Performance Evaluation of South African Coals ... - ACS Publications

Article pubs.acs.org/EF

Performance Evaluation of South African Coals under Oxy-Fuel Combustion in a Fluidized Bed Reactor H. I. Mathekga,†,‡ B. O. Oboirien,*,† A. Engelbrecht,† B. C. North,† and K. Premlal‡ †

CSIR Materials Science and Manufacturing, PO Box, 395, Pretoria 0001, South Africa Tshwane University of Technology, Department of Chemical, Metallurgical and Materials Engineering, Pretoria 0183, South Africa

‡

ABSTRACT: In this study, the experimental results of the oxy-fluidized combustion of three different South African coals (sub-bituminous) are presented. The coal samples were denoted Coal A, B, and C. Three combustion atmospheresair, oxy (21% O2/79% CO2), and oxy (30% O2/70% CO2)were studied. A total of 18 tests were conducted in a bubbling fluidized bed reactor at 850 and 925 °C. The results obtained showed that the highest carbon burnout was obtained at 30% O2/CO2, followed by air, and last at 21% O2/CO2. Coals A and B had a higher carbon burnout than Coal C. There was a marginal difference of 1% in the carbon burnout for both Coals A and B in all the conditions except at 21% O2/CO2 at 925 °C, when the carbon burnout of Coal A was 5% higher than that of Coal B. Carbon burnout in Coal C was about 20% lower than those of Coals A and B. Carbon burnout had a good correlation with the char particle temperature at all the combustion conditions studied for the three coals. The carbon burnout only had a good correlation with the vitrinite random reflectance at 21% O2/CO2 at 925 °C for the three coals. The highest char particle temperature (1250 K) was obtained in Coal B at 30% oxy-combustion (30% O2/CO2) at a 925 °C bed temperature, and it had the highest carbon burnout of 99.34%. The lowest char particle particle temperature (1167 K) was obtained in Coal C at 21% oxy-combustion (21% O2/CO2) at a 850 °C bed temperature, and it had the lowest carbon burnout of 65.81%. There was a higher fuel-nitrogen conversion in air combustion than in oxy-fuel combustion than for all the coals. Fuel-S conversion ratios to SOx indicated that air combustion had relatively higher conversions than oxy-fuel, with the highest conversion of 35% observed for Coal A at 925 °C.

1. INTRODUCTION South Africa holds extensive coal deposits estimated to be the eighth largest in the world; hence, it is heavily dependent on coal to meet its electrical and thermal energy demands.1 However, burning fossil fuels also inevitably produces greenhouse gases, predominantly carbon dioxide (CO2), which to some extent contributes to climate change.2 Carbon capture and storage (CCS) is currently viewed as a viable technology that can reduce the emission of CO2 from coal-fired power plants.3 As of the year 2010, about 43% of CO2 emissions were from the combustion of coal.4 The importance of coal as a source of electricity generation worldwide is set to continue, with about 44% of global electricity fuelled by coal in 2030.4 Notably, CCS technologies being developed for the capture of CO2 from coalfired power plants include integrated gasification combined cycle (IGCC), amine scrubbing, and oxy-combustion. The CO2 capture process in IGCC involves the gasification of coal to produce syngas (CO and H2) and the steam reforming of the syngas gas to produce concentrated CO2 that is easier to be captured. Amine scrubbing involves the separation of CO2 in coal flue gas with amine solvent. Lastly, the CO2 capture process in oxy-combustion involves the concentration of CO2 by recycling of the flue gas back into the bioler.5 Oxy-fuel combustion technology is currently being applied in demonstration plants and with an aim to be used in commercial coal power plants.6 In oxy-fuel combustion, recycle flue gas and pure O2 of about 95% is used for combustion as opposed to conventional fuel combustion that uses air. Air has a high volume of N2 gas that ultimately dilutes the CO2 concentration in the flue gas. This results in a flue gas with a low © XXXX American Chemical Society

