- Email: [email protected]
Effect of particle size on particulate matter emissions during biosolid char combustion under air and oxyfuel conditions
Fuel 232 (2018) 251–256
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Full Length Article
Effect of particle size on particulate matter emissions during biosolid char combustion under air and oxyfuel conditions Sui Boon Liawa, Xujun Chena, Yun Yua, Mário Costab, Hongwei Wua,
T
⁎
a
Discipline of Chemical Engineering, Western Australian School of Mines: Minerals, Energy and Chemical Engineering, Curtin University, GPO Box U1987, Perth, WA 6845, Australia b IDMEC, Mechanical Engineering Department, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal
A R T I C LE I N FO
A B S T R A C T
Keywords: Biosolid Particulate matter Particle size Char Oxyfuel combustion
In an industrial scale furnace, solid fuels can be fired in a wide range of particle sizes. This study aims to investigate the effect of the particle size on particulate matter (PM) emissions during combustion under air and oxyfuel (30% O2/70% CO2) conditions. Biosolid chars prepared from the pyrolysis of three different biosolid particle sizes at 1300 °C were burned in a drop tube furnace. The experimental results indicate that a shift from small to large char particles leads to > 55% reduction in PM1–10 emission when combustion occurs in air, likely due to less intense char fragmentation experienced by the large char particles. Such a reduction originates from a decrease in Mg, Ca, P, Si, Al and some trace elements (V, Co, Cu, Zn and Mn) release as PM1–10. In contrast, PM1 emission is not affected by the char particle size. However, under oxyfuel conditions, the PM1–10 emission from the small char particles is ∼65% of that released during combustion in air. In addition, the PM1–10 emission from large char particles only reduced by ∼27% when compared to that from small char particles. This observation is likely to result from the coalescence of ash particles to form PM with particle sizes > 10 µm due to the increase surface mobility caused by repeated formation and decomposition of CaCO3 in the ash.
1. Introduction Biosolid is a byproduct from wastewater treatment processes and has generated significant environmental issues for disposal, mainly due to its high contents of toxic organic compounds and heavy metals [1]. Among the available technologies for biosolids utilisation, combustion is one of the most effective strategies not only to achieve a large volume reduction but also to recover the energy in the biosolid [2]. However, combustion of various fuels are known to release particulate matters (PMs) that have potential hazard to the human health and to the environment (soil, air, and water) [3–9]. This is of particular concerns because biosolid contains high content of ash, resulting in PM significant emissions of during combustion [8,9]. A large portion of trace metals (i.e., Cu, Cr, Ni, Pb, V, Co, Cd and As) present in biosolid are also released in the PM during combustion [10]. Therefore, it is of critical importance to understand the emissions of PM and trace elements during the combustion of biosolid. Usually, biosolid is fired under air conditions in boilers to generate electricity [2]. Recently, there has been an increasing need to develop oxyfuel combustion technology in order to facilitate carbon capture and storage and, thereby, to achieve negative carbon emissions from
⁎
Corresponding author. E-mail address: [email protected] (H. Wu).
https://doi.org/10.1016/j.fuel.2018.05.163 Received 1 March 2018; Received in revised form 24 May 2018; Accepted 29 May 2018 0016-2361/ © 2018 Elsevier Ltd. All rights reserved.
biosolid-fired power plants [11]. Although PM emissions from biosolid combustion under air conditions are relatively well documented in the literature [2,8,12,13], studies on oxyfuel combustion of biosolid are scarce. Surprisingly, the release of trace elements in the PM during the combustion of biosolid under oxyfuel conditions is also seldom reported. In order to address this problem, this work presents a systematic study on the emissions of PM (including major and trace elements) during the combustion of biosolid in both air and oxyfuel environments. A distinct feature of this work is the emphasis placed on the effect of the particle size on PM emissions that, to the best of our knowledge, has not been considered before. Furthermore, to eliminate the possible effect on the PM emissions of the combustion of the volatiles [14], biosolid chars were first produced from pyrolysis at 1300 °C, and subsequently used for the combustion experiments under air and oxyfuel conditions. 2. Materials and methods 2.1. Sample preparation Biosolid, provided by a waste water treatment facility in Western Australia, was dried overnight in an oven at 105 °C before it was ground
Fuel 232 (2018) 251–256
S.B. Liaw et al.
Table 1 Properties of the biosolid size fractions of 38–60 µm, 60–90 µm and 120–150 µm and its respective chars prepared from pyrolysis at 1300 °C, which are referred to as small char particle, medium char particle and large char particle, respectively. Char samples
38–60 µm biosolid 60–90 µm biosolid 120–150 µm biosolid Small char particle Medium char particle Large char particle a b c d e
Moisture (wt%, ada)
Proximate analysis (wt%, dbb)
Table 2 Inorganic elements content of the biosolid chars. Small char particle
Ultimate analysis (wt% dafe)
ash
VMc
FCd
C
H
N
O
5.5
20.4
68.9
10.6
53.54
10.90
6.79
28.77
5.2
20.4
70.5
9.1
53.41
11.02
6.84
28.73
4.5
20.2
70.0
9.8
55.08
11.07
7.13
26.72
1.2
84.0
5.2
10.9
98.61
0.15
0.37
0.86
0.9
84.3
4.7
11.1
99.21
0.11
0.34
0.34
0.6
83.9
5.4
10.7
99.33
0.10
0.40
0.17
Air-dried. Dry-basis. Volatile matter. Fixed carbon. Dry and ash free basis.
Large char particle
Major elements (wt%, dry basis) Na 0.642 ± 0.053 K 0.642 ± 0.016 Mg 5.692 ± 0.045 Ca 13.486 ± 0.057 Al 3.323 ± 0.036 Fe 1.655 ± 0.034 Si 8.569 ± 0.030 Ti 0.797 ± 0.004 P 11.649 ± 0.160 S 0.204 ± 0.002
0.727 ± 0.095 0.867 ± 0.010 4.597 ± 0.101 13.470 ± 0.312 3.460 ± 0.014 1.710 ± 0.024 7.814 ± 0.035 0.973 ± 0.024 11.574 ± 0.252 0.220 ± 0.002
0.591 ± 0.100 0.679 ± 0.018 3.825 ± 0.028 13.007 ± 0.053 3.753 ± 0.060 1.747 ± 0.007 9.056 ± 0.030 1.022 ± 0.008 9.502 ± 0.030 0.298 ± 0.002
Trace elements (mg/kg, dry basis) As 4.075 ± 0.091 V 27.501 ± 0.120 Cr 273.921 ± 36.0 Co 21.748 ± 0.258 Cd 2.045 ± 0.030 Pb 7.607 ± 0.109 Zn 659.925 ± 16.4 Mn 523.917 ± 13.8 Cu 1755.5 ± 37.4
4.041 ± 0.028 34.141 ± 3.943 207.1 ± 16.7 25.181 ± 2.822 2.046 ± 0.087 6.789 ± 3.395 697.165 ± 4.8 522.624 ± 3.1 1778.0 ± 10.3
3.312 ± 0.477 31.402 ± 4.346 344.9 ± 47.5 19.389 ± 1.295 1.617 ± 0.188 8.404 ± 1.049 545.563 ± 12.5 545.388 ± 17.2 2131.5 ± 50.8
2.2. Char combustion experiments The biosolid char combustion experiments under air and oxyfuel (30% O2/70% CO2) conditions were carried out in a DTF equipped with a sampling system that comprises a cyclone and a 13-stage Dekati lowpressure impactor (DLPI) coupled with a backup filter [14,16]. During the char combustion experiments, ∼0.03 g/min of each char, entrained in a stream of 1 L/min of air or oxyfuel gas, was fed into the DTF at 1300 °C. A second flow of 4.6 L/min of air or oxyfuel gas was introduced in the DTF to assist the char combustion process. The value of λ (expressed as the ratio of the actual air/fuel ratio to the stoichiometric air/fuel ratio) was ∼30 and the residence time of the char particles in the isothermal zone of the DTF was ∼1.2 s. The sampling temperature was set at 115 °C in this study. The ash and PM particles with aerodynamic diameter of > 10 μm were first removed from the flue gas by the cyclone, and then the flue gas was directed to the DLPI for size-segregated collection. In this study, PMs with aerodynamic diameters < 0.1 µm, 0.1–1 µm, < 1 µm, 1–10 µm and < 10 µm are termed as PM0.1, PM0.1–1, PM1, PM1–10 and PM10, respectively. All experiments were carried out at least in triplicates. It should be noted that complete combustion was achieved for all combustion experiments as thermogravimetric analysis showed the absence of unburned carbon in the PM samples, and total organic carbon analysis showed that leachates from washing PM samples in pure water for 24 h contained negligible amounts of organic carbon.
