Characteristics of Fundamental Combustion and NO_x Emission Using Various Rank Coals

ABSTRACT

Eight types of coals of different rank were selected and their fundamental combustion characteristics were examined along with the conversion of volatile nitrogen (N) to nitrogen oxides (NO_x)/fuel N to NO_x. The activation energy, onset temperature, and burnout temperature were obtained from the differential thermogravimetry curve and Arrhenius plot, which were derived through thermogravimetric analysis. In addition, to derive the combustion of volatile N to NO_x/fuel N to NO_x, the coal sample, which was pretreated at various temperatures, was burned, and the results were compared with previously derived fundamental combustion characteristics. The authors' experimental results confirmed that coal rank was highly correlated with the combustion of volatile N to NO_x/fuel N to NO_x.

IMPLICATIONS

Low-rank coals are abundant in several regions throughout the world. However, despite their quantity and relatively low market price, these low-rank coals have not been used because of a lack of existing studies. Therefore, this paper presents the characteristics of NO_x emission and fundamental combustion on coal combustion with coals having different rank to investigate the effect of coal rank on coal combustion.

INTRODUCTION

Pollutants such as carbon monoxide (CO), unburned hydro-carbon, soot, sulfur oxides, and nitrogen oxides (NO_x) are emitted by burning coal. In particular, NO_x react with hydrocarbon to generate ozone and photochemical oxides such as peroxyacetyl nitrate [CH₃(CO)OONO] under the presence of sunlight. Moreover, NO_x have become the main cause of acid rain based on conversion to nitric acid and nitrates.1 Although NO_x are produced during combustion by the oxidation of nitrogen within the fuel, they may also form from molecular nitrogen existing in the air, such as thermal NO_x. Therefore, NO_x possess characteristics different from those of sulfide, chloride, heavy metals, and others, which are the other pollutants commonly found in fuel.

In general, NO_x refer to nitric oxide (NO), nitrogen dioxide (NO₂), and nitrous oxide (N₂O),2 with NO₂ and N₂O being commonly ignored because they are generated in such small quantities.3 NO_x generated by the combustion of coal are classified into thermal NO_x, caused by the thermal fixation of nitrogen in the air,4 prompt NO_x, produced by reaction with radicals (C and CH₂) produced from the fuel,5 and fuel NO_x, produced from N components in the fuel,6 depending on the origin of the nitrogen and mechanism of production. Among them, fuel NO_x accounted for 75–95% of all NO_x, whereas NO_x oxidized from volatile N and char N constituted 50–70% and 20–30%, respectively. Therefore, reducing the emission of NO_x during coal combustion is dependent on the regulation of fuel NO_x rather than thermal NO_x. The amount of fuel NO_x can vary significantly, depending on coal reactivity, nitrogen content, coal composition, and the combustion conditions.7 Accordingly, research has focused on methods to effectively control NO_x using flue gas recirculation, burner out of service, over-fired air, reburning, fuel and air staging, low NO_x burner, 2-step combustion, and others. However, the characteristics of fuel NO_x generation depending on the type of coal (NO_x generation caused by volatile N and char N) as opposed to the combustion method or conditions have not been extensively considered.

Therefore, eight types of coal were selected to examine the characteristics of fundamental combustion as well as NO_x generation from the ratio of volatile N to NO_x and char N to NO_x.

EXPERIMENTAL PROCEDURES

Experimental Methods

A simplified diagram of the experimental device used in this research is shown in Figure 1. Heating was generated at a constant velocity using a muffle furnace, and the device comprised a gas injection part, combustion part, and analyzing part. The whole gas supply pipe was manufactured from stainless steel, and the part into which coal is flowed during combustion was specially manufactured as a Y-type pipe combined with two-way and three-way valves. The combustion part was a special quartz pipe with an i.d. of 8 mm, an o.d. of 10 mm, and a height of 70 mm. Quartz wool was used to fix the coal. In addition, the quartz pipe was specially treated so that the wool would not be shoved toward the bottom upon increased pressure. A K-type thermocouple was mounted in contact with the top part of the quartz wool to measure the burning temperature, which was controlled by the same type of thermocouple as in the combustion furnace. The flow rate of injected O₂/Ar from the cylinder was controlled using a mass flow controller (MKS Instruments, Colorado Springs, CO). To measure cocentration of emitted NO_x, NO and NO₂ were measured using a TESTO 360 analyzer (Testo AG, Lenzkirch, Germany), and H₂O was completely removed before injection into the analyzer.

