RSC Advances

View Article Online

PAPER View Journal | View Issue

Distinctive oil shale pyrolysis behavior in indirectly

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Cite this: RSC Adv., 2017, 7, 21467

heated fixed bed with internals
Open Access Article. Published on 18 April 2017. Downloaded on 11/1/2023 9:11:15 AM.

Lanxin Lin,ab Dengguo Lai,ab Zhen Shi,ab Zhennan Hanab and Guangwen Xu *ac

Intrinsic characteristics of oil shale pyrolysis in a fixed bed reactor with internals have been investigated in
this study. Mounting particularly designed internals in fixed bed reactor improved shale oil production to up
to 90% yield by Fischer assay. Comparing particle characteristics at different radial positions of the reactors
with and without internals demonstrated that, inside the particle bed with internals, the product flow was
regulated to move from the high-temperature zone (outer) to low-temperature zone (center), which
reduced the secondary reactions of released volatiles. Terminating oil shale pyrolysis at central particle
bed temperatures of 150, 300, 450, and 550  C showed that the contents of vacuum gas oil and heavy
oil in the shale oil produced had increased from 9.63 wt% to 53.29 wt%. The volatile contents of particles
in the inner layer of the reactor slightly increased in the early stage of pyrolysis and, in turn, decomposed
to form pyrolysis products as the temperature was raised. The adsorption or condensation of liquids on
the surface of particles gradually increased from the outer region to the central region of the reactor
due to the regulated product flow direction and low temperatures in the central zone of the reactor
causing heavy components to condense. Increasing the degree of pyrolysis was also found to decrease
Received 25th February 2017
Accepted 3rd April 2017
the alkene, aromatic, and cycloalkane contents in shale oil, but increased those of alkane and
heteroatomic compounds. These results demonstrate that adopting internals into oil shale pyrolysis
DOI: 10.1039/c7ra02347b
optimized the product flow direction and selectively directed secondary reactions to occur for heavy
rsc.li/rsc-advances volatile species only.

Lurgi–Ruhrgas process in Germany, and the Fushun process in

1. Introduction China.4–9 Of these, Tosco-II, Galoter, Enet-280, ATP, and Lurgi–
Alternative and renewable energy sources have received Ruhrgas processes adopt a solid heat carrier method to heat oil
increasing attention in recent years as fossil fuel consumption shale particles and achieve relatively high oil yields. However,
has grown. Oil shale is among the most promising alternative the resultant liquid products usually exhibit high dust intake
energy sources because of its enormous reserves and plentiful due to inherent technical defects. Many studies on oil-dust
aliphatic organic compounds.1 Available oil shale utilization characteristics and separation have been conducted, but
patterns have been established in several countries since the a comprehensive solution for the dust-intake problem in
1960s. Combusting oil shale to generate power could supply pyrolysis had yet to be developed.10 Gas heat carrier retorting
domestic industrial users, with retorted shale oil considered technology is employed in the Petrosix, Kiviter, and Fushun
a potential resource to offset the future crude oil scarcity. The retorting processes. However, these methods only process large-
utilization of retorted shale oil requires a comprehensive utili- particle feedstock. Excessive air pumped into a Funshun
zation mode, as reported by many studies.2–6 furnace causes oil combustion in the pyrolysis stage, lowering
Several representative oil shale retorting processes have been the shale oil yield to 65% by Fischer assay. Process analysis is
developed and are still in service, such as in situ underground applied to optimize operating parameters and improve oil
retorting and Tosco-II technology in the USA, the Kiviter process production efficiency.11 An in situ retorting process has been
in Estonia, the Petrosix process in Brazil, the Galoter process in proposed for deeply buried oil shale, but this technology
Russia, Enet-280 technology in Germany and Estonia, the requires long-term verication for application.4 Therefore, no
Alberta Taciuk process (ATP) in Australia and Canada, the existing retorting processes provide satisfactory pyrolysis
performance for oil shale, especially for particle sizes below 10
a
State Key Laboratory of Multi-Phase Complex Systems, Institute of Process
mm. In particular, innovative retorting technologies are
Engineering, Chinese Academy of Sciences, Beijing 100190, China. E-mail: gwxu@ required for small oil shale particles to achieve high oil yield
ipe.ac.cn; Fax: +86 10 82629912; Tel: +86 10 82544886 and quality simultaneously.
b
University of Chinese Academy of Sciences, Beijing 100049, China Particularly designed internals, comprising a central gas
c
Institute of Industrial Chemistry and Energy Technology, Shenyang University of collection pipe and a number of metallic plates, have been
Chemical Technology, Shenyang 110142, China

This journal is © The Royal Society of Chemistry 2017 RSC Adv., 2017, 7, 21467–21474 | 21467
 View Article Online

RSC Advances Paper

employed to improve oil shale pyrolysis production in indirectly 2. Experimental section

heated xed bed reactors.12–14 An innovative reactor with inter-
nals was evaluated by comparing the heating rate of oil shale 2.1 Materials and apparatus
particles and pyrolysis product distribution. With the heating Oil shale was obtained from Huadian oil shale mine, and was
rate to particles improved, especially in the central region, the sieved to a particle size of 0–5 mm for use in experiments. The
resultant shale oil yield was afforded in 90% oil yield by Fischer properties of Huadian oil shale are shown in Table 1. The dry-
assay. Raising the reactor temperature from 600  C to 1000  C base oil content was 12.78 wt% by Fischer Assay. Silica sand
beneted shale oil production, which was in contrast to almost (1–2 mm) was used in the study to replace a portion of the oil
all other reports in the area, which show higher pyrolysis shale particles, which had been pretreated at 900  C for 3 h.
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

temperatures lead to lower oil yield. In conventional indirectly A conventional xed bed reactor without internals was used
heated xed bed reactors, it is generally difficult to achieve high as reference in this work. The pyrolysis investigation was con-
Open Access Article. Published on 18 April 2017. Downloaded on 11/1/2023 9:11:15 AM.

oil yields at high reaction temperatures due to unavoidable ducted in a reactor mounted with a central collection pipe and
signicant secondary reactions of volatiles.15–18 In contrast, four metallic plates. The plates were welded perpendicular to
devised internals can reduce cracking in the reactor to achieve the xed bed reactor wall. Fig. 1 shows schematic diagrams of
a higher shale oil yield at higher heating or reaction tempera- the two reactors used. The reactors had an inner diameter of 200
tures. Matching the product ow eld with temperature distri- mm, and effective volumes for loading oil shale of around 5600
bution in the reactor using internals has been speculated to mL. From the wall to the center of the reactor (Fig. 1), both
avoid excessive cracking of primary products and improved reactors were demarcated into three layers (a–c) in a radial
pyrolysis production. Therefore, further understanding of the direction. Three temperature measurement points were present
detailed pyrolysis characteristics of the new reactor and its in layers a–c to record real-time heating curves inside these
superiority are required to guide industrial applications of the layers. The experimental system was the same as used in our
technology. previous work,14 consisting of four parts: pyrolysis, liquid
Many studies have focused on the behavior and kinetics of collection, pressure control, and gas sampling.
oil shale pyrolysis, which are needed for reasonable analysis
and optimization guidance.15–22 Al-Ayed et al. adopted second-
2.2 Pyrolysis and analysis
order kinetics to model liquid accumulation15 and rst-order
kinetics to t the total weight loss of oil shale in a xed bed.16 The furnace was heated to 850  C and held until the pyrolysis
In contrast, various parameters, such as heating rate, nal test was completed. For each individual experiment, dried
temperature, particle size, density, and additional minerals, Huadian oil shale (4900 g) was loaded into the test reactor. In
have been investigated to clarify their effects on oil shale some tests, the oil shale particles in the outboard layer were
pyrolysis at various scales.22–32 A heating rate of 10  C min 1 for replaced with silica sand of equal volume. Aer connecting to
oil shale particles is optimal for obtaining a high oil yield.26–29 the experimental system, the reactors were placed into the
The internals adopted in the preceding reactor gave an obvious heated furnace and quickly covered with silica wool. The pres-
improvement on heating rate.12–14 Nonetheless, this improve- sure in the reactor was maintained at 1 atm using a vacuum
ment was not as effective at altering pyrolysis behavior in pump. Pyrolysis products were exhausted from the top of the
indirectly heated reactors.12,13 Lan et al. obtained a maximal reactor and cooled in a downstream condenser pipe. The cooled
shale oil yield at a reaction-stopping temperature of 550  C.30 liquid product was collected in a bottle. Residual low-boiling
Lai et al. demonstrated that the oil yield decreased with components in the gaseous product were adsorbed using
increasing temperature by conducting an oil shale pyrolysis three acetone bottles. The pyrolysis gas was then passed
series with and without ash.31 Zhang et al. investigated the through a wet gas meter to measure the volume and then
impact of residence time on the pyrolysis product in a uidized successive through a few clean bottles to remove dust, moisture,
bed.33 These studies were all performed in reactors without the and hydrogen sulde. Every 5 min, a gas sample was taken for
aforementioned internals.34 Pyrolysis behavior in xed bed compositional analysis by GC. When the central temperature
reactors with internals and the working mechanism of internals measurement point reached 150  C, 300  C, 450  C, and 550  C,
have yet to be reported. Therefore, a targeted fundamental the reactor was removed from the furnace to terminate pyrolysis
investigation of oil shale pyrolysis in this innovative xed bed and the reactor was cooled in air.
reactor with internals was conducted in this study. Subsequent product treatment was conducted for all pyrol-
ysis experiments conducted at central temperatures of 150  C,

Table 1 Properties of Huadian oil shale tested

Proximate analysis wt% Ultimate analysis wt%

A V FC C H N Oa S Fischer Assay wt% HHV kcal kg 1

66.27 28.10 5.62 21.77 2.58 0.64 2.09 0.96 12.78 2158.88
a
O content was obtained from the difference of balance.

21468 | RSC Adv., 2017, 7, 21467–21474 This journal is © The Royal Society of Chemistry 2017
 View Article Online

Paper RSC Advances

capture the real-time temperature distribution in the tested

reactors with or without internals.
Fig. 2 compares the heating curves for oil shale in the layers
a–c. The heating furnace temperature for the tests was 800  C.
Overall, reactor B, which had internals, gave a much shorter
heating time than reactor A to reach a temperature of 550  C at
the measured bed center. This implied faster heating in the
reactor with internals, in agreement with our previous
studies.12–14 Fig. 2 also shows that the difference between the
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

heating curves of the two reactors was largest in layer c and

Fig. 1 Schematic illustrations of particle layers a–c in reactors A and B. smallest in layer a. This meant that the internals mainly facili-
Open Access Article. Published on 18 April 2017. Downloaded on 11/1/2023 9:11:15 AM.

tated heating to particles in the inner (core) bed.

Layer a, nearest the reactor wall, was quickly heated to 700  C
300  C, 450  C and 550  C. All yields were given relative to the in 20 min, and subsequently exceeded 800  C. This formed
dry mass of treated oil shale. a high-temperature zone in layer a in the early stage of the
Aer cooling to room temperature, oil shale char from the experiment, resulting in fast oil shale pyrolysis. At this stage the
three layers (a–c in Fig. 1) was sampled for further analysis. temperatures in layers b and c were below 200  C, which were
Accordingly, the char samples were denoted a–c. The char yield too low to give oil shale pyrolysis. There was little difference
was calculated from the product weight relative to the mass of between the two reactors in the initial (rst 10 min) temperature
loaded oil shale. Condensate (or adsorbate) on 1.0 mL char, rise characteristics from outboard layer a to central layer c.
sieved to 0.2–1.0 mm, was dissolved (with the char) in 50 mL As the experiment continued, the heating rates of layers b and
petroleum ether (boiling range: 90–110  C, AR). The obtained c in reactor B (with internals) gradually became higher than those
solution was scanned using an ultraviolet spectrophotometer in reactor A. The average heating rate of oil shale in layer c was
(TU-1901) at 1 nm s 1. Char samples were smashed to a powder 4  C min 1 in reactor A, but was elevated to about 10  C min 1 in
below 0.1 mm in size for proximate analysis. reactor B.26–29 The devised internals enhanced heating by aligning
Liquid from the scrubbing bottles was mixed with the the eld of temperature, from high to low, with the ow direc-
acetone solutions from the washing pipelines and vessels, and tion.12–14 The central gas collection pipe guided the high-
rotary evaporation was performed to remove acetone. The temperature gaseous products to ow from the outboard to the
resultant residual liquid was blended with shale oil in the liquid central zone, with heat taken from the gas contributing to heat-
collection bottle of the condenser. Intake water in the blended ing of the core bed.12,13 Meanwhile, the metallic plates added
liquid was rst separated by oil-water delamination, and the extra heating surfaces to promote heat transfer from the reactor
remaining moisture in the oil phase was measured by a stan- wall to the particle bed.14 At the end of the test, layer b in reactor A
dard method using toluene as an azeotrope. This allowed the has been heated to a higher temperature than that in reactor B,
water and oil yields to be estimated. The dehydrated shale oils indicating that a higher temperature area had been created
collected from different central temperatures at termination inside reactor A. Therefore, the employed internals greatly
(150, 300, 450, and 550  C) were analyzed to obtain the distri- improved heating efficiency from the bed wall to the center and
butions of distillation fractions according to boiling points. A altered the temperature gradient in the bed.
GC 7890A instrument (Agilent) was employed to perform
simulated distillation fraction analysis. The chemical compo-
sitions of the four oil samples were determined using a GC-MS 3.2 In-bed product ow direction
QP 2010 Ultra instrument (Shimadzu). All analytical methods A decreasing temperature gradient is present from the outboard
used for shale oil in this study have been detailed in our to the center of an indirectly heated xed bed reactor. This
previous work.13,14 The pyrolysis gas yield was calculated from
the measured gas volume (in a gas meter) and the gas compo-
sition was obtained using a micro GC (Agilent 3000A).

3. Results and discussion

3.1 Temperature distribution
Particularly designed internals have been successfully applied
to improve the oil yield and heating efficiency in an indirectly
heated xed bed reactor in our previous work.12–14 Taking the
heating curve at or near the reactor center to characterize the
heating efficiency of the particle (oil shale) bed in the reactor
showed that the heating rate of the particles at the measured
positions was nearly doubled when using internals. Three Fig. 2 Heating curves for layers a–c in reactors with and without
temperature measurement points in layers a–c were recorded to internals.

This journal is © The Royal Society of Chemistry 2017 RSC Adv., 2017, 7, 21467–21474 | 21469
 View Article Online

RSC Advances Paper

gradient has a critical impact on the in-bed secondary reactions internals in reactor A, the pyrolysis products owed outwards
of pyrolysis products. If the products owed through the high- toward the reactor wall. In contrast, Fig. 3 shows that the loaded
temperature outboard, excessive product cracking would silica sand remained white in reactor B with internals,
occur. In contrast, secondary reactions are likely to be weak in demonstrating that the generated pyrolysis products did not
the reactor center. Herein, the product ow direction in the ow into the outer region next to the reactor wall. This veried
reactors with and without internals was claried to obtain that the use of particularly designed internals in a xed bed
better understanding of the system. pyrolyzer resulted in the internal ow of pyrolysis products from
A char bed gives a smaller pressure drop than an oil shale the wall zone to the central zone around the gas collection pipe.
bed of the same thickness.13 As pyrolysis progresses in an Samples of oil shale or char were taken from layers a–c at
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

indirectly heated xed bed, a char layer is formed on the different reaction stages or reaction times and analyzed to
outboard or annulus, causing the gaseous pyrolysis products to further understand the product ow direction in reactor B with
Open Access Article. Published on 18 April 2017. Downloaded on 11/1/2023 9:11:15 AM.

ow toward the reactor wall. The pyrolysis products must internals. Table 2 summarizes the results of proximate analysis
undergo a large amount of secondary reactions in the outboard for all samples, in which the reaction stage is shown by different
zone to greatly decrease shale oil production. The internals temperatures reached at the monitored central position (rep-
potentially reduce this problem by redirecting the ow of resenting the bed center). At all pyrolysis stages, from the outer
pyrolysis products into the pyrolysis reactor.12–14 Silica sand layer a to central layer c, the ash and xed carbon contents of
(1.0–2.0 mm) has better gas permeability than oil shale. the analyzed samples decreased, but their volatile contents
Therefore, silica sand was used to imitate char by replacing oil conversely increased. This agreed with the temperature gradient
shale particles with silica sand in the outer layer of the reactor. in the reactor that high-temperature in the outer layer caused
This also ensured that no chemical change would occur in the a high degree of pyrolysis. Meanwhile, as the reaction pro-
sand layer, while volatile products were only generated by gressed (measured temperature increased) the volatile contents
pyrolyzing oil shale in the central zone. Observing the color of all samples tended to decrease gradually, but those of layers
change of the silica sand allowed the ow direction of volatiles b and c initially increased slightly or did not vary until the
formed in the bed center to be determined. measured central temperature was over 300  C. As the ow of
Fig. 3 shows top-down photos inside the beds aer pyrolysis products was from the outer wall to center, species with high
tests in reactors (A) without and (B) with internals. The outer boiling points (or heavy species) would be condensed and
layer lled with silica sand had obviously different colors. Sand adsorbed on the surface of inner-layer particles when their
particles were blackened in reactor A, also having a deeper color temperatures were low, such as below 300  C. This explained the
at inner positions close to the central oil shale layer. This was slight rise in volatile content of samples from layers b and c. The
caused by the deposition of unreacted matters, including condensation of heavy species also improved the quality of the
carbon, formed from pyrolysis. Therefore, without mounting tar product. Once the temperature of the inner-layer particles
was high enough, oil shale pyrolysis and secondary reactions of
this condensed matter had to occur. Therefore, the central gas
collection pipe realized selective cracking of condensed
matters, while other non-condensable matter was directly
released as nal products without signicant secondary reac-
tions occurring when passing through the reactor space.
For UV analysis, unsaturated functional groups and hetero-
atoms in the pyrolysis products must induce light absorption in
UV region. This absorbance is proportional to the chromophore
concentration, indicating the amount of condensate deposited
on the particles. Fig. 4 compares the UV spectra of the conden-
sate on 1 mL of particles (0.2–1 mm) taken from layers a–c, and
the original oil shale. The absorption peak at around 225 nm,
indicating the C]C bond of a large alkene, was used as the
Fig. 3 Evidence of secondary cracking through color change on comparison reference. For each of the reaction temperatures
particles in the vicinity of the reactor wall for reactors A and B.