and sieved to yield three size fractions of 37–60 µm, 60–90 µm, 120–150 µm for the pyrolysis experiments. Table 1 presents the properties of the prepared biosolid. The pyrolysis experiments were conducted in a drop tube furnace (DTF) described in detail elsewhere [15]. Briefly, the DTF was preheated to 1300 °C, and then samples of the three biosolid size fractions were in turn entrained by 1 L/min of ultra-high purity nitrogen into the DTF through a water-cooled feeding probe using a sample feeder system. An additional stream of 1 L/min of nitrogen was used to prevent backflow of the volatiles. The char produced was sampled using a water-cooled sample probe. A stream of 1 L/min of helium was injected at the probe tip to quench the hot flue gas, while extra nitrogen was introduced to give a total gas flow of 10 L/min before the char particles were collected in a Dekati cyclone. Henceforth, the biosolid chars prepared from pyrolysis of the biosolid size fractions of 37–60 µm, 60–90 µm and 120–150 µm are termed as small, medium and large char particles, respectively. Fig. 1 shows the cumulative particle size distribution of the three biosolid chars, and Tables 1 and 2 show their main properties.
100
2.3. Sample analysis
80
Cumulative (%)
Medium char particle
The proximate analysis of the biosolid chars was carried out in a thermogravimetric analyser (TGA, Mettler Toledo Star 1) according to the time-temperature program described in ASTM E870-82. The C, H and N contents were determined by an elemental analyser (PerkinElmer 2400 CHNS/O Series II). The sulphur content in the biosolid chars were determined by an improved Eschka method [17]. The quantification of the contents of major (Na, K, Mg, Ca and P) and trace (As, Cr, Cu, Mn, Pb, Ti, and Mn) elements were made by using inductively coupled plasma-optical emission spectroscopy (ICP-OES, PerkinElmer Optima 8300) and inductively coupled plasma-mass spectroscopy (ICP-MS, PerkinElmer NexION 350D), respectively, with the biosolid chars being first ashed and then digested in a concentrated HNO3/HF mixture (detailed procedure described elsewhere [18]). Likewise, the major and trace elements in PM were also quantified using ICP-OES and ICP-MS,
60
40
Small particle Medium particle Large particle
20
0 0
50
100
150
Char particle diameter
200
250
m)
Fig. 1. Cumulative particle size distribution of the three biosolid chars. 252
Fuel 232 (2018) 251–256
15
Combustion in air
Combustion in oxyfuel
(a)
(b)
12 9
0.5
Small particle Medium particle Large particle
0.4
Macroporosity
dm/dlogDp (mg/g char, db)
S.B. Liaw et al.
6 3 0 0.01
0.1
1
10 0.01
0.1
1
0.3 0.2 0.1
10
Aerodynamic diameter ( m)
Yield (mg/g char, db)
8
0.0
(c)
(d) Small particle Medium particle Large particle
6
B
C
D
E
A
B
C
D
Large Particle
indicated in Fig. 3, the macroporosities of the chars decrease with char particle size, from ∼0.41 for the small char particles to ∼0.28 for the large char particles. This is reasonable as during pyrolysis of large biosolid particles, the volatiles generated within the large particles would experience more reactions with the hot char as the volatiles diffuse through the pores within the pyrolysing particles, leading to more deposition of secondary char or soot within the pore structure [24]. This subsequently leads to partial blockage of char pore structure and reduces the macroporosity of the char particles. Such an effect would be much less intensive during pyrolysis of small biosolid particles. Therefore, the combustion of small char particles that have a higher macroporosity results in a higher PM1–10 yield compared to that from the combustion of large char particles that have lower macroporosity. In addition, the small char particles have a significantly shorter gas diffusion path length and larger external surface area as compared to the larger char particles. This implies that an increase in particle size from small to large can promote a change in the combustion regime from chemical control to diffusion control [25], which can greatly affect the char particle temperature and, thus, its fragmentation intensity. Furthermore, smaller particles may experience a greater degree of thermal shock during combustion [22]. Consequently, the combustion of large-size char sample results in a > 55% reduction in PM1–10 yield.
2
A
Medium Particle
Fig. 3. Macroporosities of the three biosolid chars with different sizes (small, medium and large) used in this study.
4
0
Small Particle
E
A: PM0.1 B: PM0.1-1 C: PM1 D: PM1-10 E: PM10 Fig. 2. PSDs and yields of PM from the combustion of the three biosolid chars (small, medium and large particles) under air (panel a and c) and oxyfuel (panel b and d) conditions.
respectively, after digestion in a concentrated HNO3/HF mixture. The contents of Si, Al and Fe in the biosolid chars and PM samples were determined using the ash borate fusion method described in AS 1038.14.1-2003 [19] with the fusion bead dissolved in dilute nitric acid for analysis in ICP-OES. For determination of char macroporosity, the biosolid char was set in a resin pellet that was then polished and coated with platinum for cross-sectional imaging using a scanning electron microscope (SEM), followed by image analysis according to a previous method [20]. An image analysis software, Digimizer, was employed for analysing the acquired SEM images and determining the macroporosity of each char sample via analysing at least 100 char particles for each sample. The particle size of the biosolid chars was determined through image analysis of, at least, 200 particles using the same image analysis software.
3.2. Major elements in PM during char biosolid combustion 3. Result and discussion Figs. 4a–i and 5a–i show the distributions and the yields, respectively, of the major elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under air conditions. The PM1 released from the biosolid chars combustion in air consists mainly of volatile alkali species (Na and K) and P, likely in form of (Na,K)PO3 [14]. However, some refractory elements such as Si, Al and Fe are also observed in PM1, likely due to vaporisation of metal oxides under local reducing condition formed during char combustion [26]. The elemental yield data indicates that the yields of these elements in PM1 from the biosolid chars combustion are insensitive to the char particle size. In addition, the PM1–10 released mainly consists of P, Ca, Mg, Si, Al and Fe. A very small amount of Na and K is also observed in the PM1–10, most likely due to their release as aluminosilicate compounds during fragmentation. Figs. 3a–i and 4a–i show a clear reduction in the yields of Mg, Ca, P and Si in PM1–10, which are the major elements in the char where its oxides contributed > 70% to the total ash in the char. This further suggests that an increase in the char particle size leads to less intense char fragmentation and thus lower yields of these elements in the PM1–10 released. However, the decrease in the yields of Al and Fe in the PM1–10 is minimal probably because these elements are less abundant in the char as compared to P, Ca, Mg and Si.