concentration (10%) of CO2, which makes it difficult to capture and sequestrate.5 This is not the case for oxy-fuel combustion where a flue gas of about 90% CO2 is produced. A significant amount of the avaliable literature on oxy-fuel combustion are in pulverized coal (PC) combustion. Fluidized bed (FB) combustion could serve as alternative process for oxycombustion. An oxy-FB combustor has an advantage over a PC oxy-fuel combustor such as in the burning of multiple fuels. There is also the advantage of the reduction of the size of the boiler and lower emissions and better temperature management in the boiler.7−9 Another major benefit of oxy-FB combustion is the capability to capture SO2 in situ by addition of limestone to the furnace.10 There are few experimental data on oxy-fluidized bed combustion of South African coals. Oboirien et al.11 carried out a techno-economic assessment study of oxy-fuel combustion technology for six coal power plants in South Africa and reported that the CO2 emission rate could be decreased 10 times when oxy-fuel combustion is used in the coal power plants. This analysis was based on a pulverized oxy-combustion process. The aim of this study was to evaluate the performance of some South African coals under oxy-combustion conditions in a bubbling fluidized bed (BFB) combustor. Three SA coals which are of interest to Eskom, which is the major electricity producer in South Africa, were selected. The selected coals were part of the coal previously studied by Oboirien et al.11 in a Received: February 23, 2016 Revised: July 11, 2016

A

DOI: 10.1021/acs.energyfuels.6b00430 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

2.2.3. Combustion Test. Combustion test was started by adding 300 g of silica sand bed material. This constituted the inert bed. The bed was then heated while fluidized with compressed air. The superficial gas velocity was 1.02 m/s and was used to for the all tests. The interior of the furnace is fitted with heating elements to control and maintain temperature within the length of the tube. The bed temperature was initially set at an intermediate temperature of 500 °C on the control panel, and then raised to 925 °C. After the bed temperature reached the set point, a mixture of coal and silica sand (ratio 1:1) was fed into the furnace at a steady feed rate of about 200 g/h through a screw feeder for the combustion process. The coal feeding rate was controlled to keep the O2 concentration at 6 ± 2% vol (dry basis) at the furnace discharge. Ash was manually drained though a drain valve beneath the furnace to maintain pressure drop within the bed. Fraction of the flue gas was channelled to the gas analyzer through gas cleaning system for analysis at a rate of approximately 3 L/min. Gas concentration readings were then recorded at 15 min intervals. The process was then repeated at a different temperature set point of 850 °C. The process was started by adding exactly 300 g of silica sand material and fluidized with air while heating up the furnace. After the set point temperature was reached, predetermined flows of the oxidant gas O2 and CO2 were supplied from cylinders by rotameter to simulate oxy-fuel combustion conditions. Gas concentrations were monitored by the gas analyzer to validate CO2/O2 feed ratios prior to feeding the coal. 2.2.4. Gas Analysis. A Servomex 4200 industrial online gas analyzer was employed to measure flue gas concentrations in the laboratory. The equipment uses infrared technology to quantify CO2, CO, and CH4 concentrations, while paramagnetic sensors are used for oxygen, since O2 is a paramagnetic gas. It is attracted to magnetic fields and almost all other gases are diamagnetic; they are repelled by magnetic fields. The limitation with the current state of this equipment is that it can only read CO2 concentration between 0 and 50%, while, during oxy-fuel combustion, CO2 is generally above 90%. Additionally, the gas analyzer is currently not equipped to analyze NOx and SOx emissions. As a result, gas samples were collected in gas bags for comprehensive analyses, which were carried out by CHEMTECH LABORATORY SERVICES CC at Pretoria, South Africa. The methods used for gas analysis were: EPA Method 1A, NIOSH Method 6014, and NIOSH Method 6004. 2.2.5. Coal Burnout Analysis. The coal burnout analysis involves high temperature burnoff of unreacted char in the ash sample. The ash sample which was collected after the combustion process was screened using 850 μm aperture size screens to separate silica sand inert bed material from ash together with unburned char, since the sand particles used were between 350 and 650 μm, while coal was between 850 and 1200 μm. Subsequently, accurately weighed (∼10 g) ash was then placed in a Nabertherm burnoff furnace to combust all residual char in the ash at 925 °C. Figure 2a,b shows the images of residual ash before and after char burnoff, respectively. In Figure 2a, it can clearly be observed that unreacted char particles were present in the ash as

techno-economic study of oxy-fuel combustion technology for six coal power plants stations in South Africa. The oxycombustion characteristics were determined in a laboratoryscale bubbling fluidized bed.

