Article pubs.acs.org/EF
Trends in NOx Emissions during Pulverized Fuel Oxy-fuel Combustion Michitaka Ikeda,*,† Dobrin Toporov,‡ Dominik Christ,§ Hannes Stadler,§ Malte Förster,§ and Reinhold Kneer§ †
Energy Engineering Research Laboratory, Central Research Institute of the Electric Power Industry (CRIEPI), Nagasaka 2-6-1, Yokosuka, Kanagawa 2400196, Japan ‡ Uhde GmbH, Friedrich-Uhde-Straße 15, 44141 Dortmund, Germany § Institute of Heat and Mass Transfer, Rheinisch-Westfälische Technische Hochschule (RWTH) Aachen University, Eilfschornsteinstraße 18, 52056 Aachen, Germany ABSTRACT: This work presents the trends in NOx emissions during pulverized fuel (PF) oxy-fuel combustion using the coal combustion test facility (100 kW) at RWTH Aachen University. In this facility, three modes of pulverized coal combustion have been considered, namely, (1) firing in air (named as air mode), (2) firing in a mixture of oxygen and carbon dioxide (named as O2/CO2 mode) by varying the O2 concentration in the O2/CO2 mixture, and (3) firing in a mixture of oxygen and recycled flue gas (RFG) (named as O2/RFG mode). It was found that NOx emissions in the O2/CO2 mode were about 20% lower than those in the air mode. This causes high temperatures in the burner vicinity to form large amounts of thermal NOx in the air mode. On the other hand, NOx emissions in the O2/RFG mode are considerably reduced by approximately 50%. This is due to the fact that NOx contained in RFG is supplied back by a secondary stream to the flame, and thus, it is destructed by reduction gas in volatile matter. Moreover, in this investigation, the gas stream condition from the swirl burner is controlled. It was found that a decrease in the amount of primary stream leads to not only improved combustion but also an increase in NOx emissions for all combustion modes. A higher secondary stream leads to higher NOx emissions in both the air and O2/CO2 modes. In the case of the O2/RFG mode, however, it is clarified that NOx emissions decrease with an increasing secondary stream. The reason is that a higher secondary stream causes a higher circulating flow in the near burner region. It leads to an increase in NOx emissions because combustion is improved and a large amount of fuel NOx is formed. In the case of the O2/RFG mode, as the amount of secondary stream increases, a large amount of NOx in the secondary stream, which includes that in RFG, is destructed by volatile matter in the flame. From these results, a method of predicting NOx emissions based on the mathematical model that is calculated by multiplex analysis has been applied to oxycoal conditions. Thus, the trends in NOx emissions obtained in a pilot-scale facility can be translated to industrial-scale boilers.
1. INTRODUCTION Coal is an important energy resource for meeting the increasing demand for electricity because it exists in more abundant reserves than other fossil fuels, such as oil and gas. At present, the main method of coal use is pulverized coal combustion. In coal combustion, a large amount of pollutants exhaust. Nitrogen oxide (NO), nitrogen dioxide (NO2), and nitrous oxide (N2O), known collectively as NOx, are atmospheric pollutants that contribute to the production of photochemical smog and acid rain. The amount of NOx exhausted during coal combustion is the most significant among other fossil fuels, owing to the high nitrogen content of the fuel itself. All combustion processes result in the formation of nitrogen oxides (NOx), by either fixation of atmospheric nitrogen (the thermal or prompt N mechanism for NO formation) or conversion of nitrogen-containing structures in the fuel (the fuel N mechanisms). The formation mechanisms of NOx from the combustion of coal have been extensively studied.1−3 Knowledge of the conversion of nitrogen in coal volatiles and char is the key to understanding NOx formation from coals. Previous research on coal N and NOx has shown that there is no simple relationship between coal N and NOx emissions and that combustion stoichiometry has a significant influence upon the relative amounts of NO produced under fuel-rich conditions. As a consequence, combustion modifications © 2012 American Chemical Society
effectively control NO; fuel and air stagings result in lower NO formation largely because the volatile N is converted to N2. There are two opposing factors that affect NOx emissions during coal combustion. One is the oxidation of fuel N by oxygen and other oxidizing agents. The other is the reduction of already produced NOx by reducing agents, such as hydrocarbons from pyrolysis of the volatile matter in homogeneous reactions and resident char in heterogeneous reactions.4−6 The NOx emissions are controlled in most industrial coal burners by adjusting injection air conditions.1,7,8 A point of view of low NOx combustion is commonly the NOx reduction effect in the flame, where the O2 concentration is very low. NOx is destructed in the flame by controlling swirl flows. NOx emissions could be considerably reduced by applying both methods: installing a low NOx burner and controlling the staged air injection ratio. The other method of reducing NOx emissions is blending combustion of high volatile fuel. During two types of coal-blending combustion, it has already been known that both emissions of NOx and unburned carbon become low as the volatile content Special Issue: 7th International Symposium on Coal Combustion Received: November 15, 2011 Revised: February 10, 2012 Published: February 13, 2012 3141
dx.doi.org/10.1021/ef201785m | Energy Fuels 2012, 26, 3141−3149