3.1. Particle size distribution of PM during char biosolid combustion Fig. 2a–c show the particle size distributions (PSDs) and the yields of the PM from the combustion of the three biosolid chars (small, medium and large particles) under air conditions. The PSDs from the combustion of these chars exhibit a bimodal distribution with a course mode of ∼6.863 µm and a fine mode of ≤0.01 µm. The PM10 emitted is dominantly (> 50%) PM1–10 as majority of alkali metal, volatile S and P were released as volatiles during pyrolysis to yield the char samples. While the differences in the PM1 emitted from the combustion of the three biosolid chars are negligible, it is clear that an increase in char particle size leads to a substantial reduction in PM1–10 emission. There is a ∼12% reduction in PM1–10 emission between the combustion of small and medium char particles, whereas a > 55% reduction is observed for the combustion of the large char particles as compared to the combustion of the small char particles. Such a reduction can be attributed to char fragmentation intensity during char combustion, which is governed by char particle size and macroporosity [21,22]. Previous studies demonstrated that the combustion of char with a higher macroporosity leads to a higher yield of fine particulates [21,23]. As 253
Fuel 232 (2018) 251–256
S.B. Liaw et al.
Combustion in air
1.8
(a) Na
Combustion in oxyfuel
Combustion in air 1.2 (a) Na
(j) Na Small particle Medium particle Large particle
1.2
1.8
(b) K
0.4 0.0 1.2 (b) K
(k) K
0.4
0.6 1.8
(c) Mg
0.0 1.2 (c) Mg
(l) Mg
0.4
0.6 1.8
(d) Ca
0.0 1.2 (d) Ca
(m) Ca
(e) P
Yield (mg/g char, db)
dm/dlogDp (mg/g char, db)
0.4
0.6 1.8
(n) P
1.2 0.6 0.0 1.8
(f) Al
(o) Al
(g) Fe
(p) Fe
0.4 0.0 1.2 (f) Al
0.0 1.2 (h) Si (h) Si
(q) Si
0.4 0.0 1.2 (i) Ti (i) Ti
(r) Ti
0.4
0.6
0.0
0.0 0.1
1 10 0.01 0.1 Aerodynamic diameter ( m)
(r) Ti
0.8
1.2
0.01
(q) Si
0.8
0.6 1.8
(p) Fe
0.4
1.2
0.0
(o) Al
0.8
0.6 1.8
0.8
0.0 1.2 (g) Fe
1.2
0.0
(n) P
0.4
0.6 1.8
0.0 1.2 (e) P
0.8
1.2
0.0
(m) Ca
0.8
1.2
0.0
(l) Mg
0.8
1.2
0.0
(k) K
0.8
1.2
0.0
Small particle Medium particle Large particle
0.8
0.6 0.0
Combustion in oxyfuel (j) Na
1
A
B
C
D
E
A
B
C
D
E
A: PM0.1 B: PM0.1-1 C: PM1 D: PM1-10 E: PM10
10
Fig. 5. Yields of the major elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under air (panel a–i) and oxyfuel (panel j–r) conditions.
Fig. 4. Distributions of the major elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under air (panel a–i) and oxyfuel (panel j–r) conditions.
biosolid chars (small, medium and large particles) under air conditions. Volatile trace elements such as As, Cr, Cd and Pb contributed mainly to the formation of PM1, forming little PM1–10. It is interesting to note that although Zn is a volatile trace element [14], a portion of Zn remained in the char and thus contributed to the PM1–10 formation during the
3.3. Trace elements in PM during char biosolid combustion Figs. 6a–i and 7a–i show the distributions and the yields, respectively, of the trace elements in PM from the combustion of the three 254
Fuel 232 (2018) 251–256
S.B. Liaw et al.
0.6
Combustion in air
Combustion in oxyfuel (j) As
(a) As
Small particle Medium particle Large particle Combustion in air Combustion in oxyfuel 0.4 (a) As (i) As
Small particle Medium particle Large particle
0.4
0.3
0.2 0.0
0.2 (b) V
0.1
(k) V
0.4
0.0 0.3 (b) V
0.2
0.2 0.1
0.0 (c) Cr 15
(l) Cr
0.0 (c) Cr 15
10
5 (d) Co
(m) Co
0 0.20 (d) Co
0.2
0.15
0.1
0.10
(l) Co
0.05
0.0 30 (e) Cu
(n) Cu
Yield ( g/g char, db)
dm/dlogDp ( g/g char, db)
0.3
(k) Cr
10
5 0
(j) V
20 10 0 0.2 (f) Cd
(o) Cd
0.00 20 (e) Cu
(m) Cu
15 10 5 0 0.12 (f) Cd
(n) Cd
0.09
0.1
0.06 0.03
0.0 (g) Pb 2
0.00 (g) Pb 1.5
(p) Pb
(o) Pb
1.0
1
0.5
0 32 (h) Zn
0.0 (h) Zn 15
(q) Zn
24 16
10
8
5
0 9 (i) Mn
0
(r) Mn
3
6
(p) Zn
(h) Mn
(q) Mn
2
3
1
0
0
0.01
0.1
1
10 0.01
0.1
1
10
Aerodynamic diameter ( m)
A
B
C
D
E
A
B
C
D
A: PM0.1 B: PM0.1-1 C: PM1 D: PM1-10 E: PM10
E
Fig. 7. Yields of the trace elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under air (panel a–i) and oxyfuel (panel j–r) conditions.
Fig. 6. Distributions of the trace elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under air (panel a–i) and oxyfuel (panel j–r) conditions.
the release of Zn as PM1–10. In addition, semi-volatile trace elements such as Cu and V contributed to both PM1 and PM1–10 formation. Co and Mn are non-volatiles trace elements and only contributed to the PM1–10 formation. Similar to the major elements, the yields of trace elements such as V, Co, Cu, Zn and Mn contributed to PM1–10. PM1–10 emission also reduces with increasing biosolid char particle size due to the less intense char fragmentation during combustion in air. However,
biosolid chars combustion. Our previous study showed almost complete release of Zn during pyrolysis of biosolid at 1000 °C [14] as Zn compounds (i.e. ZnO) begin to decomposed at temperature > 600 °C to form gaseous elemental Zn [27]. However, under the higher pyrolysis temperature of 1300 °C used in this work, Zn can react with silicate, Fe and Mn in the char [28]. Subsequent biosolid char combustion leads to 255
Fuel 232 (2018) 251–256
S.B. Liaw et al.
4. Conclusions
the char particle size has little influence on the formation of PM1 from trace elements except for Pb. Higher Pb is released from the combustion of large char particles compared to that from small char particle. The exact mechanisms responsible for such an observation are unknown. However, one possibility is that Pb maybe react with aluminosilicate and retained in coarse ash collected in cyclone so that the combustion of smaller char particles, which provide more surface for such reactions, can lead to less release of Pb as PM1 [29].
This work investigated the effect of the biosolid char particle size on the release of PM10 during combustion under air and oxyfuel conditions. While a change in the char particle size does not affect the yield of PM1, increasing the char particle size from small to large leads to > 55% reduction in PM1–10 formation, likely due to less intense char fragmentation during the combustion of the large char particles. This reduces the formation of Mg, Ca, P, Al, Si and, to a lesser extent, Fe in PM1–10. An increase in the char particle size also reduces the release of trace elements such as V, Co, Cu, Zn and Mn as PM1–10. Switching the combustion atmosphere from air to oxyfuel substantially retards the formation of PM1–10, especially for the biosolid char with the small particle size, possibly due to the high coalescence tendency caused by repeated formation and decomposition of CaCO3 in CO2 rich atmospheres to form coarse ash particles > 10 µm. Consequently, the effect of the char particle size on PM1–10 release is less significant. Although an increase in the char particle size reduces the release of Mg, Ca, P, Al, Si and Fe in PM1–10, such reductions are not observed for trace elements probably due to their low relative content in the biosolid char.
3.4. Impact on the results of the oxyfuels conditions Under oxyfuel conditions, Fig. 2b and 2d show that the PM from the combustion of the three biosolid chars also exhibit a bimodal distribution with a coarse mode of ∼6.863 µm and a fine mode of ≤0.01 µm. However, the yield of PM1–10 is less than that under air conditions, particularly for small and medium char particles. The yield of PM1–10 from the combustion of large char particles under oxyfuel conditions is only slightly less than that from combustion in air, which is in line with data reported previously [14]. However, the yield of PM1 is not affected by the char particle size. This observation also indicates that combustion under oxyfuel conditions can greatly reduce the formation of PM1–10 from solid fuels with small particles. Although the increase in the char particle size also leads to a decrease in the PM1–10 yield, the yield of PM1–10 from the combustion of large char particles under oxyfuel conditions only drop by ∼27% as compared to that from the combustion of small char particles under oxyfuel conditions. The reduction is significantly less than the > 55% reduction observed in the combustion under air conditions. Such phenomena may be attributed to the higher ash coalescence tendency during combustion of the biosolid chars under oxyfuel conditions as compared to combustion in air. A previous study reported that in CO2 rich atmospheres, CaCO3 can repeatedly formed and decomposed to CaO, catalysing the surface mobility [30]. The char used in present study has a high (> 13%) Ca content. The CO2 rich atmosphere under oxyfuel conditions can lead to significant sintering and coalescence of ash, leading to the formation of coarse particles and reduction in PM1–10 formation. Since the small char particles experience more intense char fragmentation, its combustion under oxyfuel conditions lead to a more significant reduction in PM1–10 as compared to the oxy-combustion of the large char particles, where such effect is minimal. Figs. 4j–r and 5j–r show the distributions and the yields, respectively, of the major elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under oxyfuel conditions. Similar to biosolid char combustion in air, PM1 mainly consists of Na, K and some P, while Mg, Ca, Al, Fe, Si and P are dominant in PM1–10. The increased tendency in ash coalescence led to lower yields of Mg, Ca, P and Si in PM1–10 under oxyfuel conditions as compared to combustion in air. However, the reduction in the Al and Fe yields is less significant and within the experimental uncertainty. Figs. 6j–r and 7j–r show the distributions and the yields, respectively, of the trace elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under oxyfuel conditions. Similar to the biosolid chars combustion in air, trace elements such as As, Cu, Cr, Cd, V and Pb mainly contributed to the PM1 formation, while V, Co, Cu, Zn and Mn contributed to the formation of PM1–10. Although the increase in char particle size reduces the amounts of V, Co, Cu and Zn in PM1–10, such reduction is not observed, except for Mn, most likely as a result of the low content of these elements in the char and less significant effect of the char particle size on PM1–10 released.
Acknowledgements The authors would like to acknowledge the partial support received from the Australian Research Council through its Discovery Projects Scheme. The authors would like to thank Jinxiu Cao for her assistance in part of the experimental work. X. Chen is also grateful to Curtin International Postgraduate Research Scholarship for supporting his PhD study. M. Costa also acknowledges Fundação para a Ciência e a Tecnologia for the sabbatical leave Grant SFRH/BSAB/128154/2016. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]
[20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]
256
Fytili D, Zabaniotou A. Renewable Sustainable Energy Rev 2008;12:116–40. Werther J, Ogada T. Prog Energy Combust Sci 1999;25:55–116. Sheng C, Lu Y, Gao X, Yao H. Energy Fuels 2007;21:435–40. Denison RA, Silbergeld EK. Risk Anal 1988;8:343–55. Xu M, Yu D, Yao H, Liu X, Qiao Y. Proc Combust Inst 2011;33:1681–97. Li G, Li S, Dong M, Yao Q, Guo CY, Axelbaum RL. Fuel 2013;106:544–51. Patiño D, Crespo B, Porteiro J, Villaravid E, Granada E. Powder Technol 2016;288:164–75. Ninomiya Y, Zhang L, Sakano T, Kanaoka C, Masui M. Fuel 2004;83:751–64. Seames WS, Fernandez A, Wendt JOL. Environ Sci Technol 2002;36:2772–6. Liaw SB, Rahim MU, Wu H. Energy Fuels 2016;30:5766–71. Toftegaard MB, Brix J, Jensen PA, Glarborg P, Jensen AD. Prog Energy Combust Sci 2010;36:581–625. Wang Q, Zhang L, Sato A, Ninomiya Y, Yamashita T, Dong Z. Proc Combust Inst 2009;32:2701–8. Zhang L, Ninomiya Y. Proc Combust Inst 2007;31:2847–54. Liaw SB, Wu H. Energy Fuels 2017;31:2317–23. Gao X, Wu H. Energy Fuels 2011;25:2702–10. Gao X, Wu H. Energy Fuels 2011;25:4172–81. Rahim MU, Gao X, Wu H. Energy Fuels 2013;27:6992–9. Zhang M, Gao X, Wu H. Energy Fuels 2013;27:6823–30. Standards Australia, Higher rank coal ash and coke ash—Major and minor elements—Borate fusion/flame atomic absorption spectrometric method. In Coal and coke—Analysis and testing, AS 1038.14.1-2003: 2003. Wu H, Bryant G, Benfell K, Wall T. Energy Fuels 2000;14:282–90. Mitchell RE, Akanetuk AEJ. Symp (Int) Combust 1996;26:3137–44. Senneca O, Urciuolo M, Chirone R. Fuel 2013;104:253–61. Helble JJ, Sarofim AF. Combust Flame 1989;76:183–96. Bridgwater AV, Meier D, Radlein D. Org Geochem 1999;30:1479–93. Field MA. Combust Flame 1969;13:237–52. Quann RJ, Sarofim AF. Symp (Int) Combust 1982;19:1429–40. Frandsen F, Dam-Johansen K, Rasmussen P. Prog Energy Combust Sci 1994;20:115–38. He YD, Zhai YB, Li CT, Yang F, Chen L, Fan XP, et al. Environ Technol 2010;31:567–74. Linak WP, Wendt JOL. Prog Energy Combust Sci 1993;19:145–85. Chenevert BC, Kramlich JC, Nichols KM. Symp (Int) Combust 1998;27:1719–25.
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Full Length Article
Effect of particle size on particulate matter emissions during biosolid char combustion under air and oxyfuel conditions Sui Boon Liawa, Xujun Chena, Yun Yua, Mário Costab, Hongwei Wua,
T
⁎
a
Discipline of Chemical Engineering, Western Australian School of Mines: Minerals, Energy and Chemical Engineering, Curtin University, GPO Box U1987, Perth, WA 6845, Australia b IDMEC, Mechanical Engineering Department, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal
A R T I C LE I N FO
A B S T R A C T
Keywords: Biosolid Particulate matter Particle size Char Oxyfuel combustion
In an industrial scale furnace, solid fuels can be fired in a wide range of particle sizes. This study aims to investigate the effect of the particle size on particulate matter (PM) emissions during combustion under air and oxyfuel (30% O2/70% CO2) conditions. Biosolid chars prepared from the pyrolysis of three different biosolid particle sizes at 1300 °C were burned in a drop tube furnace. The experimental results indicate that a shift from small to large char particles leads to > 55% reduction in PM1–10 emission when combustion occurs in air, likely due to less intense char fragmentation experienced by the large char particles. Such a reduction originates from a decrease in Mg, Ca, P, Si, Al and some trace elements (V, Co, Cu, Zn and Mn) release as PM1–10. In contrast, PM1 emission is not affected by the char particle size. However, under oxyfuel conditions, the PM1–10 emission from the small char particles is ∼65% of that released during combustion in air. In addition, the PM1–10 emission from large char particles only reduced by ∼27% when compared to that from small char particles. This observation is likely to result from the coalescence of ash particles to form PM with particle sizes > 10 µm due to the increase surface mobility caused by repeated formation and decomposition of CaCO3 in the ash.
1. Introduction Biosolid is a byproduct from wastewater treatment processes and has generated significant environmental issues for disposal, mainly due to its high contents of toxic organic compounds and heavy metals [1]. Among the available technologies for biosolids utilisation, combustion is one of the most effective strategies not only to achieve a large volume reduction but also to recover the energy in the biosolid [2]. However, combustion of various fuels are known to release particulate matters (PMs) that have potential hazard to the human health and to the environment (soil, air, and water) [3–9]. This is of particular concerns because biosolid contains high content of ash, resulting in PM significant emissions of during combustion [8,9]. A large portion of trace metals (i.e., Cu, Cr, Ni, Pb, V, Co, Cd and As) present in biosolid are also released in the PM during combustion [10]. Therefore, it is of critical importance to understand the emissions of PM and trace elements during the combustion of biosolid. Usually, biosolid is fired under air conditions in boilers to generate electricity [2]. Recently, there has been an increasing need to develop oxyfuel combustion technology in order to facilitate carbon capture and storage and, thereby, to achieve negative carbon emissions from
⁎
Corresponding author. E-mail address: [email protected] (H. Wu).
https://doi.org/10.1016/j.fuel.2018.05.163 Received 1 March 2018; Received in revised form 24 May 2018; Accepted 29 May 2018 0016-2361/ © 2018 Elsevier Ltd. All rights reserved.
biosolid-fired power plants [11]. Although PM emissions from biosolid combustion under air conditions are relatively well documented in the literature [2,8,12,13], studies on oxyfuel combustion of biosolid are scarce. Surprisingly, the release of trace elements in the PM during the combustion of biosolid under oxyfuel conditions is also seldom reported. In order to address this problem, this work presents a systematic study on the emissions of PM (including major and trace elements) during the combustion of biosolid in both air and oxyfuel environments. A distinct feature of this work is the emphasis placed on the effect of the particle size on PM emissions that, to the best of our knowledge, has not been considered before. Furthermore, to eliminate the possible effect on the PM emissions of the combustion of the volatiles [14], biosolid chars were first produced from pyrolysis at 1300 °C, and subsequently used for the combustion experiments under air and oxyfuel conditions. 2. Materials and methods 2.1. Sample preparation Biosolid, provided by a waste water treatment facility in Western Australia, was dried overnight in an oven at 105 °C before it was ground
Fuel 232 (2018) 251–256
S.B. Liaw et al.
Table 1 Properties of the biosolid size fractions of 38–60 µm, 60–90 µm and 120–150 µm and its respective chars prepared from pyrolysis at 1300 °C, which are referred to as small char particle, medium char particle and large char particle, respectively. Char samples
38–60 µm biosolid 60–90 µm biosolid 120–150 µm biosolid Small char particle Medium char particle Large char particle a b c d e
Moisture (wt%, ada)
Proximate analysis (wt%, dbb)
Table 2 Inorganic elements content of the biosolid chars. Small char particle
Ultimate analysis (wt% dafe)
ash
VMc
FCd
C
H
N
O
5.5
20.4
68.9
10.6
53.54
10.90
6.79
28.77
5.2
20.4
70.5
9.1
53.41
11.02
6.84
28.73
4.5
20.2
70.0
9.8
55.08
11.07
7.13
26.72
1.2
84.0
5.2
10.9
98.61
0.15
0.37
0.86
0.9
84.3
4.7
11.1
99.21
0.11
0.34
0.34
0.6
83.9
5.4
10.7
99.33
0.10
0.40
0.17
Air-dried. Dry-basis. Volatile matter. Fixed carbon. Dry and ash free basis.
Large char particle
Major elements (wt%, dry basis) Na 0.642 ± 0.053 K 0.642 ± 0.016 Mg 5.692 ± 0.045 Ca 13.486 ± 0.057 Al 3.323 ± 0.036 Fe 1.655 ± 0.034 Si 8.569 ± 0.030 Ti 0.797 ± 0.004 P 11.649 ± 0.160 S 0.204 ± 0.002
0.727 ± 0.095 0.867 ± 0.010 4.597 ± 0.101 13.470 ± 0.312 3.460 ± 0.014 1.710 ± 0.024 7.814 ± 0.035 0.973 ± 0.024 11.574 ± 0.252 0.220 ± 0.002
0.591 ± 0.100 0.679 ± 0.018 3.825 ± 0.028 13.007 ± 0.053 3.753 ± 0.060 1.747 ± 0.007 9.056 ± 0.030 1.022 ± 0.008 9.502 ± 0.030 0.298 ± 0.002
Trace elements (mg/kg, dry basis) As 4.075 ± 0.091 V 27.501 ± 0.120 Cr 273.921 ± 36.0 Co 21.748 ± 0.258 Cd 2.045 ± 0.030 Pb 7.607 ± 0.109 Zn 659.925 ± 16.4 Mn 523.917 ± 13.8 Cu 1755.5 ± 37.4
4.041 ± 0.028 34.141 ± 3.943 207.1 ± 16.7 25.181 ± 2.822 2.046 ± 0.087 6.789 ± 3.395 697.165 ± 4.8 522.624 ± 3.1 1778.0 ± 10.3
3.312 ± 0.477 31.402 ± 4.346 344.9 ± 47.5 19.389 ± 1.295 1.617 ± 0.188 8.404 ± 1.049 545.563 ± 12.5 545.388 ± 17.2 2131.5 ± 50.8
2.2. Char combustion experiments The biosolid char combustion experiments under air and oxyfuel (30% O2/70% CO2) conditions were carried out in a DTF equipped with a sampling system that comprises a cyclone and a 13-stage Dekati lowpressure impactor (DLPI) coupled with a backup filter [14,16]. During the char combustion experiments, ∼0.03 g/min of each char, entrained in a stream of 1 L/min of air or oxyfuel gas, was fed into the DTF at 1300 °C. A second flow of 4.6 L/min of air or oxyfuel gas was introduced in the DTF to assist the char combustion process. The value of λ (expressed as the ratio of the actual air/fuel ratio to the stoichiometric air/fuel ratio) was ∼30 and the residence time of the char particles in the isothermal zone of the DTF was ∼1.2 s. The sampling temperature was set at 115 °C in this study. The ash and PM particles with aerodynamic diameter of > 10 μm were first removed from the flue gas by the cyclone, and then the flue gas was directed to the DLPI for size-segregated collection. In this study, PMs with aerodynamic diameters < 0.1 µm, 0.1–1 µm, < 1 µm, 1–10 µm and < 10 µm are termed as PM0.1, PM0.1–1, PM1, PM1–10 and PM10, respectively. All experiments were carried out at least in triplicates. It should be noted that complete combustion was achieved for all combustion experiments as thermogravimetric analysis showed the absence of unburned carbon in the PM samples, and total organic carbon analysis showed that leachates from washing PM samples in pure water for 24 h contained negligible amounts of organic carbon.