2. EXPERIMENTAL SECTION 2.1. Materials. Silica sand of size fraction between 0.3 and 0.65 mm was used as bed material (98% SiO2 and 0.18% Fe2O3). A collection of three South African coals, Coal A (New Vaal), Coal B (Grootegeluk), and Coal C (Khutala), which some are currently being used as feed fuel in South African power plants by ESKOM, were used in this study. The results of the coal analysis are presented in Table 2, in section 3. Oxygen and carbon dioxide from gas cylinders were used to simulate an oxy-combustion atmosphere, and compressed air for air-firing mode, while nitrogen gas was used to purge air in the ash collection vessel and in the lines to the gas analyzer when switching combustion atmospheres. 2.2. Experimental Work. 2.2.1. Coal Sample Preparation. Dried coal samples were crushed using a jaw crusher to −1.4 mm size fraction. The samples were then screened to obtain the desired size fraction of +0.85−1.4 mm. The reason for selecting this size range was to minimize the solids elutriation, as well as the effect of particle size as reactivity varies greatly with particle size. 2.2.2. Reactor Description. The main body of the reactor consists of a (i) furnace (5 cm ID and 180 cm height), (ii) preheating section, (iii) gas distributor, (iv) gas cleaning system, and (v) gas sampling system. The temperature of the reactor was regulated by electric heaters in the sections. The reactor is also equipped with a screw coal feeder, which allows for a steady coal feed rate to be set between 2 and about 50 g/min. A schematic representation of this system is shown in Figure 1.

Figure 1. A schematic diagram of the fluidized bed reactor.

Figure 2. (a) Ash sample before carbon burnoff. (b) Ash sample after carbon burnoff. B


Article

Energy & Fuels

composition and the flow properties are presented in Table 3. As presented in Table 2, the heat load between 0.8 and 1.1 kWth was achieved during the operation. This was found to have marginal influence on the results. There was no direct correlation between oxygen concentration in the flue gas and NOx concentration. Similar observations were observed by Lasek et al.9 for both air and oxy-combustion conditions. 3.1. Carbon Burnout. 3.1.1. Effect of Combustion Atmosphere. The results on the effect of different combustion atmospheres on carbon burnout are presented in Figure 3. Carbon burnout in oxy combustion conditions of 21% O2/CO2 was lower than that in air for 850 and 925 °C. The difference in combustion efficiencies can be attributed to the distinct gas properties of N2 and CO2 in different combustion atmospheres. Oxygen diffusivity is lower in CO2 than N2, thus better mass transfer rates in air combustion. This observation was also made by Duan et al.,14 who noted that substituting N2 with CO2 reduces the combustion rate. However, an increase in the O2 concentration in oxy-combustion from 21% to 30% under oxycombustion significantly improved combustion performance for all tested coal; this is due to the increased partial pressure of O2. These results are similar to those reported by Roy and Bhattacharya.7 They reported on the effect of oxygen on the combustion efficiency of Australian brown coal in a BFB reactor under oxy-combustion conditions. They assessed the O2 concentration up to 30% and observed an increase in the combustion efficiency. The carbon burnouts in Coals A and B were similar. There was a marginal difference of 1% in the carbon burnout for all the conditions except at 21% O2/CO2 at 925 °C, where the carbon burnout in Coal A was 5% higher than that in Coal B. However, there was a significant difference in the carbon burnout in Coal C. It was about 20% lower for all the conditions tested. These observations were hypothesized to be as a result of the difference in vitrinite random reflectance of the coals and the coal particles temperature during combustion. 3.1.2. Effect of Rank Parameter. Vitrinite random reflectance (Rr), is a parameter used in measuring rank.15 Coal A is an inertinitie rich coal and has an Rr value of 0.54, and Coals B and C are vitrintite rich coals and have Rr values of 0.65 and 0.66, respectively. A lower Rr value indicates a lower ranking of a coal. The results of carbon burnout at different conditions were plotted against the vitrinite random reflectance for the different coals and are presented in Figure 4a,b. Figure 4a is the carbon burnout at 850 °C, while Figure 4b is carbon burnout at 925 °C. The results show that there was a better correlation between the vitrinite random reflectance and carbon burnout at 925 °C than at 850 °C. At 850 °C, there was no clear difference between the carbon burnout of Coals A and B. The carbon burnout of the relatively lower rank coal (Coal A) was higher than that for the higher rank coals for all combustion atmospheres at 925 °C, except at 30% O2/CO2, where Coal B had the highest conversion. This is in agreement with earlier experimental results reported by Engelbrecht et al.,15 who gasified four South African bituminous coals of varying ranks (including Coals A and B) in a fluidized bed gasifier at 925 °C. They found that the fixed carbon burnout of the lower rank coals (Coal A) was higher than that for the higher rank coals. Thus, this observation can probably be attributed to gasification rate of these coals. This indicates that, for Coal A during oxy-combustion, the char gasification rate was higher; hence, the overall carbon burnout was enhanced. In oxy-combustion, char conversion is a function of char-O2 reaction (oxidation) and char-CO2 reaction (gasification).16

