A combined experimental and theoretical study of micronized coal reburning

Abstract

Micronized coal reburning (MCR) can not only reduce carbon in fly ash but also reduce NO_x emissions as compared to the conventional coal reburning. However, it has two major kinetic barriers in minimizing NO_x emission. The first is the conversion of NO into hydrogen cyanide (HCN) by conjunction with various hydrocarbon fragments. The second is the oxidation of HCN by association with oxygen-containing groups. To elucidate the advantages of MCR, a combination of Diffuse Reflection Fourier Transform Infrared (FTIR) experimental studies with Density Functional Theory (DFT) theoretical calculations is conducted in terms of the second kinetic barrier.

FTIR studies based on Chinese Tiefa coal show that there are five hydroxide groups such as OH-π, OH-N, OH-OR₂, self-associated OH and free OH. The hydroxide groups increase as the mean particle size decreases expect for free OH. DFT calculations at the B3LYP/6-31 G(d) level indicate that HCN can be oxidized by hydroxide groups in three paths, HCN+OH→HOCN+H (path 1), HCN+OH→HNCO+H (path 2), and HCN+OH→CN+H₂O (path 3). The rate limiting steps for path 1, path 2 and path 3 are IM2→P1+H (170.66 kJ/mol activated energy), IM1→IM3 (231.04 kJ/mol activated energy), and R1+OH→P3+H₂O (97.14 kJ/mol activated energy), respectively. The present study of MCR will provide insight into its lower NO_x emission and guidance for further studies.

Keywords

hydroxyl radicals / Fourier transform infrared spectroscopy (FTIR) / density functional theory (DFT) / homogeneous reaction mechanism / NO_x

Introduction

Nitrogen oxide emissions from stationary source combustors are large contributors to a number of environmental hazards, including acid rain, high ground-level ozone concentrations, and elevated fine particulate level. Since1980s, researchers from many countries have studied this problem and have developed several successful low NO_x combustion technologies in which the reburning technology is involved[1]. In principle, the reburning process can be divided into three zones in series (shown schematically in Fig. 1), namely: ① the primary combustion zone where approximately 80% of heat is released under fuel lean conditions and NO_x is produced, depending mainly on the fuel, burner type and operating conditions; ② the reburning zone where the reburning fuel, typically about 20% of total thermal input, is injected downstream of the primary combustion zone to create a fuel rich burning zone; the NO_x formed in the primary zone reacts with hydrocarbon radicals from the reburning fuel to produce N₂ and other nitrogenous intermediates, such as hydrogen cyanide (HCN) and NH₃, and ③ the burnout zone where additional combustion air is added to oxidize any remaining fuel fragments and ensure complete combustion.

Gaseous and solid fuels are usually considered in choosing reburning fuels. Compared with gaseous fuels, solid fuels, especially coals, are more economical reburning fuels and easy to be used in coal-burning power plants. More recently, there have been some investigations in which pulverized coals or coal chars were used as reburning fuels [2-6].

In coal reburning systems, the most important variables to minimize NO_x emission are quality and fineness of the coal, air stoichiometry in the primary combustion zone, and air stoichiometry in the reburn zone (SR2). Burch et al. [7,8] and Chen and Ma [9] studied the effect of different reburning fuels on NO reduction and demonstrated that high volatile coal is the ideal reburning fuel, such as bituminous coal and lignitous coal, which can provide more hydrocarbon radicals to reduce NO_x into N₂. Smart and Morgan[10] suggested that optimum SR2 was in the range of 0.75–0.85 independent of reburning fuel type. Liu et al. [4] evaluated thirteen coals and a char as reburning fuels in an isothermal drop-tube reactor system with an interesting finding that increasing the air to fuel ratio in either the primary-zone or the reburning-zone reduces the efficiency of NO reduction and that coals having a high volatile yield on heating are better reburning fuels than lower-volatile coals.

Nowadays, the challenge is to minimize carbon loss while minimizing NO_xemissions, although various investigations have been conducted on coal reburning to minimize NO_xemissions. If reburn coal particle size is the same as that of primary fuel, the unburned carbon (UBC) in fly ash will increase [1]. Micronized coal reburning (MCR) can not only reduce carbon in fly ash but also reduce NO_x emissions as compared to the conventional coal reburning. As a result, it gradually attracted the attention of numerous researchers [1,11]. For macro explanation [12-16], ignition time and temperature decreased when coal particle size was reduced. It is easier to achieve a higher burnout ratio. Moreover, the demanded burnout time is also reduced. Particle size is likely to affect the amount of NO_x formation in pulverized coal flame because of its effect on volatile release time within the particles. In addition, small particles heat up faster and obtain higher temperature than large ones so that the rate of release of volatile matter and nitrogen early in a flame is higher for the smaller particles. But for microscopic explanation, so far not one systematic investigation has been performed yet.

