure characterization Coal char is a typical porous material and contains micropores (less than 2 nm), mesopores (2–50 nm) and macropores (more than 50 nm). The Brunauer–Emmett–Teller (BET) surface areas and porosities of the
Table 1 Proximate and Ultimate Analysis Data for the Two Original Parent Coals and their Chars. Samples
Coal J Char JO Char JF Char JS Coal Y Char YO Char YF Char YS a
Proximate analysis (wt% db)
Ultimate analysis (wt% d)
VM
FC
Ash
C
H
N
S
Ob
23.47 6.44 1.54 1.71 29.45 8.84 2.95 3.07
41.84 44.44 47.18 45.23 28.68 40.45 42.01 40.86
34.69 49.11 51.10 52.88 41.87 51.08 55.04 58.07
49.82 45.86 45.79 44.66 40.36 43.97 41.74 39.06
3.28 0.81 0.17 0.18 3.29 0.84 0.18 0.16
1.03 1.04 0.74 0.81 0.74 0.99 0.74 0.72
0.93 0.76 0.83 0.85 1.19 1.07 1.39 1.45
10.25 2.42 1.37 0.62 12.55 2.05 0.91 0.54
Subscripts O, F and S represent the original, fast secondary pyrolysis and slow secondary pyrolysis char samples, respectively. b Calculated by the difference. c Average diameter of char particles.
485
Diameter (µm)c
N/C atomic ratio
– 85.5 89.1 95.8 – 103.6 104.3 104.9
0.021 0.027 0.016 0.018 0.0183 0.023 0.017 0.037
Fuel 211 (2018) 484–491
J. Xu et al.
gas flow rate and reaction temperature were 10 mg, 2 L/min, and 1373 K, respectively, using 6%, 10%, and 20% O2/Ar as the reaction gas, respectively. The test temperature was set at 1373 K based on the heat endurance of the reactor material of construction. The primary gaseous products (CO, CO2 and NO) were analyzed online using Fourier transform infrared spectroscopy (accuracy of 0.01%). Other nitrogen compounds, such as HCN and N2O, were also detected at low temperatures, but their concentrations were negligible at 1373 K. Only the conversion of char N to NO was taken into consideration, that is, not including “thermal N” or “prompt N” conversion routines, because N2 in the air did not participate in the combustion trials. The char N/NO conversion αNO , was calculated using the equation
Table 2 Comparison of the Thermal Histories of Chars. Samples Char preparation
JO, YO
JF, YF
JS, JS
Pyrolysis temperature Secondary pyrolysis Heating rate
1173 K
1373 K
NO Fast, ∼104K/s 2s None
Yes Fast, ∼104 K/s
Heated from Room temperature to 1373 K Yes Slow, 10 K/s
Residence timea Thermal history
a
30 min Reheating/recooling
approximately 4 h Reheating/re-cooling
The time span that the sample remained in the reactor.
Q
αNO = chars were determined via the volumetric adsorption method with a Micromeritics ASAP 2010 instrument based on nitrogen adsorption at 77 K. The surface areas of the micropores, mesopores and macropores were obtained using the t-plot method from the N2 adsorption isotherms. In this study, the sum of mesopore and macro pore surface areas were compared with the microporous surface area. The average char particle sizes were determined using a Malvern laser granulometric particle size analyzer (Mastersizer 2000).