highlighted by the red circle, while, in Figure 2b, all unreacted char was burnt-off. Carbon burnout percentage was calculated by subtracting the mass fraction of the unburned char in the ash as follows

⎡ ⎛ A ⎞⎤ Carbon burnout (%) = ⎢1 − ⎜ c ⎟⎥ × 100 ⎢⎣ ⎝ A i ⎠⎥⎦

(1)

where Ac is the carbon content of the residual ash and Ai is the residual ash from the combustion test. This was assuming that no coal ashes were elutriated out of the reactor during the tests. Combustible matter conversions such as fuel nitrogen and sulfur conversion to NOx and SOx were calculated based on the eqs 2−5 as proposed by Czakiert et al.12 and modified by Lupiáñez et al.13 Oxy-fuel combustion has been adapted to account for CO2 supplied in the feeding gas: [SO2 ]fg

S → SO2 (AF) =

104[CO2 ]fg + [CO]fg

Sfuel /Cfuel

(2)

[SO2 ]fg

S → SO2 (OF) =

104[CO2 ]fg − [CO2 ]in + [CO]fg

Sfuel /Cfuel

(3)

[NOx ]fg

N → NOx (AF) =

104[CO2 ]fg + [CO]fg

Nfuel /Cfuel

(4)

[NOx ]fg

N → NOx (OF) =

104[CO2 ]fg − [CO2 ]in + [CO]fg

Nfuel /Cfuel

(5)

3. RESULTS AND DISCUSSION The following section details the experimental findings and in depth analysis of the combustion characteristics of three South African sub-bituminous coals under oxy-combustion atmospheres in a bubbling fluidized bed. These results from oxycombustion were also compared with those from air combustion. Proximate and petrographic analyses are shown in Table 1, experimental test matix is presented in Table 2, and the ash Table 1. Proximate, Ultimate, and Petrographic Analyses of the Coals Tested coal sample proximate analysis (wt %) inherent moisture content volatile matter content ash fixed carbon (by difference) gross calorific value (MJ/kg) ultimate analysis (wt %) carbon hydrogen nitrogen total sulfur carbonate oxygen (by difference) petrographic analysis (wt %) vitrinite inertinite liptinite mineral matter rrandom

Coal A

Coal B

Coal C

2.1 22.7 37.5 37.7 17.37

1.4 25.0 32.3 41.3 19.82

2.1 23.0 32.8 42.1 21.12

45.54 2.59 1.07 0.60 2.31 8.29

50.60 3.06 1.11 0.93 2.58 8.02

50.70 2.68 1.19 0.75 2.60 7.18

15.3 63.5 4.6 16.7 0.54

42.8 28.3 3.9 25.1 0.65

45.5 22.1 1.6 31.1 0.66 C


Article

Energy & Fuels Table 2. Experimental Test Matrix test 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

sample ID. Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal

combustion mode

[O2]inlet (%)

set temp. (°C)

load (kWth)

[O2]outlet (%)

CO (vppm)

NOx (mg/m3)

SO2 (mg/m3)