Fourier transform infrared spectroscopy (FTIR) is currently one of the most powerful and versatile techniques for the characterization of coal and related sedimentary materials [17-19]. Quantum chemistry calculations, especially density functional theory (DFT), which can obtain substantially accurate information that cannot be detected experimentally and reproduce the experimental phenomena observed in combustion, have been proven to be of great practicality in understanding reaction mechanisms [20-23]. In the present work, an FTIR spectroscopy has been applied to the characterization and quantitative structural study of Chinese Tiefa (TF) coal to show the surface functional groups variation with coal particle size, and quantum chemical calculations have been carried out to complete a theoretical study of the HCN oxidation in an attempt to give a microscopic explanation for MCR.

Experimental study

Experimental method

The TF coal with the external moisture removed was used as coal samples, which were pulverized into four different mean particle sizes using a jet mill. As analyzed by a Malvern MAM5004 Laser Mastersizer made in UK, the resulting mean particle sizes were TF_6.82 μm, TF_10.21 μm, TF_15.55 μm, and TF_25.10 μm. The ultimate and proximate analyses of the TF bituminous coal are tabulated in Table 1. Approximately 80 mg pulverized KBr were taken and placed in agate mortar with few pulverized coal (the coal/KBr ratio was approximately 1:40). The agate mortar was dried under vacuum with P₂O₅ for 72 hours to remove the absorbed water which might induce spectral perturbations at 3400 and 1630 cm^-1. The FTIR spectra of the coal samples were carried out under the same condition and recorded by co-adding 64 scans on an EQUNOX55, FTIR-Raman spectroscopy (BRUKER Co.).

Experimental results

The FTIR spectra of coal with four different particle sizes are sketched in Fig. 2. The examination of 3000–3700 cm^-1 zone reveals a progressive increase in hydroxide groups with a decrease in particle size [24].

Selected zones of the FTIR spectra (3000–3700 cm^-1) were studied by curve-fitting analysis using Lorentz/Gauss combination with a commercially available data-processing program (PeakFit software). The results are displayed in Fig. 3. Five accepted hydroxide groups (OH-π, OH-N, OH-OR₂, self-associated OH and free OH) are found in the TF coal. Five peak positions observed in the present work are close to those estimated by Miura et al. [19] (as listed in Table 2) except for the OH-N peak. Previous study indicated that this peak enlarged with the increase of N content in the sample. It should be assigned to OH-N. Figure 4 depicts the effect of particle size on five hydroxide groups which demonstrates that OH-π, OH-N, OH-OR₂, self-associated OH on micronized coal surface are accumulated severely, yet the amount of free OH remains constant.

It seems to be almost impossible to remove water effectively especially when an attempt is made to estimate the amount of OH groups from the OH stretching absorption bands around 3000 to 3700 cm^-1, since strong and broad water adsorption bands appear in the region [19]. It is also important to point out that interactions of water with the very fine particles are likely to be very rapid. But, the FTIR experimental results are obviously significant to investigate the hydroxide groups variation with coal particle size because all the TF coal is stored under the same condition and the interactions of water with the coal still maintain when micronized coal is injected into the flame.

Computational study

Computational method

All equilibrium and transition state structures of the proposed reaction channels are fully optimized by Becke’s three-parameter hybrid DFT-Hartree-Fock method [25] using Lee-Yang-Parr’s correlation functional (B3LYP) and the 6-31 G(d) basis set [26,27]. Previous researchers have reached the agreement that B3LYP/6-31 G(d) can maintain a reasonable balance between the accuracy of final results and the computational cost. In addition, it was shown that spin contamination at the B3LYP level of theory is reasonably small [20,21]. Frequency calculation are made at the same level on the optimized structures to characterize them as either equilibrium structures (all real harmonic vibrational frequencies) or transition state(one and only one imaginary vibrational mode corresponding to the reaction coordinate). Intrinsic reaction coordinate (IRC) calculations are executed to confirm the right connection between each transition state and corresponding intermediates or products. To obtain more reliable energy estimates for the potential energy surface of each possible channel, single point energy calculations are performed on CCSD/6-31 G(d) and zero point vibrational energies (ZPE) are taken into account. The Gaussian 03 package is used in this work [28]. Figure 5 presents the optimized geometries of all reactants, intermediates, transition states and products, while Fig. 6 illustrates the imaginary vibration model of each transition state.