Figure 1. Schematic diagram of the combustion system.

The combustion experiment used a specially manufactured experimental device previously reported by the authors' group8 instead of a drop tube furnace9 to perform B1 combustion (batch combustion method 1). B1 combustion raised the temperature of the combustion furnace in a constant rate of 10 °C/min within a combustion furnace filled with coal. When the coal was burning, the combustion gas was measured. In addition, to distinguish the conversion of char N to NO_x from that of volatile N to NO_x, the coal sample was burned again after drying at various temperatures using argon gas.

Coal Characterizations

Eight types of coals of various ranks were selected for use as samples. In addition, elementary analysis of the coals was performed using a TruSpec (LECO Corporation, St. Joseph, MI) and SC-432DR (LECO Corporation), proximate analysis with an automatic proximate analyzer (LECO Corporation), and ash analysis using an inductively coupled plasma optical emission spectrometer (PerkinElmer Life and Analytical Sciences, Waltham, MA). As shown in Table 1, anthracite was ∼10% volatile matter (VM), whereas bituminous coal and pitch coal and lignite were ∼30 and 30 to 50%, respectively. Anthracite had the highest content of ash (greater than 40%), whereas bituminous coal had ∼15% and pitch coal-lignite was in the range of 5 to 20%. The coal used in each combustion experiment was sifted using 40- to 50-sized mesh to obtain uniform combustion.

Table 1. Physical analysis for various coals

Coals (country)	NAR	Total H2O	H2O	FC	VM	Ash	C	H	O	N	S	Ash	Rank
Anthracite (CHN)	3040	5.02	3.3	40.42	10.74	45.54	47.37	1.27	2.56	0.14	1.56	47.1	Anthracite
Anglo (IND)	6515	13.77	3.85	49.29	30.89	15.97	73.73	4.43	2.98	1.71	0.54	16.61	Bituminous
C&A (Austria)	6417	7.52	4.01	49.42	31.31	15.26	72.77	4.29	4.9	1.63	0.51	15.9	Bituminous
Berau (Austrua)	6155	23.61	12.04	42.56	39.68	5.72	68.70	5.23	17.58	1.5	0.49	6.5	Sub-bituminous, lignite
Shinwha (China)	6673	13.11	8.53	50.75	34.92	5.8	75.77	4.85	11.58	1.15	0.31	6.34	Sub-bituminous, lignite
SM (India)	5304	26.34	12.6	40.16	42.7	4.54	52.03	5.21	26.5	0.92	0.15	5.19	Sub-bituminous, lignite
KBB1 (India)	896	34.21	15.73	44.81	38.01	1.45	46.34	3.85	21.89	0.62	0.01		Sub-bituminous, lignite
KBB9 (India)	584	26.53	12.46	16.82	51.71	19.01	38.73	3.9	20.87	0.66	0.01		Sub-bituminous, lignite
Notes: NAR = net as received; VM = volatile matter; FC = fixed carbon.

Activation Energy and Differential Thermogravimetry

The analysis of activation energy was complex because the reaction velocity at which coals are burned varies, depending on the temperature of the combustion reaction. Therefore, the activation energy was analyzed by following the method suggested by Serageldin and Pan.10

The reaction velocity equation of the coal combustion reaction, such as response of a could be represented as

(1)

where a is the amount of A decomposed at time t, n is the reaction degree, and k is the reaction velocity constant. The constant of reaction velocity is represented according to Arrhenius theory:

(2)

where A is the frequency factor, E is activation energy, and T is temperature.

The linear heating rate can be represented as

(3)

If eqs 1, 2, and 3 are integrated, they become

(4)

The right-hand side of eq 4 is not an exact integral. Thus, if u = E/RT is substituted instead and the following relation is used,

eq 4 will become

(5)

If a log is taken, it becomes eq 6 for n ≠ 1.

(6)

If n=1, it becomes eq 7:

(7)

is almost constant in the temperature range at which the reaction occurs. If 1/T of eq 7 are plotted, the activation energy can be calculated from the slope, which is represented by −E/RT. Thermogravimetric analysis (TGA) of coal is very complex because of the presence of various components. Therefore, the activation energy was computed by the method suggested by Serageldin and Pan.10 A linear interval appeared in the Arrhenius plot performed for TGA. If the activation energy at each interval is E, the total activation energy can be expressed as

(8)

where M_n is the weight ratio of the coal sample in the temperature range of each interval.

