G Model

CATTOD-10499; No. of Pages 8 ARTICLE IN PRESS

Catalysis Today xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Catalysis Today
journal homepage: www.elsevier.com/locate/cattod

Integration of catalytic cracking and hydrotreating technology for

triglyceride deoxygenation
Hui Wang a , Hongfei Lin a , Peng Feng a , Xue Han a , Ying Zheng a,b,∗
a
Department of Chemical Engineering, University of New Brunswick, 15 Dineen Drive, Fredericton, New Brunswick, E3B5A3, Canada
b
School of Engineering, The University of Edinburgh, Colin Maclaurin Road, Edinburgh, EH93DW, UK

a r t i c l e i n f o a b s t r a c t

Article history: This study investigated the integration of catalytic cracking − hydrotreating technology for the deoxy-
Received 3 July 2016 genation of waste cooking oil (WCO), and compared the integrated technology with hydrotreating alone
Received in revised form 31 October 2016 on oxygen removal and the level of hydrogen consumption. The results showed that the performance of
Accepted 1 December 2016
acid treated kaolin (ATK) was the best in the catalytic cracking section compared to that of the petroleum
Available online xxx
commercial catalyst (CC) and the alumina catalyst. The proposed catalytic cracking − hydrotreating inte-
grated technology presented in this work is a potential methodology for commercialization to achieve
Keywords:
organic liquid products (OLPs) with low oxygenate content and low hydrogen consumption, which can
Catalytic cracking – hydrotreating
integrated technology
reduce the oxygenate content from 87.6% to 6.6% and the hydrogen consumption from 8.8 mol to 2.4 mol
Hydrotreating technology per mol of tryglyceride.
Triglyceride deoxygenation © 2016 Elsevier B.V. All rights reserved.
Acid treated kaolin
Hydrogen consumption

1. Introduction fully compatible with petroleum products. Drop-in biofuels elimi-

nate the short shelf-life defect of both ethanol and biodiesel.
Reliance on petroleum oil results in high greenhouse gas (GHG) Generally, triglycerides contain 8–15 wt% oxygen, which results
emissions. The combustion of petroleum produces one-third of in a low heating value. Research on how to remove oxygen from
total carbon dioxide emissions in the world, which significantly triglycerides to produce drop-in biofuels has been documented
contributes to global climate change. In 2015, Canada announced [4–6]. At present, two technologies have become popular for
its goal to reduce GHG emissions by 30% by 2030 from the level in refining bio-oil: catalytic cracking without hydrogen and catalytic
2005, 731 metric megatons [1]. Because of the low GHG emissions hydrotreating with hydrogen. In oil refineries, catalytic cracking is
of renewable fuels, Canada has already established a renewable fuel the most widely used process to turn the heavy fraction of crude oil
content of 5% in gasoline by 2010, and 2% in diesel and heating oil by into gasoline and other lighter hydrocarbons. Studies have shown
2012 [2]. Therefore, the focus is turning towards improving the uti- that catalytic cracking can effectively degrade large molecules in
lization of renewable energy resources [3]. As a typical renewable fatty acids into smaller compounds which fall mostly in the gaso-
source, triglycerides can be upgraded into drop-in diesel/gasoline line boiling point range [7]; oxygen can be removed in the forms
fuels, which resemble petroleum derived diesel/gasoline and are of CO, CO2 or H2 O [8]. Previous research concerning the use of
petrol-intended catalysts in the fluid catalytic cracking (FCC) of
triglycerides or model compounds has shown undesirable results:
extensive coking (even at temperatures of 350 ◦ C) [9], low liquid
Abbreviations: WCO, waste cooking oil; ATK, acid treated kaolin; CC, petroleum
commercial catalyst; OLP, organic liquid product; GHG, greenhouse gas; FCC, fluid
yields (40–60%) with high contents of aromatics (50–71%) [10], and
catalytic cracking; GC/MS, gas chromatograph/mass spectrometer; DMDS, dimethyl irreversible catalyst deactivation [11–13]. In addition, although cat-
disulphide; ATK-HDO, the integrated process of catalytic cracking with ATK and alytic cracking can partially eliminate oxygen from triglycerides,
hydrotreating; CC-HDO, the integrated process of catalytic cracking with CC and there are still massive amounts of unsaturated compounds and rel-
hydrotreating; Alumina-HDO, the integrated process of catalytic cracking with
atively small amounts of oxygenates that remain in the products −
alumina and hydrotreating; HDO, the hydrotreating only process; GC, gas chro-
matograph; GC/FID, gas chromatograph with a flame ionization detector; COx, CO the consequence is that the upgraded liquid products still cannot
and CO2 . be directly used as fuel.
∗ Corresponding author at: Department of Chemical Engineering, University of The hydrotreating of triglycerides to produce non-oxygenates
New Brunswick, 15 Dineen Drive, Fredericton, New Brunswick, E3B5A3, Canada. has been extensively studied and implemented for commercial
E-mail address: [email protected] (Y. Zheng).

http://dx.doi.org/10.1016/j.cattod.2016.12.009
0920-5861/© 2016 Elsevier B.V. All rights reserved.

Please cite this article in press as: H. Wang, et al., Integration of catalytic cracking and hydrotreating technology for triglyceride
deoxygenation, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.009
G Model
CATTOD-10499; No. of Pages 8 ARTICLE IN PRESS
2 H. Wang et al. / Catalysis Today xxx (2016) xxx–xxx

purposes. The products obtained from the hydrotreating of triglyc- water to remove Cl− (tested with an AgNO3 solution) and dried
erides are generally long-chain paraffin with high melting points overnight in an oven at 120 ◦ C.
− this easily leads to blockage in the reaction system. To solve this Alumina was prepared from beohmite (a sample provided by
problem, Neste oil developed the only commercialized technology, SASOL). Beohmite was dissolved in deionized water and then HCl
combining hydrotreating and isomerization processes [14]. The was added by drops. The mixture was put in a water bath at
hydrotreating process works for saturating carbon double bonds 60–65 ◦ C; the pH value of this mixture was then adjusted at 3–4,
and deoxygenation, and the isomerization process can make nor- and the colloidal substances were produced. Then the solution was
mal paraffin isomerize to reduce the melting point of products. dried in an oven at 105 ◦ C for 5 h. The dried sample was calcined in
However, as triglycerides usually contain 8–15% of oxygen that a muffle oven at 700 ◦ C for 2 h [30].
must be removed, a large amount of hydrogen (the calculated For cracking catalyst preparation, 40 g of adhesives (Hydrophilic
®
results is 15 mol hydrogen per mol triglycerides assumed with 3 fumed silica, AEROSIL 380) dissolved in 80 g of concentrated HNO3
carbon carbon double bonds) is likewise needed, which leads to was used as binder. Then ATK, CC, or alumina was added into the
high costs [15]. Therefore, the hydrogenation process can only be binder solution and stirred until distributed evenly. The mixture
implemented under government subsidy support. Furthermore, as was introduced into the extruding machine to shape the catalysts
the hydrogen is mainly produced from methane reforming pro- that were then dried at room temperature overnight and further
cess, availability depends on the treatment of petroleum, which dried at 120 ◦ C for 2 h. The dried catalysts were subsequently cal-
results in GHG emissions. Traditional sulfided hydrotreating cata- cined at 550 ◦ C for 2 h. The calcined catalysts were crushed and
lysts are easily deactivated in several days due to a loss of sulfur sieved to 40–80 mesh.
and the existence of by-product water generated from hydrotreat- In order to get relatively stable catalyst activities, all fresh crack-
ing [16–20]. For all these reasons, it is a real challenge to implement ing catalysts were subjected to hydrothermal treatment in steam at
the hydrotreating of triglycerides on a commercial scale [16,21]. 700 ◦ C for 5 h (the flow rate of water was set to 0.95 ml/min, calcu-
Previously reported studies focused on either hydrotreating or lated according to ASTM D 4463) before being evaluated. The flow
catalytic cracking triglycerides but rarely on both [22–25]. The chart and the information of the FCC reaction system can be found
investigated co-processing for bio-oil upgrading in the literature in the literature [31]. For a typical catalytic cracking evaluation, cat-
is the hydroprocessing followed by catalytic cracking [16,26,27], alysts and nitrogen were introduced to the FCC reactor to form a
which had drawbacks of low product quality (with high content stable fluidization before WCO and steam were sent to the reac-
oxygen and unsaturated hydrocarbons), low product yield, and high tor. The inlet water (3 ml/min) flowed through a steam generator
hydrogen consumption. The intention of the current study is to and the vaporized steam was introduced to the reactor. The crack-
address the challenge of upgrading triglycerides by combining cat- ing process started once the feedstock vapours came into contact
alytic cracking with hydrotreating technology, and by identifying a with the hot cracking catalysts. The product vapours exited from
kind of catalyst with high performance for deoxygenation. the top of the reactor and flowed through a condenser. Liquid prod-
In this study, the performances of acid treated kaolin (ATK), ucts were collected from the bottom of the separator. Gas products
petroleum commercial cracking catalyst (CC), and the alumina were retrieved via the displacement of brine. The water phase was
catalysts for the deoxygenation of WCO were investigated in an separated from the OLP phase before being hydrotreated. Catalytic
FCC reactor, and the resulting OLPs were then hydrotreated in a cracking experiments were carried out under the following condi-
batch reactor with a commercial supported CoMoS catalyst. This tions: temperature was 470 ± 3 ◦ C, the mass ratio of catalyst to feed
integrated technology was compared with the hydrotreating tech- was 3.3 ± 0.3, and weight hourly space velocity was 10.5 ±1.2 h−1 .
nology. This is the first time that the mild catalytic cracking and Assumption: 1) the FCC reactor used in the current study is an ideal
hydrotreating technology has been applied in biofuel upgrading. plug flow reactor; 2) the steam is an ideal gas under the reaction
The ultimate goal is to enhance the product quality and to reduce conditions; 3) the reaction temperature, the reaction pressure, and
the hydrogen consumption. the fluidized state are uniformly distributed inside the reactor; and
4) the reaction pressure is at atmospheric. The calculated residence
time was 3.85 s.
2. Experimental

2.1. Feed
2.2.2. Hydrotreating process
Fatty acid compositions of WCO, which are listed in Table 1, The catalyst used for the hydrotreating section was an industrial
were determined using a GC/MS (Shimadzu GCMS-QP5000) after commercial supported CoMoS catalyst (Liaoning Haitai Sci-Tech
WCO was trans esterified. The carbon double bond positions were Development Co., Ltd., China). A 25 ml batch reactor was used in
tested at Purdue University: a tandem Mass Spectrometer (AB Sciex the hydrotreating section. Before the experiment, 0.15 g catalyst
4000 Q-trap) was used to identify the positions of carbon dou- sample underwent drying and pre-sulfiding. Drying was performed
ble bonds following the Paterno–Buchi reaction method, and the at 300 ◦ C in a muffle oven for 4 h. Pre-sulfiding was performed at
detailed testing procedure was the same as in the literature [28]. 200 psi using a mass ratio of catalyst to dimethyl disulfide (DMDS)
The WCO used in this work mainly consists of three fatty acids: at 1:10. The temperature program was: 220 ◦ C (rate: 20 ◦ C/min),
palmitic acid (7.5%), oleic acid (71.3%), and linoleic acid (17.2%). 280 ◦ C (rate: 1 ◦ C/min; held for 2 h), and 320 ◦ C (rate: 1 ◦ C/min; held
for 2 h).
Dodecane (a mixture of isomers, purchased from Fisher Scien-
2.2. Catalyst preparation and performance evaluation tific) was used as the solvent in hydrotreating section. Typically,
10% feed (1.0 g) in dodecane was introduced into the reactor after
2.2.1. Catalytic cracking process catalyst pre-sulfiding experiment. The reaction conditions were as
Kaolin was purchased from Fisher Scientific. ATK was prepared follows: temperatures of 275 ◦ C, 300 ◦ C, and 325 ◦ C; pressures of
from natural kaolin and HCl solution [29]. One hundred grams of 250 psi and 500 psi; a rotation speed of 230 round per minute;
natural kaolin, calcined at 800 ◦ C for 1 h, was added into a 300 ml and a reaction time of 2 h. Blank experiments with solvent only
solution containing 21.9 g HCl, and the reaction took place at 92 ◦ C were conducted for removing the effect of solvent. This integrated
and continued for 12 h. The mixture was then washed with distilled process of catalytic cracking with ATK (CC or alumina) and then

Please cite this article in press as: H. Wang, et al., Integration of catalytic cracking and hydrotreating technology for triglyceride
deoxygenation, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.009
G Model
CATTOD-10499; No. of Pages 8 ARTICLE IN PRESS
H. Wang et al. / Catalysis Today xxx (2016) xxx–xxx 3

Table 1
Composition of fatty acids in WCO.

Fatty acid (wt.%)(n:m(o))a 16:0 16:1 18:0 18:1(9) 18:2(9,12) ≥20

WCO 7.5 0.5 2.5 71.3 17.2 1.1

a
n: m (o), n is the number of carbon atoms in fatty acids; m is the number of C C double bonds in fatty acids; o denotes position of carbon double bond.

hydrotreating was named ATK-HDO (CC-HDO or alumina-HDO). used in the current study is not suitable for aromatics and cyclic
Only hydrotreating was named HDO. hydrocarbon hydrogenation into chain hydrocarbons.
As shown in Fig. 2, under a pressure of 250 psi, when the reac-
tion temperature reached 300 ◦ C, there was an obvious increase in
2.3. Product analyses
alkene content but no obvious changes in other hydrocarbon con-
tents. This increase illustrates that alkenes are the products of the
CO and CO2 were analysed by a combination of a GC (Varian
oxygenates hydrotreatment. When the temperature increased up
GC-3400, Varian, Inc., USA) equipped with an Agilent HP-PLOT Q
to 325 ◦ C from 300 ◦ C, the contents of the alkenes and cycloalkenes
column (19095P-QO4), and a Residue Gas Analyzer (RGA, Stanford
decreased and the contents of alkanes and cycloalkanes increased.
Research Systems, Inc., USA). CO and CO2 in gas products were iden-
It reveals that under the pressure of 250 psi, the hydrogenation
tified by the ratios of mass to charge (m/z) at 12 and 44, respectively.
depth of alkenes and cycloalkenes into alkanes and cycloalkanes
The composition of hydrocarbons in the gas products was
was obviously enhanced when the reaction temperature increased.
determined by gas chromatograph with a Flame Ionization Detec-
The same trend can be seen in Fig. 2 under a pressure of 500 psi.
tor (GC/FID, SHIMADZU GC-17A, Shimadzu Scientific Instruments,
As can be seen in Fig. 2, under a reaction temperature of 275 ◦ C,
Japan), which was equipped with a middle range polarity column
when the hydrogen pressure increased from 250 psi to 500 psi,
(19095P-S25).
the contents of alkanes, alkenes, cycloalkanes, and cycloalkenes
The chemical compositions of the organic liquid products were
barely changed. Under a reaction temperature of 300 ◦ C, when the
analysed by gas chromatograph (SHIMADZU GC-17A) coupled with
hydrogen pressure increased from 250 psi to 500 psi, the contents
mass spectrometer (SHIMADZU MS-QP5000) (GC/MS, Shimadzu
of alkenes and cycloalkenes decreased, whereas the content of
Scientific Instruments, Japan), equipped with an Agilent HP-5MS
alkanes significantly increased. This phenomenon indicates that at
column.
300 ◦ C, the hydrogenation depth of alkenes and cycloalkenes into
alkanes was obviously enhanced when the hydrogen pressure was
3. Results and discussion increased from 250 psi to 500 psi. Under a reaction temperature of
325 ◦ C, when the hydrogen pressure increased from 250 psi to 500
3.1. Investigation of hydrotreating reaction parameters psi, the contents of alkenes and cycloalkenes continued decreas-
ing, whereas the contents of alkanes and cycloalkanes increased,
3.1.1. Overall reactions especially that of cycloalkanes.
In this paper, FCC is the technology expected to be “not” similar Based on the above description, the temperature of 300 ◦ C is a
to traditional FCC (main reactions during petroleum FCC process sensitive point for the hydrogenation of alkenes and cycloalkenes.
are C C and C H cracking), but expected to remove the majority Below this temperature, it is difficult for alkenes and cycloalkenes
of oxygen in triglycerides, to modify the liquid product distribu- to be hydrogenated. Meanwhile, under or above this point, alkenes
tion, and to receive a high liquid yield, and the function of the and cycloalkenes are largely hydrogenated to alkanes. This has been
hydrotreating process is to eliminate residual oxygen and to sat- also proved with different pressures. The hydrogenation capabil-
urate alkenes and cycloalkenes in the liquid products with a low ities of cycloalkenes were significantly enhanced when reaction
hydrogen consumption. temperature increased up to 325 ◦ C.
Hydrotreating of the ATK upgraded OLPs was conducted under For both the deoxygenation and hydrogenation of WCO dur-
different conditions. Our previous results showed that the degree ing the integrated process, the suitable hydrotreating temperature
of deoxygenation increased when the temperature was higher than was 325 ◦ C and the suitable hydrogen pressure was 500 psi. Under
250 ◦ C; then it decreased and the extensive cracking increased these reaction conditions, the contents of oxygenates, cycloalkenes,
when the reaction temperature reached 365 ◦ C [31]. Elliot observed and alkenes in ATK-HDO upgraded OLP were lowered to 6.6 wt%,
the same phenomenon with hydrotreating temperatures above 0.3 wt%, and 8.0 wt%.
340 ◦ C [32]. The suitable hydrotreating conditions are at a tempera-
ture in the range of 275 ◦ C to 325 ◦ C and at a pressure of 500 psi [31]. 3.1.2. Deoxygenation
The chemical and hydrocarbon compositions of OLPs are shown in ATK upgraded OLP had high contents of ketones (11.6%) and
Figs. 1 and 2, respectively. fatty acids (7.5%), low content of alcohols (1.1%), and no other oxy-
As shown in Fig. 1, under a pressure of 250 psi, when the temper- genates. More than 60% of the C O containing compounds in the
ature increased, the chain hydrocarbon content increased and the ATK upgraded OLP were cyclic ketones.
oxygenate content decreased, but there were no significant changes The oxygenates in ATK-HDO upgraded OLPs under pressures of
in cyclic hydrocarbon and aromatic contents. 250 psi and 500 psi are shown in Fig. 3. Even though there were
As can be seen in Fig. 1(d), it is clear that the oxygenate content no obvious variations in the total oxygenate contents in the OLPs
did not significantly change under different pressures when the (at the same temperature and different pressures, see Fig. 1(d)),
temperature remained constant, which indicates that the reaction the specific oxygen structures varied. The deoxygenation reactions
temperature rather than the hydrogen pressure was the key factor of the hydrotreating process are shown in Scheme 1. The cyclic
for deoxygenation during the integrated process. The deoxygena- ketones could be hydrotreated into alcohols or phenols (reaction
tion was shown to be insensitive to pressure (250–500 psi) within 1), and the produced alcohols or phenol could experience inter-
the testing ranges. The chain hydrocarbon, the cyclic hydrocarbon, molecular dehydration to produce non-oxygenates (reaction 2), or
the aromatics, and the total oxygenate contents were stable under could form C O C through intra-molecular dehydration (reac-
the same reaction temperature but different pressures of 250 psi tion 3). The C O C could be hydrotreated into alcohols, at the
and 500 psi, respectively. The reason is that the CoMoS catalyst same time, releasing non-oxygenates (reaction 4). In addition, the

Please cite this article in press as: H. Wang, et al., Integration of catalytic cracking and hydrotreating technology for triglyceride
deoxygenation, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.009
G Model
CATTOD-10499; No. of Pages 8 ARTICLE IN PRESS
4 H. Wang et al. / Catalysis Today xxx (2016) xxx–xxx

60 40

50 47.7 48.3
30 28 28.9
42.5

Compositions (wt %)
Compositions (wt %)

39.6 26.8 25.9 25.6

40 24.4
34.8 23
33.4 33

30 20
250 psi 250 psi

500 psi 500 psi

20
10
10

0 0
Feed 275 300 325 Feed 275 300 325
Reaction temperature (oC) Reaction temperature (oC)

(a) Chain hydrocarbons (c) Aromatics

25
40

19.5
20 18.7
Compositions (wt %)

16.9 30 28.5

Compositions (wt %)
16.5
15.1 14.9 25.3
15 23.9
13

250 psi 20
10 15.8 16.5 250 psi
500 psi
500 psi

5 10
6.9 6.6

0
0
Feed 275 300 325
Feed 275 300 325
Reaction temperature (oC) Reaction temperature (oC)

(b) Cyclic hydrocarbons (d) Oxygenates

Fig. 1. The chemical compositions of organic liquid products from catalytic cracking (over acid treated kaolin) and hydrotreating integrated technology under different
reaction conditions.

(2) content of C O decreased and that of −OH increased, whereas

Non-oxygenates
Inter -H2O there were no significant changes in the contents of C O C and
(1) −COOC . It indicates that the hydrogenation of C O into OH
-C=O -OH
+ H2 (Reaction 1 in Scheme 1) was obviously promoted. Under a reaction
(3)
-C-O-C- temperature of 325 ◦ C, when the hydrogen pressure increased from
Intra -H2O 250 psi to 500 psi, the content of −OH continuously increased and
(4) the contents of −C O C and −COOC decreased. It reveals that
the hydrogenation of −C O C and COOC into OH (Reaction
+ H2
4 in Scheme 1) was enhanced at this temperature.
Scheme 1. Deoxygenation reactions of hydrotreating process.
Based on the results with different temperatures and pressures,
C O conversion into OH started with a lower temperature while
−C O C conversion into −OH required a higher temperature. As
fatty acids could be reduced into ketones or aldehydes and could
Dupond et al. [33] proposed, the C O hydrogenation was followed
also be reacted with alcohols or phenols to produce esters.
by the C O bond breaking.
The fatty acids were not present in all hydrotreated products,
but were in the hydrotreating feed; therefore, the −COOH was
not shown in Fig. 3. This indicates that the conversion of COOH 3.2. Comparison of integrated technology with hydrotreating
was 100% after being hydrotreated in all hydrotreating trails. This technology
leads to an increase in the content of C O from the hydrotreating
of COOH in the hydrotreated products during deoxygenation at 3.2.1. Chemical composition distributions of integrated OLPs
275 ◦ C. The OLPs upgraded by the different catalysts were all
As shown in Fig. 3, with the temperature increasing, the con- hydrotreated under a temperature of 325 ◦ C and a pressure of 500
tents of C O and OH decreased, while the content of C O C psi, while the WCO was hydrotreated under the same conditions,
increased. It demonstrates that under a pressure of 250 psi, the to be used as a reference. The chemical compositions of OLPs from
deoxygenation of C O and OH was significantly enhanced the integrated process and the hydrotreating process are shown in
among all deoxygenation processes when temperature increased. Fig. 4.
Under a reaction temperature of 275 ◦ C, when the hydrogen Under the said conditions, the contents of oleic acid and alkanes
pressure increased from 250 psi to 500 psi, the contents of all oxy- were 77.7% and 5.7% in the HDO upgraded OLP, respectively. And
genates barely changed. Under a reaction temperature of 300 ◦ C, the total oxygenate content was as high as 89.6%, much higher than
when the hydrogen pressure increased from 250 psi to 500 psi, the in the other OLPs from integrated technology. Therefore, there was

Please cite this article in press as: H. Wang, et al., Integration of catalytic cracking and hydrotreating technology for triglyceride
deoxygenation, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.009
G Model
CATTOD-10499; No. of Pages 8 ARTICLE IN PRESS
H. Wang et al. / Catalysis Today xxx (2016) xxx–xxx 5

50 25

40.3
40 20 19.2
36.2
Compositions (wt %)

Compositions (wt %)
29
30 15

250 psi 250 psi

20 17.4 17.3 17.9 17.3 500 psi 10
500 psi
7.1

10 5
2.3 2.3
1 1.5
0
0 0
Feed 275 300 325 Feed 275/250 300/250 325/250
Reaction temperature (oC) Reaction temperature (oC)

(a) Alkanes (c) Cycloalkanes

25 20
22.3
16.4
15.9
20 14.9
15
Compositions (wt %)

Compositions (wt %)
16.9
16 15.7 12.8
15 11.5
13.5
11.5 9.5
10
250 psi 250 psi
10
8 500 psi 500 psi

5
5

0.3
0 0
Feed 275 300 325 Feed 275/250 300/250 325/250
Reaction temperature (oC) Reaction temperature (oC)

(b) Alkenes (d) Cycloalkenes

Fig. 2. The hydrocarbon compositions of organic liquid products from catalytic cracking (over acid treated kaolin) and hydrotreating integrated technology under different
reaction conditions.

a gap in the deoxygenation ability between two technologies of the gas hydrocarbon yield (see Table 2). The decarboxylation and decar-
hydrotreating and the integrated technology. bonylation reactions were more prevalent in alumina than they
The oxygenate content in CC-HDO upgraded OLP was 17.9%, were in ATK and CC, which can account for the high yield of COx (CO
while it was 6.3% in the alumina-HDO or ATK-HDO upgraded OLPs. and CO2 ). Compared to deoxygenation through dehydration, lower
Concerning the hydrocarbon types in the OLPs, the alumina-HDO liquid yields were obtained during deoxygenation through decar-
OLP was primarily composed of chain hydrocarbons (including bonylation or decarboxylation. Generally speaking, coke deposition
24.3% unsaturated), whose content was as high as 85.8%. The CC- less than 10 wt.% is acceptable for fluid catalytic cracking process
HDO OLP was mainly composed of aromatics (50.4%). The excessive [34]. The coke yields distributed from 4.8% to 6.0%, which is much
concentrations of long-chain hydrocarbons in the liquid products lower than the industrial acceptable level.
lead to properties of a higher cloud point and freezing point. The oil The liquid and gas product yields of the hydrotreating process
products with high aromatics content cannot be fully combusted are shown in Table 3. The liquid product yields reached 95–100%,
and release more PM 2.5 when being burnt, so aromatics content whereas the gas product yields were 0.2–2.1%. This reveals that
should be limited. there was not much cracking in the hydrotreating process under
The composition of the ATK-HDO OLP was somewhere in the the investigated conditions. It should be noted that the inte-
middle, composed of 40.3% chain hydrocarbons, 27.6% aromatics, grated technology and the HDO favoured hydrodecarbonylation
and 25.5% cyclic hydrocarbons. Based on all these conclusions, ATK- and hydrodecarboxylation, respectively.
HDO was selected as the best option for creating drop-in biofuels. Organic liquid product yield is a key factor in the current study.
The total organic liquid products yielded were 67.3%, 72.9%, 62.4%,
and 98.8% during CC-HDO, ATK-HDO, alumina-HDO, and HDO pro-
3.2.2. Product distributions cesses. Compared to the integrated technology, HDO yielded a
The total mass received from both the catalytic cracking and much higher total liquid. One of the reasons is due to low deoxy-
the hydrotreating processes under all conditions were more than genation; the oxygenate content in HDO liquid products is still
95 wt%. The product distribution results of both processes are 89.6%. The oxygen content in WCO is 10.6%; therefore, the high-
shown in Tables 2 and 3. est hydrocarbon liquid yield is still lower than 90% if the oxygen
The organic liquid product yield, gas product yield, and the oxy- is removed completely. However, the liquid products from the
gen removal of catalytic cracking processes are shown in Table 2. All triglyceride hydrotreating are generally long-chain paraffin with
three cracking catalysts achieved an oxygen removal of more than high melting points, which results in the blocking of the reac-
90%, and ATK and alumina even had their deoxygenation capabili- tion system. This problem can be solved by changing the product
ties around 93.8%. The liquid product yields of CC and alumina were compositions in two methods: changing the hydrotreating pro-
lower than that of ATK. For CC, this was due to the over-cracking cess into hydrocracking process, or combining hydrotreating with
that occurred when WCO was upgraded, which resulted in a high

Please cite this article in press as: H. Wang, et al., Integration of catalytic cracking and hydrotreating technology for triglyceride
deoxygenation, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.009
G Model
CATTOD-10499; No. of Pages 8 ARTICLE IN PRESS
6 H. Wang et al. / Catalysis Today xxx (2016) xxx–xxx

20 3.5
3.1
17.1
3
16 15.2 2.6

Compositions (wt %)
Compositions (wt %)

2.5
11.6
12 2 1.9

1.5
250 psi 250 psi
1.5
8 500 psi
6.3 500 psi
1
0.6
4
0.5 0.3
1.2
0 0 0
0 0
Feed 275 300 325 Feed 275 300 325
Reaction temperature (oC) Reaction temperature (oC)

(a) -C=O (c) -C-O-C-

12 2.5
10.6

10 1.9
2

Compositions (wt %)
Compositions (wt %)

8
1.5
1.3
5.8 5.8 6 1.2
6 5.6 1.1
250 psi 1 250 psi
1
500 psi 500 psi
4
2.7
0.5
2
1.1
0 0
0 0
Feed 275 300 325 Feed 275 300 325
Reaction temperature (oC) Reaction temperature (oC)

(b) -OH (d) -COOC-

Fig. 3. The oxygenate structures of organic liquid products from catalytic cracking (over acid treated kaolin) and hydrotreating integrated technology under different reaction
conditions.

80 60
50.4
61.5
60 45
Compositions (wt %)
Compositions (wt %)

44.3
40 30 25.6
Alkanes
26.3 24.3
Alkenes
20 15
5.7 4.5
4 2.8 0.9
0.8
0 0
CC-HDO Alumina-HDO ATK-HDO HDO CC-HDO Alumina-HDO ATK-HDO HDO
Reaction temperature (oC) Reaction temperature (oC)

(a) Chain hydrocarbons (c) Aromatics

25 100
89.6

20 19.2
80
Compositions (wt %)

Compositions (wt %)

15 60

Cycloalkanes
10 40
Cycloalkenes
17.9
5 3.4 2.7 20
1.2 0.9 0.9 6.1 6.6
0.3 0
0 0
CC-HDO Alumina-HDO ATK-HDO HDO CC-HDO Alumina-HDO ATK-HDO HDO
Reaction temperature (oC) Reaction temperature (oC)

(b) Cyclic hydrocarbons (d) Oxygenates

Fig. 4. Chemical compositions of organic liquid products from different technologies at 325 ◦ C under 500 psi.

Please cite this article in press as: H. Wang, et al., Integration of catalytic cracking and hydrotreating technology for triglyceride
deoxygenation, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.009
G Model
CATTOD-10499; No. of Pages 8 ARTICLE IN PRESS
H. Wang et al. / Catalysis Today xxx (2016) xxx–xxx 7

Table 2
Product distribution and oxygen removal of catalytic cracking process.

Cracking catalyst Yield (wt.%) Oxygen removal (wt.%)

OLP COx a Hydrocarbon gases Coke Othersb

CC 67.5 5.6 7.2 6.0 9.2 90.8

ATK 74.4 3.4 1.4 4.8 11.7 93.8
Al2 O3 65.9 9.2 3.8 5.0 12.3 93.9
a
COx : CO and CO2 .
b
Others: yields of H2 O and total carbons of carbon-containing compounds dissolved in water.

Table 3
Product distribution of hydrotreating process.

Yields (wt.%) CC-HDO ATK-HDO Alumina-HDO HDO

Hydrocarbon gases 0.40 1.95 0.40 0.02

CO 0.08 0.12 0.13 0.01
CO2 0.05 0.02 0.01 0.19
OLP 99.7 98.0 94.7 98.8

isomerization processes [14]. The former has drawbacks of con- 20.0

suming hydrogen and providing low liquid yield. The latter has the

Hydrogen consumption (mol/mol WCO)

drawbacks of consuming hydrogen and increasing both additional
capital and operation costs. Compared to other technologies, mild 15.0 13.8
catalytic cracking over ATK has the advantages of higher oxygen
removal, no hydrogen consumption, and producing good composi-
tions of liquid products, and relatively higher liquid yields. 10.0 8.8

3.2.3. Hydrogen consumption

The hydrogen concentrations were higher than 99% for all gas 5.0 4.0 3.9
products. Assuming that the gas products were the ideal gases, the 2.4
hydrogen consumption was calculated by the ideal gas law using
the difference of hydrogen amount before and after reaction. The 0.0
CC-HDO ATK-HDO Alumina-HDO HDO Calculation*
hydrogen consumption of the two processes is shown in Fig. 5(a).
It can be seen that hydrogen consumption for CC-HDO, ATK-HDO, Upgrading process
alumina-HDO, and HDO are 4.0, 2.4, 3.9, and 8.8 mol hydrogen per
mol WCO. Compared to the proposed integrated catalytic crack- (a) H2 consumption based on per mol feed
ing − hydrotreating technology (0.0256 mol hydrogen for removing
99.7
1 g oxygen with ATK-HDO), Miguel Mercader et al. [26] needed 100.0
H2 consumption (mol/mg liquid products)

more hydrogen (0.0268–0.0338 mol hydrogen for removing 1 g

oxygen) when using the co-processing of hydrotreatment and cat-
80.0
alytic cracking. Richard et al. [35] reported a hydrogen consumption Non-oxygenated liquid products
(1.4–3.5) similar to that of ATK-HDO when they conducted research Total organic liquid products
on mild hydrotreating of bio-oil, but the oxygen conversion was not 60.0
higher than 75.4%.
Grange et al. [36] investigated the hydrogen consumption for the
40.0
deoxygenation of different types of oxygen-containing compounds
and summarized that the consumption of hydrogen was 2–8 per
oxygen-containing compound group, which is much higher than 20.0
10.4
the hydrogen consumption in the proposed integrated technology 8.4 6.9
4.0
7.7 7.2
3.8
(0.8–1.3 per COOC ).
0.0
Even though the total oxygenate content was still high at 89.6%, CC-HDO ATK-HDO Alumina-HDO HDO
the hydrogen consumption of HDO, 8.8 mol hydrogen per mol Upgrading process
WCO, was much higher than that of the integrated process under
the same hydrotreating condition (2.4–4.0 mol hydrogen per mol (b) H2 consumption based on per mg liquid products
WCO), which indicates the integrated technology could reduce the
amount of hydrogen consumption needed, especially for ATK-HDO. Fig. 5. Hydrogen consumptions during hydrotreating process using different
This also demonstrates that the hydrogen consumption would be upgrading technologies. *Calculation: oxygen removed and C C double bonds sat-
urated completely based on the compositions of WCO.
much higher than the calculated amount, 13.8 mol per mol WCO, if
the oxygen was completely removed by HDO only.
To take into account the differences in the liquid product yields Loss of sulfur is one of the main reasons for deactivation of
from different technologies, the amounts of hydrogen consumed CoMoS hydrotreating catalyst during the triglyceride hydrodeoxy-
per unit of non-oxygenate organic liquid products/per unit of total genation process [32]. However, the integration technology can
organic liquid products are shown in Fig. 5(b). To limit the hydro- effectively help to keep the sulfur on CoMoS as 90+% of oxygen
gen consumption based on both per mg of non-oxygenate liquid was removed through the FCC process; and the concentration of
products and per mg total organic liquid products, ATK-HDO was oxygen in hydrotreating feed was much reduced from 10.6 wt% to
the best option under investigated conditions. less than 1 wt%. At the level of 0.1 wt% of oxygen (10% FCC upgraded

