Microwave drying of a low-rank sub-bituminous coal

Minerals Engineering xxx (2013) xxx–xxx

Contents lists available at ScienceDirect

Minerals Engineering journal homepage: www.elsevier.com/locate/mineng

Microwave drying of a low-rank sub-bituminous coal C.A. Pickles ⇑, F. Gao, S. Kelebek Robert M. Buchan Department of Mining, Queen’s University, Kingston, Ontario K7L 3N6, Canada

a r t i c l e

i n f o

Article history: Available online xxxx Keywords: Microwaves Drying Low-rank sub-bituminous coal Permittivities Models Effective diffusion coefficients

a b s t r a c t The moisture contents of coals are often too high and need to be reduced before further processing. In this study, the application of microwave radiation as an alternative energy source for the drying of a subbituminous coal was investigated. Firstly, the permittivities of the coal were evaluated as a function of temperature and frequency. Secondly, the drying kinetics were studied in a 2.45 GHz microwave system and the effects of incident microwave power, sample mass and initial moisture contents were determined. The results demonstrated that microwave drying had several advantages over conventional drying such as increased drying rates and lower final moisture contents. In some tests, magnetite was added as a susceptor to increase the drying rates. Thirdly, the drying data were fitted to ten exponential decay models, and although reasonable agreement was observed with all the models, the best fit was obtained with the Midilli–Kucuk model. Finally, the effective diffusion coefficients of moisture and also the activation energy of the diffusion process were estimated and used to further elucidate the mechanism of microwave drying. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction Mined high-rank coals (HRC’s) can have moisture contents in the range of about one to six percent. Some HRC’s are cleaned, usually by wet techniques and these coals can have moisture levels between 12% and 25%. For a given set of conditions, the moisture content of the coal increases with decreasing particle size, because the larger surface area of the fine coal enhances its capacity to retain moisture. Furthermore, low-ranked coals (LRC’s) can have moisture contents above 25%. Currently, LRC’s are not widely exploited but their use is expected to increase as a result of the growing demand for energy. The presence of high levels of moisture results in higher transportation costs, greater energy requirements, increased off-gas volume, lower efficiencies, increased maintenance costs and increased friability of the coal, which interferes with separation, blending and pneumatic transportation. Additionally, dusting is enhanced and the potential for spontaneous combustion increases (Sevi, 1995). Consequently, the efficient removal of water is of significant importance in coal processing. Conventional mechanical dewatering techniques such as centrifuges and vacuum filters can reduce the moisture content of the coal to between 10% and 15%, but this still may not meet the quality specifications for subsequent processing. Thermal drying is the most widely employed method to lower the moisture content. In these types of processes, heat energy is transferred to the surface by convection and to the interior by conduction. Industrially, drying is generally achieved by contacting the ⇑ Corresponding author. Tel.: +1 613 533 2759; fax: +1 613 533 6597. E-mail address: [email protected] (C.A. Pickles).

wet coals with hot gases from a combustion process. The most common types of dryers are rotary dryers, multi-louver dryers, fluidized-bed dryers, screw conveyor dryers and flash dryers. Generally, traditional coal drying processes have environmental issues and in some cases potentially explosive stack mixtures. Microwave radiation has a number of potential advantages for the processing of coal such as desulphurization (Viswanathan, 1990; Uslu and Atalay, 2003; Rowson and Rice, 1990; Hayashi et al., 1997) and in particular for drying. Conventional drying systems rely on heat transfer to the surface of the material followed by conduction of heat through the particles. This is a slow process and depends on the size of the particles, the properties of the material being heated and the process conditions. In order to heat the interior of the material and thus remove the trapped water, it may be necessary to overheat the surface. On the other hand, microwave radiation can provide a volumetric heating technique, in which the electromagnetic radiation transfers the energy into the interior of the particle. This facilitates the rapid removal of the water and also potentially lowers final moisture levels. Additionally, since the microwave absorption characteristics of water are superior to those of the coal, selective heating may be possible. Furthermore, the drying process could be self-limiting, since as the water is removed, then the amount of microwave energy absorbed decreases. Consequently, it may be possible to minimize overheating of the surface of the coal, which often occurs in conventional drying. One of the earliest studies on the microwave drying of coals was performed by Lindroth (1986) at the U.S. Bureau of Mines. Bituminous, sub-bituminous and lignite coals were dried on a conveyer belt microwave oven at 2.45 GHz and 12 kW. Drying efficiencies,

0892-6875/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.mineng.2013.10.011

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011

2

C.A. Pickles et al. / Minerals Engineering xxx (2013) xxx–xxx

near the theoretical limit of 1.54 kg/kW h, could be achieved. Standish et al. (1988) reported that the rate of moisture removal from a brown coal was one to two orders of magnitude faster than conventional convective drying. Seehra et al. (2007) investigated the dewatering of a fine coal slurry sample by both conventional thermal and microwave drying at 2.45 GHz and 800 W. The comparative thermogravimetric analysis (TGA) results clearly showed a significant advantage of microwave drying, in terms of reducing the drying time by a factor of nearly ten. Zimmermann and Niemann-Delius (2007) showed that the microwave drying rates were extremely fast and some cracking of the particles was observed. Chatterjee and Misra (1991) developed a numerical model, which when combined with electromagnetic and thermal models, allowed calculation of the electromagnetic energy absorption profile. From this, the temperature distributions in the coal particles during drying could be predicted as a function of time, power and frequency. Harrison and Rowson (1997) demonstrated that the Bond Work Index could be reduced by 30% after short exposures of the coal to microwaves. The reduction in the relative Bond Work Index was attributed to the cracking initiated around pyrite grains and pressures generated by the superheating of water in the porous coal structure. Marland et al. (2000) reported an approximate 50% reduction of the Bond Work Index after microwave treatment. Again, they suggested that gaseous evolution (water and volatile matter) as well as gangue mineral expansion were the possible causes for the improved grindability. Lester and Kingman (2004) demonstrated that microwave radiation produced physical changes, such as cracks and fissures in coal, even for short processing times. Toraman and Depci (2007) treated a lignite coal with microwaves and found that the rapid expansion of moisture resulted in cracking and improvements in the grindability. In the absence of a magnetic field, microwaves generate an electric field in the material of interest and the interaction of this electric field with a given material is fundamentally determined by the complex permittivity (e), which is defined by the following equation:

e ¼ e0  je00

ð1Þ

Here e0 is the real part of the permittivity and is referred to as the dielectric constant, e00 is the imaginary part of the permittivity and is called the loss factor or dielectric loss, pffiffiffiffiffiffiffiand j is the imaginary component in the +j-axis direction ðj ¼ 1Þ. The dielectric constant (e0 ) determines the penetration depth of the applied electric field into the irradiated material. The dielectric loss (e00 ) controls the amount of microwave energy converted to heat in the material. The permittivities are dependent on the mobility of the dipoles within the structure, and therefore are functions of temperature, frequency and composition. Although knowledge of the permittivities is useful for understanding the interaction of microwaves with any given material, the actual microwave processing of a material is much more complex. Other factors which influence the interaction are; the thermal conductivity of the sample, the heat capacity of the sample, the geometry of both the sample and the microwave cavity, the bulk density, the power level, the particle size, the sample mass or sample size, the presence of susceptors or coupling agents and the occurrence of chemical reactions or phase changes. Although considerable research has been performed on the potential advantages of microwave drying of coal, there is a paucity of information regarding the underlying fundamental processes. In the present work, firstly permittivity studies were performed on the Highvale coal, in order to determine the effects of frequency and temperature. Secondly, the microwave drying rates of the coal were studied as a function of incident microwave power, sample mass, initial moisture content and magnetite additions and also compared to conventional drying. Thirdly, the microwave drying data were fitted to a number of thin-layer exponential decay