τ
1000MN · 60 × 22.4 · ∫0 (CNO × 10−6) dτ m ·fN
(2)
where t is the experimental run time (s), Mn is the molar mass of elemental nitrogen (g/mol), Q is the volumetric flow (m3/s) of the reaction gas, CNO is the concentration of NO (ppm), m is the mass of char (mg) and fN is the percentage of nitrogen in the solid obtained from the ultimate analysis. 3. Results and discussion
2.3. Reactivity measurements
3.1. Effects of the intrinsic and apparent reactivity of char on NO emissions
Char intrinsic oxidation rate was measured with a thermogravimetric analyzer (TGA; SDTA851e, Mettler Toledo, Switzerland) at 20% O2/N2 atmosphere and a fixed reaction temperature of 773 K to ensure that there were no diffusion limitations. Kinetic measurements of chars over the temperature range of 773–873 K showed that each specimen reacted under conditions of chemical control, equivalent to regime I. In all of the experimental trials, approximately 5 mg of char sample was employed, which covered the bottom of the crucible. The gas flow rate was 100 mL/min and a linear heating rate of 40 °C/min was employed. When the reactor reached the test temperature, the inert atmosphere switched to20% O2/N2 atmosphere till the accomplishment of the isothermal combustion. The sample mass was continuously recorded by a data acquisition system until completion of the char conversion. All reported data for mass change are given on a dry, ash-free basis; each measurement was corrected for buoyancy using a blank run under the same heating condition. In this paper, the initial non-dimensional conversion rate (at X = 0.05) of the char was adopted to be the representative intrinsic reactivity determined from TGA data and the apparent reactivity measured in the HTF, as has been previously reported in the literature [22]. The apparent reactivity, R (s−1), is defined as
dX R=⎛ ⎞ ⎝ dt ⎠
Table 1 provides proximate and ultimate analysis data for chars obtained using different pyrolysis procedures. These data show that the average char particle diameter increases slightly between the original and secondary pyrolysis material; this behavior can be attributed to the release of a fraction of the remaining volatile content of char during the second pyrolysis. Nitrogen is preferentially retained in char during the first pyrolysis, whereas the N/C atomic ratio generally reduced during the secondary pyrolysis. The more release of volatile matters increases carbon contents and makes coal matrix ordered. This resulted in increased fixed carbon contents followed by decreased N/C ratio, that is, N release and carbon increase both effects to N/C ratio decrease. Previous studies [16,23–25] have reported that the thermal history of char, including the pyrolysis conditions, has a strong influence on the reactivity of the material. In this study, therefore, the reactivity of chars that had experienced different thermal histories was assessed to explore the subsequent effects on NO release. Figs. 2 and 3 show the evolution of the reaction rates of the bituminous and lignite chars during isothermal combustion at 773 K under 20 % oxygen in nitrogen, as determined from TGA data for the kinetic control regime I. In the literature, combustion rates at 773 K have been extensively used as a comparative measurement of the intrinsic reactivity of char [26]. These data indicate that following secondary pyrolysis, the reactivity of char is decreased dramatically. The maximum mass loss rates of the original pyrolysis chars are approximately 1.5 and 4.5 times higher than those of the secondary pyrolysis chars in the case of bituminous (type J) and lignite (type Y) coals. With regard to the bituminous chars JS and JF, the reactivity of the fast secondary pyrolysis char is higher than that of the slow secondary pyrolysis material, while the intrinsic reactivity of the lignite chars, YS and YF, are almost the same. These results suggest that secondary pyrolysis (either fast or slow) leads to a pronounced thermal annealing effect on the reactivity, possibly because of a more highly ordered char structure [16]. The relationship between the representative intrinsic reactivity and char N/NO conversion during combustion trials at high temperature (1373 K) and under O2 levels from 6% to 20% is summarized in Fig. 4. These data indicate that a lower intrinsic reactivity is associated with increases in char N/NO conversion. Similar results have been previously reported in the literature [25,26]. Radovic [27] found that chars that exhibit high reactivity under oxygen also show the highest
(1)
where X is the degree of char conversion, defined as X = 1−w / w0 , when ∫t (CO2 + CO) dt
using TGA data, and when X = 0 w ·C using HTF data. Here, the o carbon mass ratio is greater than 0.95, w0 is the combustible mass (daf) of the initial char, w is the combustible mass of the char at reaction time t, and C is the carbon fraction (daf) in the initial char. 2.4. Isothermal combustion and char N/NO conversion Experiments of char combustion were carried out in HTF reactor, a schematic of which is shown in Fig. 1. It consisted of three parts: a gas distribution system, a quartz reactor with an internal diameter of 34 mm and length of 1200 mm, and a gas analysis system. To maintain a completely inert atmosphere in the reactor during the fast pyrolysis process, deoxygenation equipment and a seal were used to remove trace oxygen and prevent air from leaking into the furnace, respectively. During the char combustion tests, the char sample mass, reaction 486
Fuel 211 (2018) 484–491
J. Xu et al.
Horizontal tube furnace reactor
Gas system
Analyzer system
To vent Mixed
To vent
Switch valve
Thermal couple Gas inlet Gas outlet
Mass flowmeter Deoxidation tube
Filter
Gas analyzer
Cooling water Ammeter and Controller Switch Voltmeter
O2
Laptop
Ar
Fig. 1. Schematic of horizontal tube furnace reactor.
