Improving Combustion Characteristics and NOx Emissions of a Down

ARTICLE pubs.acs.org/est

Improving Combustion Characteristics and NOx Emissions of a Down-Fired 350 MWe Utility Boiler with Multiple Injection and Multiple Staging Min Kuang, Zhengqi Li,* Shantian Xu, and Qunyi Zhu School of Energy Science and Engineering, Harbin Institute of Technology, 92 West Dazhi Street, Harbin 150001, P.R. China

bS Supporting Information ABSTRACT: Within a Mitsui Babcock Energy Limited down-fired pulverized-coal 350 MWe utility boiler, in situ experiments were performed, with measurements taken of gas temperatures in the burner and near the right-wall regions, and of gas concentrations (O2 and NO) from the near-wall region. Large combustion differences between zones near the front and rear walls and particularly high NOx emissions were found in the boiler. With focus on minimizing these problems, a new technology based on multiple-injection and multiple-staging has been developed. Combustion improvements and NOx reductions were validated by investigating three aspects. First, numerical simulations of the pulverized-coal combustion process and NOx emissions were compared in both the original and new technologies. Good agreement was found between simulations and in situ measurements with the original technology. Second, with the new technology, gas temperature and concentration distributions were found to be symmetric near the front and rear walls. A relatively low-temperature and high-oxygen-concentration zone formed in the near-wall region that helps mitigate slagging in the lower furnace. Third, NOx emissions were found to have decreased by as much as 50%, yielding a slight decrease in the levels of unburnt carbon in the fly ash.

’ INTRODUCTION NOx is an extremely toxic pollutant harmful to human health and detrimental to the atmosphere. Its main source derives from primary emissions of coal-fired power plants into the air.1,2 Reserves of anthracite and low-volatile coal are abundant and globally distributed. With their low volatile content, anthracite and low-volatile coal present difficulties in ignition and burnout. Down-fired boilers are specifically designed to burn anthracite and low-volatile coal, but unfortunately, a larger quantity of NOx is produced in these boilers than in tangential-fired furnaces and wall-arranged boilers. Four types of down-fired utility boilers are manufactured: the Foster Wheeler (FW), the Babcock & Wilcox (B&W), the Stein, and the Mitsui Babcock Energy Limited (MBEL) down-fired boilers. The Supporting Information summarizes these four types and the main differences among these. Most of these boilers suffer similarly from problems of poor stability, low burnout, and high NOx emissions. For FW boilers, reports have appeared on aerodynamic characteristics, combustion, slagging, and NOx reductions by overfire-air application.35 Fan et al. investigated the combustion characteristics and NOx formation within a 300 MWe B&W down-fired boiler by a numerical simulation approach.6 Burdett carried out industrial tests to investigate the effects of air staging on NOx emissions from a 500 MWe Stein down-fired boiler unit.7 However, little research has been reported on MBEL down-fired boilers, which r 2011 American Chemical Society

suffer similarly from problems of large differences in combustion characteristics between front- and rear-wall zones,8 high NOx emissions typically in the range 11001500 mg/m3 at 6% O2 but sometimes as high as 1700 mg/m3,8,9 and serious slagging in the lower furnace.10 According to previous results obtained from industrial tests on a full-scale furnace and cold airflow experiments within a small-scale model of a 300 MWe MBEL boiler, Li et al. discovered that a deflected flow field appearing in the lower furnace resulted in large combustion differences between zones near the front and rear walls, and that NOx emissions could only be reduced by 13% if the staged-air damper opening were fully opened from an initial 30% level.8 Moreover, an adjustment of the feed direction of staged-air to a 45° declination, instead of the original horizontally fed direction, inhibits formation of this deflected flow field.11 To obtain significant reductions of NOx, modifications of the combustion system are generally less costly. Of these modifications, air staging and overfire air (OFA) are more mature and widely used. However, the resulting low-NOx combustion environment by application of deep air staging or OFA usually leads to Received: October 27, 2010 Accepted: March 15, 2011 Revised: March 15, 2011 Published: March 23, 2011 3803

dx.doi.org/10.1021/es103598f | Environ. Sci. Technol. 2011, 45, 3803–3811

Environmental Science & Technology

ARTICLE

Figure 1. Schematics of the furnace and combustion system of the down-fired boiler (dimensions in mm).