and sieved to yield three size fractions of 37–60 µm, 60–90 µm, 120–150 µm for the pyrolysis experiments. Table 1 presents the properties of the prepared biosolid. The pyrolysis experiments were conducted in a drop tube furnace (DTF) described in detail elsewhere [15]. Briefly, the DTF was preheated to 1300 °C, and then samples of the three biosolid size fractions were in turn entrained by 1 L/min of ultra-high purity nitrogen into the DTF through a water-cooled feeding probe using a sample feeder system. An additional stream of 1 L/min of nitrogen was used to prevent backflow of the volatiles. The char produced was sampled using a water-cooled sample probe. A stream of 1 L/min of helium was injected at the probe tip to quench the hot flue gas, while extra nitrogen was introduced to give a total gas flow of 10 L/min before the char particles were collected in a Dekati cyclone. Henceforth, the biosolid chars prepared from pyrolysis of the biosolid size fractions of 37–60 µm, 60–90 µm and 120–150 µm are termed as small, medium and large char particles, respectively. Fig. 1 shows the cumulative particle size distribution of the three biosolid chars, and Tables 1 and 2 show their main properties.
100
2.3. Sample analysis
80
Cumulative (%)
Medium char particle
The proximate analysis of the biosolid chars was carried out in a thermogravimetric analyser (TGA, Mettler Toledo Star 1) according to the time-temperature program described in ASTM E870-82. The C, H and N contents were determined by an elemental analyser (PerkinElmer 2400 CHNS/O Series II). The sulphur content in the biosolid chars were determined by an improved Eschka method [17]. The quantification of the contents of major (Na, K, Mg, Ca and P) and trace (As, Cr, Cu, Mn, Pb, Ti, and Mn) elements were made by using inductively coupled plasma-optical emission spectroscopy (ICP-OES, PerkinElmer Optima 8300) and inductively coupled plasma-mass spectroscopy (ICP-MS, PerkinElmer NexION 350D), respectively, with the biosolid chars being first ashed and then digested in a concentrated HNO3/HF mixture (detailed procedure described elsewhere [18]). Likewise, the major and trace elements in PM were also quantified using ICP-OES and ICP-MS,
60
40
Small particle Medium particle Large particle
20
0 0
50
100
150
Char particle diameter
200
250
m)
Fig. 1. Cumulative particle size distribution of the three biosolid chars. 252
Fuel 232 (2018) 251–256
15
Combustion in air
Combustion in oxyfuel
(a)
(b)
12 9
0.5
Small particle Medium particle Large particle
0.4
Macroporosity
dm/dlogDp (mg/g char, db)
S.B. Liaw et al.
6 3 0 0.01
0.1
1
10 0.01
0.1
1
0.3 0.2 0.1
10
Aerodynamic diameter ( m)
Yield (mg/g char, db)
8
0.0
(c)
(d) Small particle Medium particle Large particle
6
B
C
D
E
A
B
C
D
Large Particle
indicated in Fig. 3, the macroporosities of the chars decrease with char particle size, from ∼0.41 for the small char particles to ∼0.28 for the large char particles. This is reasonable as during pyrolysis of large biosolid particles, the volatiles generated within the large particles would experience more reactions with the hot char as the volatiles diffuse through the pores within the pyrolysing particles, leading to more deposition of secondary char or soot within the pore structure [24]. This subsequently leads to partial blockage of char pore structure and reduces the macroporosity of the char particles. Such an effect would be much less intensive during pyrolysis of small biosolid particles. Therefore, the combustion of small char particles that have a higher macroporosity results in a higher PM1–10 yield compared to that from the combustion of large char particles that have lower macroporosity. In addition, the small char particles have a significantly shorter gas diffusion path length and larger external surface area as compared to the larger char particles. This implies that an increase in particle size from small to large can promote a change in the combustion regime from chemical control to diffusion control [25], which can greatly affect the char particle temperature and, thus, its fragmentation intensity. Furthermore, smaller particles may experience a greater degree of thermal shock during combustion [22]. Consequently, the combustion of large-size char sample results in a > 55% reduction in PM1–10 yield.
2
A
Medium Particle
Fig. 3. Macroporosities of the three biosolid chars with different sizes (small, medium and large) used in this study.
4
0
Small Particle
E
A: PM0.1 B: PM0.1-1 C: PM1 D: PM1-10 E: PM10 Fig. 2. PSDs and yields of PM from the combustion of the three biosolid chars (small, medium and large particles) under air (panel a and c) and oxyfuel (panel b and d) conditions.
respectively, after digestion in a concentrated HNO3/HF mixture. The contents of Si, Al and Fe in the biosolid chars and PM samples were determined using the ash borate fusion method described in AS 1038.14.1-2003 [19] with the fusion bead dissolved in dilute nitric acid for analysis in ICP-OES. For determination of char macroporosity, the biosolid char was set in a resin pellet that was then polished and coated with platinum for cross-sectional imaging using a scanning electron microscope (SEM), followed by image analysis according to a previous method [20]. An image analysis software, Digimizer, was employed for analysing the acquired SEM images and determining the macroporosity of each char sample via analysing at least 100 char particles for each sample. The particle size of the biosolid chars was determined through image analysis of, at least, 200 particles using the same image analysis software.
3.2. Major elements in PM during char biosolid combustion 3. Result and discussion Figs. 4a–i and 5a–i show the distributions and the yields, respectively, of the major elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under air conditions. The PM1 released from the biosolid chars combustion in air consists mainly of volatile alkali species (Na and K) and P, likely in form of (Na,K)PO3 [14]. However, some refractory elements such as Si, Al and Fe are also observed in PM1, likely due to vaporisation of metal oxides under local reducing condition formed during char combustion [26]. The elemental yield data indicates that the yields of these elements in PM1 from the biosolid chars combustion are insensitive to the char particle size. In addition, the PM1–10 released mainly consists of P, Ca, Mg, Si, Al and Fe. A very small amount of Na and K is also observed in the PM1–10, most likely due to their release as aluminosilicate compounds during fragmentation. Figs. 3a–i and 4a–i show a clear reduction in the yields of Mg, Ca, P and Si in PM1–10, which are the major elements in the char where its oxides contributed > 70% to the total ash in the char. This further suggests that an increase in the char particle size leads to less intense char fragmentation and thus lower yields of these elements in the PM1–10 released. However, the decrease in the yields of Al and Fe in the PM1–10 is minimal probably because these elements are less abundant in the char as compared to P, Ca, Mg and Si.