0.45
1.2x10 -3
Char JO
20% O2
0.40
Char JF Char JS
10% O2 6% O2
0.35
dx/dt(s-1)
8.0x10 -4
20% O2 10% O2
ĮNO
0.30
4.0x10 -4
6% O2 0.25 0.20 0.15
0.0
0.0
0.2
0.4
0.6
0.8
0.10
1.0
Conversion X
4.0x10-4
8.0x10-4
1.2x10-3
1.6x10-3
2.0x10-3
-1
Intrinsic reactivity(s )
Fig. 2. Reaction rates of bituminous chars at 773 K, 20% O2/Ar.
Fig. 4. Char N to NO conversions under 20%, 10% and 6% O2/Ar at 1373 K as functions of the char representative intrinsic reactivity at 773 K. White symbols: bituminous chars (Char J); black symbols: lignite chars (Char Y).
Char YO
2.0x10-3
Char YF Char YS
NO reduction rates. This effect is believed to result from the highly reactive nature of these chars, leading to an increased extent of reduction of the primary product, NO, and a lower char N/NO conversion ratio. The results in this figure also suggest that when using a relatively high temperature that promotes diffusion control, the intrinsic reactivity of the char is still important for determining the extent of NO reduction on the carbon matrix in the chars. However, at 1373 K, there is a weak correlation between the apparent char reaction rates measured at different bulk O2 concentrations in the HTF and the NO evolution rates, as shown in Fig. 5. The data show what may be an inverse correlation, although a significant scattering of the points is observed. At high reaction temperatures, the char-O2 reaction is limited by O2 diffusion; thus, not all of the pore surfaces are involved in the nitrogen conversation reactions. The larger pores in the char (especially the mesopores and macropores) become even more important for O2 mass
-3
dx/dt(s-1)
1.5x10
1.0x10-3
5.0x10-4
0.0 0.0
0.2
0.4
0.6
0.8
1.0
conversion X Fig. 3. Reaction rates of lignite chars at 773 K, 20% O2/Ar.
487
Fuel 211 (2018) 484–491
J. Xu et al. 0.45
then slightly decreases with increasing O2 concentration seen in Fig. 5. Previous reports [29–32] have sometimes been inconsistent, although there is a general agreement that if the reduction of NO occurs in the vicinity of the external surfaces of the char particles, then more NO will be released because the NO that formed in the pores will escape to the surface rapidly. In this scenario, the contribution of the NO reduction reaction on the pore surfaces will be minimal. Thus, increasing the oxygen concentration to provide a higher oxygen partial pressure will promote O2 penetration into the particles, and more of the NO that is generated will be reduced as a result of the improved access to more reactive pore surfaces, as shown in Fig. 6. Based on the above analysis, at high temperatures and in the O2 diffusion control regime, the rate-determining step will be the diffusion of oxygen and desorption of products from the external surface of the char or its mesopores and macro pores [20,33]. Reactions will primarily occur at the external surfaces of the particles; as a result, increasing the oxygen concentration may have little effect on the rate of O2 diffusion into the char pores and therefore have little impact on the NO reduction on the internal pore surfaces. These reasons are believed to account for the observation that the effect of oxygen on char N conversion to NO was weak.
0.40 0.35
0.25
20% O2
0.20
10% O2
0.15
20% O2
6% O2 10% O2 6% O2
0.10 -2
1.2x10
-2
1.6x10
-2
-2
2.0x10
2.4x10
2.8x10-2
-1
Apparent reactivity(s ) Fig. 5. Char N to NO conversions under 20%, 10% and 6% O2/Ar at 1373 K as functions of the char apparent reactivity at 1373 K. White symbols: bituminous chars (Char J); black symbols: lignite chars (Char Y).