high carbon levels in fly ash that is financially disadvantageous for power plant management.1214 For a down-fired boiler, there have to date been only two reports on low-NOx combustion retrofitting in FW boilers, one called “vent-to-OFA”15 and the other called “combined high efficiency and low-NOx technology”.3 A detailed description of the two technologies is given in the Supporting Information. For MBEL down-fired boilers, no technology has as yet been reported on significant reductions in NOx emissions, nor anything comprehensive enough that would also eliminate asymmetric combustion and minimize slagging in the lower furnace. In this paper, we propose a new deep-air-staging technology specifically for a MBEL down-fired pulverized-coal 350 MWe utility boiler, and present numerical simulations of the pulverized-coal combustion process and NOx emissions that verify that this technology improves combustion and reduces NOx emissions.

’ METHODOLOGY Utility Boiler. Figure 1 presents schematics of the furnace and combustion system with the prior MBEL art. Introductory material about the furnace, combustion system and air distribution can be found in the Supporting Information.

In Situ Experimental Measurements. In situ experiments on the full-scale boiler were performed at full load with measurements of gas temperatures taken in both the burner and right-wall regions, and of gas concentrations (O2 and NO) in the near-wall region. During experiments, coal sampling was performed at the coal hopper exits. The determination of total nitrogen in the coal sample was conducted using the semimicro Kjeldahl method, with measurement errors of less than 2%. Coal characteristics and operating conditions during experiments are listed in Table 1. The methods used in measurement taking are described in the Supporting Information, much of which has already appeared in literature.8 The New Technology. As illustrated in Figure 2, the method that realizes this new technology is as follows: (1) Fuel-rich coal/ air mixture is channeled vertically down through nozzles centered over the furnace, while fuel-lean coal/air mixture is similarly injected near the front and rear walls. The fuel-rich coal/air mixture nozzles are short and rectangular; these are arranged side-by-side in pairs so as to postpone the decay in the fuel-rich coal/air mixture. (2) Two rows of short secondary-air ports are located on arches across the breadth of the furnace, viz. one row of so-called inner secondary-air ports positioned between the fuel-rich and fuel-lean coal/air mixture nozzles and another row of so-called outer secondary-air ports located near to the front 3804

dx.doi.org/10.1021/es103598f |Environ. Sci. Technol. 2011, 45, 3803–3811


ARTICLE

Table 1. Coal Characteristics, Operational Parameters, Experimental Data, and Simulation Results coal characteristics proximate analysis, wt % (as received)

ultimate analysis, wt % (as received)

volatile matter

ash

moisture

fixed carbon

net heating value (kJ/kg)

C

H

S

N

O

9.66

31.50

6.7

52.14

20500

54.25

2.51

1.35

0.68

3.01

operation parameters, experimental details, and simulation results case coal feed rate

experiment

simulation (original)

simulation (MIMSC)

42.62

42.62

42.62

(kg/s) fuel-rich flow

fuel-lean flow

temperature (°C)

138

138

138

flow rate (kg/s)

25.50

25.50

25.50

velocity (m/s)

11.02

11.02

15.00

coal concentration (kg/kg)

1.50

1.50

1.50

temperature (°C)

138

138

138

flow rate (kg/s)

25.50

25.50

25.50

velocity (m/s)

23.39

23.39

23.39

coal concentration

0.17

0.17

0.17

temperature (°C)

323

323

323

flow rate (kg/s) velocity (m/s)

276.94 50.90

276.94 50.90

162.30 50.90

temperature (°C)

323

323

323

flow rate (kg/s)

53.46

53.46

96.96

velocity (m/s)

25.45

25.45

50.90

temperature (°C)

323

323

(kg/kg) secondary air

staged-air

near-wall air

OFA

flow rate (kg/s)

6.44

6.44

velocity (m/s)

10

10

temperature (°C) flow rate (kg/s)

323 77.56

velocity (m/s)

50.90

flow rate of main steam (ton/h)