drying models. Finally, the drying rate data were utilized to determine the diffusion coefficients and the activation energy for the process. This information was utilized to further understand the microwave drying process. 2. Experimental 2.1. Raw materials The sub-bituminous coal sample was obtained from the Highvale Mine in Alberta, Canada. The coal was processed in a manner so as to produce a homogeneous sample for the drying tests. A total of 27.5 kg of the coal with an average particle size of over 25 mm was reduced to a size range of 2.36–4.75 mm (8–4 mesh) using a laboratory jaw crusher followed by a gyratory crusher. Then the coal sample was passed through a roll crusher and screened with a 10 mesh (1.7 mm) sieve. The oversize coal was repeatedly returned to the roll crusher until the entire sample passed through the 10 mesh screen. The sample was homogenized in a mixing drum for 20 min and placed into two sealed containers to limit moisture changes as much as possible. The homogeneous sample was screened to provide a control size fraction of 12 + 16 mesh ( x = 435 lm), plus an additional size fraction of 200 + 270 mesh ( x = 63.5 lm). The samples were sealed in plastic bags to minimize moisture changes. The proximate and ultimate analyses of the coal are shown in Table 1. The coal has medium level volatile matter and fixed carbon, with relatively low ash and sulfur contents. The moisture content of the 12 + 16 mesh ( x = 1435 lm) as-received coal was 12.5%. In order to investigate the effect of initial moisture content, an additional two coal samples were prepared with the following moisture contents: partially dried; 10.5% ( x = 1435 lm) and hydrated; 21.3% ( x = 1435 lm). 2.2. Microwave and conventional drying systems A schematic diagram of the microwave drying system is shown in Fig. 1. A programmable Sylvania SM80704 microwave oven with a maximum power output of 800 W was used. The magnetron produces microwaves with a frequency of 2.45 GHz from a 120 V to 60 Hz AC power supply. The multimode cavity dimensions were: 30 cm in length, 30 cm in width and 23 cm in height. The samples were suspended from an Ohaus Adventurer™ Pro balance, which had an accuracy of 0.01 g and provided a continuous record of the mass of the sample as a function of drying time. The coal sample in a quartz crucible was suspended on a Teflon thread, which passed through a 6 mm diameter central hole in the top of the cavity. The quartz crucible had the following dimensions: 8.5 cm in height, 3.2 cm in diameter and 0.15 cm in wall thickness. The quartz crucible and Teflon thread used in the microwave drying system were essentially transparent to microwaves. Preliminary tests showed that in the microwave system, the coal sample would not dry at low sample masses, low incident powers and/or low irradiation times. On the other hand, the coal would evolve volatile matter, exhibit hot spots and combust at long irradiation times, high incident powers and/or large sample masses. Since the aim

Table 1 Ultimate and proximate analyses of as-received Highvale coal on a dry basis. Ultimate analysis

Mass%

Proximate analysis

Mass%

Carbon Hydrogen Nitrogen Sulfur Oxygen

61.00 3.64 0.63 0.31 17.26

VM Ash FCM Total water

33.21 16.85 49.81 12.78

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011

C.A. Pickles et al. / Minerals Engineering xxx (2013) xxx–xxx

Balance

Support

Power Control

Teflon Thread

Time Control

Quartz Crucible Coal

3

The extent of the temperature drop was simulated by placing a 30 g as-received coal sample in the conventional oven at preset temperatures of 130 °C, 150 °C, 170 °C, 190 °C and 210 °C. One thermocouple was employed to measure the temperature adjacent to the sample in the oven and another one was used to measure the sample temperature after removal from the oven. For the out of oven test, the temperature was again measured after a delay of ten seconds. As shown in Fig. 2, the difference between the sample temperature in the oven and out of the oven increased with the higher preset temperature of the conventional oven. However, even at the maximum oven temperature, the temperature differences were less than ten percent. So for the case of microwave heating, the measured temperature out of the microwave oven and the actual temperature inside the sample, should have similar differences. 2.4. Data analysis

Fig. 1. Microwave drying system.

of this study was to investigate only the drying kinetics, then the incident powers, sample masses and processing times were limited to those conditions shown in Table 2. The conventional drying system was similar to the microwave drying system except that the Sylvania SM80704 microwave oven was replaced by a Fisher IsotempÒ Model 106G gravity convection drying oven with an operating temperature range of 25–250 °C. The nominal power consumption of the oven was 600 W at 250 °C. The size of the drying chamber was 23 cm in length, 29 cm in width and 23 cm in height. The average variation of the temperature was ±3 °C during each test. The coal sample was placed in the cylindrical quartz crucible and suspended in the center of the drying chamber by a Nichrome wire. Conventional thermal drying is relatively simple and has less controllable parameters than microwave drying. The operating temperature of a conventional oven and the material mass are the two most important factors affecting the drying kinetics. A series of tests were performed with the oven temperature ranging from 130 to 210 °C at 20 °C intervals for the same sample masses used in the microwave tests as shown in Table 2. 2.3. Sample temperature The temperature of the sample in the conventional oven was measured by a type-K thermocouple adjacent to the sample. In microwave systems, in situ temperature measurements are difficult because of interference effects. Therefore, the following procedure was used to measure the sample temperature in the microwave drying system. The power to the microwave oven was turned off and the sample was removed as quickly as possible (usually in less than 10 s). Then the type-K thermocouple was inserted into the center of the coal sample and the maximum value was reported as the sample temperature. The time lag in measurement could lead to an underestimation of the actual temperature.