transfer, and thus, a large difference may exist between the mass transfer rates of NO and O2 in the pores [28]. The effects of the coal rank and oxygen concentration on NO emissions were also examined. The properties of a particular coal have long been known to have some effects on the reactivity of char, which in turn influences NO emissions, and there is a general agreement that the nitrogen fractional conversion ratio under combustion conditions is lower in the case of chars prepared from low-rank coal [4]. In the present work, the coal J chars generated higher conversions of char N to NO than the coal Y chars. This result may be attributed to the presence of more catalytically active mineral constituents or to differences in the fundamental carbon matrix structure of the lower ranked coal Y. Lowrank coal chars have a structure with a higher degree of disorder, providing more active sites for NO reduction and O2 intrinsic reactivity, whereas high-rank coal chars have a more ordered carbon lattice structure and thus have fewer sites at which NO reduction can occur. With regard to the effect of the bulk oxygen concentration on the NO evolution, more experimental conditions at 973 K and 1173 K were carried out and the results were showed in Fig. 6. In the case of bituminous chars, the fractional conversion ratio gradually decreases with increasing O2 concentration and the NO conversion values are in the following order: JF, JS and JO. By contrast, the data obtained from lignite chars show that the char N/NO conversion initially increases and
3.2. Effects of the pore structure on NO emissions Previous studies have shown [26,34,35] that the development of certain porous surface structures on char plays an important role in its reactivity by exerting a significant effect on the overall heterogeneous reactions associated with the reductions of NO and N2O. Figs. 7 and 8 summarize the variations in the char pore structures and effects on char N/NO conversion under 6–20% O2/Ar at 1373 K for both the J and Y chars. These data show that the BET values, as well as the micropore and mesopore/macro pore surface areas, tend to increase following secondary pyrolysis. Slow secondary pyrolysis is associated with the largest BET value and the greatest micropore surface area. The pore structure of char Y changed to a greater extent than that of char J after secondary pyrolysis, showing that secondary pyrolysis, whether fast or slow, causes the elimination of volatiles, resulting in changes to the pore structure and an increased BET value. Char Y exhibits, even more, devolatilization and pore structure transformation following the deep pyrolysis treatment. However, attempts to establish a correlation between the pore structure variations and the char N to NO conversion ratio generate an unexpected result: the greater the surface area that is evident from increases in the BET values as well as the micropore and
0.7
20% O2
Char JO at 973 K Char YO at 973 K
0.6
Char JO at 1173 K
20
15 0.3
0.4
ĮNO
ĮNO
6% O2
0.4
Char YO at 1173 K
0.5
10% O2
BET surface area Micropore surface area Mesopore and macropore surface area
0.5
10
0.3
0.2
0.2
0.1
Surface area(m2/g)
ĮNO
0.30
5
0.1 2
6
10
0.0
20
0 JO
O2 (%)
JF
JS
Chars of J
Fig. 6. Char N to NO conversions as functions of the O2 concentration at two temperatures.
Fig. 7. Variations in the surface areas of bituminous char (type J) and the associated effects on NO emissions (1373 K, 6 % -20 % O2).
488
Fuel 211 (2018) 484–491
J. Xu et al.
20% O2
0.4
10% O2
6% O2
BET surface area Micropore surface area Mesopore and macropore surface area
Table 3 Results of calculated effectiveness factor for test chars at 1373 K at different oxygen concentrations.
100
80
0.2 40 0.1
Surface area(m2/g)
Į NO
60
Char Char Char Char Char Char
20
YO
YF
YS
Chars of Y Fig. 8. Variations in the surface areas of lignite char (char Y) and associated effects on NO emissions (1373 K, 6 % -20 % O2).