air air air air air air oxy oxy oxy oxy oxy oxy oxy oxy oxy oxy oxy oxy

21 21 21 21 21 21 21 21 21 21 21 21 30 30 30 30 30 30

925 850 925 850 925 850 925 850 925 850 925 850 925 850 925 850 925 850

0.9 0.9 1.1 1.0 1.0 0.8 1.0 0.8 0.9 0.8 1.0 0.9 1.0 0.9 1.1 1.1 1.0 1.0

8.9 7.4 7.9 5.4 8.7 8.7 1.4 1.8 1.6 3.2 5.3 4.7 5.1 5.5 5.4 5.3 4.1 4.1

73 144 193 267 70 125 340 770 600 600 234 367 89 160 220 207 385 857

13 10 14 14 24 23 32 26 27 23 28 18 35 27 37 36 47 40

221 266 106 133 166 221 436 540 269 515 371 428 168 422 128 438 290 310

A A B B C C C C A A B B A A B B C C

gasification and the oxidation reactions are affected by intrinsic reactivity and temperature of coal. The char particle temperatures in air and oxy-fuel combustion atmospheres for the suite of tested coal were modeled to validate the significance of particle temperature on char conversion. The intrinsic rate parameters produced by Dhaneswar16 and assumptions made by Naredi17 were used for this theoretical particle temperature study. Particle temperature was computed from the energy balance equation (eq 8) and was solved using NDSolve in Wolfram Research, Inc., Mathematica, Version 8.0, Champaign, IL (2010). The equation was detailed by Dhaneswar.16

Table 3. Ash Properties coal/mineral oxides (%)

Coal A

Coal B

Coal C

SiO2 TiO2 Al2O3 Fe2O3 MgO CaO Na2O K2O P2O5 SO3 MnO (calculated) deformation temperature softening temperature hemisphere temperature flow temperature

55.24 1.59 30.26 3.29 1.07 4.70 0.36 0.73 0.41 2.57 0.01 1500 1500 1500 1500

58.74 1.37 26.94 4.79 1.09 3.49 0.28 0.88 0.47 2.59 0.02 1460 1470 1480 1480

53.46 1.65 29.91 3.04 1.69 5.44 0.43 0.84 0.59 2.57 0.01 1460 1460 1480 1490

mpCp

dTp dt

= (qcr·Δhf ) − σεpA p(Tp4 − Tw4 ) − hA p(Tp + Tw )

(6)

where mp is the particle mass and Tp (K) is the temperature of the char particle. Cp is the specific heat of the char particle and was estimated using a correlation reported by Eisermann et al.:18

Coal C had the lowest conversion during both air and oxy-fuel combustion. 3.2. Particle Temperature Model. The char particle temperature is an important variable that affects the char combusmathsize="9pt"tion rate. According to Dhaneswar,16 both the

Cp = − 0.218 + 2.807 × 10(−3)T − 1.758 × 10(−6)T 2

(7)

The right-hand side of eq 6 consists of the reaction, radiation, and convective terms, respectively.

Figure 3. Carbon burnout (%) versus combustion atmosphere for different coals at 850 and 925 °C. D


Article

Energy & Fuels

Figure 4. (a) Coal burnout (%) versus rank parameter (Rr) at 850 °C. (b) Coal burnout (%) versus rank parameter (Rr) at 925 °C.

3.2.1. The Reaction Term: qcr·Δhf. qcr is the burning rate of the char in kg/m2·s and is calculated from the following equation qcr = mp·R i·η·fmc ·(Pg)n

that the char particle is spherical, and the correlation used is given below

(8)

Nu =

where Ri is the intrinsic reaction rate. Pg (atm) is the partial pressure of the reacting gas. Futhermore, n is the reaction order for the char-O2 reaction and was taken as 0.5 and 0.8 for the charCO2 reaction in this study. η is the effectiveness factor estimated from the Thiele modulus approach. f mc is the maceral correction factor, calculated by fmc = 1.68 × Vit − 0.68 × In

h·Dp k

= 2.0 + 0.6·Rep0.5·Pr1/3

(10)

where Rep is the Reynolds number of a particle and Pr is the Prandtl number. The char particle temperature prediction for the three coals for air, oxy21% (21% O2/CO2), and oxy30% (30% O2/CO2) atmospheres were computed at 850 °C (1123 K) and 925 °C (1198 K) bed temperatures (Tb). The respective particle temperatures are presented in Table 4. The correlation of the

(9)

Table 4. Particle Temperature Estimations

Vit and In are the vitrinite and inertinite percentages in the parent coal, respectively. Maceral correction factor was used to account for the influence of maceral composition on char reactivity. 3.2.2. The Radiation Term: σεpAp(T4p − T4w). ε is the particle emissivity and was assumed to be 0.5 for the purpose of calculations, and σ is the Stefan−Boltzmann constant (5.67 × 10−8 W/m2 K4). Ap is the surface area (m2) of the char particle, and Tw is the wall temperature (K), which was taken to be equal to the bed temperature (Tb). 3.2.3. The Convection Term: hAp(Tp + Tw). The convective heat transfer coefficient, h, was calculated on the assumption