In this study, the effect of the variables on the moisture fraction and also the drying rate were investigated. Each experiment was repeated and the average values were utilized. The moisture fraction (X) is a dimensionless term, and is the ratio of the mass of residual water left in the sample to the total original mass of water in the sample as follows:

X¼

Mt  Me M  Mo  Me Mo

ð2Þ

where Mt is the moisture content at any time, t (s), Mo is the initial moisture content and Me is the equilibrium moisture content, which is the moisture content after drying for an infinite time at a specific temperature and humidity level. The values of Me are very small as compared to Mt and Mo, and therefore in order to simplify the calculation, Me is assumed to be zero. The drying rate (DR) is defined as the change in moisture fraction in a unit of time (t) as follows:

DR ¼

dX dt

ð3Þ

2.5. Permittivity studies The real and the imaginary permittivities of the as-received coal were measured using the cavity perturbation technique. This technique is based on the measurement of the quality factor Q and also the change in the resonant frequency in a high electric field cavity both with and without the sample (Hutcheon et al., 1992). For a sample with a small mass, the shifts of the Q factor and the resonant frequency are related to the complex susceptibility (v) of the sample, which is also a function of the complex permittivities (v = e  1). Therefore, the real and the imaginary permittivities can then be calculated from the measured values. About 0.5 grams of the material was compacted into a briquette and the sample was heated at 3 °C/min from room temperature up to about 650 °C. The tests were performed under stagnant argon at four different frequencies. 3. Results and discussion 3.1. Permittivity studies

Table 2 Microwave drying parameters. Time (s)

Power

Mass

160 W 400 W 560 W

5g

10 g

20 g

30 g

1800 1800 1380

1800 1200 900

1800 720 540

1500 480 360

The real and imaginary permittivities of the Highvale coal in an argon atmosphere are shown as a function of temperature for the indicated frequencies and two particle sizes of  x = 1435 lm and  x = 63.5 lm in Figs. 3a and 3b, respectively. At room temperature, the permittivity values are relatively low and this is in general agreement with the values reported by other researchers (Nelson et al., 1980). This would indicate that this sub-bituminous coal is

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011

4

C.A. Pickles et al. / Minerals Engineering xxx (2013) xxx–xxx 300 Sample Temperature in Furnace Sample Temperature out of Furnace

SAMPLE TEMPERATURE (°C)

280 260 240 220 200 180 160 140 120 100 80 80

100

120

140

160

180

200

220

240

260

280

300

FURNACE TEMPERATURE (°C) Fig. 2. Comparison of measured out of furnace sample temperatures with in furnace sample temperatures for conventional heating.

8 912 MHz, x=1435 μm 1499 MHz, x=1435 μm 1977 MHz, x=1435 μm 2466 MHz, x=1435 μm 912 MHz, x=63.5 μm 1499 MHz, x=63.5 μm 1977 MHz, x=63.5 μm 2466 MHz, x=63.5 μm

REAL PERMITTIVITY ( ε')

7 6 5 4 3 2

STAGE 2

STAGE 1

1 0

50

STAGE 3

STAGE 4

100 150 200 250 300 350 400 450 500 550 600 650

TEMPERATURE (°C) Fig. 3a. Real permittivities of Highvale coal as a function of temperature at various frequencies and particle sizes of  x = 1435 lm and  x = 63.5 lm.

IMAGINARY PERMITTIVITY (ε'')

1.4 912 MHz, x=1435 μm 1499 MHz, x=1435 μm 1977 MHz, x=1435 μm 2466 MHz, x=1435 μm 912 MHz, x=63.5 μm 1499 MHz, x=63.5 μm 1977 MHz, x=63.5 μm 2466 MHz, x=63.5 μm

1.2 1.0 0.8 0.6 0.4 0.2

temperature to 100 °C, there is a small increase in the permittivities, which reach a maximum at about 100 °C. It is known that the water in coal can exist in five different states: interior adsorbed water, capillary water, interparticle water, surface adsorbed water and adhesion water (Karr, 1978). Each of these types of water would be bound to the substrate to differing degrees. Water that is strongly bonded cannot respond to an electromagnetic field as easily as free water (Jones and Or, 2003). According to Bockris et al. (1966), the first monolayer closest to the surface is the most tightly bound and the permittivities increase to that of free water at three monolayers. The relationship between the well-known decrease in the permittivity of free water with increasing temperature and the associated increase in permittivity due to the release of water from a lower to a higher rotational state has been termed the thermodielectric effect (Or and Wraith, 1999). The permittivities of the bound water may be closer to those of the solid matrix. For the case of coal, it is likely that as the temperature increases, the relative proportion of water that is in the more loosely bonded states increases, which results in increases in the permittivities. Additionally, the permittivities decreased slightly with increasing frequency. In Region 2, above the boiling point of water, from about 100– 250 °C, the water in the sample can readily evaporate and also the sample density is decreasing, which both result in reductions in the permittivities. In Region 3, from about 250 °C to about 450 °C, the permittivities decrease only slightly as the volatile matter in the coal is evolved and there is some further reduction in density. It is known that the permittivities of the volatile hydrocarbons are small and therefore their loss does not have a significant effect on the permittivities of the remaining sample (Nelson et al., 1980). In Region 4, above about 450 °C, coal undergoes significant chemical and physical changes. Aliphatic and hydrogen-bonded phenolic OH groups are removed by distillation and decomposition processes. With increasing temperature, C–C cross-linking becomes more prevalent among the aromatic systems as the aliphatic groups continue to be removed with increasing temperature (Brown, 1955). Above about 550 °C graphitization has been observed and the electrical resistance drops rapidly with increasing temperature (Davis and Anvil, 1935). Consequently, the conductivity of the char begins to increase rapidly with temperature and as a result, the permittivities also increase very rapidly. This exponential rate of increase is responsible for the phenomenon known as thermal runaway. As a consequence of this effect, the coal particles can overheat and in the presence of air, the coal particles would combust very quickly. Again, as observed in all regions, the permittivities decreased with increasing frequency. With regards to the effect of decreasing particle size there was only a slight decrease in the permittivities, despite a large reduction in particle size from  x = 1435 lm to  x = 63.5 lm. This would indicate that the amount of surface area and hence the presence of species on the surface is not an important factor in determining the permittivities. 3.2. Microwave heating behavior studies

0.0 0

100

200

300

400

500

600

TEMPERATURE (°C)

The amount of microwave power that is absorbed per unit volume (or dissipation density) (PA) can be expressed by the following equation:

Fig. 3b. Imaginary permittivities of Highvale coal as a function of temperature at various frequencies and particle sizes of  x = 1435 lm and  x = 63.5 lm.

PA ¼ 2pf e00 jE2 j

not a particularly good microwave absorber despite the presence of considerable moisture. With regards to the effect of temperature, it can be seen that the behaviors of the permittivities can be characterized by four different regions. In Region 1, from room

where f is frequency and |E| is electric field strength within the material. For a given frequency and electric field strength, the dielectric loss determines the amount of microwave power absorbed and as a consequence the temperature of the material. The rate of heating (dT/dt) of the material is determined by the following equation:

ð4Þ

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011

C.A. Pickles et al. / Minerals Engineering xxx (2013) xxx–xxx

dT 5:7  1011 PA ¼ dt Cp

ð5Þ

where Cp is the heat capacity of the material being heated. At room temperature, the loss factor for coal is about 0.1 and about 12 for water. Thus, according to Eq. (4), the absorbed power is about 120 times higher for water than for coal and therefore the water in the coal should heat more rapidly. This calculation assumes the water in coal behaves like bulk or free water but as discussed previously, the water in coal exists in five states and these types of water are bound to the coal, to varying degrees, and this can significantly reduce the loss factor of the water. Fig. 4 shows the as-received coal temperature as a function of drying time for sample masses of 10 g and 30 g at both 200 W and 500 W. For the 10 g sample at 200 W and 500 W, the sample temperature increased rapidly during the first 300 s and then the rate of temperature rise became slower and then began to level off. The 30 g sample at 200 W behaved in a similar manner but at 500 W the temperature increased rapidly after about 600 s to about 800 °C and subsequently continued to increase. Hot spots appeared inside the sample and the coal showed evidence of combustion. For the 30 g sample microwaved at 560 W as shown in Fig. 4, the initial heating rate was about 0.5 °C/s. At such high heating rates, it would be expected that the rate of water removal would be high and this could affect the structure of the coal. Figs. 5 and 6 show Scanning Electron Microscope (SEM) micrographs of the as-received coal and the microwaved coal, respectively. As shown in Figs. 5a and 5b, the as-received coal showed some minor evidence of cracking, presumably as a result of the crushing and grinding process. In this case, the cracks were relatively fine and only a few particles exhibited cracks. However, the microwaved coal demonstrated much more extensive fracturing and two types were observed: (1) relatively large random cracks as shown in Fig. 6a and (2) horizontal cracks roughly parallel to each other as shown in Fig. 6b. In both cases, the cracking is attributed to the rapid release of moisture and/or volatiles. Although the exact mechanisms are not clear, it is possible that the large random cracks are due to the growth of cracks, which previously existed as a result of crushing and grinding. On the other hand, the parallel cracks are due to the opening up of the layered structure in the coal along existing cleats. This phenomenon has been previously reported by Kumar et al. (2011), who utilized short bursts of high-energy microwave radiation to create both new fractures and enhance the cleat aperture size in a similar bituminous coal.

900 10g, 200 W 10g, 500 W 30g, 200 W 30g, 500 W

800

TEMPERATURE (°C)

700 600 500 400 300 200 100 0 0

200

5

400

600

800

1000

1200

1400

1600

TIME (secs) Fig. 4. Final sample temperature of microwaved Highvale coal as a function of time for various microwave powers and sample masses.

Fig. 5a. Highvale coal after crushing and grinding at low magnification.

3.3. Comparison of conventional and microwave drying Fig. 7a shows the moisture fraction of the coal as a function of time for both conventional and microwave drying for the asreceived coal sample ( x = 1435 lm). The sample mass was 30 g, the conventional drying temperature was 150 °C and the microwave power was 560 W. It can be seen that the shapes of these curves are similar but there are some differences. In conventional drying, the incubation period is relatively long, typically up to 900 s, due to the time needed to raise the sample temperature to the oven operating temperature. In microwave drying, the incubation period is short, typically less than 120 s, as the microwave radiation is rapidly absorbed by the sample. In conventional drying, very long times are required to remove the water and even at 3000 s, the moisture fraction was still about 0.31. At extended drying times the moisture fraction began to level off at about 0.2. However, in microwave drying almost all of the water could be removed very rapidly, with a moisture fraction of only 0.014 at 300 s. Generally, the observed high microwave drying rates have been attributed to a number of phenomena including; enhanced capillarity due to the inverted temperature profile (Perkin, 1990) and also microwave pumping due to the high internal pressures, resulting from the generation of gas bubbles in the solid matrix (Lyons et al., 1972). Fig. 7b shows the smoothed drying rates for both conventional and microwave drying as a function of moisture fraction for the same conditions as in Fig. 7a. For the drying of a non-hygroscopic porous material it would be expected that there would be three drying stages: an incubation period, a constant rate drying period and a falling rate period. A constant rate drying period is observed when the free water removed at the surface is replaced by capillary action such that the drying conditions are stable. In both conventional and microwave drying of coal such behavior is not observed and this would indicate that capillary action is not effective. As mentioned previously, water can exist in coal in five different states (Karr, 1978). In the early stages of the drying process for coal, the free water is evaporated first, followed by the more strongly bound water and as a consequence, the drying rate falls. In conventional drying, the initial drying rates are low due to the relatively long incubation period. However, as the sample temperature increases, then the drying rate increases and reaches a maximum, as the more loosely bonded water is removed. As this type of water is depleted, the remaining water becomes more difficult to remove and the drying rate decreases. The rate continuously decreases as it becomes more and more difficult for the water to

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011

6

C.A. Pickles et al. / Minerals Engineering xxx (2013) xxx–xxx 1.0

MOISTURE FRACTION (X)

Microwave, 560W Conventional, 150oC

0.8

0.6

0.4

0.2

0.0 0

1000

2000

3000

4000

5000

6000

7000

8000

TIME (secs) Fig. 7a. Moisture fraction versus time for microwave drying of 30 g of Highvale coal at 560 W and also conventional drying at 150 °C. Fig. 5b. Highvale coal after crushing and grinding at high magnification.

Fig. 6a. Microwaved Highvale coal particle showing the more irregular type of cracking.

Fig. 6b. Microwaved Highvale coal particle showing the horizontal cracking along cleats.

reach the surface and evaporate. In contrast, in microwave drying, the incubation period is very short and very high rates of water removal can rapidly be achieved. Again there is a maximum in the drying rate and subsequently the rate decreases and it can be noted the maximum rate in both conventional and microwave drying occurs at a moisture fraction of about 0.8. Thus in both cases, the highest drying rate occurs at the highest moisture content and also the incubation period persists until a moisture

fraction of about 0.8 is reached. It is noteworthy that the maximum peak in microwave drying is narrower than in conventional drying, indicating that microwave drying is more effective at removing the more loosely bonded water but once this type of water is removed then the removal of the more strongly bonded water is significantly more difficult. This could be attributed to the decrease in permittivities, once the freer water is removed, as discussed previously. In conventional drying the decrease in rate is more gradual, indicating that there is not such a sharp transition from the removal of free water to more tightly bonded water due to a more uniform and possibly higher temperatures throughout the sample. In microwave drying, the drying rate can be maintained at a relatively high value even as the moisture fraction approaches zero. However, in conventional drying, the drying rate approaches zero at a moisture fraction of about 0.1. These results demonstrate that the high microwave drying rates in the present research are mainly due to the ability of microwaves to create a higher energy density in the sample, rather than the ability to selectively heat the contained water. Fig. 8 shows the average drying rates in both conventional and microwave drying, for the 10 g coal sample. For the case of conventional drying, the maximum drying rate is plotted against drying temperature, while for microwave drying, the rate is plotted against incident microwave power. In conventional drying, the drying rate increased with temperature and with power for microwave drying. The maximum drying rate in microwave drying varied from one order to two orders of magnitude higher, than for conventional drying, depending on the conditions chosen. It is noteworthy that the microwave drying rates obtained at the lowest power level (160 W) were still larger than the values obtained for the highest temperature (210 °C) in conventional drying. The microwave drying results showed that the amount of water removed increased with microwave power. A higher power density results in increased energy absorption by the coal matrix and this can lead to enhanced heat and mass transfer at higher power levels and thus higher drying rates. Fig. 9 shows the effect of sample mass on the average drying rate for both conventional drying at 150 °C and microwave drying at 400 W. In conventional drying, the rate was low and decreased slightly with increasing mass, while in microwave drying, the average drying rate increased dramatically with increasing sample mass. For the sample mass of 10 g, the drying rates were one order of magnitude higher than for conventional drying, while at 30 g the rates were two orders of magnitude higher. In conventional drying, the amount of power available for absorption by the sample is constant and therefore increasing the sample mass reduces the