mesopore/macro pore surface areas leads to a greater degree of NO conversion, a result that contradicts several previous reports [28,36,37] that showed that smaller surface areas were associated with higher conversion.Visona and Stanmore [34] conducted an analysis of the sensitivity of char N to NO conversion to variations in the pore internal surface area and found that the conversion remained unchanged upon increasing the internal area by an order of magnitude at 1750 K. However, another study by Spinti [35] found the same result as in the present study, showing that the char with the highest char N to NOx conversion ratio under FB conditions had the highest surface area. In the present study, the high temperature (1373 K) combustion conditions resulted in transportation limitations that affected O2 and NO diffusion into the pores, such that the char-O2 or char-NO reactions occur less frequently in micro pores than in mesopores/macropores. Therefore, under such conditions, mesopores/macropores not only transport gaseous reactants (O2 and NO) into the pores but also provide reaction surfaces. The extent to which O2 (or NO) diffuses into the pores determines the O2-char and NO-char rates. The accessible pore surface area Sa can be defined as the surface area that is most effective for the reaction of O2 as follows:
Sa = SBET ·ηp
ηp =
(3)
r
rm
ηp
rm
ηp
rm
ηp
2.02 8.88 3.17 2.01 3.97 4.15
0.02277 0.02109 0.02042 0.01995 0.02201 0.02018
0.0113 0.0024 0.0064 0.0099 0.0055 0.0049
0.01495 0.02038 0.01528 0.01563 0.01636 0.01296
0.0074 0.0023 0.0048 0.0078 0.0041 0.0031
0.01165 0.01192 0.01184 0.01288 0.01561 0.01246
0.0058 0.0013 0.0037 0.0064 0.0039 0.0030
rm r
(6)
0.5
20% O2 10% O2 0.4
6% O2 Fitting
R=0.9 0.3
ĮNO
(4)
where R is the gas constant and PO2 is the partial pressure of oxygen. The activation energy E and pre-exponential factor A were calculated by multiple linear regressions using TGA data over the temperature range of 773–873 K. To establish the conditions for diffusion of oxygen in the pores, the effectiveness factor, ηp, was used to account for diffusion resistance in the pores during the overall char combustion process. The overall reaction rate can be expressed as follows:
rm = ηp Aint e−Eint /RT (1−X) PO2
6% O2/Ar
The results of the effectiveness factor under different O2 concentrations were presented in Table 3, and the accessible surface area was calculated by Eq. (3) seen in Fig. 9. Fig. 9 summarizes the relationship between the conversion of char N to NO and the accessible pore surface area of the char. Experimental data of test chars were obtained at 1373 K of the O2 concentrations 6%, 10% and 20%. The fitting was well done (R = 0.9) when just char-N conversion vs the accessible surface area was in consideration. An apparent correlation is found: increasing the accessible surface area results in less NO being released. This shows that the accessible surface area may profoundly influence the conversion of char-N to NO during the char combustion. This occurs because there is a reduced possibility of NO reduction on the micropore surfaces because of the diffusion limitation compared with the reaction at a lower temperature. However, note that the value
where SBET is the BET surface area respectively measured by the N2 adsorption t-plot method and ɳp is an effectiveness factor that accounts for the diffusion resistance in the pores during the overall char combustion process [38]. In this study, the first-order oxidation model was used to fit the intrinsic kinetics data measured in TGA, because this model was widely used in investigations of char intrinsic kinetics. The reaction rate is expressed as:
r= Ae−E / RT (1−X)PO2
JO JF JS YO YF YS
10% O2/Ar
effects of inter-particles and the external mass transfer coefficient. The overall reaction rate can be considered as the rate influenced by pore diffusion, when the external mass transfer resistance was calculated to be less than 0.01% of the total mass resistance according to a procedure found in the literature [39] using a correlation that was valid for a fixed bed with inert and active particles. The apparent reaction rate in the HTF, ra, was calculated based on the time integrals of the measured CO2 and CO molar quantities. The initial gasification rate (x = 0.05) was employed as a representative rate because the char structures are believed to be the same at the initial gasification stage. And then the effectiveness factor was calculated as the ratio of the observed rate to the intrinsic rate at the test temperature as [40].
0
0.0
20% O2/Ar
Samples
0.3
0.2
0.1
0.0 0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Sa(m2/g)
(5)
where the subscripts m and int represent experiments performed with and without diffusional (intrinsic) effects, respectively. In this study, the char particles were mixed with 1 g quartz sand to eliminate the
Fig. 9. Char N to NO conversions at 1373 K, 6 % -20 % O2/Ar as functions of the accessible pore surface area. White symbols: bituminous chars (Char J); black symbols: lignite chars (Char Y).
489
Fuel 211 (2018) 484–491
J. Xu et al.
1.6
JO(6% O2)
Char JF
JO(20% O2)
Char JS 1.2
White symbol
80
16
JO(20% O2)
Char YO
14
SBET(m2/g)
Char YF
1.0
18
JO(6% O2)
Char YS
0.8
12
60
10 8
40
0.6
6
Black symbol
4
20
0.4
Sa*10(m2/g)
1.4
NO/(CO+CO2)*100
20
100
Char JO
2
0.2
0
0
0.2
0.4
0.6
0.8
0.0
1.0
0.2
0.4
0.6
0.8
1.0
Conversion X
Carbon conversion X
Fig. 11. Variations in surface area with the conversion of bituminous char JO at different oxygen concentrations at 1373 K. White symbols: the BET surface area; black symbols: the accessible pore surface area.