Table 2 Proximate analysis for oil shale or char taken from different radial positions (a–c)

Ash [wt%] Volatile [wt%] Fixed carbon [wt%]

Central temperature [ C] a b c a b c a b c

150 67.43 63.84 63.37 25.89 31.10 31.33 6.68 5.07 5.02
300 74.63 64.30 63.78 17.88 31.52 31.17 7.49 6.05 5.57
450 80.03 76.32 70.70 10.52 15.09 21.97 9.45 8.59 7.35
550 82.38 82.53 81.35 7.05 8.21 9.78 10.57 9.27 8.88

21470 | RSC Adv., 2017, 7, 21467–21474 This journal is © The Royal Society of Chemistry 2017
 View Article Online

Paper RSC Advances

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.
Open Access Article. Published on 18 April 2017. Downloaded on 11/1/2023 9:11:15 AM.

Fig. 4 UV analysis of condensate on particles in different particle layers a–c from pyrolysis terminated at different central temperatures.

compared, the absorbance of observable peaks increased from shale oil and pyrolysis gas increased, while that of char
outer layer a to central layer c, with the original oil shale having decreased. For the tabulated time, the variation in yield with
the lowest absorbance. This indicated that more chemicals were time was almost linear for all such products. Pyrolysis water
present on particles in the inner low-temperature layers, further production showed little variation during the last stage of
verifying the product ow from the high-temperature outer layer reaction because it resulted from water evaporation and
to the bed center and into the central gas collection pipe. UV decomposition of hydroxy groups at low temperatures. These
absorbance for samples from a single layer of the reactor man- features of production were different from those in a conven-
ifested a similar variation with pyrolysis temperature (or time). tional xed bed reactor (without internals), which gives obvious
Absorbance was elevated to a maximum when the plotted central peak production at temperatures of 480–530  C.22–27 This
temperature was about 450  C, and then sharply dropped to difference was caused by oil shale pyrolysis in the reactor with
a very low value by the end of pyrolysis (scale is different in four internals occurring layer-by-layer from outside to inside, and
sub-plots of Fig. 4). This evidenced the secondary reactions of the ow and temperature elds in the reactor being controlled
species condensed on the particle surface. During pyrolysis, non- to form the desired direction and gradient, respectively.
condensable products directly ow into the central gas collection The shale oil product can be characterized according to
pipe, but most heavy species with high boiling points must be boiling points as fractions of gasoline (<180  C), diesel (180–350

condensed on the low-temperature surface of particles in the C), vacuum gas oil (VGO, 350–500  C) and heavy oil (>500  C).
inner layer. As the reaction progresses, the inner bed reached Fig. 5 shows the variations in fractions of these distillates with
a high temperature, causing re-evaporation and cracking of the pyrolysis temperature at the measured central point and the
condensed components into further pyrolysis products via yield at each pyrolysis stage at a range of measured central
secondary reactions. Therefore, in the reactor with internals, the temperatures. In Fig. 5, the fraction was the oil obtained in
heavy species experienced, in succession, generation, condensa- a test at room temperature to the mentioned individual central
tion, and re-evaporation or cracking, while light species were temperature, while the yields are for the shale oil product ob-
directly discharged as products. This demonstrated that inter- tained at each temperature range, including 0–150  C, 150–
nals resulted in secondary reactions selectively on only heavy 300  C, 300–450  C and 450–550  C. Using the temperature
species, affording oil production with high yields (due to no range 150–300  C as an example, two pyrolysis experiments were
secondary reaction of light species) and quality (only cracking of
heavy species).
Table 3 Pyrolysis product distributions using different reaction times
3.3 Product generation characteristics in the reactor with internals

Previously, we reported pyrolysis product characteristics based Yield [wt%]

on the nal yields and qualities of shale oil and pyrolysis Central
gas.12–14 This section further characterizes production by Time [min] temperature [ C] Char Oil Water Gas
analyzing oil and gas products at different reaction times or
20 150 97.55 0.60 0.46 1.20
temperatures at the measured central point. Table 3 shows the 38 300 91.13 2.86 2.50 3.82
product distribution at measured central temperatures of 150, 56 450 80.87 7.71 4.22 8.32
300, 450 and 550  C. As the reaction progressed, the yields of 74 550 75.31 11.45 4.29 11.74

This journal is © The Royal Society of Chemistry 2017 RSC Adv., 2017, 7, 21467–21474 | 21471
 View Article Online

RSC Advances Paper

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.
Open Access Article. Published on 18 April 2017. Downloaded on 11/1/2023 9:11:15 AM.

Fig. 5 Variations in (a) oil fraction and (b) yield with changing central temperature in the reactor with internals.

conducted in this range to obtain shale oil yields from room

temperature to 150  C and 300  C. Then, by subtraction, the
plotted yield was estimated for the temperature range of 150–
300  C, as shown in Fig. 5(b).
Fig. 5(a) clearly shows that no heavy oil was generated by
pyrolysis at temperatures below 300  C at the measured central
point, but the ultimate heavy oil content reached 2.86 wt% of oil
gained at a central temperature of 550  C. The total VGO yield
was 0.57 wt% up to 300  C and 5.20 wt% at 300–550  C. The oils
obtained at measured central temperatures of 450  C and 550  C
had VGO contents that reached 44.25 wt% and 50.43 wt%,
respectively. Therefore, heavy components, including VGO and Fig. 6 GC-MS spectra of shale oils from pyrolysis terminated at
heavy oil, were generated mainly in late stage of pyrolysis at different central temperatures in the reactor with internals.
temperatures above 300  C. The fraction of gasoline in the shale
oil product sharply decreased from 53.53% to about 13.0%
when the measured temperature was raised from 150  C to relative intensities of groups C6–C9 but increased those of C19,
550  C. However, the absolute gasoline yield at each reaction C21–C24, C36, and C40. This result complied with the
stage was low, uctuating around 0.4 wt%. The heavy products preceding distillation analysis, but was the opposite of our
(VGO + heavy oil) decreased the gasoline fraction at high general knowledge regarding oil composition variation with
temperatures. Diesel distillate was produced mainly at central pyrolysis temperature in normal xed bed reactors.19,23 Deep
temperatures above 300  C, with the highest yield (1.9 wt%) cracking of long-chain hydrocarbons in high-temperature wall-
obtained at 300–450  C. The diesel content remained at about annulus was expected to cause obvious decomposition of
40% in oils obtained using all tested central temperatures. species with carbon numbers above 17. However, the action of
Diesel production is thought to be caused by the pyrolysis of oil internals caused these high-C species to be released and
shale and secondary racking of condensate at 400–550  C. moderately decomposed in the late stage of pyrolysis.
Therefore, the benecial actions from our mounted internals From GC-MS analysis, the oil components were classied
enabled the production of gasoline and diesel (light oils) in the into ve types: alkanes, alkenes, aromatics, heteroatom
entire temperature range up to 550  C, whereas the production compounds, and cyclohydrocarbons. Their area percentages
of VGO and heavy oil (heavy species) mainly occurred at high- were calculated by integrating peaks of the GC-MS spectrum.
temperature reaction stages (>300  C). In the reactor with The area percentage was proportional to the content of the
internals, the condensation of heavy matter on inner-layer indicated chemicals. Fig. 7 shows that the area percentage of
particles had to occur during earlier reaction stages (<300  C aliphatic hydrocarbons (alkane and alkene) was above 80% in
at the measured central point). This was a critical process to all oils, indicating the aliphatic hydrocarbon base of the shale
ensure selective cracking to heavy species. oil composition. The area percentages for aromatic and
Fig. 6 shows the GC-MS spectra of oil samples obtained at the heteroatom compounds were both around 10%, while that of
tested different central temperatures shown in Fig. 5. The cyclohydrocarbon was below 5%. The variation in relative areas
carbon numbers of components were marked to identify alkane with central temperature (pyrolysis stage) revealed that
peaks. In the rst spectrum (150  C), the relative intensity continued pyrolysis at high temperatures reduced the alkenes,
decreased with increasing carbon number. Raising the central aromatics, and cyclohydrocarbons in oil, but gave more alkanes
temperature (300, 450, and 550  C) gradually elevated the rela- and heteroatom compounds. Considering the chemical struc-
tive intensities of carbon numbers above 15. Further pyrolysis ture of these components, the molecular weights of the alkenes,
and re-evaporation/cracking of heavy species lowered the aromatics, and cyclohydrocarbons were much lower than those

21472 | RSC Adv., 2017, 7, 21467–21474 This journal is © The Royal Society of Chemistry 2017
 View Article Online

Paper RSC Advances

cracking. Meanwhile, light components were directly exhausted

via the central channel as pyrolysis products. Because the
volatiles released from oil shale particles owed into a low-
temperature region, where only heavy species underwent the
aforementioned selective secondary reactions, the shale oil
yield continuously increased with increasing heating tempera-
ture and maintained a high light content even at higher heating
temperatures. Terminating oil shale pyrolysis at central
temperatures of 150, 300, 450, and 550  C further demonstrated
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

that, in the reactor with internals, light species (gasoline +

diesel) were produced throughout the entire pyrolysis period,
Open Access Article. Published on 18 April 2017. Downloaded on 11/1/2023 9:11:15 AM.

Fig. 7 Major chemical components of shale oils from pyrolysis

terminated at different central temperatures in the reactor with whereas heavy species (VGO + heavy oil) were mainly generated
internals. in the late stage. Pyrolysis in the reactor with internals was also
shown to produce more alkenes and aromatics in the early
reaction stage at low temperatures, but more alkanes in the late
of the alkanes. The carbon numbers of most alkenes, aromatics, stage at higher temperatures.
and cyclohydrocarbons were below 19, 11, and 16, respectively.
Fig. 6 shows the carbon number distribution of the alkanes Acknowledgements
obtained, indicating that high molecular weight alkane
components were present. C19H37N, C27H56O, and C28H58O The study was nanced by the National Basic Research Program
were detected as the major heteroatomic compounds, with of China (2014CB744303), National Natural Science Foundation
molecular weights higher than the alkenes, aromatics, and (91534125), and Strategic Priority Research Program of Chinese
cyclohydrocarbons obtained. The boiling point of oil compo- Academy of Sciences (XDA07010100).
nents increases with higher molecular weights. Therefore,
alkanes and heteroatomic compounds would condense in the References
earlier stage of pyrolysis and re-evaporate in the late stage to
give a gradually higher percentage, as shown in Fig. 7. 1 J. R. Dyni, Oil Shale, 2003, 20, 193–252.
2 W. Taciuk, Oil Shale, 2013, 30, 1–5.
3 K. Bredow, Oil Shale, 2003, 20, 81–92.
4. Conclusions 4 J. G. Speight, Shale oil production process, Elsevier, New York,
2012.
Pyrolysis of Huadian oil shale in xed bed reactors with or 5 X. M. Jiang, Z. G. Cui, C. Yan, D. C. Liu, J. R. Qiu and J. B. Li,
without internals was conducted to investigate the working Fuel, 2002, 81, 793–797.
mechanism of internals in improving pyrolysis performance. 6 J. Qian, J. Wang and S. Li, Proceedings of the 18th International
Measuring heating curves of the oil shale particles at the bed Offshore and Polar Engineering Conference, ISOPE, Vancouver,
center claried that the particularly designed internals 2008, pp. 19–20.
increased heat transfer from the reactor wall to the central layer 7 A. R. Brandt, Energy Fuels, 2009, 23, 6253–6258.
of oil shale by 30–50%. The internals were further shown to 8 X. M. Jiang, X. X. Han and Z. G. Cui, Energy, 2007, 32, 772–
regulate the ow of volatile products in the reactor from the 777.
annular high-temperature zone to the central low-temperature 9 I. Öpik, N. Golubev, A. Kaidalov, J. Kann and A. Elenurm, Oil
region. This caused silica sand loaded in the annular layer of Shale, 2001, 18, 99–108.
the reactor with internals to maintain its white color, whereas in 10 P. Chaturong and P. Adisak, Powder Technol., 2013, 245, 233–
the reactor without internals the silica sand was blackened, 240.
demonstrating a certain degree of coking of the volatiles owing 11 S. Y. Yang, J. Zhang, Q. C. Yang and Y. Qian, Energy Fuels,
through the sand particles. Furthermore, because of the regu- 2014, 28, 5557–5564.
lation of volatile ow by internals, UV analysis of the liquid 12 C. Zhang, R. C. Wu and G. W. Xu, Energy Fuels, 2014, 28, 236–
washing particles demonstrated that the particles taken from 244.
the central layer of the reactor with internals contained more 13 L. X. Lin, C. Zhang, H. J. Li, D. G. Lai and G. W. Xu, Fuel
deposited surface chemicals, conrming the condensation of Process. Technol., 2015, 138, 147–155.
heavy components on these particles, which were at relatively 14 L. X. Lin, D. G. Lai, E. W. Guo, C. Zhang and G. W. Xu, Fuel,
low temperatures during the early stages of pyrolysis. The 2016, 163, 48–55.
condensed heavy components were re-evaporated or cracked in- 15 O. S. Al-Ayed, M. R. Suliman and N. A. Rahman, Appl. Energy,
turn when the central bed temperature was further elevated 2010, 87, 2273–2277.
during late-stage pyrolysis. Consequently, by regulating volatile 16 O. S. Al-Ayed, M. Matouq, Z. Anbar, A. M. Khaleel and E. Abu-
ow in the reactor from high temperature to low temperature, Nameh, Appl. Energy, 2010, 87, 1269–1272.
secondary reactions were selectively restricted to heavy species 17 H. Bar, R. Ikan and Z. Aizenshtat, J. Anal. Appl. Pyrolysis,
via a successive process of condensation, evaporation, and 1988, 14, 49–71.