Computational results

Three paths (path 1, path 2 and path 3) involving HCN oxidation are proposed, giving birth to HOCN(P1) +H, HNCO(P2) +H and CN(P3) +H₂O, as exhibited in Fig. 7.

OH initiates an attack on HCN in two ways: ① an attack on the C atom to form IM1. ② an attack on the H atom, climbing a barrier of TS5 to generate P3. For path 1, OH initiates an attack on the C atom, resulting in the weakness of C-H bond strength (C-H bond changing from 0.107 to 0.109 nm) and C-N bond strength (C-N bond changing from 0.116 to 0.126 nm), produced with 110.27 kJ/mol exothermic. This exothermic energy makes it easy for IM1 to climb the barrier of TS1 (34.13 kJ/mol) accompanied with H rotation to give rise to IM2. The energy barrier for IM2→P1+H (a C-H bond cleavage) is 170.66 kJ/mol. HOCN (P1) and H are generated with 42.01 kJ/mol endothermic surmounted in path 1. For path 2, the previously generated IM1 can undergo H shift via TS3 to form IM3, a synchronous breakage of C-H bond and conformation of N-H bond. The calculated energy barrier for this step is 231.04 kJ/mol. From IM3, the evolution of the system through TS4 with an energy barrier of 133.90 kJ/mol for the breakage of O-H band leads to HNCO (P2) and H. Path 2 releases 49.88 kJ/mol. For path 3, OH launches an attack on H in HCN to transform into CN and H₂O via TS5. This step has an energy barrier of 97.14 kJ/mol and is 49.88 kJ/mol endothermic.

Figure 8 is an energetic sketch of the stationary points for path 1, path 2 and path 3. The activated energy for IM1→IM2 and IM2→P1+H is 34.13 kJ/mol and 170.66 kJ/mol respectively. Obviously, the rate constant for path 1 depends heavily on IM2→P1+H. The activated energy for IM1→IM3 and IM3→P2+H in path 2 is 231.04 kJ/mol and 133.90 kJ/mol respectively. A rate-controlling step, IM1→IM3, is received. R1+OH→P3+H₂O reaction is the rate-controlling step for path 3. Path 1 competes with path 2, and the activated energy for IM1 to climb along path 1 is significantly lower than that for IM1 to climb along path 2.

Discussion

Many researchers studied reburning mechanisms through elementary chemical reactions [1], and found that, for different reburning fuels, there were two major kinetic pathways to control the efficiency of reburning, i.e.,

(1)\mathrm{C}{\mathrm{H}}_i+\mathrm{N}\mathrm{O}\to \mathrm{H}\mathrm{C}\mathrm{N}+

(2)\mathrm{H}\mathrm{C}\mathrm{N}+\mathrm{O},\mathrm{O}\mathrm{H}\to {\mathrm{N}}_{}+

Based on the mechanisms and sensitivity analysis of Miller and Bowman [29], it seems reasonable to view reburning as possessing two major kinetic barriers. The first barrier is the conversion of NO into HCN, by combination with various hydrocarbon fragments (CH_i) as in

(3)\mathrm{C}+\mathrm{N}\mathrm{O}\to \mathrm{C}\mathrm{N}+\mathrm{O}

(4)\mathrm{C}\mathrm{H}+\mathrm{N}\mathrm{O}\to \mathrm{H}\mathrm{C}\mathrm{N}+\mathrm{O}

(5)\mathrm{C}{\mathrm{H}}_{}+\mathrm{N}\mathrm{O}\to \mathrm{H}\mathrm{C}\mathrm{N}\mathrm{O}+\mathrm{H}

The second major kinetic obstacle appears to be oxidation of HCN via one of the following reactions:

(6)\mathrm{H}\mathrm{C}\mathrm{N}+\mathrm{O}\mathrm{H}\to \mathrm{H}\mathrm{O}\mathrm{C}\mathrm{N}+\mathrm{H}

(7)\mathrm{H}\mathrm{C}\mathrm{N}+\mathrm{O}\mathrm{H}\to \mathrm{H}\mathrm{N}\mathrm{C}\mathrm{O}+\mathrm{H}

(8)\mathrm{H}\mathrm{C}\mathrm{N}+\mathrm{O}\to \mathrm{N}\mathrm{C}\mathrm{O}+\mathrm{H}