Fig. 10. Isothermal combustion gas profiles from char conversion at 1373 K in 20% O2/ Ar. White symbols: bituminous chars; black symbols: lignite chars.
of Sa calculated from the char-O2 reaction may not accurately reflect the reaction surface area of the char-NO and char-O2 reactions because of the higher effective diffusion coefficient of NO and the lower reaction rate (the char-NO reduction rate may less than one tenth of the char-O2 reaction rate). However, it is reasonable to expect that a greater penetration of O2 into the pores and the subsequent reaction with a larger pore surface area indicates that NO will be simultaneously reduced on the larger pore surface area while NO diffuses into the bulk gas phase.
250
20 YO(6% O2)
18
YO(20% O2) YO(6% O2)
200
16
YO(20% O2)
White symbol
SBET(m2/g)
14
3.3. Conversion of char N to NO with char conversion
150
12 10
100
8
Sa*10(m2/g)
0.0
Black symbol 6
Fig. 10 shows the isothermal combustion profiles in the HTF for the various chars at 1373 K in 20% O2/Ar as a function of carbon conversion. These data demonstrate that during the first stage of the burn-off, the NO/(CO + CO2) ratio remains essentially constant, whereas, at the end of the combustion, there is an apparent sharp increase in the ratio. Tullin et al. [41] proposed that this phenomenon can be attributed to the formation of NO during char oxidation, a portion of which is reduced as it diffuses out of the pores. Because the diameter of the char particles decreases with increasing carbon conversion, there will be a greater or lesser possibility of a NO reaction based on the extent of char conversion. However, Harding [26] explained that at low temperatures, it is more likely that changes in the surface area rather than in particle diameter will affect the availability of surface sites for NO reduction. Harding suggested that the reason for the rapid increase in the NO/(CO + CO2) ratio results from nitrogen retention in the char and a lower extent of reduction in the pores or from decreasing char surface area at the end of the burn-off. An apparent change in the pore structure of the char with increasing carbon conversion was identified in the present study, and Figs. 11 and 12 show the changes in the BET surface area and accessible surface area with burn-off of the original chars. Here, given the effectiveness factor remains constant during char oxidation. From these data, it is evident that both the BET surface area and accessible surface area initially increased because of the opening of previously closed pores and widening of existing pores up to a maximum value during the early burn-off stage, followed by a decrease as a result of the formation of mesopores and macropores. Hence, the observed dramatic increase in the NO/(CO + CO2) ratio in the latter stages of char conversion can be attributed to the much lower reaction surface area available for NO reduction. Moreover, during the initial oxidation stage (at less than 45% conversion), the NO/(CO + CO2) ratio does not change significantly. This may be attributed to the increase in the BET and accessible surface area. Different O2 concentrations were used to examine the manner in which the surface structure developed, and almost identical trends in
50
4 2
0
0 0.0
0.2
0.4
0.6
0.8
1.0
Conversion X Fig. 12. Variations in surface area with the conversion of lignite char YO at different oxygen concentrations at 1373 K. White symbols: the BET surface area; black symbols: the accessible pore surface area.
pore surface development were found. It may, therefore, be reasonable to assume that the surface structure of the secondary pyrolysis chars develops in the same manner, although data supporting this contention are not presented here. Note that although the accessible surface areas of the chars during burn-off are presented and these values tend to correlate with the overall variations in char N/NO conversion, these values should not be considered as the exact values of the reactive surface areas for the char to NO reaction since the effectiveness factor changes with the burn-off in actual combustion process.
4. Conclusions The release of nitrogen during the combustion of coal chars generated using different pyrolysis processes was studied. Following secondary pyrolysis of the char, variations in the reactivity and pore structure were found to influence the char N to NO conversion. There is a better correlation of the char N/NO conversion with intrinsic reactivity at 773 K than with apparent reactivity at 1373 K, indicating that increases in the intrinsic reactivity may allow NO to react with char at a higher reduction rate to form other nitrogen species (such as N2), such that the conversion of char N to NO decreases. Moreover, highrank bituminous coal char yielded more NO because it had a less accessible pore surface area after pyrolysis. The extent of char N/NO 490
Fuel 211 (2018) 484–491
J. Xu et al.
[15] Senneca O, Salatino P, Masi S. Heat treatment-induced loss of combustion reactivity of a coal char: the effect of exposure to oxygen. Exp Thermal Fluid Sci 2004;28:735–41. [16] Lu L, Kong C, Sahajwalla V, Harris D. Char structural ordering during pyrolysis and combustion and its influence on char reactivity. Fuel 2002;81:1215–25. [17] Simons GA. The role of pore structure in coal pyrolysis and gasification. Prog Energy Combust Sci 1983;9:269–90. [18] Gale TK, Bartholomew CH, Fletcher TH. Effects of pyrolysis heating rate on intrinsic reactivities of coal chars. Energy Fuels 1996;10:766–75. [19] Arenillas A, Rubiera F, Parra JB, Pis JJ. Influence of char structure on reactivity and nitric oxide emissions. Fuel Process Technol 2002;77:103–9. [20] Xu J, Sun R, Ismail TM, Sun S, Wang Z. Effect of oxygen concentration on NO formation during coal char combustion. Energy Fuels 2017. [21] Fei J, Sun R, Yu L, Liao J, Sun S, Kelebopile L, et al. NO emission characteristics of low-rank pulverized bituminous coal in the primary combustion zone of a drop-tube furnace. Energy Fuels 2010;24:3471–8. [22] Cai H-Y, Güell A, Chatzakis IN, Lim J-Y, Dugwell D, Kandiyoti R. Combustion reactivity and morphological change in coal chars: effect of pyrolysis temperature, heating rate and pressure. Fuel 1996;75:15–24. [23] Zygourakis K. Effect of pyrolysis conditions on the macropore structure of coalderived chars. Energy Fuels 1993;7:33–41. [24] Cetin E, Gupta R, Moghtaderi B. Effect of pyrolysis pressure and heating rate on radiat a pine char structure and apparent gasification reactivity. Fuel 2005;84:1328–34. [25] Wang WX, Thomas KM, Cai HY, Dugwell DR, Kandiyoti R. NO release and reactivity of chars during combustion: the effect of devolatilization temperature and heating rate. Energy Fuels 1996;10:409–16. [26] Harding A, Brown S, Thomas K. Release of NO from the combustion of coal chars. Combust Flame 1996;107:336–50. [27] Radovic L, Illan-Gomez M, Salinas-Martinez de Lecea C, Linares-Solano A. Reduction of nitric oxide by low-rank coal chars, in. Washington, DC (United States): American Chemical Society; 1996. [28] Calo J, Suuberg E, Aarna I, Linares-Solano A, Salinas-Martínez de Lecea C, IllanGomez M. The role of surface area in the NO-carbon reaction. Energy Fuels 1999;13:761–2. [29] Winter F, Löffler G, Wartha C, Hofbauer H, Preto F, Anthony EJ. The NO and N2O formation mechanism under circulating fluidized bed combustor conditions: From the single particle to the pilot-scale. Can J Chem Eng 1999;77:275–83. [30] Hayhurst A, Lawrence A. The amounts of NOx and N2O formed in a fluidized bed combustor during the burning of coal volatiles and also of char. Combust Flame 1996;105:341–57. [31] Chambrion P, Kyotani T, Tomita A. C−NO reaction in the presence of O2. Symposium (International) on Combustion. Elsevier; 1998. p. 3053–9. [32] Ashman PJ, Haynes BS, Nicholls PM, Nelson PF. Interactions of gaseous NO with char during the low-temperature oxidation of coal chars. Proc Combust Inst 2000;28:2171–9. [33] Lorenz H, Carrea E, Tamura M, Haas J. The role of char surface structure development in pulverized fuel combustion. Fuel 2000;79:1161–72. [34] Visona S, Stanmore B. Modeling NOx release from a single coal particle II. Formation of NO from char-nitrogen. Combust Flame 1996;106:207–18. [35] Spinti JP, Pershing DW. The fate of char-N at pulverized coal conditions. Combust Flame 2003;135:299–313. [36] Johnsson JE. Formation and reduction of nitrogen oxides in fluidized-bed combustion. Fuel 1994;73:1398–415. [37] Li Y, Radovic L, Lu G, Rudolph V. A new kinetic model for the NO–carbon reaction. Chem Eng Sci 1999;54:4125–36. [38] Satterfield CN. Mass transfer in heterogeneous catalysis. RE Krieger Publishing Company; 1981. [39] Johnsson JE, Jensen A. Effective diffusion coefficients in coal chars. Proc Combust Inst 2000;28:2353–9. [40] Huo W, Zhou Z, Wang F, Wang Y, Yu G. Experimental study of pore diffusion effect on char gasification with CO2 and steam. Fuel 2014;131:59–65. [41] Tullin CJ, Goel S, Morihara A, Sarofim AF, Beer JM. Nitrogen oxide (NO and N2O) formation for coal combustion in a fluidized bed: effect of carbon conversion and bed temperature. Energy Fuels 1993;7:796–802.