On the other hand, differential thermogravimetry (DTG), which is a concept on the decreasing rate of weight, depending on the temperature. In other words, it is to plot, , where M is the weight of the coal sample and T is the time for which the unit is percent per minute.

RESULTS AND DISCUSSION

Fundamental Combustion Characteristics

High temperature TGA is widely used for the analysis of coal combustion. Cumming and McLaughlin11 reported that thermal analysis could be used to select the type of coal, whereas Serageldin and Pan10 explained the methods for deriving the activation energy from the results of the TGA of coal. As shown in Figure 2, TGA was also performed to determine the fundamental combustion characteristics of the eight types of coal. In addition, a DTG curve was derived from the result of TGA to clearly distinguish between the starting and ending burning temperature. As shown in Figure 3, the Arrhenius plot of the TGA results was constructed to determine the activation energy. Ng et al.12 analyzed pore volume depending on coal type and were able to explain the correlation between morphological characteristics and rank. Therefore, to determine the effects on the morphological and fundamental combustion characteristics of coal, the results of the Brunnauer-Emmett-Toller (BET) and Barett-Joyner-Halenda (BJH) analyses as shown in Figure 3 were arranged into Table 2. Anthracite and bituminous coal of high rank generally possess high activation energy, whereas pitch coal and lignite have low activation energy. This observation highlights a direct correlation between the starting and ending burning temperatures. It was confirmed that the starting and ending burning temperatures were in the low temperature range for pitch coal and lignite and were relatively high for bituminous coal and anthracite. Therefore, the starting and ending burning temperatures decreased as the amount of VM in the coal was increased. Furthermore, Ng et al.12 reported that the specific surface area of the BET analysis did not correlate with coal rank, whereas pore volume showed close correlation. As shown in Table 2, the specific surface areas shared similarity based on the different pore size and distribution of the coals. However, the average pore size of coal was directly related with rank because it is based on the formation mechanism.

Table 2. Combustion characteristics of each coal using TGA and morphological analysis of coals

Coals	T v (°C)	T f (°C)	Activation E (kJ/mol)	VM/(VM + FC)	Total H2O	Pore Volume (cm^3/g)	Pore Size (nm)	DTG Peak Temperature (°C)
Anthracite	310	630	197.713	0.20993	5.02	0.0069	10.717	655
C&A	327	615	137.6152	0.387836	7.52	0.0100	15.105	578
Anglo	300	610	107.289	0.385258	13.77	0.0181	11.940	572
Shinwha	300	850	46.19678	0.407611	23.61	0.0170	6.2043	661
Berau	250	550	59.71546	0.48249	13.11	0.0109	22.821	498
SM	250	490	82.29758	0.515327	26.34	0.0223	26.133	397
KBB1	250	510	52.27732	0.458947	34.21	0.0282	23.926	488
KBB9	250	490	59.33303	0.75456	26.53	0.0209	17.220	451
Notes: VM = volatile matter; FC = fixed carbon.

Figure 2. TGA for non-isothermal combustion of various coals.

Figure 3. DTG profiles (a) and Arrhenius plot (b) through the TGA for various coals.

NO_x Emission Characteristics of Coals Using B1 Combustion

The B1 combustion method was used to examine NO_x generation characteristics from fuel N of each coal type. According to the authors' previous research, B1 combustion produced data on concentration behavior depending on the burning time and temperature of target coals. The mass of total generated NO_x was then determined. Figure 4 shows the NO_x generated by the B1 combustion of coal. Because the theoretical air content of coal was not considered, NO_x emitted to burn coals using various feeding rate of gases (0.5–1.5 L/min) is represented. The results show that the same amount of NO_x was emitted as the amount of N content in the coals. The amount of emitted NO_x was similar, beginning at 1 L/min, depending on the injected gases. With an increase in the amount of fed gases, the amount of NO_x generation was increased at below 1 L/min. However, the amount of generated NO_x converged toward a certain amount above 1 L/min. This trend appeared for every type of coal used in the experiment and could be explained by the fact that the amount of fuel N dissociated at the corresponding temperature was limited.

Figure 4. Outlet NO concentration on coal combustion according to feeding flows used B1 combustion (coal: 0.2 g).