Tb = 850 °C (1123 K) fuel

combustion

Tp (K)

carbon burnout (%)

Tp (K)

carbon burnout (%)

Coal A

air oxy21% oxy30% air oxy21% oxy30% air oxy21% oxy30%

1174 1170 1178 1175 1171 1181 1172 1167 1175

91.68 84.70 96.38 92.36 84.62 97.98 70.77 65.81 84.95

1244 1239 1248 1242 1234 1250 1227 1224 1231

97.27 93.53 98.51 96.28 88.66 99.34 84.62 82.05 87.07

Coal B

Coal C

E

Tb = 925 °C (1198 K)


Article

Energy & Fuels

Figure 5. Fuel-N conversion to NOx for different combustion atmospheres and temperatures.

Table 5. Review of NOx Emissions under Oxy-Fuel Combustion in BFB Units researchers

coal types

parameters

observations

de Diego et al.20

anthracite

air and oxy-fuel 35% [O2]

Lupiáñez et al.,19,13,21

lignite, bituminous, anthracite, and blend

air, oxy, and gas staging 21−45% [O2]

Roy et al.22

brown coal

air and oxy-fuel 15−30% [O2]

this study

sub-bituminous


NO emissions were similar in air and oxy-combustion conditions. NO was increased slightly with an increase in temperature. The presence of steam lead to a decrease in NO. NO concentration in oxy-fuel higher than in air combustion. NO emissions and fuel-N conversion ratios higher at higher bed temperatures. NO emissions found to increase with increase in excess O2. Gas staging reported to be advantageous toward NOx reductions. NOx emissions lower in oxy-fuel combustion than in air combustion. NOx formation found to increase with O2 concentration in the feed gas and with temperature. Significantly lower fuel-N conversion in oxy-fuel combustion than in air combustion for all the three coal explored. The highest conversion was obtained in Coal C. This was attributed to fuel-bound nitrogen content of the parent coals.

concentration in oxy-fuel combustion. This was also observed in our study for the three coals. The highest conversions were observed for Coal C in air combustion, followed by Coal B, and last Coal A. This can be attributed to fuel-bound nitrogen content of the parent coals, where their respective nitrogen contents were 1.07, 1.11, and 1.19% for Coals A, B, and C, respectively. In this study, there was no correlation between the volatile content in the coals and the NOx emissions. A summary of comparison of results with those obtained by different researchers is presented in Table 5. 3.3.2. SOx Emissions. Figure 6 shows the fuel-S conversion to SO2 for different combustion atmospheres at 850 and 925 °C. The results show that air combustion had higher sulfur conversion ratios than the oxy-fuel atmosphere for the three coals. An increase in temperature from 850 to 925 °C lead to a decrease in fuel-S conversions. The influence of oxygen concentration on SOx emission was observed to be insignificant in this study. Czakiert et al.12 also reported lower sulfur conversions at higher temperature in oxy-combustion, irrespective of O2 concentration. Lupiáñez et al.13 reported an optimum temperature of 900 °C for effective SO2 capture. Coal A had the highest sulfur conversion, followed by C, and last B. On the contrary, Coal A had the lowest fuel-bound sulfur of 0.6%, followed by C with 0.75% and B with 0.93%. This signified that the coal with the highest sulfur content had the highest sulfur retention within the ash. This was in good agreement with the study by de las Obras-Loscertales et al.,23 who observed that sulfur capture depends on the amount of sulfur in the coal sulfur and the higher the sulfur in the coal, the higher the sulfur capture efficiency. A summary of comparison of results with those obtained by different researchers is presented in Table 6.

