- Email: [email protected]
An approach for utilization of direct coal liquefaction residue: Blending with low-rank coal to prepare slurries for gasification
Fuel 145 (2015) 143–150
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
An approach for utilization of direct coal liquefaction residue: Blending with low-rank coal to prepare slurries for gasification Dongmei Lv a,b, Wei Yuchi a, Zongqing Bai a,⇑, Jin Bai a, Lingxue Kong a, Zhenxing Guo a, Jingchong Yan a,b, Wen Li a a b
State Key Laboratory of Coal Conversion, Institute of Coal Chemistry, Chinese Academy of Sciences, Taiyuan 030001, China University of Chinese Academy of Sciences, Beijing 100049, China
h i g h l i g h t s The liquefaction residue can improve the slurryability of low-rank coals. YM and SM coal have synergistic effect with the residue in slurryability. Blending slurries with the residue and low-rank coal have strong pseudoplasticity. Static stability of blending slurries is remarkable.
a r t i c l e
i n f o
Article history: Received 15 August 2014 Received in revised form 3 December 2014 Accepted 22 December 2014 Available online 5 January 2015 Keywords: Direct coal liquefaction residue Low-rank coal Blending Slurryability
a b s t r a c t Direct coal liquefaction residue (DCLR) is a main byproduct in direct coal liquefaction process and its clean and high-efficient utilization is important. If DCLR could be gasified for hydrogen production, it will compensate hydrogen consumption during liquefaction and lower the operation cost, which makes DCLR utilization more promising. We proposed to blend DCLR with low-rank coals to prepare DCLR–coal–water slurries (DCLRCWS) as feedstock for gasification. In this work, one DCLR and four low-rank coals were used to prepare the individual DCLR–water slurries (DCLRWS), coal water slurries (CWS) and the mixed DCLRCWS at the weight ratio of 1:1. The slurryability, static stability and rheology of various slurries were investigated. The results show that DCLR and the low-rank coals are complementary in terms of slurry properties and DCLRCWS could meet the requirement of gasification process. Adding 50 wt.% DCLR apparently improves the slurryability of low-rank coals, and the maximum solid loading (Cmax) of DCLRCWS prepared with ZLNR coal and DCLR is about 10% higher than that of the corresponding CWS. The effects of coal properties, dispersant adsorption, zeta potential on preparation of highly loaded slurries were examined in terms of slurryability. Moreover, compared with DCLRWS, DCLRCWS display higher degree of pseudoplasticity and better static stability owing to the addition of low-rank coals. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Recently, interest in the development of coal hydroliquefaction has increased considerably, as driven by the surging global oil demand and increasing concerns about energy security in China. However, the high cost of coal hydroliquefaction is a main obstacle for its further development, which has brought unprecedented opportunity and challenge to research areas at the same time [1]. As a main byproduct of direct coal liquefaction process, direct coal liquefaction residue (DCLR) should be disposed and utilized reasonably. The direct coal liquefaction plants tend to gasify DCLR to ⇑ Corresponding author. Tel.: +86 351 4048967; fax: +86 351 4050320. E-mail address: [email protected] (Z. Bai). http://dx.doi.org/10.1016/j.fuel.2014.12.075 0016-2361/Ó 2014 Elsevier Ltd. All rights reserved.
supply hydrogen needed in the liquefaction process, which can realize self-sufficiency in hydrogen consumption and significantly increase the economy of the process [2]. Generally, DCLR accounts for 20–30% of the coal consumed in the process [3], but it is still not enough for the operational capacity of a single gasifier and the stability and running period of the liquefaction process should be considered as well. Thus, it is urgent to find suitable raw materials to co-gasify with DCLR for the normal operation of the whole system. Usually, DCLR is separated by reduced pressure distillation, which contains considerable heavy oil and asphaltenes. Accordingly, dry feeding gasification is not suitable for DCLR due to its low softening point (about 180 °C), leading to great difficulty in milling system, while wet feeding could be operated easily. The
144
D. Lv et al. / Fuel 145 (2015) 143–150
results of our previous work [4] indicate that the slurryability of DCLR is relatively high, its static stability determined by rod dropping method is 2 days, and the shear-thinning behavior of direct coal liquefaction residue-water slurry (DCLRWS) is not strong, which fairly accords with other reported results [5,6]. Low-rank coals account for nearly half of the coal reserve worldwide and have an advantage of low price, regarded as a promising raw material for preparation of coal water slurry (CWS) [7]. But it is assumed that low-rank coals, especially brown coal, are difficult to make highly loaded CWS. Over the years, various measures have been taken to improve the slurryability of lowrank coals by thermal treatment [8], hydrothermal dewatering process [9,10], microwave/ultrasonic irradiation [8,11], developing high-efficient dispersant [12,13] and blending with petroleum coke [14], etc. Blending DCLR with low-rank coals to prepare DCLRCWS may be an energy-efficient and environmental approach, which could not only supply sufficient feedstock to the gasification process but also have the potential to enhance the slurryability of low-rank coals. However, the studies on the blending principles of DCLR and low-rank coal as well as the properties of DCLRCWS have not been reported yet. In this work, four low-rank coals were selected to prepare CWS, whose properties were compared with that of DCLRWS. DCLR was blended with the coal samples to prepare DCLRCWS at the weight ratio of 1:1 and the slurryability, rheology and static stability of DCLRCWS were systematically investigated. The surface physical and chemical properties of DCLR and coal samples, dispersant adsorption isotherms, zeta potential, and SEM pictures of slurries were analyzed to study the interactions between DCLR and coal samples in terms of slurry properties. The blending principles of DCLRCWS can provide experimental data and theoretical basis for the integration of coal liquefaction and gasification process.
graded on Alfred particle size distribution [15]. An anionic dispersant, naphtalenesulfonate-formaldehyde condensate (NSF), was used and the dosage of dispersant is 1 wt.% on dry basis of coal and/or DCLR. For DCLRCWS, certain amount of DCLR and coal sample at the weight ratio of 1:1 was successively transferred into the bottle containing a predetermined quantity of dispersant and deionized water. And then, the mixture was agitated at 3000 r/ min for 10 min (CWS), while for 20 min (DCLRWS and DCLRCWS) for homogenization. All experiments were conducted at room temperature. 2.3. Determination of slurry properties The rheology of slurries was determined by NXS-11 rotational viscometer (Chengdu Analytical Instrument Factory, China) at the shear rate range of 10–100 s1 and the average of 6 values at shear rate of 100 s1 was defined as the apparent viscosity. The maximum solid loading of slurry was designated as the solid loading when the apparent viscosity of slurries was 1000 mPa s at 20 °C [10]. The maximum solid loading of DCLRWS, CWS and DCLRCWS is denoted as Cmax,1, Cmax,2, Cmax,3, respectively. The rod dropping method was used to evaluate the static stability of slurries with the apparent viscosity close to 1000 mPa s once a day till the appearance of hard sediment up to 14 days. 2.4. Characterization 2.4.1. Physical adsorption test The texture properties analyses of the samples were carried out with isothermal N2 adsorption at 77 K using ASAP 2020 – Physisorption Analyzer (Miromeritics, USA). All the samples were dried at 60 °C under vacuum for 24 h before physical adsorption test. 2.4.2. Chemical structure analysis The samples with Alfred particle size distribution were ground to an appropriate size and dried at 60 °C under vacuum for 24 h before FTIR measurement. FTIR spectra of the samples were obtained on an IR spectrometer (VERTEX 70, Bruker, Germany), and the cell was the 0030–102 accessory (Pike Co. Ltd., USA) with ZnSe windows. The same Al2O3 crucible was used in measuring the five samples, so the thickness of the samples is also the same with the height of the crucible. The diffuse reflection mode was selected to record the spectra within 4000–600 cm1 with the co-addition of 200 scans at 4 cm1 resolution. Chemical titration was conducted to determine the content of hydrophilic function groups and the procedure was presented by Fan et al. [16].
2. Experimental 2.1. Materials DCLR used in this work was obtained from 6 t/d direct coal liquefaction pilot plant of Shenhua Group Corporation in China. Its raw coal is Shangwan coal from Inner Mongolia. Four low-rank coal samples from Shenmu, Yima, Yuxian and Zhalainuoer were employed and denoted as SM, YM, YX and ZLNR coal, respectively. The proximate and ultimate analyses of DCLR and four coal samples are listed in Table 1. 2.2. Preparation of slurries
2.4.3. Dispersant adsorption measurement During dispersant adsorption tests, the mixture of graded samples and dispersant solution at weight ratio of 1:10 was charged into a 100 ml conical flask and shaken for 5 h at 200 rpm to make
The samples were separately crushed and sieved to several groups with different particle sizes: 280–154 lm, 157–74 lm and less than 74 lm. DCLR and four coal samples were separately
Table 1 Proximate and ultimate analyses of five samples. Proximate analysis (wt.%, ad)
DCLR YM SM YX ZLNR
Ultimate analysis (wt.%, daf)
M
A
V
C
H
N
S
O⁄
0.17 4.37 2.68 13.00 12.07
12.46 38.69 10.47 14.23 12.39
45.12 23.34 29.73 23.83 30.19
90.26 73.46 81.05 80.76 75.32
5.99 5.78 5.24 5.58 6.50
1.25 1.05 1.08 0.99 1.05
1.77 1.07 0.30 1.02 0.30
0.73 18.64 12.33 11.65 16.83
ad: air dried basis; daf: dry and ash-free basis; ⁄ By difference. M: moisture; A: ash; V: volatile; C: carbon; H: hydrogen; N: nitrogen; S: sulfur; O: oxygen. H/C: the ratio of hydrogen atom to carbon atom. MHC: the maximum moisture holding capacity.
H/C
MHC
0.80 0.94 0.78 0.83 1.04
1.0 7.5 8.5 16.6 24.7
145
D. Lv et al. / Fuel 145 (2015) 143–150
the particles to reach adsorption equilibrium. Then the suspension was filtered under reduced pressure and centrifuged to obtain the supernatant which was transferred to determine the concentration of dispersant by Shimadzu UV-3600 spectrophotometer at 292 nm. The initial concentrations of the dispersant solution are 0, 50, 100, 200, 400, 600, 800, 1000, 1200 mg/L for DCLR and the four coal samples. The dispersant adsorption was calculated from the difference between the dispersant concentration before and after adsorption, and the adsorption isotherms of two measurements can almost overlap. 2.4.4. SEM A high resolution field emission scanning electron microscope (FESEM, JSM-7001F, Japan) was employed to investigate the packing efficiency of different slurries. 2.4.5. Zeta potential 50 ml dispersant solution (0.02 g/L) and 0.1 g graded coal samples, DCLR or blended samples were put into a 100 ml conical flask and shaken for 5 h at 200 rpm, so the particles could reach adsorption equilibrium. Then the zeta potential of the suspension was measured by JS94HM Zetasizer (POWEREACH, China) under natural pH (about 6.5). Average values of duplicated tests were used and the measurement error was within ±1%. 3. Results and discussion 3.1. Chemical and physical characteristics of DCLR and four coal samples Table 1 shows that the DCLR contains less O than the coals, attributing to the removal of oxygen during hydroliquefaction. On the other hand, DCLR contains more S derived from the spent catalyst and the raw coals. In addition, it is notable that DCLR has high content of C and H, which makes it potential to be used as feedstock for gasification. The results of physical adsorption analysis are shown in Table 2. The BET surface area and pore volume of DCLR are the smallest because of the collapsing and plugging of pore structure in the raw coal during liquefaction. However, the four coal samples have more developed pore structure. Fig. 1 illustrates the infrared spectra of the five samples. The characteristic peaks including 3400 cm1 due to OAH, 2800– 3000 cm1 due to CAH stretching in methyl and methylene, 1720 cm1 due to C@O stretching vibration, 1400–1600 cm1 due to the stretching vibration of aromatic rings and 1000– 1200 cm1 due to CAOAC are found for the five samples. Compared with the coal samples, the peak strength of the oxygen-containing functional groups for DCLR are much weaker. The quantitative analysis of the functional groups was done by chemical analysis method. The results, shown in Table 2, indicate that the total acidity content of DCLR is apparently lower than that of the coal samples, which is consistent with the infrared spectra results. Compared with YM and SM coal, YX and ZLNR coal have higher content of oxygen-containing functional groups.
Fig. 1. The FTIR spectra of DCLR and four coal samples.
3.2. Adsorption behavior of dispersant on the samples Fig. 2 illustrates the adsorption isotherms of NSF dispersant on the five samples. The equilibrium adsorption quantity (U1) obtained from Langmuir isotherm curve-fitting are 2.16, 2.04, 0.90, 0.83 and 0.80 mg/g for SM, YM, DCLR, YX and ZLNR coal, respectively. Langmuir adsorption model is as follows:
C=C1 ¼ Kc=ð1 þ KcÞ:
ð1Þ
where, U is the adsorption quantity (mg/g); c is the equilibrium concentration (mg/L); K is the Langmuir equilibrium constant and stated adsorption capacity of dispersants on the coal. The equilibrium adsorption quantity of YM and SM coal is apparently higher than that of other samples, and the different adsorption behavior with the NSF dispersant is mainly related to the pore structure and the oxygen-containing functional groups content of the samples. Table 2 indicates that YM and SM coal have more developed pore structure which will supply more adsorption sites for the dispersant, thus resulting in higher dispersant adsorption capacity. The BET specific surface area and pore volume of YX and ZLNR coal are much smaller, thus resulting in fewer adsorption sites. Besides, the higher contents of oxygen-containing functional groups on the surface of YX and ZLNR coal further reduce the amount of adsorption sites [17], thus their equilibrium adsorption quantity is relatively low. The low porosity of DCLR is the main reason for its low dispersant adsorption. 3.3. Properties of CWS and DCLRCWS 3.3.1. Slurryability The apparent viscosity of slurries increases with increasing solid loading, as shown in Fig. 3. This is because the increased solid loading enhances interparticle frictional forces and the water available outside the particles decreases, and thus the sharply increased resistence of shear improves the apparent viscosity of slurries [18].
Table 2 BET specific surface area, pore volume and oxygen-containing functional groups in the five samples. Samples
Specific surface area (m2/g)
Pore volume (102 cm3/g)
Total acidity (mmol/g)
ACOOH (mmol/g)
ArAOH (mmol/g)
DCLR YM SM YX ZLNR
0.66 7.81 7.83 2.09 2.32
0.45 1.87 1.78 1.15 1.33
0.09 1.66 0.78 2.78 3.78
0.06 0.10 0.06 0.23 0.36
0.03 1.56 0.72 2.55 3.42
146
D. Lv et al. / Fuel 145 (2015) 143–150
Fig. 2. Adsorption isotherms of NSF dispersant on DCLR and four coal samples.
Fig. 3. Effect of solid loadings on the apparent viscosity of CWS, DCLRCWS and DCLRWS.
Table 3 The slurryability of different slurries. Samples
Cmax,2 (wt.%)
Cmax,3 (wt.%)
Cwa (wt.%)
D1 (wt.%)
D2 (wt.%)
YM SM YX ZLNR
65.0 63.0 56.2 51.7
70.1 69.5 64.2 61.4
69.3 68.3 64.9 62.6
5.1 6.5 8.0 9.7
0.8 1.2 0.7 1.2
Cmax,1 = 73.6; Cwa = 0.5 (Cmax,1 + Cmax,2); D1 = Cmax,3–Cmax,2; D2 = Cmax,3–Cwa.
Table 3 shows the slurryability of the slurries. Cmax is an important index to appraise the slurryability of slurries. A higher Cmax means a higher slurryability of slurries. As shown in Table 3, Cmax,1 (73.5%) of DCLRWS is apparently higher than that of the four CWS, and even 20% higher than Cmax,2 of CWS prepared with ZLNR coal. The reason is that the absorbing water capacity of DCLR is much lower due to its special properties, such as higher carbon content, less oxygen-containing functional group, underdeveloped pore structure and lower MHC. Consequently, more free water existing in DCLRWS acts as lubricant and improves its slurryability. The solid loading of the slurry fuels should be enhanced maximally to reach high heat value and to ensure efficient gasification and combustion [19]. However, it is difficult for low-rank coals to prepare highly loaded CWS and the maximum solid loadings of four coals are generally low. Therefore, it is necessary to make more efforts to improve their slurryability.
The four DCLRCWS prepared with the blending samples of DCLR and coal samples at the weight ratio of 1:1 demonstrate superior slurryability to their corresponding CWS, and all their solid loadings reach 60% meeting the requirement of gasification. In Table 3, the values of D1 are determined from the difference between Cmax,3 and Cmax,2. The maximum solid loading of CWS prepared with ZLNR coal is 51.7%, while that of corresponding DCLRCWS increases to 61.4%, and the value of D1 is the largest about 10%. Compared with the coal samples, the higher slurryability of the blending samples is mainly attributed to the positive effect of DCLR and the interaction between DCLR and coal sample. The water absorbing capacity of DCLR is low. Hence, the addition of DCLR could reduce the water absorbing capacity of the blending samples. Moreover, the representative SEM pictures (Fig. 4) illustrate that the packing efficiency of the blending slurry (YX-DCLR) and DCLR is higher than that of the corresponding CWS. This is because higher contents of hydrophilic groups in coal samples easily lead to forming loose aggregates of flocs preventing particles closing to each other. On the other hand, more hydrophobic DCLR particles enter into the loose structure and fill the voids among the coal particles, thus the blending particles could accumulate more tightly, and then resulting in higher packing efficiency. High packing efficiency means lower interparticle voidage and less free water required to fill the inter-particle voids, which is favorable for improving slurryability [20]. Based on the above considerations, the water absorbing capacity and the amount of free water decrease after blending DCLR into coal samples, which accounts for the high slurryability of DCLRCWS. Cwa is the weighted average of Cmax, 1 and Cmax, 2 at the weight ratio of 1:1, and D2 is equal to the difference of Cmax,3 and Cwa. The values of D2 always deviate from zero, suggesting that slurryability of the blending samples has no linear additivity. Besides, the values of D2 for YM-DCLR and SM-DCLR are positive, while those for YXDCLR and ZLNR-DCLR are negative. This is thought to be mainly related to the dispersant adsorption behavior and zeta potential of the samples, which will be discussed in the following section. The dispersant adsorption on the four blending samples (termed as Q0) is investigated when the dispersant concentration (1000 mg/L) is equivalent to the dispersant dosage in slurries. Qwa is the weighted average of the dispersant adsorption on DCLR and coal alone at the dispersant concentration of 1000 mg/L. Fig. 5 shows that the values of Qwa and Q0 are basically the same within the measurement uncertainties, which indicates that there is little interaction between DCLR and coal samples in the aspect of dispersant adsorption. The surface of coal is usually negatively charged with minus zeta potential. The zeta potential of particles in slurries is related with many factors, such as the nature of coal, pH value, electrolyte concentration, surface oxygen-containing functional group and dispersant adsorption [21]. Fig. 6 compares the zeta potential of the individual samples with that of the blending samples. The absolute zeta potential of DCLR is the lowest, which is mainly attributed to its fewer oxygen-containing functional groups and lower dispersant adsorption. Blending samples of YM-DCLR and SM-DCLR have higher absolute zeta potential than the corresponding coal samples, YM and SM coal, while the absolute zeta potential of blending samples YX-DCLR and ZLNR-DCLR is lower than that of YX and ZLNR coal, respectively. The maximum solid loading of slurries is directly proportional to the product of the equilibrium adsorption quantity and the absolute zeta potential [22]. Based on the analysis above, the addition of DCLR has little influence on the dispersant adsorption, while apparently affects the zeta potential which plays the dominant role in slurryability of the blending samples. If the absolute zeta potential of the blending sample is higher than that of the corresponding coal, DCLR and the coal sample have synergistic effect on the
D. Lv et al. / Fuel 145 (2015) 143–150
147
Fig. 4. The SEM photographs of CWS (YX), DCLRCWS (YX-DCLR) and DCLRWS.
Fig. 5. Comparison between dispersant adsorption on blending samples (Q0) and weighted average of dispersant adsorption on DCLR and coal sample (Qwa).
maximum solid loading of DCLRCWS, which is consistent with the positive value of D2. On the contrary, there is anti-synergistic effect, which is consistent with the negative value of D2. Consequently, the values of D2 are positive for YM-DCLR and SM-DCLR, while negative for YX-DCLR and ZLNR-DCLR. 3.3.2. Static stability The static stability is an important property for storage of the slurries. As a coarse disperse system, the solid–liquid separation of slurries might be originally expected to occur easily. This is because the DCLR/coal particles, with most regions of surface being hydrophobic, tend to form large multi-particle units that can settle rapidly facilitating separation of the phases [23]. Dispersant could reduce the tendency of sediment by adsorbing on the surface of particles. The particles adsorbing certain amounts of dispersant
Fig. 6. The zeta potential of DCLR, coal samples and blending samples.
could interact with each other by dispersant bridges and eventually form a three-dimensional network structure in slurries. Usually, the structure of a dispersant molecular has one hydrophobic tail and one hydrophilic tail. The hydrophobic tails of dispersant adsorb on the hydrophobic regions of coal particles, and meanwhile, hydrophilic tails could form interpolymer complexes with surface carboxylic acid or the hydrophilic tails of other dispersant mainly through hydrogen bonding. Thus dispersant bridges form among the particles and construct a three-dimensional network structure in slurries [24]. The structure could prevent the particles from gravity sedimentation and enhance the stabilization of slurries [25]. In addition, the type and dosage of dispersant, particle size distribution of samples, the concentration of electrolyte and the physiochemical properties of the raw materials also have very important effects on the static stability of the slurries [25,26].
148
D. Lv et al. / Fuel 145 (2015) 143–150
The static stability of DCLRWS is not satisfying and hard sediment appears within two days, which cannot afford long-distance transport and long-time storage. However, the static stability of the four CWS is remarkable. The CWS prepared with YM, YX, ZLNR coal can maintain homogeneous for more than 14 days, and hard sediment appears within four days for the CWS prepared with SM coal. Furthermore, all of the DCLRCWS have the same static stability with their corresponding CWS, which are better than that of DCLRWS. The difference of the static stability between the DCLRWS and the four CWS is mainly ascribed to the different surface properties of DCLR and coals. The dispersant adsorption and oxygen-containing functional groups of DCLR is so low that dispersant bridges among the particles are hard to form, thus there is few chances for the formation of three-dimensional network structure, which is unfavorable for improving the stability of DCLRWS. On the other hand, since the dispersant could improve the hydrophilicity of DCLR particles, the hydrophobicity of their surface is still strong leading to serious aggregation in the aqueous medium and formation of hard sediment soon. Compared with DCLR, the four lowrank coals have apparently higher dispersant adsorption (for YM and SM coal) or distinctly higher oxygen-containing functional groups (for YX and ZLNR coal), which facilitate the formation of three-dimensional network structure accounting for the enhanced static stability of the four CWS. In addition, the zeta potential of particles displays a substantial correlation with the static stability of slurries. Higher the absolute zeta potential is, more likely the slurries are to remain in stable state [27]. Therefore, the better stabilization of the four CWS could also be attributed to the higher absolute zeta potential, which could provide stronger electrostatic
repulsive forces among the coal particles and enhance the dispersion of the coal particles. The improvement in the static stability, when 50% weight percent of DCLR particles is replaced by the low-rank coal particles, might be correlated with the formation of three-dimensional network structure in the four DCLRCWS. The respective dispersant adsorption of DCLR and coals almost has no difference before and after the blending process. Thus the coal particles could still exhibit close connection with each other and even with DCLR particles through dispersant bridges in DCLRCWS, and a three-dimensional network structure could form to stabilize the slurries. Additionally, the coal particles distribute around DCLR particles in the DCLRCWS, which could weaken the hydrophobic attractive forces among DCLR particles and prevent adverse aggregation, thus resulting in enhanced static stability of the blending slurries. Moreover, the improvement in the static stability of DCLRCWS is also ascribed to the higher absolute zeta potential of blending samples compared with that of DCLR (seen in Fig. 6). 3.3.3. Rheological properties The rheological behavior of highly loaded slurries is important for their preparation, storage, transportation and atomization [28]. The effects of shear rate on viscosity and shear stress of DCLRWS, CWS and DCLRCWS are depicted in Fig. 7. The shear stress/shear rate dependence fits well with the Herschel–Bulkley model shown as follows:
s ¼ sy þ K c n
ð2Þ
where, s and sy are shear stress and yield shear stress, respectively; K is the rheological constant and n is the rheological exponential.
Fig. 7. The effect of shear rate on viscosity (a and c) and shear stress (b and d) of DCLRWS, CWS and DCLRCWS.
149
D. Lv et al. / Fuel 145 (2015) 143–150 Table 4 The rheology of DCLRWS, CWS and DCLRCWS. Samples
Solid loading (wt.%)
Rheological exponential, n
Correlation coefficient, R2
D3
DCLR YM SM YX ZLNR YM-DCLR SM-DCLR YX-DCLR ZLNR-DCLR
73.5 65.0 63.0 56.2 51.7 70.4 69.7 64.3 61.5
0.86 0.57 0.42 0.71 0.62 0.69 0.83 0.73 0.68
0.9972 0.9990 0.9968 0.9938 0.9986 0.9972 0.9961 0.9971 0.9960
– – – – – 0.17 0.03 0.13 0.18
D3 = nDCLRWS–nDCLRCWS.
For a Newtonian fluid, n = 1; for non-Newtonian fluids, n < 1 (pseudoplastic) or n > 1 (dilatant). When n < 1, the lower the n is, the stronger the shear-thinning/pseudoplastic behavior of the slurries is. The corresponding parameters of different slurries are listed in Table 4. Fig. 7 demonstrates that all types of slurries are pseudoplastic fluids, with the typical shear-thinning behavior that viscosity decreases with increasing shear rate, which is favorable for transportation of slurries in gasification. The rheological exponentials of the slurries are all less than 1, which is corresponding to the shearthinning behavior. Table 4 shows the rheological exponential of DCLRWS is the largest among the five individual slurries, suggesting that the shear-thinning behavior of DCLRWS is weaker than that of the four CWS. The rheological exponential of the DCLRWS is 0.86 which is higher than those of all the DCLRCWS, indicating that the pseudoplastic behavior of the blending slurries are strengthened after adding the low-rank coals. Moreover, the four coal samples have different influence on the rheological behavior of DCLRCWS, which may be attributed to the characteristics of the coal samples. The values of D3 are derived from the difference between the rheological exponential of DCLRWS and DCLRCWS, which reflects the different influence of the four coal samples. ZLNR coal has the highest efficiency in improving the flowability of DCLRCWS, decreasing n from 0.86 to 0.68 and other three coals, YM, YX, SM coal have decreasing efficiency in turn. Note that the value of D3 has good relationship with H/C ratio of coal samples based on ultimate analysis, that is, the larger the H/C ratio is, the higher the value of D3 is. With the highest H/C ratio, ZLNR coal has the highest value of D3, corresponding to the strongest pseudoplastic behavior of DCLRCWS.
4. Conclusion Based on the results and discussions mentioned above, satisfying DCLRCWS can be obtained by blending DCLR with the four coals at the weight ratio of 1:1, and the resultant slurries have high solid loading, good static stability and rheological behavior owing to their complementary advantage in properties. The blending principle and slurry properties are as follows: (1) In terms of slurryability, addition of the DCLR largely improves the maximum solid loadings of the CWS; and the higher MHC of the coal sample is, the higher the value of D1 is. Furthermore, slurryability of the blending samples has no linear additivity mainly due to the significant change of zeta potential after the blending process. Comparing with the coal sample, the improvement of absolute zeta potential of the blending sample has a positive impact on the slurryability of DCLRCWS. (2) All of the DCLRCWS have the same static stability with their corresponding CWS, which are better than that of DCLRWS. (3) The rheological behavior of the DCLRWS is markedly improved by blending coal samples with low metamorphic
degree. The values of D3 decrease with increasing H/C ratio of the coal samples, suggesting their improved flowability.
Acknowledgement This research was financially supported by Joint Foundation of Natural Science Foundation of China and Shenhua Group Corporation Ltd. (U1261209), National Basic Research Program of China (973 program, 2011CB201401) and Strategic Priority Research Program of the Chinese Academy of Sciences (Grant No. XDA07060100).
References [1] Yang J, Wang Z, Liu Z, Zhang Y. Novel use of residue from direct coal liquefaction process. Energy Fuels 2009;23(10):4717–22. [2] Cui H, Yang J, Liu Z, Bi J. Characteristics of residues from thermal and catalytic coal hydroliquefaction. Fuel 2003;82(12):1549–56. [3] Bai L, Nie Y, Li Y, Dong H, Zhang X. Protic ionic liquids extract asphaltenes from direct coal liquefaction residue at room temperature. Fuel Proc Technol 2013;108:94–100. [4] Lv D, YuChi W, Bai Z, Bai J, Li W. The slurryability of direct coal liquefaction residue-water slurry. J Fuel Chem Technol 2013;41(12):1437–44. [5] Gong K, Liu X, Sun H, Wang F. A new type material of coal–ater slurry gasification–coal liquefaction residue water slurry. Shanghai Gas 2009;3:3–6. [6] Luo J, Zheng H, Meng C, Yuan S, Xie X. Study on the pulping of coal liquefaction residues. Modern Chem Industry 2012;32(4):76–9. [7] Yu J, Tahmasebi A, Han Y, Yin F, Li X. A review on water in low rank coals: The existence, interaction with coal structure and effects on coal utilization. Fuel Proc Technol 2013;106:9–20. [8] Cheng J, Zhou J, Li Y, Liu J, Cen K. Improvement of coal water slurry property through coal physicochemical modifications by microwave irradiation and thermal heat. Energy Fuels 2008;22:2422–8. [9] Yu Y, Liu J, Cen K. Properties of coal water slurry prepared with the solid and liquid products of hydrothermal dewatering of brown coal. Ind Eng Chem Res 2014;53(11):4511–7. [10] Yu Y, Liu J, Wang R, Zhou J, Cen K. Effect of hydrothermal dewatering on the slurryability of brown coals. Energy Convers Manage 2012;57:8–12. [11] Guo Z, Feng R, Zheng Y, Fu X. Improvement in properties of coal water slurry by combined use of new additive and ultrasonic irradiation. Ultrason Sonochem 2007;14(5):583–8. [12] Das D, Dash U, Meher J, Misra PK. Improving stability of concentrated coal– water slurry using mixture of a natural and synthetic surfactants. Fuel Proc Technol 2013;113:41–51. [13] Zhu J, Zhang G, Li J, Zhao F. Synthesis, adsorption and dispersion of a dispersant based on starch for coal–water slurry. Colloids Surf A: Physicochem Eng Aspects 2013;422:165–71. [14] Xu R, He Q, Cai J, Pan Y, Shen J, Hu B. Effects of chemicals and blending petroleum coke on the properties of low-rank Indonesian coal water mixtures. Fuel Proc Technol 2008;89(3):249–53. [15] Funk JE. Coal–water slurry and methods for its preparations. USA Patent 4468232. [16] Fan Y, Hu H, Jin L, Zhu S, Zhang Q. Static stability and rheological behavior of lignite char–water mixture. Fuel 2013;104:7–13. [17] Gutierrez-Rodriguez JA, Purcell RJ, Aplan FF. Estimating the hydrophobicity of coal. Colloids Surf 1984;12(1):1–25. [18] Mishra SK, Senapati PK, Panda D. Rheological behavior of coal–water slurry. Energy Sources 2002;24(2):159–67. [19] Wang R, Liu J, Gao F, Zhou J, Cen K. The slurrying properties of slurry fuels made of petroleum coke and petrochemical sludge. Fuel Proc Technol 2012;104:57–66.
150
D. Lv et al. / Fuel 145 (2015) 143–150
[20] Chen R, Wilson M, Leong YK, Bryant P, Yang H, Zhang D. Preparation and rheology of biochar, lignite char and coal slurry fuels. Fuel 2011;90(4): 1689–95. [21] Marsalek R. The influence of surfactants on the zeta potential of coals. Energy Source Part A 2008;31(1):66–75. [22] Sun C, Li B, Yuchi W, Cao B. Adsorption characteristics of dispersant on coal surface and its resultant electro-chemical properties and their effect on rheological behaviour of coal water slurry. J Fuel Chem Technol 1996;24(4):323–8. [23] Biggs SR. Aggregate structures and solid-liquid separation processes. KONA Powder Part J 2006;24:41–53. [24] Tudor PR, Atkinson D, Crawford RJ, Mainwaring DE. The effect of adsorbed and non-adsorbed additives on the stability of coal–water suspensions. Fuel 1996;75(4):443–52.
[25] Zhu J, Liu J, Shen W, Wu J, Wang R, Zhou J, et al. Improving the slurrying ability of XiMeng brown coal by medium- to low-temperature thermal treatment. Fuel Proc Technol 2014;119:218–27. [26] Yuchi W, Li B, Li W, Chen H. Effects of coal characteristics on the properties of coal water slurry. Int J Coal Prep Util 2007;25(4):239–49. [27] Tridib T, Bhudeb RD. Flocculation: a new way to treat the waste water. J Phys Sci 2006;10:93–127. [28] Roh NS, Shin DH, Kim DC, Kim JD. Rheological behaviour of coal–water mixtures. 2. Effect of surfactants and temperature. Fuel 1995;74(9):1313–8.
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
An approach for utilization of direct coal liquefaction residue: Blending with low-rank coal to prepare slurries for gasification Dongmei Lv a,b, Wei Yuchi a, Zongqing Bai a,⇑, Jin Bai a, Lingxue Kong a, Zhenxing Guo a, Jingchong Yan a,b, Wen Li a a b
State Key Laboratory of Coal Conversion, Institute of Coal Chemistry, Chinese Academy of Sciences, Taiyuan 030001, China University of Chinese Academy of Sciences, Beijing 100049, China
h i g h l i g h t s The liquefaction residue can improve the slurryability of low-rank coals. YM and SM coal have synergistic effect with the residue in slurryability. Blending slurries with the residue and low-rank coal have strong pseudoplasticity. Static stability of blending slurries is remarkable.
a r t i c l e
i n f o
Article history: Received 15 August 2014 Received in revised form 3 December 2014 Accepted 22 December 2014 Available online 5 January 2015 Keywords: Direct coal liquefaction residue Low-rank coal Blending Slurryability
a b s t r a c t Direct coal liquefaction residue (DCLR) is a main byproduct in direct coal liquefaction process and its clean and high-efficient utilization is important. If DCLR could be gasified for hydrogen production, it will compensate hydrogen consumption during liquefaction and lower the operation cost, which makes DCLR utilization more promising. We proposed to blend DCLR with low-rank coals to prepare DCLR–coal–water slurries (DCLRCWS) as feedstock for gasification. In this work, one DCLR and four low-rank coals were used to prepare the individual DCLR–water slurries (DCLRWS), coal water slurries (CWS) and the mixed DCLRCWS at the weight ratio of 1:1. The slurryability, static stability and rheology of various slurries were investigated. The results show that DCLR and the low-rank coals are complementary in terms of slurry properties and DCLRCWS could meet the requirement of gasification process. Adding 50 wt.% DCLR apparently improves the slurryability of low-rank coals, and the maximum solid loading (Cmax) of DCLRCWS prepared with ZLNR coal and DCLR is about 10% higher than that of the corresponding CWS. The effects of coal properties, dispersant adsorption, zeta potential on preparation of highly loaded slurries were examined in terms of slurryability. Moreover, compared with DCLRWS, DCLRCWS display higher degree of pseudoplasticity and better static stability owing to the addition of low-rank coals. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Recently, interest in the development of coal hydroliquefaction has increased considerably, as driven by the surging global oil demand and increasing concerns about energy security in China. However, the high cost of coal hydroliquefaction is a main obstacle for its further development, which has brought unprecedented opportunity and challenge to research areas at the same time [1]. As a main byproduct of direct coal liquefaction process, direct coal liquefaction residue (DCLR) should be disposed and utilized reasonably. The direct coal liquefaction plants tend to gasify DCLR to ⇑ Corresponding author. Tel.: +86 351 4048967; fax: +86 351 4050320. E-mail address: [email protected] (Z. Bai). http://dx.doi.org/10.1016/j.fuel.2014.12.075 0016-2361/Ó 2014 Elsevier Ltd. All rights reserved.
supply hydrogen needed in the liquefaction process, which can realize self-sufficiency in hydrogen consumption and significantly increase the economy of the process [2]. Generally, DCLR accounts for 20–30% of the coal consumed in the process [3], but it is still not enough for the operational capacity of a single gasifier and the stability and running period of the liquefaction process should be considered as well. Thus, it is urgent to find suitable raw materials to co-gasify with DCLR for the normal operation of the whole system. Usually, DCLR is separated by reduced pressure distillation, which contains considerable heavy oil and asphaltenes. Accordingly, dry feeding gasification is not suitable for DCLR due to its low softening point (about 180 °C), leading to great difficulty in milling system, while wet feeding could be operated easily. The
144
D. Lv et al. / Fuel 145 (2015) 143–150
results of our previous work [4] indicate that the slurryability of DCLR is relatively high, its static stability determined by rod dropping method is 2 days, and the shear-thinning behavior of direct coal liquefaction residue-water slurry (DCLRWS) is not strong, which fairly accords with other reported results [5,6]. Low-rank coals account for nearly half of the coal reserve worldwide and have an advantage of low price, regarded as a promising raw material for preparation of coal water slurry (CWS) [7]. But it is assumed that low-rank coals, especially brown coal, are difficult to make highly loaded CWS. Over the years, various measures have been taken to improve the slurryability of lowrank coals by thermal treatment [8], hydrothermal dewatering process [9,10], microwave/ultrasonic irradiation [8,11], developing high-efficient dispersant [12,13] and blending with petroleum coke [14], etc. Blending DCLR with low-rank coals to prepare DCLRCWS may be an energy-efficient and environmental approach, which could not only supply sufficient feedstock to the gasification process but also have the potential to enhance the slurryability of low-rank coals. However, the studies on the blending principles of DCLR and low-rank coal as well as the properties of DCLRCWS have not been reported yet. In this work, four low-rank coals were selected to prepare CWS, whose properties were compared with that of DCLRWS. DCLR was blended with the coal samples to prepare DCLRCWS at the weight ratio of 1:1 and the slurryability, rheology and static stability of DCLRCWS were systematically investigated. The surface physical and chemical properties of DCLR and coal samples, dispersant adsorption isotherms, zeta potential, and SEM pictures of slurries were analyzed to study the interactions between DCLR and coal samples in terms of slurry properties. The blending principles of DCLRCWS can provide experimental data and theoretical basis for the integration of coal liquefaction and gasification process.
graded on Alfred particle size distribution [15]. An anionic dispersant, naphtalenesulfonate-formaldehyde condensate (NSF), was used and the dosage of dispersant is 1 wt.% on dry basis of coal and/or DCLR. For DCLRCWS, certain amount of DCLR and coal sample at the weight ratio of 1:1 was successively transferred into the bottle containing a predetermined quantity of dispersant and deionized water. And then, the mixture was agitated at 3000 r/ min for 10 min (CWS), while for 20 min (DCLRWS and DCLRCWS) for homogenization. All experiments were conducted at room temperature. 2.3. Determination of slurry properties The rheology of slurries was determined by NXS-11 rotational viscometer (Chengdu Analytical Instrument Factory, China) at the shear rate range of 10–100 s1 and the average of 6 values at shear rate of 100 s1 was defined as the apparent viscosity. The maximum solid loading of slurry was designated as the solid loading when the apparent viscosity of slurries was 1000 mPa s at 20 °C [10]. The maximum solid loading of DCLRWS, CWS and DCLRCWS is denoted as Cmax,1, Cmax,2, Cmax,3, respectively. The rod dropping method was used to evaluate the static stability of slurries with the apparent viscosity close to 1000 mPa s once a day till the appearance of hard sediment up to 14 days. 2.4. Characterization 2.4.1. Physical adsorption test The texture properties analyses of the samples were carried out with isothermal N2 adsorption at 77 K using ASAP 2020 – Physisorption Analyzer (Miromeritics, USA). All the samples were dried at 60 °C under vacuum for 24 h before physical adsorption test. 2.4.2. Chemical structure analysis The samples with Alfred particle size distribution were ground to an appropriate size and dried at 60 °C under vacuum for 24 h before FTIR measurement. FTIR spectra of the samples were obtained on an IR spectrometer (VERTEX 70, Bruker, Germany), and the cell was the 0030–102 accessory (Pike Co. Ltd., USA) with ZnSe windows. The same Al2O3 crucible was used in measuring the five samples, so the thickness of the samples is also the same with the height of the crucible. The diffuse reflection mode was selected to record the spectra within 4000–600 cm1 with the co-addition of 200 scans at 4 cm1 resolution. Chemical titration was conducted to determine the content of hydrophilic function groups and the procedure was presented by Fan et al. [16].
2. Experimental 2.1. Materials DCLR used in this work was obtained from 6 t/d direct coal liquefaction pilot plant of Shenhua Group Corporation in China. Its raw coal is Shangwan coal from Inner Mongolia. Four low-rank coal samples from Shenmu, Yima, Yuxian and Zhalainuoer were employed and denoted as SM, YM, YX and ZLNR coal, respectively. The proximate and ultimate analyses of DCLR and four coal samples are listed in Table 1. 2.2. Preparation of slurries
2.4.3. Dispersant adsorption measurement During dispersant adsorption tests, the mixture of graded samples and dispersant solution at weight ratio of 1:10 was charged into a 100 ml conical flask and shaken for 5 h at 200 rpm to make
The samples were separately crushed and sieved to several groups with different particle sizes: 280–154 lm, 157–74 lm and less than 74 lm. DCLR and four coal samples were separately
Table 1 Proximate and ultimate analyses of five samples. Proximate analysis (wt.%, ad)
DCLR YM SM YX ZLNR
Ultimate analysis (wt.%, daf)
M
A
V
C
H
N
S
O⁄
0.17 4.37 2.68 13.00 12.07
12.46 38.69 10.47 14.23 12.39
45.12 23.34 29.73 23.83 30.19
90.26 73.46 81.05 80.76 75.32
5.99 5.78 5.24 5.58 6.50
1.25 1.05 1.08 0.99 1.05
1.77 1.07 0.30 1.02 0.30
0.73 18.64 12.33 11.65 16.83
ad: air dried basis; daf: dry and ash-free basis; ⁄ By difference. M: moisture; A: ash; V: volatile; C: carbon; H: hydrogen; N: nitrogen; S: sulfur; O: oxygen. H/C: the ratio of hydrogen atom to carbon atom. MHC: the maximum moisture holding capacity.
H/C
MHC
0.80 0.94 0.78 0.83 1.04
1.0 7.5 8.5 16.6 24.7
145
D. Lv et al. / Fuel 145 (2015) 143–150
the particles to reach adsorption equilibrium. Then the suspension was filtered under reduced pressure and centrifuged to obtain the supernatant which was transferred to determine the concentration of dispersant by Shimadzu UV-3600 spectrophotometer at 292 nm. The initial concentrations of the dispersant solution are 0, 50, 100, 200, 400, 600, 800, 1000, 1200 mg/L for DCLR and the four coal samples. The dispersant adsorption was calculated from the difference between the dispersant concentration before and after adsorption, and the adsorption isotherms of two measurements can almost overlap. 2.4.4. SEM A high resolution field emission scanning electron microscope (FESEM, JSM-7001F, Japan) was employed to investigate the packing efficiency of different slurries. 2.4.5. Zeta potential 50 ml dispersant solution (0.02 g/L) and 0.1 g graded coal samples, DCLR or blended samples were put into a 100 ml conical flask and shaken for 5 h at 200 rpm, so the particles could reach adsorption equilibrium. Then the zeta potential of the suspension was measured by JS94HM Zetasizer (POWEREACH, China) under natural pH (about 6.5). Average values of duplicated tests were used and the measurement error was within ±1%. 3. Results and discussion 3.1. Chemical and physical characteristics of DCLR and four coal samples Table 1 shows that the DCLR contains less O than the coals, attributing to the removal of oxygen during hydroliquefaction. On the other hand, DCLR contains more S derived from the spent catalyst and the raw coals. In addition, it is notable that DCLR has high content of C and H, which makes it potential to be used as feedstock for gasification. The results of physical adsorption analysis are shown in Table 2. The BET surface area and pore volume of DCLR are the smallest because of the collapsing and plugging of pore structure in the raw coal during liquefaction. However, the four coal samples have more developed pore structure. Fig. 1 illustrates the infrared spectra of the five samples. The characteristic peaks including 3400 cm1 due to OAH, 2800– 3000 cm1 due to CAH stretching in methyl and methylene, 1720 cm1 due to C@O stretching vibration, 1400–1600 cm1 due to the stretching vibration of aromatic rings and 1000– 1200 cm1 due to CAOAC are found for the five samples. Compared with the coal samples, the peak strength of the oxygen-containing functional groups for DCLR are much weaker. The quantitative analysis of the functional groups was done by chemical analysis method. The results, shown in Table 2, indicate that the total acidity content of DCLR is apparently lower than that of the coal samples, which is consistent with the infrared spectra results. Compared with YM and SM coal, YX and ZLNR coal have higher content of oxygen-containing functional groups.
Fig. 1. The FTIR spectra of DCLR and four coal samples.
3.2. Adsorption behavior of dispersant on the samples Fig. 2 illustrates the adsorption isotherms of NSF dispersant on the five samples. The equilibrium adsorption quantity (U1) obtained from Langmuir isotherm curve-fitting are 2.16, 2.04, 0.90, 0.83 and 0.80 mg/g for SM, YM, DCLR, YX and ZLNR coal, respectively. Langmuir adsorption model is as follows:
C=C1 ¼ Kc=ð1 þ KcÞ:
ð1Þ
where, U is the adsorption quantity (mg/g); c is the equilibrium concentration (mg/L); K is the Langmuir equilibrium constant and stated adsorption capacity of dispersants on the coal. The equilibrium adsorption quantity of YM and SM coal is apparently higher than that of other samples, and the different adsorption behavior with the NSF dispersant is mainly related to the pore structure and the oxygen-containing functional groups content of the samples. Table 2 indicates that YM and SM coal have more developed pore structure which will supply more adsorption sites for the dispersant, thus resulting in higher dispersant adsorption capacity. The BET specific surface area and pore volume of YX and ZLNR coal are much smaller, thus resulting in fewer adsorption sites. Besides, the higher contents of oxygen-containing functional groups on the surface of YX and ZLNR coal further reduce the amount of adsorption sites [17], thus their equilibrium adsorption quantity is relatively low. The low porosity of DCLR is the main reason for its low dispersant adsorption. 3.3. Properties of CWS and DCLRCWS 3.3.1. Slurryability The apparent viscosity of slurries increases with increasing solid loading, as shown in Fig. 3. This is because the increased solid loading enhances interparticle frictional forces and the water available outside the particles decreases, and thus the sharply increased resistence of shear improves the apparent viscosity of slurries [18].
Table 2 BET specific surface area, pore volume and oxygen-containing functional groups in the five samples. Samples
Specific surface area (m2/g)
Pore volume (102 cm3/g)
Total acidity (mmol/g)
ACOOH (mmol/g)
ArAOH (mmol/g)
DCLR YM SM YX ZLNR
0.66 7.81 7.83 2.09 2.32
0.45 1.87 1.78 1.15 1.33
0.09 1.66 0.78 2.78 3.78
0.06 0.10 0.06 0.23 0.36
0.03 1.56 0.72 2.55 3.42
146
D. Lv et al. / Fuel 145 (2015) 143–150
Fig. 2. Adsorption isotherms of NSF dispersant on DCLR and four coal samples.
Fig. 3. Effect of solid loadings on the apparent viscosity of CWS, DCLRCWS and DCLRWS.
Table 3 The slurryability of different slurries. Samples
Cmax,2 (wt.%)
Cmax,3 (wt.%)
Cwa (wt.%)
D1 (wt.%)
D2 (wt.%)
YM SM YX ZLNR
65.0 63.0 56.2 51.7
70.1 69.5 64.2 61.4
69.3 68.3 64.9 62.6
5.1 6.5 8.0 9.7
0.8 1.2 0.7 1.2
Cmax,1 = 73.6; Cwa = 0.5 (Cmax,1 + Cmax,2); D1 = Cmax,3–Cmax,2; D2 = Cmax,3–Cwa.
Table 3 shows the slurryability of the slurries. Cmax is an important index to appraise the slurryability of slurries. A higher Cmax means a higher slurryability of slurries. As shown in Table 3, Cmax,1 (73.5%) of DCLRWS is apparently higher than that of the four CWS, and even 20% higher than Cmax,2 of CWS prepared with ZLNR coal. The reason is that the absorbing water capacity of DCLR is much lower due to its special properties, such as higher carbon content, less oxygen-containing functional group, underdeveloped pore structure and lower MHC. Consequently, more free water existing in DCLRWS acts as lubricant and improves its slurryability. The solid loading of the slurry fuels should be enhanced maximally to reach high heat value and to ensure efficient gasification and combustion [19]. However, it is difficult for low-rank coals to prepare highly loaded CWS and the maximum solid loadings of four coals are generally low. Therefore, it is necessary to make more efforts to improve their slurryability.
The four DCLRCWS prepared with the blending samples of DCLR and coal samples at the weight ratio of 1:1 demonstrate superior slurryability to their corresponding CWS, and all their solid loadings reach 60% meeting the requirement of gasification. In Table 3, the values of D1 are determined from the difference between Cmax,3 and Cmax,2. The maximum solid loading of CWS prepared with ZLNR coal is 51.7%, while that of corresponding DCLRCWS increases to 61.4%, and the value of D1 is the largest about 10%. Compared with the coal samples, the higher slurryability of the blending samples is mainly attributed to the positive effect of DCLR and the interaction between DCLR and coal sample. The water absorbing capacity of DCLR is low. Hence, the addition of DCLR could reduce the water absorbing capacity of the blending samples. Moreover, the representative SEM pictures (Fig. 4) illustrate that the packing efficiency of the blending slurry (YX-DCLR) and DCLR is higher than that of the corresponding CWS. This is because higher contents of hydrophilic groups in coal samples easily lead to forming loose aggregates of flocs preventing particles closing to each other. On the other hand, more hydrophobic DCLR particles enter into the loose structure and fill the voids among the coal particles, thus the blending particles could accumulate more tightly, and then resulting in higher packing efficiency. High packing efficiency means lower interparticle voidage and less free water required to fill the inter-particle voids, which is favorable for improving slurryability [20]. Based on the above considerations, the water absorbing capacity and the amount of free water decrease after blending DCLR into coal samples, which accounts for the high slurryability of DCLRCWS. Cwa is the weighted average of Cmax, 1 and Cmax, 2 at the weight ratio of 1:1, and D2 is equal to the difference of Cmax,3 and Cwa. The values of D2 always deviate from zero, suggesting that slurryability of the blending samples has no linear additivity. Besides, the values of D2 for YM-DCLR and SM-DCLR are positive, while those for YXDCLR and ZLNR-DCLR are negative. This is thought to be mainly related to the dispersant adsorption behavior and zeta potential of the samples, which will be discussed in the following section. The dispersant adsorption on the four blending samples (termed as Q0) is investigated when the dispersant concentration (1000 mg/L) is equivalent to the dispersant dosage in slurries. Qwa is the weighted average of the dispersant adsorption on DCLR and coal alone at the dispersant concentration of 1000 mg/L. Fig. 5 shows that the values of Qwa and Q0 are basically the same within the measurement uncertainties, which indicates that there is little interaction between DCLR and coal samples in the aspect of dispersant adsorption. The surface of coal is usually negatively charged with minus zeta potential. The zeta potential of particles in slurries is related with many factors, such as the nature of coal, pH value, electrolyte concentration, surface oxygen-containing functional group and dispersant adsorption [21]. Fig. 6 compares the zeta potential of the individual samples with that of the blending samples. The absolute zeta potential of DCLR is the lowest, which is mainly attributed to its fewer oxygen-containing functional groups and lower dispersant adsorption. Blending samples of YM-DCLR and SM-DCLR have higher absolute zeta potential than the corresponding coal samples, YM and SM coal, while the absolute zeta potential of blending samples YX-DCLR and ZLNR-DCLR is lower than that of YX and ZLNR coal, respectively. The maximum solid loading of slurries is directly proportional to the product of the equilibrium adsorption quantity and the absolute zeta potential [22]. Based on the analysis above, the addition of DCLR has little influence on the dispersant adsorption, while apparently affects the zeta potential which plays the dominant role in slurryability of the blending samples. If the absolute zeta potential of the blending sample is higher than that of the corresponding coal, DCLR and the coal sample have synergistic effect on the
D. Lv et al. / Fuel 145 (2015) 143–150
147
Fig. 4. The SEM photographs of CWS (YX), DCLRCWS (YX-DCLR) and DCLRWS.
Fig. 5. Comparison between dispersant adsorption on blending samples (Q0) and weighted average of dispersant adsorption on DCLR and coal sample (Qwa).
maximum solid loading of DCLRCWS, which is consistent with the positive value of D2. On the contrary, there is anti-synergistic effect, which is consistent with the negative value of D2. Consequently, the values of D2 are positive for YM-DCLR and SM-DCLR, while negative for YX-DCLR and ZLNR-DCLR. 3.3.2. Static stability The static stability is an important property for storage of the slurries. As a coarse disperse system, the solid–liquid separation of slurries might be originally expected to occur easily. This is because the DCLR/coal particles, with most regions of surface being hydrophobic, tend to form large multi-particle units that can settle rapidly facilitating separation of the phases [23]. Dispersant could reduce the tendency of sediment by adsorbing on the surface of particles. The particles adsorbing certain amounts of dispersant
Fig. 6. The zeta potential of DCLR, coal samples and blending samples.
could interact with each other by dispersant bridges and eventually form a three-dimensional network structure in slurries. Usually, the structure of a dispersant molecular has one hydrophobic tail and one hydrophilic tail. The hydrophobic tails of dispersant adsorb on the hydrophobic regions of coal particles, and meanwhile, hydrophilic tails could form interpolymer complexes with surface carboxylic acid or the hydrophilic tails of other dispersant mainly through hydrogen bonding. Thus dispersant bridges form among the particles and construct a three-dimensional network structure in slurries [24]. The structure could prevent the particles from gravity sedimentation and enhance the stabilization of slurries [25]. In addition, the type and dosage of dispersant, particle size distribution of samples, the concentration of electrolyte and the physiochemical properties of the raw materials also have very important effects on the static stability of the slurries [25,26].
148
D. Lv et al. / Fuel 145 (2015) 143–150
The static stability of DCLRWS is not satisfying and hard sediment appears within two days, which cannot afford long-distance transport and long-time storage. However, the static stability of the four CWS is remarkable. The CWS prepared with YM, YX, ZLNR coal can maintain homogeneous for more than 14 days, and hard sediment appears within four days for the CWS prepared with SM coal. Furthermore, all of the DCLRCWS have the same static stability with their corresponding CWS, which are better than that of DCLRWS. The difference of the static stability between the DCLRWS and the four CWS is mainly ascribed to the different surface properties of DCLR and coals. The dispersant adsorption and oxygen-containing functional groups of DCLR is so low that dispersant bridges among the particles are hard to form, thus there is few chances for the formation of three-dimensional network structure, which is unfavorable for improving the stability of DCLRWS. On the other hand, since the dispersant could improve the hydrophilicity of DCLR particles, the hydrophobicity of their surface is still strong leading to serious aggregation in the aqueous medium and formation of hard sediment soon. Compared with DCLR, the four lowrank coals have apparently higher dispersant adsorption (for YM and SM coal) or distinctly higher oxygen-containing functional groups (for YX and ZLNR coal), which facilitate the formation of three-dimensional network structure accounting for the enhanced static stability of the four CWS. In addition, the zeta potential of particles displays a substantial correlation with the static stability of slurries. Higher the absolute zeta potential is, more likely the slurries are to remain in stable state [27]. Therefore, the better stabilization of the four CWS could also be attributed to the higher absolute zeta potential, which could provide stronger electrostatic
repulsive forces among the coal particles and enhance the dispersion of the coal particles. The improvement in the static stability, when 50% weight percent of DCLR particles is replaced by the low-rank coal particles, might be correlated with the formation of three-dimensional network structure in the four DCLRCWS. The respective dispersant adsorption of DCLR and coals almost has no difference before and after the blending process. Thus the coal particles could still exhibit close connection with each other and even with DCLR particles through dispersant bridges in DCLRCWS, and a three-dimensional network structure could form to stabilize the slurries. Additionally, the coal particles distribute around DCLR particles in the DCLRCWS, which could weaken the hydrophobic attractive forces among DCLR particles and prevent adverse aggregation, thus resulting in enhanced static stability of the blending slurries. Moreover, the improvement in the static stability of DCLRCWS is also ascribed to the higher absolute zeta potential of blending samples compared with that of DCLR (seen in Fig. 6). 3.3.3. Rheological properties The rheological behavior of highly loaded slurries is important for their preparation, storage, transportation and atomization [28]. The effects of shear rate on viscosity and shear stress of DCLRWS, CWS and DCLRCWS are depicted in Fig. 7. The shear stress/shear rate dependence fits well with the Herschel–Bulkley model shown as follows:
s ¼ sy þ K c n
ð2Þ
where, s and sy are shear stress and yield shear stress, respectively; K is the rheological constant and n is the rheological exponential.
Fig. 7. The effect of shear rate on viscosity (a and c) and shear stress (b and d) of DCLRWS, CWS and DCLRCWS.
149
D. Lv et al. / Fuel 145 (2015) 143–150 Table 4 The rheology of DCLRWS, CWS and DCLRCWS. Samples
Solid loading (wt.%)
Rheological exponential, n
Correlation coefficient, R2
D3
DCLR YM SM YX ZLNR YM-DCLR SM-DCLR YX-DCLR ZLNR-DCLR
73.5 65.0 63.0 56.2 51.7 70.4 69.7 64.3 61.5
0.86 0.57 0.42 0.71 0.62 0.69 0.83 0.73 0.68
0.9972 0.9990 0.9968 0.9938 0.9986 0.9972 0.9961 0.9971 0.9960
– – – – – 0.17 0.03 0.13 0.18
D3 = nDCLRWS–nDCLRCWS.
For a Newtonian fluid, n = 1; for non-Newtonian fluids, n < 1 (pseudoplastic) or n > 1 (dilatant). When n < 1, the lower the n is, the stronger the shear-thinning/pseudoplastic behavior of the slurries is. The corresponding parameters of different slurries are listed in Table 4. Fig. 7 demonstrates that all types of slurries are pseudoplastic fluids, with the typical shear-thinning behavior that viscosity decreases with increasing shear rate, which is favorable for transportation of slurries in gasification. The rheological exponentials of the slurries are all less than 1, which is corresponding to the shearthinning behavior. Table 4 shows the rheological exponential of DCLRWS is the largest among the five individual slurries, suggesting that the shear-thinning behavior of DCLRWS is weaker than that of the four CWS. The rheological exponential of the DCLRWS is 0.86 which is higher than those of all the DCLRCWS, indicating that the pseudoplastic behavior of the blending slurries are strengthened after adding the low-rank coals. Moreover, the four coal samples have different influence on the rheological behavior of DCLRCWS, which may be attributed to the characteristics of the coal samples. The values of D3 are derived from the difference between the rheological exponential of DCLRWS and DCLRCWS, which reflects the different influence of the four coal samples. ZLNR coal has the highest efficiency in improving the flowability of DCLRCWS, decreasing n from 0.86 to 0.68 and other three coals, YM, YX, SM coal have decreasing efficiency in turn. Note that the value of D3 has good relationship with H/C ratio of coal samples based on ultimate analysis, that is, the larger the H/C ratio is, the higher the value of D3 is. With the highest H/C ratio, ZLNR coal has the highest value of D3, corresponding to the strongest pseudoplastic behavior of DCLRCWS.
4. Conclusion Based on the results and discussions mentioned above, satisfying DCLRCWS can be obtained by blending DCLR with the four coals at the weight ratio of 1:1, and the resultant slurries have high solid loading, good static stability and rheological behavior owing to their complementary advantage in properties. The blending principle and slurry properties are as follows: (1) In terms of slurryability, addition of the DCLR largely improves the maximum solid loadings of the CWS; and the higher MHC of the coal sample is, the higher the value of D1 is. Furthermore, slurryability of the blending samples has no linear additivity mainly due to the significant change of zeta potential after the blending process. Comparing with the coal sample, the improvement of absolute zeta potential of the blending sample has a positive impact on the slurryability of DCLRCWS. (2) All of the DCLRCWS have the same static stability with their corresponding CWS, which are better than that of DCLRWS. (3) The rheological behavior of the DCLRWS is markedly improved by blending coal samples with low metamorphic
degree. The values of D3 decrease with increasing H/C ratio of the coal samples, suggesting their improved flowability.
Acknowledgement This research was financially supported by Joint Foundation of Natural Science Foundation of China and Shenhua Group Corporation Ltd. (U1261209), National Basic Research Program of China (973 program, 2011CB201401) and Strategic Priority Research Program of the Chinese Academy of Sciences (Grant No. XDA07060100).
References [1] Yang J, Wang Z, Liu Z, Zhang Y. Novel use of residue from direct coal liquefaction process. Energy Fuels 2009;23(10):4717–22. [2] Cui H, Yang J, Liu Z, Bi J. Characteristics of residues from thermal and catalytic coal hydroliquefaction. Fuel 2003;82(12):1549–56. [3] Bai L, Nie Y, Li Y, Dong H, Zhang X. Protic ionic liquids extract asphaltenes from direct coal liquefaction residue at room temperature. Fuel Proc Technol 2013;108:94–100. [4] Lv D, YuChi W, Bai Z, Bai J, Li W. The slurryability of direct coal liquefaction residue-water slurry. J Fuel Chem Technol 2013;41(12):1437–44. [5] Gong K, Liu X, Sun H, Wang F. A new type material of coal–ater slurry gasification–coal liquefaction residue water slurry. Shanghai Gas 2009;3:3–6. [6] Luo J, Zheng H, Meng C, Yuan S, Xie X. Study on the pulping of coal liquefaction residues. Modern Chem Industry 2012;32(4):76–9. [7] Yu J, Tahmasebi A, Han Y, Yin F, Li X. A review on water in low rank coals: The existence, interaction with coal structure and effects on coal utilization. Fuel Proc Technol 2013;106:9–20. [8] Cheng J, Zhou J, Li Y, Liu J, Cen K. Improvement of coal water slurry property through coal physicochemical modifications by microwave irradiation and thermal heat. Energy Fuels 2008;22:2422–8. [9] Yu Y, Liu J, Cen K. Properties of coal water slurry prepared with the solid and liquid products of hydrothermal dewatering of brown coal. Ind Eng Chem Res 2014;53(11):4511–7. [10] Yu Y, Liu J, Wang R, Zhou J, Cen K. Effect of hydrothermal dewatering on the slurryability of brown coals. Energy Convers Manage 2012;57:8–12. [11] Guo Z, Feng R, Zheng Y, Fu X. Improvement in properties of coal water slurry by combined use of new additive and ultrasonic irradiation. Ultrason Sonochem 2007;14(5):583–8. [12] Das D, Dash U, Meher J, Misra PK. Improving stability of concentrated coal– water slurry using mixture of a natural and synthetic surfactants. Fuel Proc Technol 2013;113:41–51. [13] Zhu J, Zhang G, Li J, Zhao F. Synthesis, adsorption and dispersion of a dispersant based on starch for coal–water slurry. Colloids Surf A: Physicochem Eng Aspects 2013;422:165–71. [14] Xu R, He Q, Cai J, Pan Y, Shen J, Hu B. Effects of chemicals and blending petroleum coke on the properties of low-rank Indonesian coal water mixtures. Fuel Proc Technol 2008;89(3):249–53. [15] Funk JE. Coal–water slurry and methods for its preparations. USA Patent 4468232. [16] Fan Y, Hu H, Jin L, Zhu S, Zhang Q. Static stability and rheological behavior of lignite char–water mixture. Fuel 2013;104:7–13. [17] Gutierrez-Rodriguez JA, Purcell RJ, Aplan FF. Estimating the hydrophobicity of coal. Colloids Surf 1984;12(1):1–25. [18] Mishra SK, Senapati PK, Panda D. Rheological behavior of coal–water slurry. Energy Sources 2002;24(2):159–67. [19] Wang R, Liu J, Gao F, Zhou J, Cen K. The slurrying properties of slurry fuels made of petroleum coke and petrochemical sludge. Fuel Proc Technol 2012;104:57–66.
150
D. Lv et al. / Fuel 145 (2015) 143–150
[20] Chen R, Wilson M, Leong YK, Bryant P, Yang H, Zhang D. Preparation and rheology of biochar, lignite char and coal slurry fuels. Fuel 2011;90(4): 1689–95. [21] Marsalek R. The influence of surfactants on the zeta potential of coals. Energy Source Part A 2008;31(1):66–75. [22] Sun C, Li B, Yuchi W, Cao B. Adsorption characteristics of dispersant on coal surface and its resultant electro-chemical properties and their effect on rheological behaviour of coal water slurry. J Fuel Chem Technol 1996;24(4):323–8. [23] Biggs SR. Aggregate structures and solid-liquid separation processes. KONA Powder Part J 2006;24:41–53. [24] Tudor PR, Atkinson D, Crawford RJ, Mainwaring DE. The effect of adsorbed and non-adsorbed additives on the stability of coal–water suspensions. Fuel 1996;75(4):443–52.
[25] Zhu J, Liu J, Shen W, Wu J, Wang R, Zhou J, et al. Improving the slurrying ability of XiMeng brown coal by medium- to low-temperature thermal treatment. Fuel Proc Technol 2014;119:218–27. [26] Yuchi W, Li B, Li W, Chen H. Effects of coal characteristics on the properties of coal water slurry. Int J Coal Prep Util 2007;25(4):239–49. [27] Tridib T, Bhudeb RD. Flocculation: a new way to treat the waste water. J Phys Sci 2006;10:93–127. [28] Roh NS, Shin DH, Kim DC, Kim JD. Rheological behaviour of coal–water mixtures. 2. Effect of surfactants and temperature. Fuel 1995;74(9):1313–8.