To further investigate the suppression of NO_x generation below 1 L/min of injected gas, the intensity of N₂ generated under the same combustion conditions was measured using quadrupole mass (Pfeiffer Vacuum GmbH, Asslar, Germany) and is represented in Figure 5. It was confirmed that N₂ generation below 1 L/min of injected gas was relatively larger compared with conditions above 1 L/min. The authors previously reported8 that generated NO_x was decreased by the reduction of NO to N₂ by hydrocarbon intermediate generation material existing in the fuel-rich zone (λ < 1)13; in other words, the reaction progressed along path 1 in Figure 6. Likewise, the results of this research show that NO_x generation was suppressed when gases were injected at a rate below 1 L/min because of the conversion of fuel N to N₂ rather than NO. Therefore, N₂ generation by path 1 inhibits the conversion of fuel N to NO.

Figure 5. Released fuel N to N₂ intensity on coal combustion according to temperatures using B1 combustion [coal: 0.2 g of C&A coal; flow: various (21% O₂/Ar)].

Figure 6. Mechanism of NO_x formation and reduction.

Volatile N and Char N to NO_x

The previous section considered the characteristics of NO_x generation for all eight types of coal. In this section, NO_x generated from volatile N was separated from that generated from char N by examining the ratio of NO_x generated from volatile N among fuel NO_x of coals according to it. Figure 7 is the result, which shows measured NO_x generated to burn by B1 combustion, cooling to room temperature again after drying of each coal sample to increase to the desired temperature by Ar, an inert gas (after production as char at the desired temperature). A certain amount of NO_x was generated by each coal at a dry temperature of 700 °C, because only NO_x generated from char N was emitted after removal of volatile N by Ar. Therefore, NO_x emission by coal without dry processing is derived from volatile N. Each type of coal converted volatile N to NO_x/total fuel N to NO_x, as shown in Table 3. The reason NO_x emission was not compared with the amount of NO_x generated from volatile N but instead was represented as the ratio of volatile N to NO_x and total fuel N to NO_x is that a problem could arise if it is simply compared with the amount because of different N contents, depending on target coal. As shown in Table 3, volatile N to NO_x/total fuel N to NO_x was highly correlated with VM/(VM + FC), which indicates VM content rather than the H₂O content of coal. The plot in Figure 8 was constructed to clarify this concept. As a result, volatile N to NO_x/total fuel N to NO_x was directly related to the coal rank.

Table 3. Volatile matter contents and the NO emission characteristics of various coals

Coals	Volatile N to NO/Total N to NO	VM/(VM + FC)
Anthracite	0.2922	0.20993
C&A	0.3743	0.387836
Anglo	0.5314	0.385258
Shinwha	0.6282	0.407611
Berau	0.3643	0.48249
SM	0.8333	0.515327
KBB1	0.74126	0.458947
KBB9	0.65049	0.75456
Notes: VM = volatile matter; FC = fixed carbon.

Figure 7. Outlet NO (mg) on coal and char combustion according to treated temperatures using B1 combustion (coal: 0.2 g of coals, 21% O₂/Ar 1 L/min).

Figure 8. The correlation between volatile N to NO/fuel N ratio to NO and VM/(VM + FC) (a) and H₂O contents (b) in various coals.

CONCLUSION

Eight types of coal of different rank were selected to determine the fundamental combustion characteristics of the conversion of volatile N and char N to NO_x.

1.	On the basis of the Arrhenius plot and DTG curve of the TGA, the activation energy and starting and ending burning temperatures were derived. Bituminous coal and anthracite of high rank showed relatively high activation energies as well as high starting and ending burning temperatures, whereas pitch coal and lignite showed low activation energy as well as low starting and ending burning temperatures.
2.	When subjected to B1 combustion, emitted NOx was closely related to the N content of coal. In addition, the effect of injected gases showed that NOx emission was suppressed because of the formation of N2 in a reduced atmosphere under air-deficient conditions. Therefore, it was confirmed that the amount of generated NOx converged toward a certain level under air-rich conditions.
3.	The NO conversion ratio of volatile N to fuel N for the coal samples showed similar characteristics, but the conversion of fuel N to NOx converted as the dry temperature was increased. In other words, the conversion of char N to NOx was identical for every type of coal. In addition, the volatile N/fuel N ratio showed a linear relationship with coal rank as opposed to H2O content.