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011

7

C.A. Pickles et al. / Minerals Engineering xxx (2013) xxx–xxx -0.0200

-0.0006

Microwave, 560W Conventional,150oC

-0.0175

-0.0004 -0.0125 -0.0100

-0.0003

-0.0075 -0.0002

DRYING RATE (dX/dt)

DRYING RATE (dX/dt)

-0.0005 -0.0150

-0.0050 -0.0001 -0.0025 0.0000 1.0

0.0000 0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

MOISTURE FRACTION (X) Fig. 7b. Drying rate versus moisture fraction of Highvale coal for microwave and conventional drying for the conditions in Fig. 7a.

drying rate. On the other hand, as the sample mass increases in microwave drying, the amount of power absorbed by the sample increases. Additionally, the heat is generated internally and is transferred by conventional thermal conduction from the inside to the outside of the sample. At the surface, heat can be transferred by convection and/or radiation. For a sample with a larger mass, with internal heat generation, the outer portion of the sample can serve to reduce the heat loss, particularly for a material with low thermal conductivity. This combination of higher power absorption and reduced heat loss, results in higher drying rates. As discussed previously, the moisture content of the coal was adjusted by either pre-drying the coal or by adding additional moisture. This gave initial coal moisture contents of 10.5%, 12.5% and 21.3% for the pre-dried, as-received and hydrated coal samples, respectively. The sample mass was 10 g ( x = 1435 lm) and the incident microwave power was 560 W. The moisture fraction is shown as a function of time in Fig. 10a, the corresponding smoothed drying rates are shown as a function of moisture fraction in Fig. 10b and the final sample temperature is shown in Fig. 10c. With increasing initial moisture content, it can be seen that the incubation time decreases and the maximum drying rate occurs at higher moisture fractions. These results demonstrate that an increasing amount of initial moisture results in increased microwave absorption. Also, the drying rates increased with initial

moisture content, except at long drying times and hence low moisture fractions, where the drying rates decreased with increasing initial moisture content. From Fig. 10c it can be seen that the final sample temperature decreases slightly with increasing moisture content. As a result of the lower sample temperatures, the drying rate decreases with increasing initial moisture content at moisture fractions lower than about 0.2. It is possible that as the initial moisture content increases, then more of the microwave energy is utilized to remove the water and less energy is available for heating the sample. Additionally, since the same sample mass was utilized in each test then the final samples mass is lower, for higher initial moisture contents, and this could also result in decreased microwave absorption and hence lower temperatures. In order to improve the microwave absorption characteristics of the coal and hence the drying kinetics, some magnetite was added to the coal in varying amounts. It is well known that magnetite has excellent microwave absorption characteristics and is considered to be hyperactive. Also, after drying, the magnetite could potentially be removed by magnetic separation. Fig. 11 shows the average microwave drying rate over a period of 600 s as a function of the percentage of magnetite added. The sample mass was 30 g and the microwave power was 400 W. It can be seen that for a magnetite addition of ten percent, the drying rate increased by over 25%. At a magnetite addition of 25%, the drying rate had

POWER (W) 200

300

400

500

600 -0.0025

CONVENTIONAL MICROWAVE

AVERAGE DRYING RATE (dX/dt)

AVERAGE DRYING RATE (dX/dt)

100 -0.004

-0.003

-0.002

-0.001

0.000 140

MICROWAVE - 400W CONVENTIONAL - 150oC

-0.0020

-0.0015

-0.0010

-0.0005

0.0000 150

160

170

180

190

200

210

220

TEMPERATURE (°C) Fig. 8. Average drying rate of 10 g of Highvale coal as a function of power for microwave drying and temperature for conventional drying.

5

10

15

20

25

30

35

SAMPLE MASS (g) Fig. 9. Average drying rate as a function of sample mass for Highvale coal for microwave drying (400 W) and conventional drying (150 °C).

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011

8

C.A. Pickles et al. / Minerals Engineering xxx (2013) xxx–xxx 1.0

-0.0016 10.48% H2O

-0.0014

21.26% H2O

DRYING RATE (dX/dt)

MOISTURE FRACTION (X)

12.54% H2O

0.8

0.6

0.4

0.2

-0.0012

-0.0010

-0.0008

-0.0006

-0.0004

0.0 0

50

100

150

200

250

300

350

400

0

Fig. 10a. Moisture fraction versus time for microwave drying of 10 g of Highvale coal at 560 W for various initial moisture contents.

-0.025 12.54 %H2O

-0.020

DRYING RATE (dX/dt)

10

15

20

25

30

35

Fig. 11. Effect of magnetite additions on the average drying rate for microwave drying of 30 g of Highvale coal at 400 W.

decay equation. Therefore, the moisture fraction should change exponentially with the drying time and this relationship can be expressed by the following equation:

10.48 %H2O 21.26 %H2O

X ¼ a expðktÞ

ð6Þ

-0.015

-0.010

-0.005

0.000 1.0

0.8

0.6

0.4

0.2

0.0

MOISTURE FRACTION, X Fig. 10b. Drying rate versus moisture fraction for microwave drying of 10 g of Highvale coal at 560 W for various initial moisture contents.

300

SAMPLE TEMPERATURE (°C)

5

MAGNETITE ADDITION (%)

TIME (secs)

280

260

240

220

where X is the moisture fraction, t is time, a is the pre-exponential factor and k is the exponential variable. To determine the most suitable equation for drying a coal sample, the calculated moisture fractions were fitted to ten different models as presented in Table 3. These equations are modifications of the decay equations with some additional coefficients and/or terms. These models were initially developed for drying thin layer materials, in which the temperature distribution was uniform, however they have successfully been applied to other materials, in which the temperature distribution was not uniform. In order to investigate the temperature distribution in the coal sample in the present work, the vertical temperature gradient was measured after processing a 30 g sample for 600 s at 400 W. From Fig. 12, it can be seen that the temperature was the lowest at the surface, increased to a maximum at 1.5 cm, then decreased slowly with increasing depth. The temperature was relatively uniform and only varied about 25 °C from the average value of 174 °C. Statistical and non-linear regression analyses were carried out using Sigma-Plot (version 8.02) to estimate the statistical parameters for each drying model. Analysis of the variance (ANOVA) is used to test the ‘‘adequacy of fit’’. Both the correlation coefficient

Table 3 Mathematical models considered for describing the drying curves. 160W 400W 560W

200

180 8

10

12

14

16

18

20

22

24

Model name

Model equation

Reference

Lewis Page Modified page

X ¼ expðktÞ n X ¼ expðkt Þ n X ¼ expððktÞ Þ

Henderson and pabis Logarithmic

X ¼ a expðktÞ

Diffusion approach Verma Two-term exponential Simplified fick diffusion Midilli–Kucuk

X ¼ a expðktÞ þ ð1  aÞ expðkbtÞ

Lewis (1921) Page (1949) Overhults et al. (1973) Henderson and Pabis, 1961 Yagcioglu et al. (1999) Yaldız and Ertekin (2001) Verma et al. (1985) Sharaf-Elden et al. (1980) Diamante and Munro (1991) Midilli et al. (2002)

MOISTURE CONTENT (%) Fig. 10c. Final sample temperature versus moisture content for microwave drying of 10 g of Highvale coal at various powers.