conversion generally decreased as the oxygen concentration was increased, an effect that was attributed to the significant depth of oxygen diffusion and the subsequent greater degree of NO reduction on the surface. As a result of the diffusion limitation of this process, the conversion of char N to NO was well correlated with the accessible pore surface area, demonstrating that at higher temperatures, due to transportation limitations of O2 and NO in the pores, the accessible pore surface area make the primary contribution to the char N reaction surface area. Char N/NO conversion was also found to thus, a greater fraction of the already formed NO was evidently reduced on the increased accessible pore surface area. In addition, the NO/(CO + CO2) ratio was observed to sharply increase with final char burn-off in combustion trials. This may also be attributed to a decrease in both the BET surface area and accessible pore surface area, which reduces the amount of NO formed during the combustion process. Acknowledgments The financial support of the National Natural Science Foundation of China (No. 51476046) and the Innovative Research Groups of the National Natural Science Foundation of China (Grant No. 51421063) is gratefully acknowledged. References [1] Hill S, Glarborg P, Dam-Johansen K. Influence of process parameters on nitrogen oxide formation in pulverized coal burners. Prog Energy Combust Sci 1997;23:349–77. [2] Hill S, Smoot LD. Modeling of nitrogen oxides formation and destruction in combustion systems. Prog Energy Combust Sci 2000;26:417–58. [3] Thomas KM. The release of nitrogen oxides during char combustion. Fuel 1997;76:457–73. [4] Wang W, Brown SD, Hindmarsh CJ, Thomas KM. NOx release and reactivity of chars from a wide range of coals during combustion. Fuel 1994;73:1381–8. [5] Molina A, Eddings EG, Pershing DW, Sarofim AF. Char nitrogen conversion: implications to emissions from coal-fired utility boilers. Prog Energy Combust Sci 2000;26:507–31. [6] Glarborg P, Jensen A, Johnsson JE. Fuel nitrogen conversion in solid fuel fired systems. Prog Energy Combust Sci 2003;29:89–113. [7] Aarna I, Suuberg EM. A review of the kinetics of the nitric oxide-carbon reaction. Fuel 1997;76:475–91. [8] Sung Y, Moon C, Eom S, Choi G, Kim D. Coal-particle size effects on NO reduction and burnout characteristics with air-staged combustion in a pulverized coal-fired furnace. Fuel 2016;182:558–67. [9] Hayashi Ji, Hirama T, Okawa R, Taniguchi M, Hosoda H, Morishita K, et al. Kinetic relationship between NO/N2O reduction and O2 consumption during flue-gas recycling coal combustion in a bubbling fluidized-bed. Fuel 2002;81:1179–88. [10] Yan X, Che D, Xu T. Effect of rank, temperatures and inherent minerals on nitrogen emissions during coal pyrolysis in a fixed bed reactor. Fuel Process Technol 2005;86:739–56. [11] Arenillas A, Rubiera F, Pis JJ, Jones JM, Williams A. The effect of the textural properties of bituminous coal chars on NO emissions. Fuel 1999;78:1779–85. [12] Tsubouchi N, Abe M, Xu C, Ohtsuka Y. Nitrogen release from low rank coals during rapid pyrolysis with a drop tube reactor. Energy Fuels 2003;17:940–5. [13] Park D-C, Day SJ, Nelson PF. Nitrogen release during reaction of coal char with O2, CO2, and H2O. Proc Combust Inst 2005;30:2169–75. [14] Suuberg E. Thermally induced changes in reactivity of carbons. Fundamental issues in control of carbon gasification reactivity. Springer; 1991. p. 269–305.
491