1080

NOx in flue gas (mg/m3 at 6% O2 dry)

1682

1840

923

carbon in fly ash (%)

3.07

5.56

5.51

and rear walls. Accordingly, secondary air is fed through arches in a two-stage manner and a first combustion stage forms in the burner region. The near-wall air is then canceled by the outer secondary air adjacent to the front and rear wall. (3) Staged-air slots are arranged uniformly on the front and near walls across the breadth of the furnace, and a high staged-air ratio (25%) is established to lower the stoichiometry to close to 0.65 in the zone above the staged-air. Next, a second combustion stage is formed. Again, staged-air is fed at a set declination of 45°. (4) In the secondary-air box near the furnace throat, direct-flow OFA ports are arranged symmetrically on the front and near arches across the breadth of the furnace with a declination of 40°, so as to lower the stoichiometry to nearly 1.0 in the zone below the OFA. With this technique, a third combustion stage forms in the furnace throat region. It is necessary to point out that the proposed technology is based on the injection mechanism derived from the Bernoulli’s principle, and defined as follows: The injection mechanism is such that the higher the airflow velocity, the lower the airflow static pressure. Thus for the two parallel airflows injected into the furnace, a staticpressure difference arises that depends on the difference in airflow

velocities. Owing to the static-pressure difference, the lower-velocity airflow is deflected toward and gradually mixes with the highervelocity airflow. Carried by the higher-velocity airflow, the lowervelocity airflow can now penetrate further into the furnace. Here, the technology solving the above-mentioned problems is based on three levels of air injection and comprehensive deep-staged combustion. Because of the presence of multiple airflow injections under deep air staging, we call the new technology “multi-injection multi-stage combustion” (MIMSC). To understand more clearly the underlying principle behind MIMSC technology, Figure 2 presents a schematic diagram of the multi-injection airflow trajectories. For a combustion system with this technology, the design parameters, such as airflow rate and velocity, are listed in Table 1. Numerical Simulation Method. A commercial computational fluid dynamics code (Fluent version 6.3.26),16 was used in conducting the simulations. In this work, gas turbulence was specifically taken into account by the so-called realizable kε model. The Lagrangian stochastic tracking model was applied to analyze the gas/particle flow field, while calculations of gas/ particle two-phase coupling employed the particle-source-in-cell 3805



ARTICLE

Figure 2. Schematics of the combustion system of the down-fired boiler applying the “multi-injection multi-stage combustion” (MIMSC) technology.

method. Radiation was described using the P-1 model, and devolatilization was modeled with the two-competing-rate Kobayashi model. The combustion of volatiles was modeled by employing probability density function theory, and char combustion was modeled by employing a diffusion/kinetics model. More detailed descriptions of these models can be found in the literature.17 The formation of NOx includes thermal-NOx and fuel-NOx, but little prompt-NOx. Here, only the production of NO was taken into account, because NOx emitted into the atmosphere from combusting fuels consists mostly of NO, with there being much lower concentrations of NO2 and N2O. The concentration of thermal-NOx was calculated using the extended Zeldovich mechanism (specifically, N2 þ O f NO þ N, N þ O2 f NO þ O, N þ OH f NO þ H).18 The fuel-NOx concentration was calculated using De Soete’s model.19 The formation of prompt-NOx was neglected in calculations. Table S1 in the Supporting Information presents details of the mathematical model and parameters used in the numerical simulations. A reflection symmetry in the furnace design allows simulations to be performed over a computation domain corresponding to the right half of the furnace. The operational parameters were set to simulate those at the full load, with coal characteristics that are the same as those in the in situ experiments (Table 1). A summary of grid divisions, calculation methods, boundary conditions, and other important settings in the simulations can be found in the Supporting Information.