3.1. Particle size distribution of PM during char biosolid combustion Fig. 2a–c show the particle size distributions (PSDs) and the yields of the PM from the combustion of the three biosolid chars (small, medium and large particles) under air conditions. The PSDs from the combustion of these chars exhibit a bimodal distribution with a course mode of ∼6.863 µm and a fine mode of ≤0.01 µm. The PM10 emitted is dominantly (> 50%) PM1–10 as majority of alkali metal, volatile S and P were released as volatiles during pyrolysis to yield the char samples. While the differences in the PM1 emitted from the combustion of the three biosolid chars are negligible, it is clear that an increase in char particle size leads to a substantial reduction in PM1–10 emission. There is a ∼12% reduction in PM1–10 emission between the combustion of small and medium char particles, whereas a > 55% reduction is observed for the combustion of the large char particles as compared to the combustion of the small char particles. Such a reduction can be attributed to char fragmentation intensity during char combustion, which is governed by char particle size and macroporosity [21,22]. Previous studies demonstrated that the combustion of char with a higher macroporosity leads to a higher yield of fine particulates [21,23]. As 253
Fuel 232 (2018) 251–256
S.B. Liaw et al.
Combustion in air
1.8
(a) Na
Combustion in oxyfuel
Combustion in air 1.2 (a) Na
(j) Na Small particle Medium particle Large particle
1.2
1.8
(b) K
0.4 0.0 1.2 (b) K
(k) K
0.4
0.6 1.8
(c) Mg
0.0 1.2 (c) Mg
(l) Mg
0.4
0.6 1.8
(d) Ca
0.0 1.2 (d) Ca
(m) Ca
(e) P
Yield (mg/g char, db)
dm/dlogDp (mg/g char, db)
0.4
0.6 1.8
(n) P
1.2 0.6 0.0 1.8
(f) Al
(o) Al
(g) Fe
(p) Fe
0.4 0.0 1.2 (f) Al
0.0 1.2 (h) Si (h) Si
(q) Si
0.4 0.0 1.2 (i) Ti (i) Ti
(r) Ti
0.4
0.6
0.0
0.0 0.1
1 10 0.01 0.1 Aerodynamic diameter ( m)
(r) Ti
0.8
1.2
0.01
(q) Si
0.8
0.6 1.8
(p) Fe
0.4
1.2
0.0
(o) Al
0.8
0.6 1.8
0.8
0.0 1.2 (g) Fe
1.2
0.0
(n) P
0.4
0.6 1.8
0.0 1.2 (e) P
0.8
1.2
0.0
(m) Ca
0.8
1.2
0.0
(l) Mg
0.8
1.2
0.0
(k) K
0.8
1.2
0.0
Small particle Medium particle Large particle
0.8
0.6 0.0
Combustion in oxyfuel (j) Na
1
A
B
C
D
E
A
B
C
D
E
A: PM0.1 B: PM0.1-1 C: PM1 D: PM1-10 E: PM10
10
Fig. 5. Yields of the major elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under air (panel a–i) and oxyfuel (panel j–r) conditions.
Fig. 4. Distributions of the major elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under air (panel a–i) and oxyfuel (panel j–r) conditions.
biosolid chars (small, medium and large particles) under air conditions. Volatile trace elements such as As, Cr, Cd and Pb contributed mainly to the formation of PM1, forming little PM1–10. It is interesting to note that although Zn is a volatile trace element [14], a portion of Zn remained in the char and thus contributed to the PM1–10 formation during the
3.3. Trace elements in PM during char biosolid combustion Figs. 6a–i and 7a–i show the distributions and the yields, respectively, of the trace elements in PM from the combustion of the three 254
Fuel 232 (2018) 251–256
S.B. Liaw et al.
0.6
Combustion in air
Combustion in oxyfuel (j) As
(a) As
Small particle Medium particle Large particle Combustion in air Combustion in oxyfuel 0.4 (a) As (i) As
Small particle Medium particle Large particle
0.4
0.3
0.2 0.0
0.2 (b) V
0.1
(k) V
0.4
0.0 0.3 (b) V
0.2
0.2 0.1
0.0 (c) Cr 15
(l) Cr
0.0 (c) Cr 15
10
5 (d) Co
(m) Co
0 0.20 (d) Co
0.2
0.15
0.1
0.10
(l) Co
0.05
0.0 30 (e) Cu
(n) Cu
Yield ( g/g char, db)
dm/dlogDp ( g/g char, db)
0.3
(k) Cr
10
5 0
(j) V
20 10 0 0.2 (f) Cd
(o) Cd
0.00 20 (e) Cu
(m) Cu
15 10 5 0 0.12 (f) Cd
(n) Cd
0.09
0.1
0.06 0.03
0.0 (g) Pb 2
0.00 (g) Pb 1.5
(p) Pb
(o) Pb
1.0
1
0.5
0 32 (h) Zn
0.0 (h) Zn 15
(q) Zn
24 16
10
8
5
0 9 (i) Mn
0
(r) Mn
3
6
(p) Zn
(h) Mn
(q) Mn
2
3
1
0
0
0.01
0.1
1
10 0.01
0.1
1
10
Aerodynamic diameter ( m)
A
B
C
D
E
A
B
C
D
A: PM0.1 B: PM0.1-1 C: PM1 D: PM1-10 E: PM10
E
Fig. 7. Yields of the trace elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under air (panel a–i) and oxyfuel (panel j–r) conditions.
Fig. 6. Distributions of the trace elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under air (panel a–i) and oxyfuel (panel j–r) conditions.
the release of Zn as PM1–10. In addition, semi-volatile trace elements such as Cu and V contributed to both PM1 and PM1–10 formation. Co and Mn are non-volatiles trace elements and only contributed to the PM1–10 formation. Similar to the major elements, the yields of trace elements such as V, Co, Cu, Zn and Mn contributed to PM1–10. PM1–10 emission also reduces with increasing biosolid char particle size due to the less intense char fragmentation during combustion in air. However,
biosolid chars combustion. Our previous study showed almost complete release of Zn during pyrolysis of biosolid at 1000 °C [14] as Zn compounds (i.e. ZnO) begin to decomposed at temperature > 600 °C to form gaseous elemental Zn [27]. However, under the higher pyrolysis temperature of 1300 °C used in this work, Zn can react with silicate, Fe and Mn in the char [28]. Subsequent biosolid char combustion leads to 255
Fuel 232 (2018) 251–256
S.B. Liaw et al.
4. Conclusions
the char particle size has little influence on the formation of PM1 from trace elements except for Pb. Higher Pb is released from the combustion of large char particles compared to that from small char particle. The exact mechanisms responsible for such an observation are unknown. However, one possibility is that Pb maybe react with aluminosilicate and retained in coarse ash collected in cyclone so that the combustion of smaller char particles, which provide more surface for such reactions, can lead to less release of Pb as PM1 [29].