particle temperatures and carbon burnouts are also presented in Table 3. There is a good correlation between the char particle temperature and carbon burnout. Where there was a marginal difference in the carbon burnout, there was also a marginal difference in the char particle temperature, and where there was a significant difference in the carbon burnout, there was a significant difference in the char particle temperature. A higher char particle temperature resulted in a higher carbon burnout in all the conditions. The highest char particle temperature was at 30% oxy-combustion (30% O2/CO2), and the lowest was at 21% oxy-combustion (21% O2/CO2). For the three coals, the highest char particle temperature (1250 K) was in Coal B at 30% oxy-combustion (30% O2/CO2) at 925 °C (1198 K) bed temperature, and it had the highest carbon burnout of 99.34%. The lowest char particle temperature (1167 K) was at 21% oxy-combustion (21% O2/CO2) at 850 °C for Coal C and the lowest carbon burnout of 65.81%. 3.3. Pollutants Emissions. 3.3.1. NOx Emissions. The nitrogen conversion ratios to NOx (NO plus NO2) were calculated using eqs 4 and 5. The results are presented in Figure 5. The fuel nitrogen (fuel-N) conversion ratios were significantly higher in air conversion than in oxy-fuel for all fuels at both temperatures. This observation was also made by other authors.8,13 This was attributed to the absence of nitrogen in the oxy-fuel combustion atmosphere and the low NOx formation was mainly from the fuel nitrogen. Furthermore, the fuel nitrogen (fuel-N) conversion increased with an increase in temperature and oxygen concentration in oxy-combustion. Lupiáñez et al.19 also reported that fuel-N conversions to NOx to increased with an increase in temperature and oxygen F


Article

Energy & Fuels

Figure 6. Fuel-S conversion to SOx for different combustion atmospheres and temperatures.

Table 6. Review of SOx Emissions under Oxy-Fuel Combustion in BFB Units researchers

coal types

de las Obras-Loscertales et al.23

lignite and anthracite

oxy 27−45% [O2]

parameters

observations

Lupiáñez et al.21,13

lignite and anthracite

air, oxy, and gas staging 21−45% [O2]

Roy et al.22

brown coal


this study

sub-bituminous


Sulfur retention depends on the amount of sulfur in the coalthe higher the sulfur in coal, the higher the retention. Higher emission in lignite and lower emission in anthracite. SO2 emissions increased in anthracite when gas-staging was used. Higher fuel-S conversion to SO3 in air combustion than in oxy-fuel. An increase in oxygen concentration lead to an increase in fuel-S conversion to SO3. Fuel-S conversion ratios to SOx indicated that air combustion had relatively higher conversions than oxy-fuel, with the highest conversion of 35% observed for Coal A at 925 °C.

• The lowest particle temperature (1167 K) was obtained in Coal C at 21% oxy-combustion (21% O2/CO2) at 850 °C bed temperature, and it had the lowest carbon burnout of 65.81%. The following conclusions were drawn for NOx and SOx emissions:

4. CONCLUSION This study reports on the experimental analysis of oxy-fluidized bed combustion of three SA coals in a bubbling fluidized bed reactor. The results include carbon burnout and NOx and SOx emissions. Oxy-combustion tests were conducted at 21% O2/CO2 and 30% O2/CO2 and at two different temperatures of 850 and 925 °C. The results were compared with those obtained from air combustion. The char particle temperature was modeled for all the coals to validate the significance of particle temperature on char conversion. On the basis of the results obtained, the following conclusions were drawn: • The highest carbon burnout was observed at 30% O2/CO2, followed by air, and last 21% O2/CO2 at both temperatures (850 and 925 °C). • Coals A and B had a higher carbon burnout than Coal C. • Carbon burnout in Coal C was about 20% lower than those of Coals A and B • There was a marginal difference of 1% in the carbon burnout in Coals A and B in all the conditions. • An exception was at 21% O2/CO2 at 925 °C, when the carbon burnout of Coal A was 5% higher than that of Coal B. • The carbon burnout had a good correlation with the particle temperature at all the combustion conditions studied for the three coals. • The carbon burnout only had a good correlation with vitrinite random reflectance at 21% O2/CO2 at 925 °C for the three coals. • The highest particle temperature (1250 K) was obtained in Coal B at 30% oxy-combustion (30% O2/CO2) at 925 °C bed temperature, and it had the highest carbon burnout of 99.34%.

■

• Fuel nitrogen (fuel-N) conversion ratios were significantly higher in air combustion than in oxy-combustion for all the coals at both temperatures. • Fuel-N conversions to NOx were found to increase with an increase in temperature and oxygen concentration in oxy-fuel combustion. • Similar to NOx emissions, air combustion had higher sulfur conversion ratios than the oxy-fuel atmosphere for all the coals. • Fuel-S conversions were observed to decrease with an increase in temperature.