’ RESULTS AND DISCUSSION In Situ Experimental Results and Model Validation in Simulations. Figure 3 presents distribution plots of gas tem-

peratures and component concentrations acquired from in situ

experiments and simulations set to prior MBEL technology. As seen in Figure 3a, due to secondary air feeding, gas temperatures recorded within 1.0 m from the outlet along the sight port pipe are all below 600 °C; these then rise quickly with increasing distance, reaching 900 °C at 2.0 m. This indicates a delay in the ignition of pulverized coal. This is because the fuel-rich coal/air mixture nozzles and the secondary air ports are closely positioned in an alternating arrangement. After leaving the port outlets, fast secondary air mixes rapidly with the slow fuel-rich coal/air mixture, and thus affects ignition.8 Figure 3b shows clearly that the combustion characteristics in the lower furnace are asymmetric, with gas near the front wall (ports 1 and 2) being clearly hotter than those near the rear wall (ports 3 and 4). An explanation of this phenomenon can be found in the following discussion on the asymmetric flow seen in Figure 4a. As shown in Figure 3c, O2 concentrations are a little higher in the zone below the burner (at distances beyond 0.8 m from the wing wall, Figures 1 and 3c) than those near the wing wall; the reverse situation occurs with NO concentrations. Both O2 and NO concentrations are at relatively high levels; viz. average values of 9% and 850 ppm, respectively. This is because in this region not far from the fuel-rich coal/air mixture nozzle outlet (at nearly 2.6 m, see Figure 1), pulverizedcoal is at an early stage of ignition and burns in an oxygen-rich atmosphere, because of the rapid mixing of secondary air with the fuel-rich coal/air mixture immediately after leaving the port outlets.8 Thus, a large quantity of fuel-NOx is produced, although the highest gas temperature in the zone is only about 1200 °C. As listed in Table 1, NOx emissions are as high as 1682 mg/m3 at 6% and carbon in fly ash is at a low level of about 3%, although asymmetric combustion occurs. The high NOx emissions and asymmetric combustion are in accordance with the common problems already mentioned in the Introduction. 3806



ARTICLE

Figure 3. Results of in situ measurements and comparison with calculated values.

Simulation results based on prior MBEL technology were compared with those acquired in in situ experiments. As shown in Figure 3, calculated gas temperatures in the burner region and near the right wall, as well as O2 and NO concentrations, are all consistent with measured values. The temperatures near the right wall show similar asymmetric combustion characteristics to those of the lower furnace. The calculated NOx emissions listed in Table 1 also show good agreement with those measured, with a 9.4% discrepancy in NOx concentrations in flue gas. However, the result for carbon in fly ash is not satisfactory, with an absolute difference of 2.49% between measured and calculated values of 3.07% and 5.56% respectively. As the dominant component of NOx, NO in flue gas was 1578 mg/m3 (6% O2 dry) in the in situ measurement and 1722 mg/m3 (6% O2 dry) in simulations, with a calculation error of 9.1%. That is, the level of accuracy achieved in the computations is very good in view of the fact that they are not using parameters that exactly match the coal used. This indicates that the models adopted in the present study are suitable for investigating the effects of MIMSC technology on pulverized-coal combustion processes and NOx formation within the furnace. Comparison of Simulated Results with Prior MBEL and MIMSC Technologies. The calculated flow field, gas temperature field, and O2 and NOx concentration fields with the original and the MIMSC technologies are extracted and presented in panels ad, respectively, of Figure 4. These are taken along the

longitudinal cross-section intersecting the vertical centerline of one of the fuel-rich coal/air mixture nozzles. As shown in Figure 4a, there is for the original technology a deflected flow field in the lower furnace, with the downward coal/airflow near the rear wall clearly not penetrating as deep as that near the front wall. The downward coal/airflow near the front wall penetrates deep into the middle and lower parts of the dry bottom hopper region, from where it deflects upward along the right side of the hopper, filling most of the area in the hopper. After reaching a height level with the staged-air slots near the rear wall, this deflected coal/airflow emanating from the zone near the front wall is redirected upward toward the furnace center, and then deflects toward the throat region near the front arch, below the coal/airflow extruding from the rear arch and wall. Due to the high momentum flux of this downward coal/air flow, the initial horizontally directed staged-air near the front wall is, after leaving the port outlet, redirected downward along the left side of the hopper. Lifted by upward gas below the staged-air slot outlet region and suppressed by downward coal/airflow, staged air in the rear-wall zone maintains its initial horizontal direction at a small distance from the slot outlet and is then redirected slightly downward. Blocked by the staged air and lifted by the upward gas near the staged-air zone, the downward coal/airflow near the rear wall deflects upward toward the furnace center after reaching the upper part of the staged-air zone. A recirculation zone of hot gas 3807