(9)\mathrm{H}\mathrm{C}\mathrm{N}+\mathrm{O}\to \mathrm{N}\mathrm{H}+\mathrm{C}\mathrm{O}

For the first barrier, the salient features of MCR can be interpreted. Small particles yield various hydrocarbon fragments (CH_i) to form HCN in the fuel-rich condition. Besides, small particles heat up faster and release more volatile when injected into the reburning zone. For the second barrier, the advantages of MCR can be explained as follows: ① The superficial hydroxide groups (OH-π, OH-N, OH-OR₂ and self-associated OH) increase with the decrease of coal particle size. ② OH concentrations in the gas phase will be higher in the reburning zone if small particles are injected into a flame, since OH groups on the coal surface are easier to be broken than the carbon skeleton. ③ OH as a main oxygen-containing functional group has a strong chemical activity to enliven micronized coal and to furnish favorable condition for HCN oxidation via path 1, path 2 and path 3. Once oxidation of HCN is accomplished, the subsequent conversion of products to N atoms is rapid. N atoms are then recycled to form NO or react with NO into form N₂.

Conclusions

A compact experimental study based on Chinese Tiefa bituminite was investigated using FTIR spectra to find out the effect of particle size on surface hydroxide groups. A systematic theoretical study using a nonlocal hybrid B3LYP DFT was conducted to investigate the HCN oxidation in the reburning zone. The combination of FTIR experimental studies with DFT theoretical calculations on micronized coal reburn is attempted to provide an insight into the characteristics of lower NO_x emissions on MCR from a microscopic point of view. Based on the data acquired from the experiments and calculations, the following conclusions can be reached:

1) Surface hydroxide groups (OH-π, OH-N, OH-OR₂, self-associated OH) experience a tremendous variation with the decrease of coal particle size. Instead, the amount of free OH remains unvaried.

2) HCN can be oxidized by OH in three paths. Rate constant for path 1, path 2 and path 3 depends heavily on the steps of IM2→P1+H(170.66 kJ/mol activated energy), IM1→IM3(231.04 kJ/mol activated energy), and R1+OH→P3+H₂O(97.14 kJ/mol activated energy), respectively.

3) MCR is convenient to create a higher OH-containing condition for HCN oxidation via path 1, path 2 and path 3. Once oxidation of HCN is accomplished, NO will be reduced easily and quickly.

References

[1] Smoot L D, Hill S C, Xu H. NO_x control through reburning. Progress in Energy and Combustion Science, 1998, 24(5): 385-408

[2] Spliethoff H, Greul U, Rüdiger H, Hein K R G. Basic effects on NO emissions in air staging and reburning at a bench-scale test facility. Fuel, 1996, 75(5): 560-564

[3] Zhong B J, Shi W W, Fu W B. Effects of fuel characteristics on the NO reduction during the reburning with coals. Fuel Processing Technology, 2002, 79(2): 93-106

[4] Liu H, Hampartsoumian E, Gibbs B M. Evaluation of the optimal fuel characteristics for efficient NO reduction by coal reburning. Fuel, 1997, 76(11): 985-993

[5] Li S, Xu T M, Zhou Q L, Tan H Z, Hui S E, Hu H L. Optimization of coal reburning in a 1 MW tangentially fired furnace. Fuel, 2007, 86(7,8): 1169-1175

[6] Luan T, Wang X D, Hao Y Z, Cheng L. Control of NO emission during coal reburning. Applied Energy, 2009, 86(9): 1783-1787

[7] Burch T E, Tillman F R, Chen W Y, Lester T W, Conway R B, Sterling A M. Partitioning of Nitrogenous species in the fuel-rich stage of reburning. Energy & Fuels, 1991, 5(2): 231-237

[8] Burch T E, Chen W Y, Lester T W, Sterling A M. Interaction of fuel nitrogen with nitric oxide durning reburning with coal. Combustion and Flame, 1994, 98(4): 391-401

[9] Chen W Y, Ma L. Effect of heterogeneous mechanisms during reburning of nitrogen oxide. AIChE Journal. American Institute of Chemical Engineers, 1996, 42(7): 1968-1976

[10] Smart J P, Morgan D J. The effectiveness of multi-fuel reburning in an internally fuel-staged burner for NO_x reduction. Fuel, 1994, 73(9): 1437-1442

[11] Maly P M, Zamansky V M, Ho L, Payne R. Alternative fuel reburning. Fuel, 1999, 78(3): 327-334

[12] Zhang C Q, Jiang X M, Wei L H, Wang H. Research on pyrolysis characteristics and kinetics of super fine and conventional pulverized coal. Energy Conversion and Management, 2007, 48(3): 797-802