approximately doubled. Thus, magnetite acts as a susceptor and can result in significant improvements in the drying kinetics. 3.4. Mathematical modeling of microwave drying process The curves of moisture fraction versus time in conventional and microwave drying can usually be fitted to the typical mathematical

X ¼ a expðktÞ þ c

X ¼ a expðktÞ þ ð1  aÞ expðgtÞ X ¼ a expðktÞ þ ð1  aÞ expðkatÞ X ¼ a expðcðt=L2 ÞÞ X ¼ a expðkðtn ÞÞ þ bt

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011

9

C.A. Pickles et al. / Minerals Engineering xxx (2013) xxx–xxx Table 4 Analysis of variance results for microwave drying of 30 g of Highvale coal at both 160 W and 560 W. Model

160 W

560 W

2

Lewis Page Modified page Henderson and Pabis Logarithmic Wang and singh Verma Two-term exponential Simplified fick diffusion Midilli–Kucuk

R

SEE(±)

RSS

F-ratio

R2

SEE(±)

RSS

F-ratio

0.9716 0.9728 0.9716 0.9803 0.9890 0.9687 0.9730 0.9719 0.9803 0.9908

0.0466 0.0457 0.0467 0.0389 0.0291 0.0490 0.0456 0.0464 0.0389 0.0266

0.6521 0.6240 0.6521 0.4518 0.2517 0.7181 0.6196 0.6450 0.4518 0.2107

10257.7 10697.3 10223.5 14887.7 13436.9 9255.91 5369.55 10338.5 7418.94 10686.3

0.9655 0.9892 0.9655 0.9861 0.9892 0.9804 0.9655 0.9913 0.9861 0.9921

0.0582 0.0329 0.0587 0.0373 0.0331 0.0443 0.0592 0.0294 0.0377 0.0286

0.2032 0.0639 0.2032 0.0822 0.0635 0.1157 0.2032 0.0512 0.0822 0.0467

1681.14 5383.31 1653.12 4171.37 2661.97 2947.07 812.552 6741.07 2050.33 2377.87

(R) and the coefficient of determination (R2) measure how well the regression model describes the data. Values near one indicate that the equation provides a good description of the relation between the independent variables and the response. Other statistical parameters such as the standard error of the estimate (SEE), the residual sum of squares (RSS) and the F-ratio were also used to evaluate the goodness of fit of the models. A higher quality of fit was associated with lower values of SEE and RSS and higher values of the F-ratio. The statistical analysis results are given in Table 4 for microwave powers of 160 W and 560 W. The Midilli–Kucuk Model exhibited the best fit and this result is in agreement with the results of other authors for the microwave drying of various materials (McKinn et al., 2005; Akpinar, 2006; McKinn, 2006). As yet there is no physical explanation as to why one model fits better than another. 3.5. Effective diffusion coefficients and activation energy Fick’s second law of diffusion is often employed to describe the drying of various materials as follows:

@cðx; tÞ @2c ¼D 2 @t @x

ð7Þ

where c is concentration, x is distance and D is the diffusion coefficient. Assuming that the moisture moves only by diffusion, there is no shrinkage, the diffusion coefficients are constant and the water evaporates only from the free surface of the sample, then an analytical solution for Fick’s second law can be obtained as follows (Crank, 1975):

X¼

1 8 X



1

p2 n¼0 ð2n þ 1Þ2

exp ð2n þ 1Þ2 p2

Deff t L

2

 ð8Þ

Fig. 13 shows such a plot of ln X versus t for the results obtained in the present work. After excluding the data for the incubation period, the results were fitted to straight lines and the values of the calculated average effective diffusion coefficients are shown in Fig. 14 as a function of the mass to power ratio (m/P). It can be seen that the values increased with increasing power and ranged from 6.45  107 m2/s at a mass to power ratio of 0.1875 W/ g to 2.84  106 m2/s at a mass to power ratio of 0.0536 W/g. There is very limited information in the literature on the diffusion coefficients of moisture in coals. However, the average effective diffusion coefficients in the present work are 1–2 orders of magnitude higher than those reported in the literature for a lignite coal, which ranged from about 2  108 to 8  10–8 m2/s for moisture fractions ranging from 0 to 1 and temperatures from 25 °C to 200 °C (Pakowski et al., 2011). These in turn, are higher than those reported for other porous materials such as fruit and vegetable tissue, which are typically in the range of 1011 to 109 m2/s, even under microwave conditions (Doymaz, 2005). The higher average effective diffusion coefficients in the present work can be attributed to both the porous nature of the coal and also the ability of microwaves to generate high energy densities and temperatures in the coal, which promotes both liquid and vapor diffusion, particularly at high power levels. It can be noted, that Eq. (10) has been evaluated using Fourier number (Fo) as follows:

F o ¼ Deff

X¼

8

p2

8

p2

exp p2

Deff t

ð9Þ

L2

Taking natural logarithms (ln) of both sides gives the following straight-line relationship:

 lnðXÞ ¼ ln

8

p2



 t

p2 Deff L2

 ð10Þ

exp½p2 F o 

ð13Þ

300

250

TEMPERATURE (°C)

X¼



ð12Þ

L2

Substituting in Eq. (9) gives:

where n is a positive integer, Deff is the effective diffusivity (m2/s) and L is the thickness of the sample (m). The solution to this equation can be approximated for long drying times by considering only the first term in the series, that is n = 0 as follows:



  t

200

150 AVERAGE

100

50

Thus, from a plot of ln X versus t, yields a slope (m) from which the average effective diffusion coefficient can be estimated as follows: 2

Deff ¼ m

L

p2

!

0 0

1

2

3

4

5

6

DISTANCE BELOW SURFACE (cm)

ð11Þ

Fig. 12. Vertical temperature gradient in a 30 g coal sample at 400 W for 600 s.

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011

10

C.A. Pickles et al. / Minerals Engineering xxx (2013) xxx–xxx -12.6

0.0

-12.8 -0.5 -13.0 -13.2

ln Deff

ln X

-1.0

-1.5

-13.4 ln Deff = -12.24 -10.96 (m/P)

-13.6

(R2=0.9985)

-13.8

-2.0

-14.0

160W 260W 400W 560W

-2.5

-14.2

-3.0 0

100

200

300

400

500

600

-14.4 0.04

0.06

0.08

Fig. 13. Plot of ln X versus time for microwave drying of 30 g of Highvale coal at various powers.

4e-6 3e-6

Deff (m2/s)

3e-6

160W 260W 400W 560W

2e-6 2e-6 2e-6 1e-6 5e-7 0 0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

MOISTURE FRACTION (X) Fig. 14. Effective diffusion coefficient as a function of moisture fraction for 30 g of Highvale coal at various powers.

which can be rewritten as:

F o ¼ 0:101 lnðXÞ  0:0213

ð14Þ

Thus from Eqs. (12) and (14), the effective diffusion coefficient can be determined as a function of time and hence moisture fraction as follows:

Deff

L2 ¼ ð0:101 ln X  0:0213Þ t

0.10

0.12

0.14

0.16

0.18

0.20

m/P (g/W)

TIME (secs)

!

ð15Þ

Fig. 15 shows the effective diffusion coefficients as a function of moisture fraction for various microwave powers. Once more it can be seen that the diffusion coefficient increases with microwave power. It would be expected that the mass transport of moisture in a porous material such as sub-bituminous coal would be due to both the diffusion of liquid water and under more extreme conditions, the diffusion of water vapor, both resulting from a moisture gradient. Previous research on coal has shown that the diffusion coefficient increases significantly with temperature but only weakly with moisture fraction (Pakowski et al., 2011). In the current work, as shown in Fig. 15, the effective diffusivity increases with decreasing moisture and therefore temperature is the predominant factor. The behavior of the diffusion coefficient in Fig. 15, can be characterized by three stages: (1) an initial rapid rate of increase, (2) a slowing of the rate of increase and levelling off, and (3) a decrease or an increase in

Fig. 15. Deff values obtained from Fig. 13 as a function of the mass to power (m/P) ratio.

the rate depending on the power level. In Stage (1), the diffusion coefficient increases due to the increasing temperature, despite the decreasing moisture fraction. Eventually, in Stage (2) the heat loss is matched by the energy input, the temperature remains relatively constant and consequently the diffusion coefficient remains relatively constant. In Stage (3), the moisture fraction has decreased and at low power levels the microwave absorption decreases, the temperature drops and the diffusion coefficient drops. On the other hand, at high power levels, the temperature of the sample continues to increase and this further increases the diffusion coefficient. It is likely that the temperatures are now so high that vapor diffusion is promoted and this results in a rapid increase in the diffusion coefficient to very high levels. Even at low moisture fractions, where the water is tightly bonded, drying can still be achieved and relatively high drying rates and low moisture contents can be achieved. Generally, the activation energy for a diffusion process can be determined using the well-known Arrhenius equation from the temperature dependency of the diffusion coefficient. Since it is difficult to obtain an accurate value for the temperature in a microwave drying system, then the activation energy can be calculated using a modified Arrhenius equation (Darvishi et al., 2012). In this case, the activation energy is related to the effective moisture diffusion coefficient and the ratio of microwave power to sample mass (m/P), as follows:

Deff ¼ Do exp

  Ea m P

ð16Þ

where Deff is the effective diffusion coefficient (m2/s), Ea is the activation energy (W/g), m is the initial mass of the sample (g), P is the microwave power (W) and Do is the pre-exponential factor (m2/s). As shown in Fig. 14, a plot of ln Deff versus m/P gives a straight line with a very good fit (R2 = 0.9985) and the slope or activation energy is 10.96 W/g. This value is relatively low, in comparison to the known values for similar materials (Darvishi et al., 2012), again reflecting the ability of microwaves to promote moisture diffusion in this type of material. Ozbek and Dadali (2007) derived an alternative method for determining the activation energy using the drying rate constant from the drying model as follows:

k ¼ ko exp

  Ea m P

ð17Þ

where k is the drying rate constant (1/min) from the Midilli– Kucuk model (1/min) and ko is the pre-exponential factor (1/min). Fig. 16 shows a plot of ln k versus m/P and it can be seen

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011

C.A. Pickles et al. / Minerals Engineering xxx (2013) xxx–xxx -0.6 -0.8 -1.0

ln k

-1.2 -1.4

ln k = -0.15 - 10.91 (m/P) (R2=0.9990)

-1.6 -1.8 -2.0 -2.2 -2.4 0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

m/P (g/W) Fig. 16. Plot of the drying rate constant (k) from the Midilli–Kucuk model as a function of the mass to power ratio (m/P).

that there is a very good fit (R2 = 0.9990) and the activation energy is 10.91 (W/g), which is in excellent agreement with the value of 10.96 (W/g) as obtained above from the effective diffusion coefficients. The excellent agreement between the two activation energies would indicate that the method utilized to estimate the diffusion coefficients is valid.

11

was attributed to the porous nature of the coal and the ability of microwaves to generate both high energy densities and temperatures in the coal. (5) The effective moisture diffusion coefficients were determined as a function of moisture fraction and incident microwave power. In general, the diffusion coefficients increased with incident power. During the initial stages of the drying process, the diffusion coefficients increased and then levelled off at intermediate moisture fractions. At low moisture fractions, the diffusion coefficients increased rapidly at high powers, but decreased at low powers. Since the diffusion coefficient is mainly determined by temperature, then at high powers and low moisture fractions, diffusion of water vapor occurs and this results in very high diffusion coefficients. (6) The activation energy for the diffusion process was calculated using two modified Arrenhius equations. The first utilized the relationship between the average effective diffusion coefficients (Deff) and the mass to power ratio (m/P) and yielded a value of 10.96 W/g. The second method employed the relationship between the drying rate constants (k) from the Midilli–Kucuk model and the mass to power ratio (m/P) and produced a value of 10.91 W/g. These relatively low values indicate that moisture diffusion is relatively easy in the microwave process. The excellent agreement between the two values of the activation energy indicates that the effective diffusion coefficients are valid.

8. Conclusions Acknowledgements (1) The real and imaginary permittivities of Highvale sub-bituminous coal were determined as a function of temperature, frequency and particle size. In general, the permittivities were relatively low and the effects of frequency and particle size were relatively minor, but temperature had a major effect. From room temperature to 100 °C, the permittivities increased and this was attributed to the increasing proportion of relatively free water. Above 100 °C, the amount of water in the sample diminished significantly and the density decreased, which lowered the permittivities. Above about 550 °C, the physical and chemical properties of the coal changed dramatically and the permittivities increased rapidly with temperature. (2) The microwave drying of a sub-bituminous coal was investigated using thermogravimetric analysis (TGA). The drying rate increased with both incident microwave power and sample mass. The average microwave drying rates were one to two orders of magnitude higher than in conventional drying. At high powers and masses and at low moisture fractions, the microwave drying rates were relatively high in comparison to conventional drying and much lower moisture fractions could be achieved. The addition of magnetite resulted in significant increases in the microwave drying rates. (3) The drying data were fitted to ten exponential decay thin layer drying models. Although reasonable agreement was obtained for most of the models, the best fit was obtained with the Midilli–Kucuk model. This result is in agreement with the results of other researchers, who have applied these models to the microwave drying of other materials. (4) The average effective diffusion coefficients for moisture were calculated from the drying rate data and were found to vary in the range of 6.45  107 m2/s at a mass to power ratio of 0.1875 W/g to 2.84  106 m2/s at a mass to power ratio of 0.0536 W/g. In comparison to the values reported in the literature, these values are relatively high and this

The authors wish to acknowledge the financial support of the Natural Sciences and Engineering Research Council of Canada (NSERC). Microwave Properties North (MPN) performed the permittivity measurements. John Forster is thanked for his assistance with some of the experimental results. References Akpinar, E.K., 2006. Determination of suitable thin layer drying curve model for some vegetables and fruits. Journal of Food Engineering 73 (1), 75–84. Bockris, J.O., Gileadi, E., Muller, J., 1966. Dielectric relaxation in the electric double layer. The Journal of Chemical Physics 44, 1445–1456. Brown, J.K., 1955. Infrared studies of carbonised coals. Journal of Chemical Society, 752–757. Chatterjee, I., Misra, M., 1991. Electromagnetic and thermal modeling of microwave drying of fine coal. Minerals and Metallurgical Processing 8, 110–114. Crank, J., 1975. The Mathematics of Diffusion, second ed. Clarendon Press, Oxford, UK. Darvishi, H., Azadbakht, M., Rezaieasl, A., Farhang, A., 2012. Drying characteristics of sardine fish dried with microwave heating. Journal of the Saudi Society of Agricultural Sciences 12 (2), 121–127. Davis, J.D., Anvil, H.S., 1935. Electrical conductivity of coke. Industrial and Engineering Chemistry 27 (10), 1196–1200. Diamante, L.M., Munro, P.A., 1991. Mathematical modelling of hot air drying of sweet potato slices. International Journal of Food Science and Technology 26, 99–109. Doymaz, I., 2005. Drying characteristics and kinetics of okra. Journal of Food Engineering 69, 275–279. Harrison, P.C., Rowson, N.A., 1997. The effect of microwave and conventional heat treatment on the comminution of coal. IChemE Research Event 1, 33–36. Hayashi, J.I., Kusakabe, K., Morooka, S., 1997. The role of microwave irradiation in coal desulphurization with molten caustics. Fuel 69, 739–742. Henderson, S.M., Pabis, S., 1961. Grain drying theory I: temperature effect on drying coefficient. Journal of Agricultural Research Engineering 6, 169–174. Hutcheon, R.M., De Jong, M.S., Adams, F.P., 1992. A system for rapid measurement of RF and microwave properties up to 1400 °C. Journal of Microwave Power and Electromagnetic Energy 27 (2), 87–92. Jones, S.B., Or, D., 2003. Model effects on permittivity measurements of water content in high surface area porous media’’. Physica B 338, 284–290. Karr, C., 1978. Analytical Methods for Coal and Coal Products, vol. 1. Academic Press, NY. Kumar, H., Lester, E., Kingman, S., Bourne, R., Avila, C., Jones, A., Robinson, J., Halleck, P.M., Mathews, J.P., 2011. Inducing fractures and increasing cleat apertures in a

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011

12

C.A. Pickles et al. / Minerals Engineering xxx (2013) xxx–xxx

bituminous coal under isotropic stress via application of microwave energy. International of Journal of Coal Geology 88 (1), 75–82. Lester, E., Kingman, S.W., 2004. The effect of microwave preheating on five different coals. Fuels 83, 1941–1947. Lewis, W.K., 1921. The rate of drying of solid materials. Industrial Engineering Chemistry 13, 427–432. Lindroth, D.P., 1986. Microwave Drying of Fine Coal. US Bureau of Mines Report # 9005, US Department of Interior. Lyons, D.W., Hatcher, J.D., Sunderland, J.E., 1972. Drying of a porous medium with internal heat generation. International Journal of Heat and Mass Transfer 9, 897–905. Marland, S., Han, B., Merchant, A., Rowson, N., 2000. The effect of microwave radiation on coal grindability. Fuel 79, 1283–1288. McKinn, W.A.M., 2006. Thin-layer modelling of the convective, microwave, microwave-convective and microwave-vacuum drying of lactose powder. Journal of Food Engineering 72 (2), 113–123. Mckinn, W.A.M., McLoughlin, C.M., Magee, T.R.A., 2005. Thin-layer modelling of microwave, microwave-convective, and microwave-vacuum drying of pharmaceutical powders. Drying Technology: An International Journal 23 (3), 513–532. Midilli, A., Kucuk, H., Yapar, Z., 2002. A new model for single-layer drying. Drying Technology 20 (7), 1503–1513. Nelson, S.O., Fanslow, G.E., Bluhm, D.D., 1980. Frequency dependence of the dielectric properties of coal. Journal of Microwave Power 15, 277–282. Or, D., Wraith, J.M., 1999. Temperature effects on soil bulk dielectric permittivity measured by time domain reflectometry: a physical model. Water Resources Research 35 (2), 371–383. Overhults, D.D., White, G.M., Hamilton, M.E., Ross, I.J., 1973. Drying soybeans with heated air. Transactions of the ASEA 16, 195–200. Ozbek, B., Dadali, G., 2007. Thin-layer drying characteristics and modeling of mint leaves undergoing microwave treatment. Journal of Food Engineering 83, 541– 549. Page, G., 1949. Factors influencing the maximum rates of air drying shelled corn in thin layers. MSc Thesis, Purdue University, Indiana, USA.

Pakowski, Z., Adamski, R., Kwapisz, S., 2011. Effective diffusivity of moisture in low rank coal during superheated steam drying at atmospheric pressure. Chemical and Process Engineering 33 (1), 43–51. Perkin, R.M., 1990. Simplified modeling for the drying of a non-hygroscopic capillary porous body using a combination of dielectric and convective heating. Drying Technology 8, 931–951. Rowson, N.A., Rice, N., 1990. Desulphurization of coal using low power microwave energy. Minerals Engineering 3 (3/4), 363–371. Seehra, M.S., Kalra, A., Manivannan, A., 2007. Dewatering of fine coal slurries by selective heating with microwaves. Fuel 86, 829–834. Sevi, U., 1995. A review of spontaneous combustion of coals. Fuel Science and Technology International 13 (9), 1103–1120. Sharaf-Elden, Y.I., Blaisdell, J.L., Hamdy, M.Y., 1980. A model for ear corn drying. Transactions of the ASAE 5, 1261–1265. Standish, N., Worner, H., Kaul, H., 1988. Microwave drying of brown coal agglomerates. Journal of Microwave Power EM Energy 23, 171–175. _ Depci, T., 2007. Applications of microwave pretreatment in coal. Toraman, I.Y., Madencilik 46 (3), 43–53 (in Turkish). Uslu, T., Atalay, U., 2003. Microwave heating of coal for enhanced removal of pyrite. Fuel Processing Technology 85, 21–29. Verma, L.R., Bucklin, R.A., Endan, J.B., Wratten, F.T., 1985. Effects of drying air parameters on rice drying models. Transactions of the ASAE 28, 296–301. Viswanathan, M., 1990. Investigation on the effect of microwave pre-treatment on comminution/beneficiation and desulphurisation of coals. In: 11th International Coal Preparation Congress, Tokyo, pp. 151–155. Yagcioglu, A., Degirmencioglu, A., Cagatay, F., 1999. Drying characteristics of laurel leaves under different drying conditions. In: Proceedings of the 7th International Congress on Agricultural Mechanization and Energy in Agriculture. Adana, Turkey, pp. 565–569. Yaldız, O., Ertekin, C., 2001. Thin layer solar drying of some different vegetables. Drying Technology 19 (3), 583–596. Zimmermann, E., Niemann-Delius, C., 2007. Microwave beneficiation of brown coal. _ Górnictwo i Geoinzynieria 31 (2), 627–633.

Please cite this article in press as: Pickles, C.A., et al. Microwave drying of a low-rank sub-bituminous coal. Miner. Eng. (2013), http://dx.doi.org/10.1016/ j.mineng.2013.10.011