ARTICLE

Figure 4. Comparison with calculated flow fields, temperature fields, and gas components of the two technologies.

appears in the zone below the front and rear arches. This greatly benefits ignition of the pulverized-coal. Also appearing in the upper furnace is an asymmetric flow with a proportion of the upward flue gas flow being deflected toward the rear-wall zone. It is this deflected flow field that leads to asymmetric combustion, as revealed in in situ experiments. The dominant coal/ airflow in the front-wall zone takes a long travel path, and hence pulverized-coal particles have long residence times in the lower furnace, releasing lots of heat; in contrast, the opposite occurs for weak coal/airflow in the rear-wall zone. Thus, gas temperatures measured in the front-wall zone are higher than those near the rear wall (see Figure 3b). Moreover, airflow washes over the walls at a high speed as it moves into the region approaching the front and rear walls; this accounts for slagging on the front and rear walls. For the MIMSC technology, a perfectly symmetric flow field appears not only in the lower furnace, but also in the straight section

of the upper furnace, where downward coal/airflows near front and rear walls have been diverted upward, having just penetrated middle and lower parts of the dry bottom hopper. The coal/airflows emanating from arches penetrate downward into the lower furnace adjacent to the front and rear walls. Meanwhile, staged-air flowing downward at high speed along the dry bottom hopper wall forms a distinctive airflow layer covering the wall. This layer serves to prevent fast particulates washing over the hopper walls, thus avoiding slagging in the dry bottom hopper. In contrast with the original technology, here the recirculation zone below the arches swells up from the arch to the zone above the upper edge of the staged-air port, enabling entrainment of more high-temperature flue gas and facilitating ignition. Fast OFA jets reach the uppermost part of the lower furnace region after leaving the port outlet and then mix well with the upward gas in the throat region. An explanation of the formation of the 3808


Environmental Science & Technology asymmetric and symmetric flow fields from the respective original and MIMSC technologies is described in the Supporting Information. Figure 4b presents calculated contours of gas temperatures. For the original technology, these temperatures increase initially, but then decrease in the zone below the arches. Cashew-shaped hot regions occur in the front- and rear-wall zones. There, gas temperatures reach as high as 1900 K, but not in the central part of the furnace. In this high temperature zone, slagging may occur due to fast flowing particulates washing over the walls (Figure 4a). Because of the deflected flow field formation, asymmetric combustion appears with the high-temperature zone near the front wall penetrating the upper part of the hopper region, but that of the rear-wall being truncated above the stagedair slot. The asymmetric low-temperature contours in the hopper region are attributed to the presence of coal/airflow emanating from the front-wall zone (Figure 4a). In general, calculated gas temperatures are higher in the front half of the furnace than in the rear, in agreement with measured results (Figure 3b). The exception is the similar temperatures in the zones below the front and rear arches. This may be attributable to simulation error. For the MIMSC technology, a perfectly symmetric gas temperature field appears not only in the lower furnace, but also in the straight section of the upper furnace, because of the formation of the symmetric W-shaped flow field. Three high-temperature zones occur in the lower furnace; two smaller zones are located in the region near the front and rear walls below arches, but not far away from the nozzle outlet, and a larger wedgedshaped zone is located in the central part of the furnace extending from the throat to the lower part of the hopper region. Although the highest gas temperature in the high-temperature zones below the arches is a little lower than that in the original configuration (viz. 1800 K vs 1900 K), the negative effect on burnout would be slight, because gas temperature levels in the lower furnace are higher than those in the original configuration. The different coal combustion intensities in the burner region explain this temperature difference between the original and MIMSC technologies. With the original technology, the rapid mixing of secondary air with the fuel-rich coal/air mixture results in the combustion of pulverized coal in an oxygen-rich atmosphere. Thus, coal combustion is intense in the burner region. However, under the gradual mixing of secondary air with the fuel-rich coal/air mixture in MIMSC technology, a particularly low-O2-concentration zone (2%, see Figure 4c) through which pulverized coal passes, is just below the fuel-rich flow nozzle outlet and extends for a long distance from the nozzle outlet. Pulverized coal burns in a particularly oxygen-lean atmosphere for quite some time. Therefore, a restrained combustion develops, and the highest gas temperatures appear further away than those in the original. These changes in the calculated O2 concentration (Figure 4c), in addition to rises in gas temperatures below the arches in each technology, validate the above explanation. Secondary air ejected into the lower furnace adjacent to the front and rear walls and staged-air flowing downward along the dry bottom hopper wall, screened by the airflow layer, cause gas temperatures in the nearwall region to be lower, thus promoting a reduction in slagging in the lower furnace. Figure 4c presents calculated O2 concentrations. For the original technology, O2 concentrations also showed an asymmetric distribution along the furnace center, with low-O2-concentration regions being higher in the front-half than in the rear-

ARTICLE

half of the furnace. This is because of asymmetric combustion, for which flame penetration is deeper and more O2 is being consumed during combustion in the front-half than in the rearhalf. In the high temperature zones below the arches and near the front and rear walls, as seen in Figure 4b, O2 concentrations decrease with flame penetration depth, with the highest concentrations (812%) located near the nozzle outlet. This is a consequence of premature mixing of secondary air with the fuelrich coal/air mixture. Because secondary air carries the coal/airflow that washes over the walls (Figure 4a), relatively low O2 concentrations (68%) appear in the front and rear wall zones. Thus, a high-temperature low-oxygen region forms (Figure 4b) that facilitates slagging there. High O2 concentrations adhering to the left wall of the hopper are attributed to staged-air near the front wall, while the low concentrations over most of the right-side hopper region results from deflected flue gas emanating from the zone near the front wall. With the MIMSC technology, the O2 concentration field shows a symmetric pattern similar to that for flow and gas temperature fields. High-temperature zones, depicted in Figure 4b, all maintain the lowest O2 concentrations, as a consequence of the intensity of combustion. Because of the injection of outer secondary air, a high O2 concentration (1418%) appears in both front and rear wall zones. This is to say, low-temperature high-oxygen regions form in these zones (Figure 4b) that mitigate slagging on the front and rear walls. Particularly high O2 concentrations appear along the left and right walls of the hopper, due to the airflow layer formed by the downward moving staged-air near the front and rear walls (Figure 4a). Slagging is avoided in the hopper because of low gas temperatures and high O2 concentrations in the hopper-wall zone. In the throat region, where OFA is directed from the arches toward the furnace center and mixes with the upward flue gas, the O2 concentration distribution features a low-level center and two high-level ends. Near the front and rear walls, O2 concentrations decrease rapidly, with gases rising into the upper furnace and finally, due to reaction with the unburnt particles, fall to the same levels as found at the furnace outlet in the original configuration. Thus, burnout cannot be affected, but unfortunately, gas temperatures in the upper furnace rise about 100 K compared with the original (Figure 4b). Figure 4d presents calculated NOx concentrations. For the original technology, the NOx concentration field exhibits asymmetric patterns similar to those for airflow, gas temperature, and O2 concentration fields. Levels are high in the whole furnace, and all but one zone have concentrations greater than 1000 ppm; the exception is the left wall zone of the hopper. The highest NOx concentrations of 1800 ppm appear in the zone below the arches not far from the nozzle outlet, where the combination of highest gas temperatures and O2 concentrations produces copious quantities of thermal- and fuel-NOx. NOx concentrations decrease with flame penetration depth, since O2 concentrations also decrease and the downward coal/airflow mixes with expanding gases. Reasons for the particularly high NOx production are 3-fold: (1) The premature mixing of secondary air with the fuelrich coal/air mixture at the early stage of ignition (see O2 concentrations in Figure 4c). The respective nozzles and ports are arranged in alternating positions and are closely spaced. After leaving the port outlet, secondary air with velocities much higher than that of the fuel-rich coal/air mixture mixes rapidly with the slower flow resulting in early ignition of pulverized-coal in this 3809


Environmental Science & Technology oxygen-rich atmosphere, but produces a large amount of fuelNOx. (2) A shallow staging condition. To ensure high velocities of secondary air and thus guaranteeing a long flame travel by propelling the fuel-rich coal/air mixture into the hopper region, the staged-air ratio (viz., the ratio of staged-air mass flux to total air mass flux into the furnace) must be set to about 13% (calculated in Table 1) and shallow staging conditions must be established in the lower furnace. and (3) High levels of gas temperature in the lower furnace, especially in the zone below arches (Figure 4b) needed for good burnout, facilitates thermalNOx formation. Compared with the original technology, the NOx concentration field formed using the MIMSC technology shows a strong asymmetric pattern, with concentrations throughout the furnace being much lower, viz. values almost one-half those of the original. This is because in MIMSC, mixing of secondary air with the fuel-rich coal/air mixture is clearly postponed and a deep air staging condition is formed with a large increase in the stagedair ratio (25%, compared with the original 13%, calculated in Table 1) and the introduction of OFA. Therefore, pulverizedcoal ignites in an oxygen-lean environment for a long combustion time, thereby reducing NOx production dramatically. Zones with the highest NOx concentrations are also positioned below arches not far from the nozzle outlet. It is well-known that the Zeldovich mechanism does not appear to be a major source of NO in flames at gas temperatures below 1700 °C.17,18 Here, gas temperatures are all below 1700 °C (Figure 4b). Again, gas temperature plays a large role in thermalNOx formation, which involves O2 consumption and NO generation. Therefore, the calculated O2 concentration level in Figure 4c should be higher and the calculated NOx concentration level in Figure 4d would be lower if this block of reaction mechanisms were removed. Table 1 lists the calculated carbon in fly ash and NOx emissions. With the prior MBEL technology replaced by the MIMSC technology, calculated NOx emissions decreased by 50% (viz. from 1840 to 923 mg/m3), and it should be noted that the calculated carbon content in fly ash decreased slightly. This is because, due to the equitable configuration in secondary air, staged-air, and OFA, the downward coal/airflow is propelled deep into middle and lower parts of the hopper region, prolonging residence times of pulverized-coal particles in the lower furnace, and thus enabling efficient fuel ignition, despite a three-level stage combustion being established. In addition, adequate reaction times between OFA and unburnt particles are also beneficial. In conclusion, in applying MIMSC technology as a replacement for the MBEL prior art, the original asymmetric flow field, gas temperature, and gas component distributions that are seen in the furnace all develop a symmetric pattern. Relatively low gas temperatures and high O2 concentrations are found in the front- and rear-wall zones, as well as in the hopper wall zone, thereby mitigating slagging in the lower furnace. NOx emissions can be lowered by as much as 50%, without increasing levels of unburnt carbon in fly ash. This work proposed a MIMSC technology and confirmed through numerical simulations certain beneficial features. The applicability of the new technology will be verified in future work with the modification of in-service MBEL down-fired boilers and new designs.

’ ASSOCIATED CONTENT

bS

Supporting Information. Figure S1, Tables S1 and S2; notes introducing differences among the four types of down-fired

ARTICLE

boilers, details of the furnace and combustion system, methods for in situ experimental measurements, grid divisions and simulation parameter settings. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Tel.: þ86 451 86418854; fax: þ86 451 86412528; e-mail: [email protected].

’ ACKNOWLEDGMENT This work was sponsored by the Hi-Tech Research and Development Program of China (Program 863; Contract 2006AA05Z321). ’ REFERENCES (1) Arenillas, A.; Rubiera, F.; Pis, J. J. Nitric oxide reduction in coal combustion: role of char surface complexes in heterogeneous reactions. Environ. Sci. Technol. 2002, 36 (24), 5498–5503. (2) Yao, M. Y.; Che, D. F.; Liu, Y. H.; Liu, Y. H. Effect of volatile-char interaction on the NO emission from coal combustion. Environ. Sci. Technol. 2008, 42 (13), 4771–4776. (3) Li, Z. Q.; Ren, F.; Chen, Z. C.; Liu, G. K.; Xu, Z. X. Improved NOx emissions and combustion characteristics for a retrofitted downfired 300-MWe utility boiler. Environ. Sci. Technol. 2010, 44 (10), 3926–3931. (4) Li, Z. Q.; Ren, F.; Chen, Z. C.; Chen, Z.; Wang, J. J. Influence of declivitous secondary air on combustion characteristics of a down-fired 300-MWe utility boiler. Fuel 2010, 89 (1), 410–416. (5) Fang, Q. Y.; Wang, H. J.; Wei, Y.; Lei, L.; Duan, X. L.; Zhou, H. C. Numerical simulations of the slagging characteristics in a down-fired, pulverized-coal boiler furnace. Fuel Process. Technol. 2010, 91 (1), 88–96. (6) Fan, J. R.; Zha, X. D.; Cen, K. F. Study on coal combustion characteristics in a W-shaped boiler furnace. Fuel 2001, 80 (3), 373–381. (7) Burdett, N. A. The effects of air staging on NOx emissions from a 500 MW(e) down-fired boiler. J. Inst. Energy 1987, LX, 103–107. (8) Li, Z. Q.; Kuang, M.; Zhang, J.; Han, Y. F.; Zhu, Q. Y.; Yang, L. J.; Kong, W. G. Influence of staged-air on airflow, combustion characteristics and NOx emissions of a down-fired pulverized-coal 300 MWe utility boiler with direct flow split burners. Environ. Sci. Technol. 2010, 44 (3), 1130–1136. (9) Miao, C. X.; Che, G.; Wang, J. W.; Zhang, Z. J.; Cheng, L. Combustion optimization and debugging of 600 MW W-flame boiler. Electric Power 2003, 36 (8), 169 (in Chinese). (10) Han, K. H.; Gao, H.; Zhai, L.; Liu, J.; Wang, D. F. Cause analysis and prevention measures on the slagging in 600MW “W”-flame boiler. Boiler Technol. 2005, 36 (5), 4750 (in Chinese). (11) Kuang, M.; Li, Z. Q.; Han, Y. F.; Yang, L. J.; Zhu, Q. Y.; Zhang, J. Influence of the staged-air declination angle on flow-field deflection in a down-fired pulverized-coal 300 MWe utility boiler with direct-flow split burners. Energy Fuels 2010, 24 (3), 1603–1610. (12) Costa, M.; Azevedo, J. L.; Carvalho, M. G. Combustion characteristics of a front-wall-fired pulverized-coal 300Mwe utility boiler. Combust. Sci. Technol. 1997, 129 (1), 277–293. (13) Costa, M.; Silva, P.; Azevedo, J. L. T. Measurements of gas species, temperature, and char burnout in a low-NOx pulverized-coal fired utility boiler. Combust. Sci. Technol. 2003, 175 (2), 271–289. (14) Costa, M.; Azevedo, J. L. T. Experimental characterization of an industrial pulverized coal-fired furnace under deep staging conditions. Combust. Sci. Technol. 2007, 179 (9), 1923–1935. (15) Garcia-Mallol, J. A.; Steitz, T.; Chu, C. Y.; Jiang, P. Z. Ultra-Low NOx Advanced FW Arch Firing: Central Power Station Applications, 2nd ed.; U.S. China NOx and SO2 Control Workshop: Dalian, 2005. (16) FLUENT 6.3-user’s guide; Fluent Inc.: Lebanon, NH, 2006. 3810



ARTICLE

(17) Smoot, L. D.; Smith, P. J. Coal Combustion and Gasification; Plenum Press: New York, 1989. (18) Hill, S. C.; Smoot, L. D. Modeling of nitrogen oxides formation and destruction in combustion systems. Prog. Energy Combust. Sci. 2000, 26 (46), 417–458. (19) De Soete, G. G. Overall reaction rates of NO and N2 formation from fuel nitrogen. Proceedings of the 15th International Symposium on Combustion; Pittsburgh, PA, 1975.

3811


Improving Combustion Characteristics and NOx Emissions of a Down

Recommend Documents