Energy & Fuels
Article
increases.9,10 Also, the reburning method, which means that high volatile content fuel injects directly in the combustion zone, is effective to reduce NOx emission at the exit of the furnace.4,11,12 The reduction agents with high hydrocarbons are able to destruct NOx by homogeneous reactions. A coal including high volatile matter also has an effect on the destruction of NOx. This is influenced by not only homogeneous reactions but also heterogeneous reactions between char and carbon monoxide.13−15 The flue gas recirculation method, for instance, uses flue gas recycled from the stack and injected into the combustion air supplied from a burner. Thus, a lower oxygen partial pressure in the flame zone is obtained, and the peak flame temperature is reduced. The target percentage of recycled flue gas (RFG) to total combustion air is typically 10−15 vol %. This control technique has been used with gas- or oil-fired furnaces. The major effect of flue gas recirculation is on thermal NOx formation, with a relatively small impact on fuel NOx production. On the other hand, the CO2 concentration in the atmosphere has increased markedly in the past 100 years, thus contributing to the greenhouse effect. In the coal-fired power station, carbon dioxide (CO2) emission is the highest at the same energy production level among other fossil fuels, although the recoverable reserve of coal is abundant. Recently, CO2 capture and storage technologies applied to the coal-based electricity and heat generation sector, which is among the major sources of CO2, have gained huge interest as a promising option, which has the potential to reduce CO2 emissions drastically. This concept is usually divided into three different approaches, namely, postcombustion capture, pre-combustion capture, and oxy-fuel techniques. The current work is focused on the oxy-fuel option and especially on pulverized coal combustion in an O2/CO2 atmosphere because of easy CO2 recovery. The introduction of oxy-fuel technology provided the possibility of reducing NOx emissions from pulverized fuel (PF) boilers without deep staging of the combustion. In comparison to air-fired units, the NOx emissions generated per unit of energy during oxy-fuel combustion are approximately 70% lower depending upon the burner design, coal type, and operating conditions.16−19 This substantial decrease in NOx emission is presumed to be the result of the following mechanism.20−22 One is the reduced thermal NO, owing to the absence of atmospheric N2. The other is the reduction of recycled NO after being supplied through the flame: (1) by CH fragments from the pyrolysis of volatile matter, (2) on the char surface, which can be enhanced by an increased CO concentration, owing to high CO2 concentrations in the furnace, and (3) by the interaction between recycled NO and released fuel N. Liu and Okazaki23 suggested that heat recirculation (e.g., hot recycling) could reduce exhausted NO emission further in three ways: (1) less exhausted flue gas (high recycling ratio), (2) a low oxygen/fuel stoichiometric ratio, and (3) an increase in the flame temperature with a low oxygen ratio. The concentration of O2 in the inlet gas decreases, and those of NOx reducing agents increase simultaneously when the flue gas recycling ratio increases, thus affecting the reduction of NOx emissions directly. The conversion ratio of fuel N/NOx increases with an increasing O2 concentration in the mixture, as reported by Hu et al.21 and Liu et al.24 However, an increase in the recycling ratio (lowering O2 content) leads to improved NOx reduction efficiency.20,21 As the recycling ratio increases, the heterogeneous reaction between NO and char is promoted in the beginning stage by high CO concentrations, which are obtained as a result of the low O2 content in the inlet gas. However, the
resulting decrease in the fuel concentration has an inverse effect on the reduction of NOx because of the decrease in the hydrocarbon fragment concentrations, as observed by Hu et al.21 Hence, there is a contrary effect of the emission of fuel N on the reduction of recycled NOx. The detailed mechanisms still need to be elucidated. Park et al.25 studied the fuel N conversion during the heterogeneous reaction of bituminous coal char with O2, CO2, and H2O over a broad range of temperatures, pressures, and reactant gas connections. The results showed that char N is converted entirely into N2 when char reacts with CO2, into N2 and NO when char reacts with O2, and into HCN, NH3, and N2 when char reacts with H2O. Shaddix and Molina26 performed experiments in a downfired entrained flow reactor, measuring NOx formation at three different O2 concentrations (12, 24, and 36 vol %) in both N2 and CO2 diluents. For enhanced O2 levels, combustion in a CO2 environment led to a decrease in the amount of NOx. This was explained by the lack of thermal NOx production in the N2-free gas environment and with lower volatile flame temperatures and lower char combustion temperatures in a CO2 environment. When burning bituminous coal particles in an atmosphere of 12 vol % O2, only 1/5 of the fuel N is converted to NOx. At 24 vol % O2, approximately 1/3 of the fuel N is converted to NOx in a N2 diluent and about 1/4 is converted when in a CO2 environment. At 36 vol % O2, the conversion factors increase to 55 and 45%, respectively. For highly volatile coals, the conversion ratio of fuel N/NOx was approximately 35% for combustion in 12 vol % O2, 60% for combustion in 24% O2, and 80% for combustion in 36 vol % O2, with similar decreases in the conversion factor in a CO2 environment as obtained for bituminous coal. The results also showed that volatile-generated NOx constitutes a large fraction of the total coal NOx generated at enhanced O2 levels. This may result from the enhanced effect of O2 on the volatile flame temperature than on char combustion temperatures. The influence of CO2 upon the conversion of char N to NOx appears to be similar to that at 12 vol % O2, as also found during the combustion of raw coal particles. When NOx is recycled back to the flame, at 12 vol % O2, a net NOx reduction during coal combustion was observed, and at 24 vol % O2, there is essentially zero net NOx production. However, at 36 vol % O2, there is significant net production of NOx for both N2 and CO2 diluents. In comparison to combustion flame under the air condition, that under the oxy-fuel condition, which has the same O 2 concentration, is not stable because CO2 exists as an inert gas instead of N2 in air and the heat capacity of CO2 is higher than that of N2.27,28 At RWTH Aachen University, fundamental research on the oxy-fuel combustion method was carried out by both means numerically and experimentally. Aerodynamic measures for the stabilization of oxy-fuel swirl flames with low O2 content were conducted. The oxy-fuel burner design is investigated in detail numerically. The experiments were performed successfully and demonstrated the possibility of burning pulverized coal with lower than 21 vol % O2 in the oxy-fuel combustion.16,27 The introduction of oxy-fuel technology has provided a possibility of reducing NOx emissions from PF boilers without deep staging of the combustion. The current work focuses on the NOx formation and destruction trends during oxy-fuel combustion. However, there is data for just a small-scale facility. Therefore, a tool for the prediction of NOx emissions in a large-scale boiler is necessary. This paper describes the influence of the stream injection condition from the swirl burner upon NOx emission at a coal combustion test facility at RWTH 3142
dx.doi.org/10.1021/ef201785m | Energy Fuels 2012, 26, 3141−3149
Energy & Fuels
Article
inner diameter of 400 mm and a total length of 4200 mm. The burner is axially traversable, allowing for measurements at different distances from the burner. In this combustion test, the distance from the top is fixed at 2100 mm. Moreover, the test facility is equipped with an electric heating system with which the wall temperature can be regulated at 1073 K. A schematic of the furnace with peripheral equipment is given in Figure 2. In this facility, the gas provided from the burner is mixed with O2 and CO2. The burner geometry and its dimensions are shown in Figure 3. The burner is a swirl burner with a single annular orifice through which the primary stream and coal are supplied. The secondary stream can be swirled, whereas the amount of swirl can be adjusted. It is injected into the combustion chamber through an annulus surrounding the primary stream inlet. A tertiary stream can be injected through an annulus enclosing the quarl. Another gas stream, the staging stream, enters the furnace through an annulus at the outer diameter of the furnace.
Aachen University. From these results, the method of predicting NOx emissions based on the mathematical model, which is calculated by multiplex analysis, has been applied to oxy-fuel conditions to translate the results to industrial-scale boilers.
2. EXPERIMENTAL SECTION Figure 1 shows the coal combustion test furnace at the Institute of Heat and Mass Transfer in RWTH Aachen University. Experiments were conducted in a vertical, cylindrical, top-fired furnace with an
3. MEASUREMENT METHOD The oxidized mixture stream is split into several streams according to Table 1. The primary stream is used as a carrier fluid for the coal. In this investigation, lignite is selected for the tested coal, and it is easy to ignite from the viewpoint of spontaneous combustion. In the area of the storage bin and coal feeder, CO2, which is an inert gas, is enclosed to prevent
Figure 1. Schematic diagram of the pulverized coal combustion test furnace.
Figure 3. Schematic diagram of the pulverized coal burner.
Figure 2. Schematic diagram of the test facility at RWTH Aachen University. 3143
dx.doi.org/10.1021/ef201785m | Energy Fuels 2012, 26, 3141−3149
Energy & Fuels
Article
fired under the air condition (named as the air mode). Second, instead of N2 gas, CO2 gas is supplied and O2 gas is also supplied. Under the standard combustion condition, the O2 concentration in the blown gas from the burner is set at 21 vol %, the same as that under the air-blown condition (named as the O2/CO2 mode). Third, flue gas from the exit of the furnace is recycled and used to stream from the burner, and O2 gas is supplied in the stream from the burner (named as the O2/RFG mode). In this mode, the O2 concentration in the blown gas from the burner is also set at 21 vol %. On the basis of these three combustion modes, the influence of supplying the mixture gas condition upon NOx emission characteristics is evaluated as follows. At first, the influence of the primary stream condition, which means the existence of O2 and flow amount of the primary stream, is investigated. Second, the influence of the secondary stream condition, which means the temperature, O2 concentration, and flow amount of the secondary stream, is studied. Finally, from these results, the method of predicting NOx emissions is developed.
Table 1. Flow Parameters mass flow (kg/h) coal
O2 content (vol %)
CO2 content (vol %)
8.5,a 12, 15
0, 19a
81,a 100
313a
19.2−30.4 (24.4)a 2.4−3.8 (3.0)a 5.5−78.3 (42.7)a
18−27 (21)a 18−27 (21)a 18−27 (21)a
73−82 (79)a 73−82 (79)a 73−82 (79)a
323a−423
temperature (K)
5.7−18.4 (11.6)a
primary stream secondary stream tertiary stream straging stream a
amount of gas volume (m3N/h)
323a−423 1073a
Standard conditions.
self-ignition. In the primary stream, which means the pulverized coal carrier stream, oxidized mixture gas is used. The oxidized gas is mixed with CO2 gas from the feeder. Therefore, in the standard condition, the oxygen concentration is 19 vol % in the primary stream. The secondary stream is highly swirled, with a swirl number of 1.0 at the secondary stream inlet. The flow amount ratio of tertiary stream in this burner is just a few percentages. Also, it is injected from the outer side of the quarl. However, in the case of the actual burner in the power station, it is about half of the total gas that is injected from inner side of the quarl. From the viewpoint of the flow amount and injection method, it seems that the tertiary stream for this burner has little impact on combustion and NOx formation. In this investigation, the tertiary stream is enabled for scavenging purposes only. The remaining mixture gas was injected as the staging stream. Here, the excess oxygen ratio is defined as follows. One is the overall excess oxygen ratio. This ratio is the amount of oxygen gas to that of total stream, which is the total of primary, secondary, tertiary, and staging. The other is an excess oxygen ratio at the burner. This ratio is the amount of oxygen gas to total stream, which is primary and secondary. The overall excess oxygen ratio at the inlet of the flue stack is set at 1.2, whereas the excess oxygen ratio at the burner is varied. Table 2 shows proximate
4. RESULTS AND DISCUSSION 4.1. Influence of the Excess Oxygen Ratio at the Burner. The NOx emission characteristics of the three cases, namely, air mode, O2/CO2 mode, and O2/RFG mode, were evaluated among several conditions of the excess oxygen ratio at the burner. The O2 concentration of the burner inlet is set at 21 vol % for all of the combustion modes. In this experiment, the flow rate of the mixture gas is the same for the three combustion modes. The excess oxygen ratio at the burner is varied by controlling the coal feed rate. In the coal combustion test, both the oxygen ratio and coal feed rate are controlled to keep the stream line for primary and secondary. It is difficult to control only one item under the same conditions of the remaining two items. When the coal feed rate is changed, the temperature of the combustion flame is changed. Then, it seems that there is little impact on NOx formation because of the impact of the coal feed rate on NOx formation in the range of this investigation.29 Also, the temperature of the combustion zone is kept constant by the heating wall. The excess oxygen ratio is defined using the following equation:
Table 2. Proximate and Ultimate Analyses of Rhenish Lignite in Mass Percentages
excess oxygen ratio=λ =
Rhenish lignite moisture (wt %) ash (wt %) volatile matter (wt %) fixed carbon (wt %) fuel ratio heating value (MJ/kg) carbon (wt %) hydrogen (wt %) nitrogen (wt %) oxygen (wt %) sulfer (wt %)
9.8 4.1 43.7 42.4 0.97 20.7 60.7 4.20 0.84 20.2 0.23
mO2 /m fuel (mO2 /m fuel )stoichiometric
(1)
Here, mO2/mfuel is the mole ratio of O2 to fuel. In the case of this tested coal, mfuel is 1.23 m3 of O2/kg of coal. The furnace wall temperature is kept constant at 1073 K by electric heaters in the furnace. In this study, comparing the emission characteristics of NOx, defined using the following index, which is the conversion to NOx, is carried out. Conversion to NOx is calculated using the following equation: CR = NOx (stk)/NOx (ttl) × 100
(2)
Here, CR is conversion to NOx (%); NOx(stk) is actual NOx emissions at the flue stack (mg/MJ); and NOx(ttl) is total NOx emissions when all nitrogen components in coal are converted to NOx (mg/MJ). In this investigation, conversion to NOx at the flue stack includes not only fuel NOx but also thermal NOx. Figure 4 shows the influence of the excess oxygen ratio at the burner upon conversion to NOx. Increasing the excess oxygen ratio at the burner leads to an increase in conversion to NOx for all three combustion modes. Conversion to NOx in the O2/CO2 mode is about 20% lower than that in the air mode.
and ultimate analyses of tested coal in mass percentages. The coal used is pre-dried Rhenish lignite mined in Germany. The moisture content of the pre-dried Rhenish lignite is less than 10 wt %, and this value is almost the same as that of bituminous coal. At the inlet of the flue stack, the gas concentration, such as those of O2, CO2, NO, and NO2 in the flue gas, is measured. The measurement conditions for this investigation were operated as follows in three cases. At first, pulverized coal is 3144
dx.doi.org/10.1021/ef201785m | Energy Fuels 2012, 26, 3141−3149
Energy & Fuels
Article
Figure 4. Influence of the excess oxygen ratio at the burner upon conversion to NOx.
Figure 5. Influence of the existence of oxygen in the primary stream upon conversion to NOx.
This mainly causes high temperatures in the burner vicinity to form large amounts of thermal NOx in the air mode. Also, the flame temperature becomes low because the specific heat of CO2 is higher than that of N2. When burning the same amount of coal with the same amount of O2 in different gas mixtures, the enthalpy supplied is the same. However, because enthalpy may be expressed as a product of the specific heat capacity and temperature, the increase of the temperature becomes lower and the maximum temperature of the combustion flame decreases at the O2/CO2 mode. Therefore, it seems that the formation of fuel NOx in the O2/CO2 mode is lower than that in the air mode. This value is in good agreement with the results of previous research using inert gas, instead of N2.1 On the other hand, the conversion to NOx in the O2/RFG mode is considerably reduced by approximately 50%. These tendencies are in good agreement with the results of previous research.20 This is mainly due to the fact that NOx contained in RFG is supplied by secondary stream back to the flame and, thus, it is destructed by reduction gas in volatile matter. The staging stream flown from the gas mixture has NOx in O2/RFG mode. This stream is not injected into the combustion flame because this stream is supplied around the furnace wall. It seems that mainly NOx contained in secondary stream is reduced in the combustion flame because NOx is easy to be decomposed by hydrocarbon volatized from coal. On the other hand, staging stream is mixed after the combustion flame. In this area, even if the excess oxygen ratio is 1.2, the temperature is lower than that in the combustion flame.30 It seems that NOx formation in the mixing area of staging stream is low. Also, it is considered as the effect of the interaction between fuel N and NOx in RFG. It is already known that, as NOx is added in the combustion zone, the peak NOx concentration in the flame is reduced and conversion of fuel N to NOx becomes low.20,31,32 4.2. Primary Stream Injection Condition. In this section, the influence of the supplying condition of the oxidized mixture gas from the burner upon the NOx emission characteristics is investigated. 4.2.1. Existence of Oxygen in Primary Stream. At first, the influence of the primary stream condition is investigated. Figure 5 shows the influence of the existence of O2 in the primary stream upon the conversion to NOx under the O2/CO2 mode. In this combustion test, the NOx emission characteristics
in the case of the three levels of O2 concentrations in oxidized gas are evaluated at the same excess oxygen ratio at the burner. It was found that the conversion to NOx is almost the same value, despite the existence of O2 in the primary stream. This is due to the diffusion of pulverized coal particles at the combustion chamber as proven in previous research.16 Pulverized coal is mainly fired after mixing the secondary stream. It seemed that the influence of the existence of O2 in the primary stream is negligible because the volume of the secondary stream is larger than that of the primary stream, and both streams are mixed immediately in the burner vicinity. 4.2.2. Volume of Primary Stream. When pulverized coal is fired under the oxy-fuel condition, the flame stability becomes worse than that in the air mode under the same O2 concentration in the stream. To improve the combustion condition, the gas streamline should be controlled and the trajectory of pulverized coal particles is kept in the high-temperature region in the burner vicinity. It has already been investigated numerically and experimentally that a strong internal recirculation flow is formed in the flame.33 To increase the residence time of pulverized coal particles in the flame, the primary stream momentum ratio should be reduced. Figure 6 shows the influence of the volume of primary stream upon the conversion to NOx in the case of two excess oxygen ratios at the burner. In this study, two combustion modes, which are the air mode and O2/CO2 mode, are performed. It is clarified that the conversion to NOx decreases as the volume of primary stream increases under the two combustion modes. Conversion to NOx in the case of the air mode becomes high when the excess oxygen ratio at the burner is high. These reasons are considered as follows. When the volume of primary stream is low, the inertial force to the axial direction becomes weak and pulverized coal is easy to diffuse to the outer side of the furnace. In this region, the O2 concentration is high and the nitrogen in coal is easy to oxidize to NO. Additionally, in the air mode, a large amount of thermal NOx is also formed. It was cleared that the high primary stream affects the low NOx combustion. However, the high volume of primary stream causes low combustion efficiency. In the case of this combustion test, low combustion efficiency is not indicated. This reason is follows. It is cleared that combustion efficiency decreased linearly as the fuel ratio, which is the weight ratio of fixed carbon 3145
dx.doi.org/10.1021/ef201785m | Energy Fuels 2012, 26, 3141−3149
Energy & Fuels
Article
already been clarified numerically and experimentally that a strong internal recirculation flow is formed in the combustion flame.33 In parallel, the NOx reduction region inside the flame expands, owing to early ignition, thereby increasing the NOx reduction ratio. Therefore, increasing the O2 concentration in the secondary stream leads to a decrease in the NOx emission at the exit of the furnace. 4.3.2. Temperature of Secondary Stream. Figure 8 shows the influence of the stream temperature from the burner upon
Figure 6. Influence of the volume of primary stream upon conversion to NOx.
to volatile matter, increased.34 In this paper, we are firing lignite. The fuel ratio of lignite is much lower than that of bituminous coal. Therefore, even if the combustion condition is not good, the combustion efficiency of lignite is kept extremely high and only NOx is mainly changed. For all experiments, the CO levels were below 50 ppm. The reason for low NOx emission is that the reaction time in the furnace becomes short as the flow velocity of pulverized coal particles in the axial direction increases. Therefore, the volume of primary stream should be adjusted to confirm combustion flame stability. 4.3. Secondary Stream Injection Condition. In this section, the influence of the secondary stream condition upon the NOx emission characteristics is investigated. 4.3.1. Oxygen Concentration in Secondary Stream. Figure 7 shows the influence of the O2 concentration in the secondary
Figure 8. Influence of the temperature of secondary stream upon conversion to NOx.
the conversion to NOx. The conversion to NOx slightly increases with the increase in the inlet temperature of the secondary stream. This increase tends to be higher in the air mode than in the O2/CO2 mode. It is considered that the formation of thermal NOx increases because the flame temperature increases at the higher stream temperature. In comparison to the influence of the O2 concentration in the secondary stream upon the conversion to NOx, it is found that the conversion to NOx at the higher stream temperature is not reduced. Both the higher O2 concentration and higher temperature in the secondary stream lead to closing the ignition point to the burner, and the NOx reduction region inside the flame expands. However, the higher stream temperature also causes a higher flame temperature. Then, the amount of fuel NOx formation increases. Therefore, it is considered that the conversion to NOx does not decrease with an increasing stream temperature, despite the improvement of the ignition condition. 4.3.3. Velocity of Secondary Stream. In this section, the influence of the volume of secondary stream upon the NOx emission characteristics is discussed. Figure 9 shows the influence of the volume of secondary stream upon the conversion to NOx. In this study, two excess oxygen ratios at the burner are used for the three combustion modes. In the air mode, the NOx emission becomes high as the volume of secondary stream becomes high. This tendency is almost the same as that in the O2/CO2 mode. This is because the pulverized coal particles are easy to diffuse near the combustion chamber, owing to the high volume of secondary stream with swirl flow, and the nitrogen in coal is oxidized to NOx at the outer side of the furnace, the O2 concentration of which is higher than that at the inner side. On the other hand, it was found that the NOx emission in the case of the
Figure 7. Influence of the oxygen concentration in secondary stream upon conversion to NOx.
stream upon the conversion to NOx in the case of the O2/CO2 and O2/RFG modes. Here, the O2 concentration in the secondary, tertiary, and staging stream is controlled to the same value in the gas mixture. The increase in the O2 concentration in the secondary stream does not affect the global NOx emission. In this case, the ignition point closes to the burner and NOx formation increases in the near burner region. It has 3146
dx.doi.org/10.1021/ef201785m | Energy Fuels 2012, 26, 3141−3149
Energy & Fuels
Article
Figure 9. Influence of the volume of secondary stream upon conversion to NOx (left panel, excess oxygen ratio at the burner of 0.57; right panel, excess oxygen ratio at the burner of 0.67).
Table 3. Calculated Results for Each Cofficient To Predict Conversion to NOx mode
a1
a2
b1
b2
c1
c2
d1
d2
air O2/CO2 O2/RFG
−0.127 −0.098 −0.039
0.097 0.083 0.028
−0.91 −0.11
3.50 0.51
1.5 1.4 −3.1
0.3 1.9 6.6
0.00020 0.00007
0.378 1.110
conversion to NOx and the volume of primary stream is assumed as a linear function. (3) Although the conversion to NOx increases as the volume of secondary stream increases in the air and O2/CO2 modes, this tendency in the case of the O2/RFG mode is opposite. The relationship between the conversion to NOx and the volume of secondary stream is assumed as a linear function. (4) The conversion to NOx slightly increases as the temperature of the secondary stream increases. The relationship between the conversion to NOx and the temperature of secondary stream is assumed as a linear function. On the basis of these concepts, the prediction equation of conversion to NOx is defined as follows:
O2/RFG mode becomes low as the volume of secondary stream increases. The reason is considered as follows. The high secondary stream also leads to the diffusion of pulverized coal particles and causes a higher NOx emission near the burner, the same as in the air and O2/CO2 modes. However, the secondary stream is recirculated to the region of the combustion flame.33 In the case of the O2/RFG mode, recycled gas contains NOx in the secondary stream. As the amount of secondary stream increases, the amount of recirculated stream with NOx increases. In the combustion flame, NOx in the secondary stream is able to be destructed by the reduction gas in volatile matter.4−6 Therefore, the NOx emission is able to be reduced by increasing the volume of secondary stream. From the investigation of the influence of the secondary stream upon the NOx emission, it was clarified that both the higher O2 concentration in the secondary stream in the range for a given excess oxygen ratio and the higher volume of secondary stream lead to low NOx combustion. 4.4. NOx Prediction and Validation. From these results, the method of predicting NOx emissions is applied under oxyfuel conditions. Here, this method is analyzing experimental results with principal parameters, such as the previous one.9,35,36 The coefficient for parameters is decided by the mathematical model calculated by multiplex analysis using the minimum mean square method, and this method is the most widely used. In this prediction method, four parameters, which are the excess oxygen ratio at the burner, the volume of primary stream, the volume of secondary stream, and the temperature of oxidized gas, are selected as the base factors for the prediction of NOx. The prediction equation for each parameter is defined as follows: (1) The relationship between the conversion to NOx and the excess oxygen ratio at the burner is assumed as a quadratic function. In this investigation, the overall excess oxygen ratio at the inlet of the flue stack is set at 1.2. Therefore, conversion to NOx becomes a maximum at an excess oxygen ratio of 1.2 at the burner. (2) The relationship between the
CR = (a1(λb − 1.2)2 + a2)(b1QP + b2)(c1QS + c 2) (d1TS + d2)
(3)
Here, ai, bi, ci, and di are factors for each function. λb is the excess oxygen ratio at the burner. QP is the volume of primary stream (m3/h). QS is the volume of secondary stream (m3/h). TS is the temperature of the secondary stream. By multiplex analysis using the minimum mean square method, these factors are calculated. The calculated results for each coefficient for the three combustion modes are shown in Table 3. The calculated value is then compared to the experimented value. Figure 10 shows the comparison of the predicted conversion to NOx and the experimental one among the air, O2/CO2, and O2/RFG modes. The correlation coefficients of linear regression between predicted NOx and experimental NOx on three combustion modes are 0.90, 0.84, and 0.96 (R = 0.95, 0.92, and 0.98), respectively. In this investigation, it is clarified that the prediction value of the conversion to NOx is in good agreement with the experimental value among not only the air mode but also the O2/CO2 and O2/RFG modes. It is proven that this method of predicting the conversion to NOx is adjusted on this experimental equipment. These results were adapted for a given excess oxygen ratio and volume of each stream. The coefficient values in this 3147
dx.doi.org/10.1021/ef201785m | Energy Fuels 2012, 26, 3141−3149
Energy & Fuels
Article
Figure 10. Comparison of the predicted conversion to NOx and the experimental one among air, O2/CO2, and O2/RFG modes (left panel, air mode; middle panel, O2/CO2 mode; right panel, O2/RFG mode).
From these results, the method of predicting NOx emissions based on the mathematical model that is calculated by multiplex analysis has been applied under oxy-fuel conditions. Thus, the trends in NOx emission can be translated to industrial-scale facilities.
equation should be recalculated using past data for each boiler by the multiplex analysis method. Even if the boiler size is different, the influence of the condition of primary and secondary stream upon NOx emission should be indicated as a similar tendency as the results in this investigation. When the condition of primary and secondary stream is changed at the scaled-up industrial boiler, NOx emission will be estimated by adjusting the coefficients in this equation.
■
AUTHOR INFORMATION
Corresponding Author
*Telephone: +81-46-856-2121. Fax: +81-46-856-3346. E-mail:
[email protected].
5. CONCLUSION At a coal combustion test facility at RWTH Aachen University, the trends in NOx emissions during PF oxy-fuel combustion are investigated from the viewpoint of stream injection conditions from the swirl burner. In this investigation, three modes of pulverized coal combustion, namely, air mode, O2/CO2 mode, and O2/RFG mode, have been considered. The main results are as follows: (1) Increasing the excess oxygen ratio at the burner leads to an increase in the NOx formation for all three combustion modes. NOx emissions in the O2/CO2 mode are about 20% lower than those in the air mode. This causes high temperatures in the burner vicinity to form large amounts of thermal NOx in the air mode. On the other hand, NOx emissions in the O2/RFG mode are strongly reduced by approximately 50%. This is due to the fact that NOx contained in RFG is supplied back by the secondary stream to the flame, and thus, it is destructed by reduction gas in volatile matter. (2) A decrease in the volume of primary stream leads to not only improved combustion but also an increase in NOx emissions for all combustion modes. (3) An increase in the O2 concentration in the secondary streams does not affect the global NOx emission. In this case, the ignition point closes to the burner and NOx formation increases in the near burner region. In parallel, the NOx reduction region inside the flame expands, owing to early ignition, thereby increasing the NOx reduction ratio. (4) NOx emissions increase slightly with the increase in the inlet temperature of the secondary stream. This increase is higher in the air mode than in the O2/CO2 mode. It is considered that the formation of thermal NOx increases because the flame temperature increases at the higher stream temperature. (5) A higher volume of secondary stream leads to higher NOx emissions in both the air and O2/CO2 modes. In the case of the O2/RFG mode, however, it is found that NOx emissions decrease with increasing the volume of secondary stream.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors gratefully acknowledge the financial support from BMWi, MIWFT, and the companies RWE Power AG, E. ON AG, Hitachi Power Europe GmbH, MAN Turbo, Linde AG, and WS-Wärmeprozesstechnik GmbH.
■
REFERENCES
(1) Pershing, D. W.; Wendt, J. O. L. Proceedings of the 16th International Symposium on Combustion; The Combustion Institute, Pittsburgh, PA, 1977; pp 389−399. (2) Chen, S.; Heap, M.; Pershing, D.; Martin, G. Fuel 1982, 61, 1218−1224. (3) Wendt, J. O. L. Combust. Sci. Technol. 1995, 108, 323−344. (4) Mereb, J. B.; Wendt, J. O. L. Proceedings of the 23th International Symposium on Combustion; The Combustion Institute, Pittsburgh, PA, 1990; pp 1273−1279. (5) Spliethoff, H.; Greul, U.; Rudiger, H.; Hein, K. R. G. Fuel 1996, 75 (5), 560−564. (6) Smoot, L. D.; Hill, S. C.; Xu, H. Prog. Energy Combust. Sci. 1998, 24, 385−408. (7) Harding, N. S. Jr.; Smoot, L. D.; Hedman, P. O. AIChE J. 1982, 28 (4), 573−580. (8) Kurose, R.; Makino, H.; Suzuki, A. Fuel 2004, 83, 693−703. (9) Man, C. K.; Gibbins, J. R.; Witkamp, J. G.; Zhang, J. Fuel 2005, 84 (17), 2190−2195. (10) Arenillas, A.; Backreedy, R. I.; Jones, J. M.; Pis, J. J.; Pourkashanian, M.; Rubiera, F.; Williams, A. Fuel 2002, 81 (5), 627−636. (11) Ostberg, M.; Glarborg, P.; Jensen, A.; Johnsson, J. E.; Pedersen, L. S.; Johansen, K. D. Proceedings of the 27th International Symposium on Combustion; The Combustion Institute, Pittsburgh, PA, 1998; pp 3027−3035. 3148
dx.doi.org/10.1021/ef201785m | Energy Fuels 2012, 26, 3141−3149
Energy & Fuels
Article
(12) Singh, S.; Nimmo, W.; Gibbs, B. M.; Williams, P. T. Fuel 2009, 88, 2473−2480. (13) Liu, H.; Hampartsoumian, E.; Gibbs, B. M. Fuel 1997, 76, 985−993. (14) Zhang, H.; Fletcher, T. Energy Fuels 2001, 15, 1512−1522. (15) Liu, S.; Xu, T.; Zhou, Q.; Tan, H.; Hui, S.; Hu, H. Fuel 2007, 86, 1169−1175. (16) Toporov, D.; Bocian, P.; Heil, P.; Kellermann, A.; Stadler, H.; Tschunko, S.; Förster, M.; Kneer, R. Combust. Flame 2008, 155, 605−618. (17) Croiset, E.; Thambimuthu, K. V. Fuel 2001, 80, 2117−2121. (18) Nozaki, T.; Takano, S.; Kiga, T.; Omata, K.; Kimura, N. Energy 1997, 22, 199−205. (19) Normann, F.; Andersson, K.; Leckner, B.; Johnsson, F. Prog. Energy Combust. Sci. 2009, 35, 385−397. (20) Okazaki, K.; Ando, T. Energy 1997, 22, 207−215. (21) Hu, Y.; Kobayashi, N.; Hasatani, M. Energy Convers. Manage. 2003, 44, 2331−2340. (22) Stadler, H.; Christ, D.; Habermehl, D.; Heil, P.; Kellermann, A.; Ohliger, A.; Toporov, D.; Kneer, R. Fuel 2011, 90, 1604−1611. (23) Liu, H.; Okazaki, K. Fuel 2003, 82, 1427−1436. (24) Liu, H.; Zailani, R.; Gibbs, B. Fuel 2005, 84, 833−840. (25) Park, D. C.; Day, S. J.; Nelson, P. F. Proc. Combust. Inst. 2005, 30, 2169−2175. (26) Shaddix, C. R.; Molina, A. Proceedings of the Fall Meeting of the Western States Section of the Combustion Institute; Sandia National Laboratories, Livermore, CA, 2007. (27) Toporov, D.; Förster, M.; Kneer, R. Clean Air 2007, 8, 321−338. (28) Yamada, T.; Kiga, T.; Okawa, M.; Omata, K.; Kimura, N.; Arai, K.; Mori, T.; Kato, M JSME Int. J., Ser. B 1998, 41 (4), 1017−1022. (29) Kimoto, M.; Ikeda, M.; Makino, H.; Kiga, T. FACT (Am. Soc. Mech. Eng.) 1999, 1, 293−298. (30) Ikeda, M.; Makino, H.; Morinaga, H.; Higashiyama, K. JSME Int. J. 2004, B, 47 (2), 180−185. (31) Levy, J. M.; Chan, L. K.; Sarofim, A. F.; Beer, J. M. Proceedings of the 18th International Symposium on Combustion; The Combustion Institute, Pittsburgh, PA, 1981; pp 111−120. (32) Taniguchi, M.; Yamamoto, K.; Kobayshi, H.; Kiyama, K. Fuel 2002, 81, 363−371. (33) Heil, P.; Toporov, D.; Stadler, H.; Tschunko, S.; Förster, M.; Kneer, R. Fuel 2009, 88, 1269−1274. (34) Kurose, R.; Ikeda, M.; Makino, H.; Kimoto, M.; Miyazaki, T. Fuel 2004, 83 (13), 1777−1785. (35) Abe, R.; Sasatsu, H.; Harada, T.; Misawa, N.; Saito, I. Fuel 2001, 80, 135−144. (36) Lee, Y. H.; Han, C.; Kim, M. J. Environ. Eng. 2005, 131, 961−970.
3149
dx.doi.org/10.1021/ef201785m | Energy Fuels 2012, 26, 3141−3149