AUTHOR INFORMATION

Corresponding Author

*Tel.: +27 128412371, +27 731740431. Fax: +27 841 2135. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS

This work was funded by the CSIR South Africa and South African National Research Foundation (NRF) (Grant number TTK13060718689). Opinions expressed and conclusions are those of the authors and the funders are not liable. G


Article

Energy & Fuels

■

REFERENCES

(1) Telsnig, T.; Tomaschek, J.; Ö zdemir, E. D.; Bruchof, D.; Fahl, U.; Eltrop, L. Energy Policy 2013, 63, 168−180. (2) Akorede, M. F.; Hizam, H.; Ab Kadir, M. Z. A.; Aris, I.; Buba, S. D. Renewable Sustainable Energy Rev. 2012, 16 (5), 2747−2761. (3) Lupion, M. B.; Navarrete, P.; Otero; Cortés, V. J. Chem. Eng. Res. Des. 2011, 89 (9), 1494−1500. (4) Huaman, R. N. E.; Jun, T. X. Renewable Sustainable Energy Rev. 2014, 31, 368−385. (5) Wall, T.; Liu, Y.; Spero, C.; Elliott, L.; Khare, S.; Rathnam, R.; Zeenathal, F.; Moghtaderi, B.; Buhre, B.; Sheng, C.; Gupta, R.; Yamada, T.; Makino, K.; Yu, J. Chem. Eng. Res. Des. 2009, 87 (8), 1003−1016. (6) Mathekga, H.; Oboirien, B.; North, B. Int. J. Energy Res. 2016, 40, 878−902. (7) Roy, B.; Bhattacharya, S. Fuel Process. Technol. 2014, 117 (0), 23−29. (8) Hofbauer, G.; Beisheim, T.; Dieter, H.; Scheffknecht, G. Energy Procedia 2014, 51 (0), 24−30. (9) Lasek, J. A.; Janusz, M.; Zuwała, J.; Głód, K.; Iluk, A. Energy 2013, 62, 105−112. (10) Wang, C.; Jia, L.; Tan, Y.; Anthony, E. J. Energy Fuels 2011, 25 (2), 617−623. (11) Oboirien, B. O.; North, B. C.; Kleyn, T. Energy 2014, 66, 550− 555. (12) Czakiert, T.; Sztekler, K.; Karski, S.; Markiewicz, D.; Nowak, W. Fuel Process. Technol. 2010, 91 (11), 1617−1623. (13) Lupiáñez, C.; Guedea, I.; Bolea, I.; Díez, L. I.; Romeo, L. M. Fuel Process. Technol. 2013, 106 (0), 587−594. (14) Duan, L.; Zhao, C.; Zhou, W.; Qu, C.; Chen, X. Int. J. Greenhouse Gas Control 2011, 5 (4), 770−776. (15) Engelbrecht, A. D.; North, B. C., Hadley, T. D. Clean coal technology: Gasification of South African coals. In Triennial Industrial Fluidization South Africa (IFSA 2008) Conference, The Southern African Institute of Mining and Metallurgy, Johannesburg, South Africa, Nov 19−20, 2008; Industrial Fluidization South Africa, 2008; pp 188−202. (16) Dhaneswar, S. R. Effect of Coal Rank During Oxy-fuel Combustion: Role of Char-CO2 Reaction. Master Thesis, Pennsylvania State University, State College, PA, 2011. (17) Naredi, P. Coal combustion in O2/CO2 medium for power generation: Numerical modeling of char burnout, NOx and CO emissions. Thesis, Pennsylvania State University, State College, PA, 2009. (18) Eisermann, W.; Johnson, P.; Conger, W. L. Fuel Process. Technol. 1980, 3 (1), 39−53. (19) Lupiáñez, C.; Díez, L. I.; Romeo, L. M. Energy Fuels 2013, 27 (12), 7619−7627. (20) de Diego, L. F.; de las Obras-Loscertales, M.; Rufas, A.; GarcíaLabiano, F.; Gayán, P.; Abad, A.; Adánez, J. Appl. Energy 2013, 102 (0), 860−867. (21) Lupiáñez, C.; Díez, L. I.; Romeo, L. M. Chem. Eng. J. 2014, 256 (0), 380−389. (22) Roy, B.; Chen, L.; Bhattacharya, S. Environ. Sci. Technol. 2014, 48, 14844−14850. (23) de las Obras-Loscertales, M.; de Diego, L. F.; García-Labiano, F.; Rufas, A.; Abad, A.; Gayán, P.; Adánez, J. Fuel 2014, 137 (0), 384− 392.

H


Performance Evaluation of South African Coals ... - ACS Publications

Recommend Documents