[13] Jiang X M, Zheng C G, Yan C, Liu D C, Qiu J R, Li J B. Physical structure and combustion properties of super fine pulverized coal particle. Fuel, 2002, 81(6): 793-797

[14] Jiang X M, Zheng C G, Qiu J R, Li J B, Liu D C. Combustion characteristics of super fine pulverized coal particles. Energy & Fuels, 2001, 15(5): 1100-1102

[15] Jiang X M, Huang X Y, Liu J X, Han X X. NO_x Emission of fine- and superfine-pulverized coal combustion in O₂/CO₂ atmosphere. Energy & Fuels, 2010, 24(12): 6307-6313

[16] Maier H, Spliethoff H, Kicherer A, Fingerle A, Hein K R G. Effect of coal blending and particle size on NO_x emission and burnout. Fuel, 1994, 73(9): 1447-1452

[17] Chen C, Gao J S, Yan Y J. Observation of the type of hydrogen bonds in coal by FTIR. Energy & Fuels, 1998, 12(3): 446-449

[18] Li D T, Li W, Chen H K, Li B Q. The adjustment of hydrogen bonds and its effect on pyrolysis property of coal. Fuel Processing Technology, 2004, 85(8-10): 815-825

[19] Miura K, Mae K, Li W, Kusakawa T, Morozumi F, Kumano A. Estimation of hydrogen bond distribution in coal through the analysis of OH stretching bands in diffuse reflectance infrared spectrum measured by in-situ technique. Energy & Fuels, 2001, 15(3): 599-610

[20] Sendt K, Haynes B S. Density functional study of the chemisorption of O₂ on the zig-zag surface of graphite. Combustion and Flame, 2005, 143(4): 629-643

[21] Sendt K, Haynes B S. Density functional study of the reaction of carbon surface oxides: The behavior of ketones. The Journal of Physical Chemistry A, 2005, 109(15): 3438-3447

[22] Kyotani T, Tomita A. Analysis of the reaction of carbon with NO/N₂O using Ab initio molecular orbital theory. Journal of Physical Chemistry B, 1999, 103(17): 3434-3441

[23] Zhu R S, Lin M C. The NCO+NO reaction revisited: Ab initio MO/VRRKM calculations for total rate constant and product branching ratios. Journal of Physical Chemistry A, 2000, 104(46): 10807-10811

[24] Sevilla M, Fuertes A B. Chemical and structural properties of carbonaceous products obtained by hydrothermal carbonization of saccharides. Chemistry (Weinheim an der Bergstrasse, Germany), 2009, 15(16): 4195-4203

[25] Becke A D. A new mixing of Hartree-Fock and local density functional theories. Journal of Chemical Physics, 1993, 98(2): 5648-5652

[26] Lee C, Yang W, Parr R G. Development of the Colle-Salvetti correction-energy formula into a functional of electron density. Physics Letters. Part B, 1988, 37: 785-789

[27] Miehlich B, Savin A, Stoll H, Preuss H. Results obtained with the correlation energy density functionals of becke and Lee. Yang and Parr. Chemical Physics Letters, 1989, 157(3): 200-206

[28] Frisch M J, Trucks G W, Schlegel H B, Scuseria G E, Robb M A, Cheeseman J R, Zakrzewski V G, Montgomery J A Jr, Stratmann R E, Burant J C, Dapprich S, Millam J M, Daniels D, Kudin K N, Strain M C, Farkas O, Tomasi J, Barone V, Cossi M, Cammi R, Mennucci B, Pomelli C, Adamo C, Clifford S, Ochterski J, Petersson G A, Ayala P Y, Cui Q, Morokuma K, Malick D K, Rabuck A D, Raghavachari K, Foresman J B, Cioslowski J, Ortiz J V, Baboul A G, Stefanov B, Liu G, Liashenko A, Piskorz P, Komaromi I, Gomperts R, Martin R L, Fox D J, Keith T, Al-Laham M A, Peng Y, Nanayakkara A, Gonzalez C, Challacombe M, Gill P M W, Johnson B, Chen W, Wong M W, Andres J L, Gonzalez C, Head-Gordon M, Replogle E S, Pople J A. Gaussian03 Revision D 01, Gaussian Inc: Wallingford CT, 2004

[29] Miller J A, Bowman C T. Mechanism and modeling of nitrogen chemistry in combustion. Progress in Energy and Combustion Science, 1989, 15(4): 287-338

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg