Contents lists available at ScienceDirect

Fuel

journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Pyrolysis of coal hydroliquefaction residue in a dual loop reaction system

Chao Wang, Shaoping Xu⁎, Guangyong Wang, Yahui XiaoState Key Laboratory of Fine Chemicals, Institute of Coal Chemical Engineering, School of Chemical Engineering, Dalian University of Technology, No. 2 Linggong Road,Dalian 116024, China

G R A P H I C A L A B S T R A C T

A R T I C L E I N F O

Keywords:Coal hydroliquefaction residuePyrolysisFluidized bedCaking

A B S T R A C T

To utilize the coal hydroliquefaction residue (CHR), a dual loop reaction system (DLRS) has been developed andupon which the pyrolysis of CHR was conducted under atmospheric pressure. The DLRS combines three reactors,i.e. a fluidized bed pyrolyzer, a radial-flow moving bed filter and a riser combustor followed by a particlesgrading cyclone. These three reactors form two parallel circulating loops, i.e. pyrolysis loop and filtration loopwith small and big quartz sand particles as circulating bed materials, respectively. CHR mixed with quartz sandwas injected into the fluidized bed pyrolyzer rapidly by a gas driven feeder, which makes it possible to disposethe CHR continuously. Under optimized blending ratio of CHR/quartz sand (1:4) and pyrolysis temperature from500 °C to 550 °C, a dust-free pyrolysis tar with the yield of 20 wt% could be obtained, which was about two timesof the Fischer Assay yield. The hexane soluble part (HS) and the hexane insoluble but toluene soluble part(asphaltene, A) account for nearly 94wt% of the tar. Almost all the HS in the CHR was transferred into the tarduring the pyrolysis.

1. Introduction

Coal-to-liquid technology has been attractive to many countrieswith local supply shortage of oil but abundant coal. With the rapidlygrowing demand for transportation fuels and the increasing concernsabout energy security, China has made great effort on the developmentof coal hydroliquefaction. China Shenhua’s Direct Coal Liquefaction

Project is operating the largest demonstration plant of coal hydro-liquefaction in the world [1,2]. The coal hydroliquefaction residue(CHR) as the main by-product discharged from the vacuum distillationunit of this plant accounts for 1/3 of coal feed. It contains high boilingpoint liquefied oils, asphaltenes, unreacted coal, mineral matter andliquefaction catalysts [3]. The oils and asphaltenes composed mainly ofpolycyclic aromatic hydrocarbons account for nearly half of the CHR, so

http://dx.doi.org/10.1016/j.fuel.2017.10.078Received 9 July 2017; Received in revised form 12 October 2017; Accepted 16 October 2017

⁎ Corresponding author.E-mail address: spxu@dlut.edu.cn (S. Xu).

Fuel 212 (2018) 448–454

0016-2361/ © 2017 Published by Elsevier Ltd.

MARK

both economic and environmental considerations will drive the re-covery and utilization of CHR [4].

CHR could be used through combustion, gasification and pyrolysisand so on [5–15]. Though CHR is of high calorific value, it is not sui-table to be burnt to generate heat and electricity due to its high sulfurcontent. The high content of the spent catalysts and the minerals-richunreacted coal makes the CHR an unsatisfactory feedstock for gasifi-cation. Pyrolysis, instead, could pick up the high-value oils and as-phaltenes and would be a favorite route for CHR conversion. Researchon the pyrolysis of CHR focused on the basic pyrolysis characteristics ofthe CHR itself by general method such as TGA and fixed bed reactor[16]. The co-pyrolysis of CHR and lignite in a fixed bed reactor wasmade to improve the tar yield [17], but high blending ratio of CHR tolignite would hinder the release of volatiles. The co-pyrolysis of CHRand lignite in a rotary kiln was investigated to understand the in-hibitory effect on lignite pulverization [18]. Due to low softening pointand high caking property, CHR melted and captured the fine particles toreinforce the structure of particles during the co-pyrolysis process.However, bonding on the furnace wall cannot be avoided which mayaffect the normal operation of the reactor.

To pyrolyze CHR continuously, dual fluidized bed (DFB) mightprovide a favorable condition because the fluidization with the circu-lating bed materials would alleviate the caking of CHR. DFB has beenapplied in pyrolysis of variety of fuels [19–23] except for CHR. Thefeeding of CHR into DFB could be a key problem raised by its lowsoftening point and high caking property. In addition, the fine particlesderived from minerals-rich unreacted coal and spent catalysts entrainedin pyrogas during pyrolysis would cause the additional operations ofdust separation and purification of the pyrolysis product. For this issue,moving bed filtration could be one of the most promising technologies,which has been applied in some advanced coal conversion technologies[24–27].

In this study, a dual loop reaction system (DLRS) coupled withfluidized bed pyrolysis and moving bed filtration has been developedand successfully applied to the pyrolysis of CHR. By blending CHR withquartz sand before being fed into the pyrolyzer with a gas driven feeder,the phenomenon of caking to bulks could be alleviated, thus making itpossible to operate the reaction system continuously and steadily. Theeffects of pyrolysis temperature and the blending ratio of CHR/quartzsand have been studied. Moreover, the pyrolysis tar of CHR has beenanalyzed.

2. Experimental section

2.1. The samples

The CHR used was provided by Shanghai Research Institute of ChinaShenhua Coal to Liquid and Chemical Corporation. The proximateanalysis and Fischer Assay analysis of the sample were conducted ac-cording to Chinese standard GB/T 212-2008 and GB/T 480-2010, re-spectively. Ultimate analysis was made by a vario EL elemental ana-lyzer. The thermal gravimetric analysis (TGA) was performed using aDTU-2B thermal balance.

The solvents extraction of CHR was carried out in a Soxhlet ex-tractor by GB/T 30044-2013 standard, in which the hexane soluble part(defined as hexane soluble, HS), the hexane insoluble but toluene so-luble part (asphaltene, A), the toluene insoluble but tetrahydrofuran(THF) soluble part (preasphaltene, PA) and the THF insoluble part(THFIS) were obtained.

The caking property of CHR was measured by the following method.The sample of CHR ground and sieved to 60–80 mesh, 0.5 g was mixedwith quartz sand of the same particle size (CHR/quartz sand massblending ratio: 1:0–1:10) in a crucible. The crucible with cap was thenput in a muffle furnace at temperature of 900 °C ± 5 °C for 7min toremove the volatile matter. After cooling down to room temperature,the residual char cake was dropped from a fixed height of 30 cm and the

fragmentized sample was sieved using a 20-mesh sieve. The massfraction of those particles larger than 850 μm (20 mesh) is a measure ofthe caking property of CHR.

2.2. Apparatus and test procedure

The principle of DLRS is shown in Fig. 1. It is composed of threereactors, i.e. a fluidized bed pyrolyzer where pyrolysis of solid rawmaterial takes place, a radial-flow moving particulate bed filter wherefine particles in pyrogas from the pyrolyzer is filtered, and a risercombustor where the pyrolysis char is burned with air. These threereactors form two parallel loops, i.e. the combustor-pyrolyzer loop,where small quartz sands as heat carrier particles (S-HCPs) are circu-lating, and the combustor-filter loop, where big quartz sands as heatcarrier and dust capturing particles (B-HCPs) are circulating. The twoparallel loops are connected by the riser combustor followed by aparticles grading cyclone (PGC). Through the PGC, the mixed sizesparticles are segmented into two different size ranges using the gradinggas. A scrubber unit after cyclone is set up to recover the entrained dustand the sulfur dioxide from the flue gas of char burning.

Fig. 2 presents the schematic diagram of the lab-scale experimentalfacility of DLRS. The pyrolyzer is a fluidized bed reactor which con-sists of a lower zone of 56 mm i.d. and 80mm height and an upperzone of 98 mm i.d. and 190 mm height. The feeder of the pyrolyzerconsists of a screw feeder combined with a gas driven feeder. The gasdriven feeder is a concentric double layer tube with an inner tube of4 mm i.d. and 5mm o.d. and an outer tube of 8 mm i.d. and 10mmo.d. The inner tube is for CHR feeding with N2 as carrier gas. Theannular space between the inner tube and the outer tube is filled withroom-temperature deionized water as cooling agent to keep the CHRfrom softening during feeding. The moving particulate bed filter is agas–solid radial cross flow moving bed reactor which has an annularbed of 28 mm i.d., 100 mm o.d. and 250mm height. The riser com-bustor is a fast fluidized bed reactor of 20 mm i.d. and 2.6 m height.All the reactors are made of SUS 310S stainless steel and externallyheated by electrical furnaces. The operating parameters includingtemperature, reactor pressure and differential pressure between re-actors and gas flow rate were monitored in a smart touch monitor andrecorded by a data acquisition computer.

For each experiment, about 3.3 kg fine particle quartz sand(0.15–0.25mm) and 5.0 kg coarse particle quartz sand (0.43–0.85mm)were added to DLRS. The bed material circulating rates of both loopswere controlled at 6 kg/h. Nitrogen gas was preheated and fed into thebottom of pyrolyzer through the gas distributor as the fluidizing gas.

Fig. 1. Principle of the DLRS reaction system (B-HCPs: Big Heat Carrier Particles; S-HCPs:Small Heat Carrier Particles; PGC: Particles Grading Cyclone).

C. Wang et al. Fuel 212 (2018) 448–454

449

The superficial gas velocity in the upper zone of the fluidized bed is1.5–3.7 times of the minimum fluidization velocity of bed materialparticles. When all the reactors reached to the desired temperatures,CHR (60–80 mesh) mixed with quartz sand of the same size in a massblending ratio from 1:0 to 1:6 was fed into the pyrolyzer at a feedingrate of 50–100 g/h by the feeder. The cooling water of the feeder wascontrolled at about 100 g/h. During feeding, the water was vaporizedand injected into the fluidized bed and finally recovered together withthe pyrolysis water from CHR. The fine particles entrained in pyrogasfrom the pyrolyzer was captured in the filter. The pyrogas then passedthrough several ice water cooled condensers in series and finallythrough a THF absorption trap at −15 °C to capture the condensablecomponents in it. The incondensable gas passed through a wet flow-meter to record the gas volume and then was sampled using gas sam-pling bag and analyzed by a gas chromatography GC-7900 equippedwith a thermal conductivity detector (TCD) and a flame ionizationdetector (FID). The gas was sampled every 10min and the final gascompositions excluding nitrogen are averaged values of at least threesamplings in the steady operation stage. After experiment, the con-densers were rinsed with THF. The substances scrubbed from the con-densers were vacuum filtrated using a Buchner funnel. The weight gainof the filter paper is composed of THFIS of the condensates and the fineparticles passed through the filter. The filter paper was then burned inthe muffle furnace. The remaining mass after burning is defined as dustof the pyrolysis tar. It accounts approximately 0.1–0.3 percent of thetar. The recovered mixture of tar and water after filtration was ex-tracted with dichloromethane (DCM) to get the tar. The obtained tarwas weighted and stored for further extraction analysis using n-hexaneand toluene solvent. The water from the pyrolysis of CHR could not bemeasured appropriately because it is about two orders of magnitudelower than that of the cooling water. The operation conditions aresummarized in Table 1.

2.3. Characterization of the tar

Solvent extraction of the tar was conducted the same way as that ofCHR. The compositions of HS from the CHR and the tar were analyzedby GC/MS (Agilent 6890N gas chromatograph coupled with Agilent5975 mass detector).

3. Results and discussion

3.1. Basic properties of CHR

The basic properties of CHR are listed in Table 2. It is shown thatCHR is carbon-rich and ash-rich with relatively low softening point. Thetar yield obtained from the Fischer assay analysis is 11.0 wt%.

Fig. 3 shows the distribution of solvent extraction components ofCHR. The HS which contains the lowest average molecular weight

Fig. 2. Schematic of the DLRS reaction system.

Table 1Operation conditions of the DLRS system.

Size (mm), weight of bed material and circulating rate incombustor-pyrolyzer loop (kg)/(kg/h)

0.150–0.250/3.3/6

Size (mm), weight of bed material and circulating rate incombustor-filter loop (kg)/(kg/h)

0.425–0.850/5.0/6

Feeding gas volume flow rate (m3/h) 0.05U0/Umf* 1.5–3.7Grading gas volume flow rate (m3/h) 0.9–1.1Combustion rising air volume flow rate (m3/h) 5–6Pyrolyzer temperature (°C) 450–600Filter temperature (°C) 500Combustor temperature (°C) 850Particle tank temperature (small and big) (°C) 800

* U0, Umf denote the superficial gas velocity and the minimum superficial velocity ofbed materials based on upper zone diameter respectively.

C. Wang et al. Fuel 212 (2018) 448–454

450

molecules accounts for 10.5 wt% of CHR. The A and PA composed ofhighly aromatic mixtures account for 36.9 wt% and 8.7 wt%, respec-tively. In other words, the CHR contains more than 50 wt% of ex-tractable components which is responsible for its relatively low soft-ening point. The CHR was characterized to be a non-Newtonianpseudoplastic fluid at temperature 483—523 K [12,29,30].

Fig. 4 shows the TG and DTG analysis profiles of CHR under N2atmosphere at the heating rate of 20 °C/min. At about 460 °C there is adevolatilization peak on the DTG profile. Two gentle wide shoulderpeaks around 140 °C and 800 °C are attributed to the weight loss ofwater or light hydrocarbons and the decomposition of minerals, re-spectively.

Upon heating, CHR undergoes melting, boiling and expanding andfinally forms a hard coke. To alleviate the caking property of CHR, itwas mixed with quartz sand. The caking properties of the CHR-quartzsand mixture at varied blending ratio of CHR/quartz sand are shown inFig. 5. With the increase of the quartz sand addition, the caking prop-erties of the samples tend to be weaken.

3.2. Pyrolysis of CHR in DLRS

As discussed above, the caking property of CHR could be alleviatedby blending it with quartz sand. So a blend of CHR and quartz sand wasused as feedstock of the pyrolysis in DLRS. To optimize the blendingratio, tests with different blending ratio of CHR/quartz sand wereconducted in DLRS. In the condition of no quartz sand blending, theexperiments lasted less than 15min (one cycle retention time of thefluidized bed materials) due to the defluidization and the blocking ofthe circulation pipe by the bulk cakes formed from CHR and the bedmaterial quartz sand. When the quartz sand was blended with CHR atthe blending ratio from 1:2 to 1:4, a more and more steady operation ofpyrolysis could be achieved which indicates that the addition of quartzsand favored the dispersion of the CHR particles and prevented it fromcaking to unbreakable bulks. Due to intensive solids mixing and heattransfer in the fluidized bed pyrolyzer, rapid devolatilization of CHRparticles would inhibit them to be agglomerated. In consideration thatan even higher blending ratio like 1:6 or more will lower efficiency ofthe fluidized bed pyrolyzer, the CHR/quartz sand blending ratio of 1:4was adopted to conduct the further pyrolysis experiments.

Fig. 6 shows the yield of pyrolysis products and the gas compositionunder different pyrolysis temperature, in which the char yield wascalculated by difference. According to Fig. 6(a), the tar yield increasesat the lower temperature and then keeps constant at about 20 wt% atthe higher temperature. The tar yield is about two times of the FischerAssay value and is higher than those from the fixed bed pyrolysis re-ported in the literature [16,17]. Under rapid heating condition by thecirculated hot bed materials in the fluidized bed pyrolyzer, the devo-latilization of CHR particles was enhanced and the secondary reactionsof the volatiles were restrained which favored the tar yield in thepyrolysis of CHR. During the pyrolysis temperature range investigated,the gas yield shows no obvious changes and stabilizes at a low level of3–4wt%. It indicates that the increase of tar yield at the lower tem-perature might be contributed from the evaporation of those lightcomponents in CHR, i.e. HS, while the gas yielded mainly from the

Table 2Properties of CHR.

Proximate analysis (wt%, ad) Ultimate analysis (wt%, daf) Fischer Assay analysis (wt%, d) Softening point [17,28,29] (°C)

Moisture Ash Volatile C H N S O* Tar Water Char Gas*

0.18 14.76 35.11 90.25 4.98 0.83 2.92 1.02 11.0 1.5 77.5 10.0 150–180

* By difference.

Fig. 3. The distribution of solvent extraction components of CHR.

Fig. 4. The TG and DTG profiles of CHR under N2 atmosphere (heating rate: 20 °C/min;flow rate: 100ml/min).

Fig. 5. The caking properties of the CHR-quartz sand mixtures.

C. Wang et al. Fuel 212 (2018) 448–454

451

polycondensation of those heavier components attributed to A and PA.As depicted from Fig. 6(b), H2 and CO2 components are the main

pyrolysis gases especially at lower pyrolysis temperature. With the in-crease of pyrolysis temperature, H2 and CH4 concentrations in productgas increase gradually while CO and CO2 decrease. The evolution of H2is the result of polycondensation of the highly aromatic components ofCHR, i.e. A and PA, as mentioned above, while that of CH4 may comefrom the break of Ar-CH3 or Ar-CH2-Ar, and both would be enhanced atthe higher temperature. The CO and CO2 are mainly generated from thecleavage of the relatively weak CeO bond which is seen to release in-tensively at lower temperature.

3.3. Analysis of pyrolysis tar

Table 3 shows the Soxhlet extraction results of the pyrolysis tar. Thetar is composed of mainly HS and A and a small amount of PA. The totalof HS and A accounts for nearly 94 wt% of the tar obtained from thepyrolysis between 450 °C and 600 °C. When increasing the pyrolysistemperature, the HS content increases from about 50 wt% of the tar at450 °C to 60 wt% at 600 °C while the PA keeps constant at around5–6wt%. Table 4 compares the Soxhlet extraction of the tar on the massbasis of CHR with that of CHR. It can be found that the HS yield of thetar is near that of CHR, while the A and PA yields are much lower. Inassociation with the discussion above, it could be concluded that themajor part of HS in the CHR was transferred into the tar. A largeamount of A and PA in the CHR were transformed into THFIS as theresults of polycondensation, which is in accordance with the high charyield as shown in Fig. 6. The polycondensation of PA might happen atthe temperature as low as 450 °C. Higher temperature facilitates thetransformation from A to HS because that the cracking of A was en-hanced [31].

The GC/MS total ion chromatograms (TIC) of the HS extracted fromCHR, 550 °C tar and 600 °C tar are shown in Fig. 7. They have similar

compositions, indicating that there is no obvious change of the HS afterthe pyrolysis of CHR. Approximately 80–100 compounds (relativecontent higher than 0.5%) were detected. The four most prominentcompounds are fluoranthene, 1-methylpyrene, methylchrysene and in-deno(1,2,3-C,D)pyrene. Fig. 8 shows the results of classification of thecomposition of HS determined by GC/MS analysis. Aromatics are themain compositions in the HS of CHR and tar in which 3–6 rings poly-cyclic aromatic hydrocarbons (30–40%), substituted polycyclic aro-matic hydrocarbons (20–30%) and partially hydrogenated polycyclicaromatic hydrocarbons (15–20%) were detected. Aliphatics account forabout 10% of total in the form of long chain alkanes (from hexadecaneto hexacosane were detected). In addition to some unknown com-pounds that cannot be ascribed through the mass spectrometry data-base, a very small amount of phenolic, ester and heterocyclic com-pounds were also detected.

4. Conclusion and perspective

Basic properties of CHR were investigated and the pyrolysis of CHRwas conducted in DLRS under atmospheric pressure. The CHR iscarbon-rich and ash-rich and has relatively low softening point andhigh caking property. The caking property of CHR could be alleviatedby blending it with quartz sand. To obtain a steady pyrolysis of CHR inDLRS, the CHR mixed with quartz sand was rapidly injected into thefluidized bed pyrolyzer by a gas driven feeder. Under optimizedblending ratio of CHR/quartz sand, i.e. 1:4, the caking of CHR to un-breakable bulks could be avoided. At the pyrolysis temperature from500 °C to 550 °C, a tar yield of 20 wt% and a gas yield of 3–4wt% wereobtained. The tar was dust-free because that the dust entrained with thevolatiles from the pyrolyzer was captured efficiently by the particulatesfilter. The Soxhlet extraction of the pyrolysis tar showed that the HSand A account for nearly 94 wt% of the tar. Almost all the HS and a

Fig. 6. The yield of pyrolysis products and the gas composition under different pyrolysis temperature (Feeding rate: 50 g/h; Particle size: 60–80 mesh; CHR/quartz sand mass blendingratio: 1:4; U0/Umf= 2.2).

Table 3The solvent extraction results of tar.

Sample HS A PA

wt%, on the mass basis of tar

Tar-450 °C 48.39 45.00 6.61Tar-500 °C 55.24 38.08 6.68Tar-550 °C 57.87 36.20 5.93Tar-600 °C 64.68 29.13 6.19

Table 4The comparison between Soxhlet extraction of tar and CHR on the mass basis of CHR.

Sample HS A PA

wt%, on the mass basis of CHR

CHR 10.49 36.92 8.71Tar-450 °C 6.93 6.46 0.95Tar-500 °C 10.79 7.57 1.33Tar-550 °C 11.67 7.30 1.20Tar-600 °C 12.47 5.62 1.19

C. Wang et al. Fuel 212 (2018) 448–454

452

small part of A and PA in the CHR were transferred into the tar duringthe pyrolysis. The GC/MS analyses of the HS extracted from CHR,550 °C tar and 600 °C tar showed that they have the same four mostprominent compounds, i.e. fluoranthene, 1-methylpyrene, methyl-chrysene and indeno(1,2,3-C,D)pyrene, and there was no obviouschange of the HS after the pyrolysis of CHR.

For pyrolysis of CHR in DLRS, there are some key issues to be re-solved in the technology development towards commercialization:

1) The contact mode between the feed and the bed materials not onlydetermines the bogging tendency and operability, but also de-termines the conversion and product yields. The fact that CHR fi-nally dries out and solidifies as pyrolysis progresses to completion isthe key. Currently, continuous operation of DLRS has been enabledby blending the feed with excess quartz sand, resulting in the cir-culating ratio varying between 60 and 120. Usually, to supply theheat required for pyrolysis by hot solids, a circulating ratio of 15–20should suffice for a temperature difference of 150 °C or larger

Fig. 7. The GC/MS total ion chromatograms ofHS (a: CHR; b: Tar-550 °C; c: Tar-600 °C).

Fig. 8. Classification of the composition of HS determined by GC/MS analysis.

C. Wang et al. Fuel 212 (2018) 448–454

453

between the combustor and pyrolyzer. Actually, excessively highcirculating ratio is both unnecessary and unfavorable for large-scaleindustrial systems because of high energy consumption to drive thesolids around the loop, plus materials erosion and particle attrition(fines) issues. An alternative solution to the caking and boggingproblem of CHR is to pump and atomizing it at suitable temperature,possibly by mixing it with some oil or solvent. In this case, therheological characteristics of CHR and design of atomizing nozzleshould be major research topics.

2) In the present study, the pyrolyzer operates at a superficial gas ve-locity (SGV) about 1.5–3.7 Umf, which is quite low compared tosimilar processes. The bed is more resistant to bogging as SGV in-creases, which suggests trade-offs between the SGV and operatingtemperature and sand/feed blending ratio. One can operate atslightly lower temperature (promising higher tar yield) with lessquartz sand addition if the pyrolyzer runs at higher SGV. An addi-tional benefit of higher SGV is that it enhances mass and heattransfer in the bed, which not only reduces the temperature gradientin the bed but also allows the fresh tar vapor to be stripped from thereacting particle surfaces as fast as possible, which in turn wouldencourage more tar vapor production due to improved vapor–liquidequilibrium conditions. Steam or nitrogen can be used as fluidizing/stripping agent. This is another point to consider for future tech-nology development.

Acknowledgements

This work was supported by the Shanghai Research Institute ofChina Shenhua Coal to Liquid and Chemical Corporation Ltd. The re-viewers’ valuable suggestions are highly appreciated.

References

[1] Rong F, Victor DG. Coal liquefaction policy in China: explaining the policy reversalsince 2006. Energy Policy 2011;39(12):8175–84.

[2] Liu ZY, Shi SD, Li YW. Coal liquefaction technologies – development in China andchallenges in chemical reaction engineering. Chem Eng Sci 2010;65(1):12–7.

[3] Cui H, Yang JL, Liu ZY, Bi JC. Characteristics of residues from thermal and catalyticcoal hydroliquefaction. Fuel 2003;82(12):1549–56.

[4] Khare S, Dell'Amico M. An overview of conversion of residues from coal liquefactionprocesses. Can J Chem Eng 2013;91(10):1660–70.

[5] Liu X, Zhou ZJ, Hu QJ, Dai ZH, Wang FC. Experimental study on co-gasification ofcoal liquefaction residue and petroleum coke. Energy Fuels 2011;25(8):3377–81.

[6] Chu X, Li W, Li B, Chen H. Gasification property of direct coal liquefaction residuewith steam. Process Saf Environ Prot 2006;84(6):440–5.

[7] Chu XJ, Li W, Li BQ, Chen HK. Sulfur transfers from pyrolysis and gasification ofdirect liquefaction residue of Shenhua coal. Fuel 2008;87(2):211–5.

[8] Li J, Yang JL, Liu ZY. Hydro-treatment of a direct coal liquefaction residue and its

components. Catal Today 2008;130(2):389–94.[9] Sugano M, Ikemizu R, Mashimo K. Effects of the oxidation pretreatment with hy-

drogen peroxide on the hydrogenolysis reactivity of coal liquefaction residue. FuelProcess Technol 2002;77:67–73.

[10] Li J, Yang JL, Liu ZY. Hydrogenation of heavy liquids from a direct coal liquefactionresidue for improved oil yield. Fuel Process Technol 2009;90(4):490–5.

[11] Bai L, Nie Y, Li Y, Dong HF, Zhang XP. Protic ionic liquids extract asphaltenes fromdirect coal liquefaction residue at room temperature. Fuel Process Technol2013;108:94–100.

[12] Li P, Zong ZM, Liu FJ, Wang YG, Wei XY, Fan X, et al. Sequential extraction andcharacterization of liquefaction residue from Shenmu-Fugu subbituminous coal.Fuel Process Technol 2015;136:1–7.

[13] Li KJ, Cheng SF, Lin HL, Zhang XW, Chang HY, Shu C, et al. Study and developmentof Shenhua direct coal liquefaction technology. Clean Coal Technol2015;21(1):50–5.

[14] Miao Q. Status and development progress of direct coal liquefaction residue ex-traction technologies. Clean Coal Technol 2015;21(1):56–60.

[15] Lv DM, Wei YC, Bai ZQ, Jin B, Kong LX, Guo ZX, et al. An approach for utilization ofdirect coal liquefaction residue: blending with low-rank coal to prepare slurries forgasification. Fuel 2015;145:143–50.

[16] Xu L, Tang MC, Duan LE, Liu BL, Ma XX, Zhang YL, et al. Pyrolysis characteristicsand kinetics of residue from China Shenhua industrial direct coal liquefaction plant.Thermochim Acta 2014;589(10):1–10.

[17] Li XH, Xue YL, Feng J, Yi Q, Li WY, Guo XF, et al. Co-pyrolysis of lignite andShendong coal direct liquefaction residue. Fuel 2015;144:342–8.

[18] Qu Y, Chu M, Shen GD, Yuan Y, Zhang Y. Inhibitory effect of coal direct liquefactionresidue on lignite pulverization during co-pyrolysis. Fuel Process Technol2016;147:57–63.

[19] Zhang JW, Wu RC, Zhang GY, Yao CB, Zhang YM, Wang Y, et al. Recent studies onchemical engineering fundamentals for fuel pyrolysis and gasification in dual flui-dized bed. Ind Eng Chem Res 2013;52:6283–302.

[20] Zhang YM, Wang Y, Cai LG, Yao CB, Gao SQ, Xu GW. Dual bed pyrolysis gasificationof coal: process analysis and pilot test. Fuel 2013;112:624–34.

[21] Wang JG, Lu XS, Yao JZ, Lin WG, Cui LJ. Experimental study of coal topping processin a downer reactor. Ind Eng Chem Res 2005;44(3):463–70.

[22] Fan XX, Lu QG, Na YJ, Liu Q. Experimental study on coal multi-generation in dualfluidized beds. J Therm Sci 2007;16(3):277–82.

[23] Lu QG, Liu Q, Na YJ, Zhao K, He J. Experimental research on pyrolysis character-istics of coal in dual fluidized beds. Proc CSEE 2010;30(8):15–9.

[24] Xiao G, Wang XH, Zhang JP, Ni MJ, Gao X, Luo ZY, et al. Granular bed filter: apromising technology for hot gas clean-up. Powder Technol 2013;244(4):93–9.

[25] Chen YS, Chyou YP, Li SC. Hot gas clean-up technology of dust particulates with amoving granular bed filter. Appl Therm Eng 2015;74:146–55.

[26] Zhang YQ, Liang P, Yu J, Zhu JL, Qin XZ. Studies of granular bed filter for dustremoval in the process of coal pyrolysis by solid heat carrier. RSC Adv2017;7(33):20266–72.

[27] Paenpong C, Inthidech S, Pattiya A. Effect of filter media size, mass flow rate andfiltration stage number in a moving-bed granular filter on the yield and propertiesof bio-oil from fast pyrolysis of biomass. Bioresour Technol 2013;139(13):34–42.

[28] Cui H, Yang JL, Liu ZY, Bi JC. Coal liquefaction residue and its gasification forhydrogen. Coal Convers 2001;24(01):15–20.

[29] Wang ZC, Ge Y, Shui HF, Ren SB, Pan CX, Kang SG, et al. Molecular structure andsize of asphaltene and preasphaltene from direct coal liquefaction. Fuel ProcessTechnol 2015;137:305–11.

[30] Ren YJ, Wei AL, Zhang DX, Zhao JC, Lin CD, Gao JS. Rheological characteristics ofcoal hydro-liquefaction residue. J Fuel Chem Technol 2007;35(3):262–7.

[31] Benito AM, Martínez MT, Fernández I, Miranda JL. Visbreaking of an asphalteniccoal residue. Fuel 1995;74(6):922–7.

C. Wang et al. Fuel 212 (2018) 448–454

454