This work investigated the effect of the biosolid char particle size on the release of PM10 during combustion under air and oxyfuel conditions. While a change in the char particle size does not affect the yield of PM1, increasing the char particle size from small to large leads to > 55% reduction in PM1–10 formation, likely due to less intense char fragmentation during the combustion of the large char particles. This reduces the formation of Mg, Ca, P, Al, Si and, to a lesser extent, Fe in PM1–10. An increase in the char particle size also reduces the release of trace elements such as V, Co, Cu, Zn and Mn as PM1–10. Switching the combustion atmosphere from air to oxyfuel substantially retards the formation of PM1–10, especially for the biosolid char with the small particle size, possibly due to the high coalescence tendency caused by repeated formation and decomposition of CaCO3 in CO2 rich atmospheres to form coarse ash particles > 10 µm. Consequently, the effect of the char particle size on PM1–10 release is less significant. Although an increase in the char particle size reduces the release of Mg, Ca, P, Al, Si and Fe in PM1–10, such reductions are not observed for trace elements probably due to their low relative content in the biosolid char.
3.4. Impact on the results of the oxyfuels conditions Under oxyfuel conditions, Fig. 2b and 2d show that the PM from the combustion of the three biosolid chars also exhibit a bimodal distribution with a coarse mode of ∼6.863 µm and a fine mode of ≤0.01 µm. However, the yield of PM1–10 is less than that under air conditions, particularly for small and medium char particles. The yield of PM1–10 from the combustion of large char particles under oxyfuel conditions is only slightly less than that from combustion in air, which is in line with data reported previously [14]. However, the yield of PM1 is not affected by the char particle size. This observation also indicates that combustion under oxyfuel conditions can greatly reduce the formation of PM1–10 from solid fuels with small particles. Although the increase in the char particle size also leads to a decrease in the PM1–10 yield, the yield of PM1–10 from the combustion of large char particles under oxyfuel conditions only drop by ∼27% as compared to that from the combustion of small char particles under oxyfuel conditions. The reduction is significantly less than the > 55% reduction observed in the combustion under air conditions. Such phenomena may be attributed to the higher ash coalescence tendency during combustion of the biosolid chars under oxyfuel conditions as compared to combustion in air. A previous study reported that in CO2 rich atmospheres, CaCO3 can repeatedly formed and decomposed to CaO, catalysing the surface mobility [30]. The char used in present study has a high (> 13%) Ca content. The CO2 rich atmosphere under oxyfuel conditions can lead to significant sintering and coalescence of ash, leading to the formation of coarse particles and reduction in PM1–10 formation. Since the small char particles experience more intense char fragmentation, its combustion under oxyfuel conditions lead to a more significant reduction in PM1–10 as compared to the oxy-combustion of the large char particles, where such effect is minimal. Figs. 4j–r and 5j–r show the distributions and the yields, respectively, of the major elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under oxyfuel conditions. Similar to biosolid char combustion in air, PM1 mainly consists of Na, K and some P, while Mg, Ca, Al, Fe, Si and P are dominant in PM1–10. The increased tendency in ash coalescence led to lower yields of Mg, Ca, P and Si in PM1–10 under oxyfuel conditions as compared to combustion in air. However, the reduction in the Al and Fe yields is less significant and within the experimental uncertainty. Figs. 6j–r and 7j–r show the distributions and the yields, respectively, of the trace elements in PM from the combustion of the three biosolid chars (small, medium and large particles) under oxyfuel conditions. Similar to the biosolid chars combustion in air, trace elements such as As, Cu, Cr, Cd, V and Pb mainly contributed to the PM1 formation, while V, Co, Cu, Zn and Mn contributed to the formation of PM1–10. Although the increase in char particle size reduces the amounts of V, Co, Cu and Zn in PM1–10, such reduction is not observed, except for Mn, most likely as a result of the low content of these elements in the char and less significant effect of the char particle size on PM1–10 released.
Acknowledgements The authors would like to acknowledge the partial support received from the Australian Research Council through its Discovery Projects Scheme. The authors would like to thank Jinxiu Cao for her assistance in part of the experimental work. X. Chen is also grateful to Curtin International Postgraduate Research Scholarship for supporting his PhD study. M. Costa also acknowledges Fundação para a Ciência e a Tecnologia for the sabbatical leave Grant SFRH/BSAB/128154/2016. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]
[20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]
256
Fytili D, Zabaniotou A. Renewable Sustainable Energy Rev 2008;12:116–40. Werther J, Ogada T. Prog Energy Combust Sci 1999;25:55–116. Sheng C, Lu Y, Gao X, Yao H. Energy Fuels 2007;21:435–40. Denison RA, Silbergeld EK. Risk Anal 1988;8:343–55. Xu M, Yu D, Yao H, Liu X, Qiao Y. Proc Combust Inst 2011;33:1681–97. Li G, Li S, Dong M, Yao Q, Guo CY, Axelbaum RL. Fuel 2013;106:544–51. Patiño D, Crespo B, Porteiro J, Villaravid E, Granada E. Powder Technol 2016;288:164–75. Ninomiya Y, Zhang L, Sakano T, Kanaoka C, Masui M. Fuel 2004;83:751–64. Seames WS, Fernandez A, Wendt JOL. Environ Sci Technol 2002;36:2772–6. Liaw SB, Rahim MU, Wu H. Energy Fuels 2016;30:5766–71. Toftegaard MB, Brix J, Jensen PA, Glarborg P, Jensen AD. Prog Energy Combust Sci 2010;36:581–625. Wang Q, Zhang L, Sato A, Ninomiya Y, Yamashita T, Dong Z. Proc Combust Inst 2009;32:2701–8. Zhang L, Ninomiya Y. Proc Combust Inst 2007;31:2847–54. Liaw SB, Wu H. Energy Fuels 2017;31:2317–23. Gao X, Wu H. Energy Fuels 2011;25:2702–10. Gao X, Wu H. Energy Fuels 2011;25:4172–81. Rahim MU, Gao X, Wu H. Energy Fuels 2013;27:6992–9. Zhang M, Gao X, Wu H. Energy Fuels 2013;27:6823–30. Standards Australia, Higher rank coal ash and coke ash—Major and minor elements—Borate fusion/flame atomic absorption spectrometric method. In Coal and coke—Analysis and testing, AS 1038.14.1-2003: 2003. Wu H, Bryant G, Benfell K, Wall T. Energy Fuels 2000;14:282–90. Mitchell RE, Akanetuk AEJ. Symp (Int) Combust 1996;26:3137–44. Senneca O, Urciuolo M, Chirone R. Fuel 2013;104:253–61. Helble JJ, Sarofim AF. Combust Flame 1989;76:183–96. Bridgwater AV, Meier D, Radlein D. Org Geochem 1999;30:1479–93. Field MA. Combust Flame 1969;13:237–52. Quann RJ, Sarofim AF. Symp (Int) Combust 1982;19:1429–40. Frandsen F, Dam-Johansen K, Rasmussen P. Prog Energy Combust Sci 1994;20:115–38. He YD, Zhai YB, Li CT, Yang F, Chen L, Fan XP, et al. Environ Technol 2010;31:567–74. Linak WP, Wendt JOL. Prog Energy Combust Sci 1993;19:145–85. Chenevert BC, Kramlich JC, Nichols KM. Symp (Int) Combust 1998;27:1719–25.