This journal is © The Royal Society of Chemistry 2017 RSC Adv., 2017, 7, 21467–21474 | 21473
 View Article Online

RSC Advances Paper

18 G. Yang, Z. H. Wang, Q. M. Xian, F. Shen, C. Sun, Y. Z. Zhang 27 F. D. Guffey and T. R. McLendon, Liq. Fuels Technol., 1984, 2,
and J. Wu, RSC Adv., 2015, 5, 40117–40125. 439–461.
19 Y. H. Sun, F. T. Bai, B. C. Liu, Y. M. Liu, M. Y. Guo, W. Guo, 28 T. H. Fletcher, R. Gillis, J. Adams and T. Hall, Energy Fuels,
Q. W. Wang, X. S. Lü, F. Yang and Y. Yang, Fuel, 2014, 115, 2014, 28, 2959–2970.
338–346. 29 J. H. Campbell, G. H. Koskinas, N. D. Stout and T. T. Coburn,
20 J. Gersten, V. Fainberg, G. Hetsroni and Y. Shindler, Fuel, In Situ, 1978, 2, 1–47.
2000, 79, 1679–1686. 30 X. Z. Lan, W. J. Luo, Y. H. Song, J. Zhou and Q. L. Zhang,
21 M. C. Torrente and M. A. GalaÂn, Fuel, 2001, 80, 327–334. Energy Fuels, 2015, 29, 7800–7806.
22 V. Dieckmann, H. J. Schenk, B. Horseld and D. H. Welte, 31 D. G. Lai, Z. H. Chen, L. X. Lin, Y. M. Zhang, S. Q. Gao and
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Fuel, 1998, 77, 23–31. G. W. Xu, Energy Fuels, 2015, 29, 2219–2226.
23 L. Ballice, Fuel Process. Technol., 2005, 86, 673–690. 32 L. Zhang, B. Zhang, Z. Q. Yang and Y. F. Yan, RSC Adv., 2014,
Open Access Article. Published on 18 April 2017. Downloaded on 11/1/2023 9:11:15 AM.

24 H. F. Guo, J. D. Lin, Y. D. Yang and Y. Y. Liu, Fuel, 2014, 118, 4, 39145–39155.

186–193. 33 H. Y. Zhang, J. Zheng, R. Xiao, D. K. Shen, B. S. Jin,
25 S. Wang, X. M. Jiang, X. X. Han and J. H. Tong, Fuel Process. G. M. Xiao and R. Chen, RSC Adv., 2013, 3, 5769–5774.
Technol., 2014, 121, 9–15. 34 H. F. Guo, Q. X. Cheng, Z. Jin, D. Wang, G. W. Xu and
26 P. T. Williams and N. Ahmad, Fuel, 1999, 78, 653–662. Y. Y. Liu, RSC Adv., 2016, 6, 102231–102248.

21474 | RSC Adv., 2017, 7, 21467–21474 This journal is © The Royal Society of Chemistry 2017