Please cite this article in press as: H. Wang, et al., Integration of catalytic cracking and hydrotreating technology for triglyceride
deoxygenation, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.009
G Model
CATTOD-10499; No. of Pages 8 ARTICLE IN PRESS
8 H. Wang et al. / Catalysis Today xxx (2016) xxx–xxx

organic liquid products in dodacane), loss of sulfur becomes min- [4] G.W. Huber, A. Corma, Angew. Chem. Int. Ed. 46 (2007) 7184–7201.
imal. Therefore, sulfur loss is not a major concern in the current [5] H. Zhang, H. Lin, W. Wang, Y. Zheng, P. Hu, Appl. Catal. B 150–151 (2014)
(2014) 238–348.
study. Even though the catalyst was deactivated due to a loss of [6] H. Zhang, H. Lin, Y. Zheng, Appl. Catal. B 160–161 (2014) 415–422.
sulfur, resulfiding is a potential way to reactivate the catalyst. [7] S.H. Ng, Y. Zheng, Y. Shi, L.H. Ding, S. Chen, S.H.D. Gieleciak, Catalytic
conversion of plant oil to clean transportation fuels, Pacifichem (2010).
[8] A. Corma, G.W. Huber, L. Sauvanaud, P. O’Connor, J. Catal. 247 (2007) 307–327.
4. Conclusions [9] I. Graca, J.M. Lopes, H.S. Cerqueira, M.F. Ribeiro, Ind. Eng. Chem. Res. 52 (2013)
275–287.
In summary, the integrated catalytic cracking − hydrotreating [10] R. Luque, J. Campelo, J. Clark, Handbook of Biofuels Production: Processes and
Technologies WP Woodhead Publishing Limited, Cambridge, 2011, pp.
technology was successfully applied to the triglyceride deoxygena- 397–412.
tion process. In the proposed process, the majority of oxygen (more [11] K.A. Rogers, Y. Zheng, ChemSusChem 9 (2016) 1750–1772.
than 90%) in triglycerides was removed by catalytic cracking, and [12] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Prieto, J. Bilbao, Ind. Eng. Chem. Res. 43
(2004) 5042–5048.
the liquid products were then hydrotreated with low hydrogen
[13] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Prieto, J. Bilbao, Energy Fuels 18 (2004)
consumption to eliminate residual oxygen and to saturate alkenes 1640–1647.
and cycloalkenes. Compared to alumina and CC, ATK was found [14] https.//en.wikipedia.org/wiki/NEXBTL#cite note-1, NEXBTL, Wikipedia, the
free encyclopedia.
to be the best catalytic cracking catalyst. During the ATK catalytic
[15] F.D. Mercader, M.J. Groeneveld, S.R.A. Kersten, C. Geantet, G. Toussaint, N.W.J.
cracking process, the liquid yield and oxygen removal were high Way, C.J. Schaverien, K.J.A. Hogendoorn, Energy Environ. Sci. 4 (2011)
at 74.4% and over 93.8%, respectively. During the ATK upgraded 985–997.
OLP hydrotreating process, the liquid yield was high at 98.0% and [16] P.M. Mortensen, J.D. Grunwaldt, P.A. Jensen, K.G. Knudsen, A.D. Jensen, Appl.
Catal. A 407 (2011) 1–19.
hydrogen consumption was low at 2.4 mol hydrogen per mol WCO. [17] D. Kubicka, J. Horacek, Appl. Catal. A 394 (2011) 9–17.
The ATK-HDO upgraded OLP was not only high quality (low con- [18] M. Badawi, J.F. Paul, S. Cristol, E. Payen, Y. Romero, F. Richard, S. Brunet, D.
tents of oxygenates (6.6 wt%), cycloalkenes (0.3 wt%), and alkenes Lambert, X. Portier, A. Popov, E. Kondratieva, J.M. Goupil, J. El Fallah, J.P.
Gilson, L. Mariey, A. Travert, F. Mauge, J. Catal. 282 (2011) 155–164.
(8.0 wt%)) but also had good hydrocarbon distribution, consisting [19] O.I. Senol, T.R. Viljava, A.O.I. Krause, Catal. Today 106 (2005) 186–189.
of 40.3% chain hydrocarbons, 27.6% aromatics, and 25.5% cyclic [20] E. Laurent, B. Delmon, J. Catal. 146 (1994) 281–291.
hydrocarbons. [21] I. Graca, J.M. Lopes, M.F. Ribeiro, F.R. Ribeiro, H.S. Cerqueira, M.B.B. de
Almeida, Appl. Catal. B 101 (2011) 613–621.
[22] N. Taufiqurrahmi, S. Bhatia, Energy Environ. Sci. 4 (2011) 1087–1112.
Acknowledgements [23] K. Liu, F.T.T. Ng, Catal. Today 149 (2010) 28–34.
[24] A.A. Lappas, S. Bezergianni, I.A. Vasalos, Catal. Today 145 (2009) 55–62.
[25] J.A. Melero, J. Iglesias, A. Garcia, Energy Environ. Sci. 5 (2012) 7393–7420.
The authors gratefully thank the Natural Sciences and [26] F. de Miguel Mercader, M.J. Groeneveld, S.R.A. Kersten, N.W.J. Way, C.J.
Engineering Research Council of Canada (NSERC strategic: 463140- Schaverien, J.A. Hogendoorn, Appl. Catal. B 96 (2010) 57–66.
2014STPGP, NSERC DISCOVERY: RGPIN-2015-03869), the Canada [27] R.H. Venderbosch, A.R. Ardiyanti, J. Wildschut, A. Oasmaa, H.J. Heeres, J. Chem.
Technol. Biotechnol. 85 (2010) 674–686.
Foundation for Innovation (#31983), and Canada Research Chairs
[28] X.X. Ma, Y. Xia, Angew. Chem. Int. Ed. 53 (2014) 2592–2596.
program (950-228053) for financial assistance. The authors much [29] C.H. Liu, Y.Q. Deng, Y.Q. Pan, S.G. Zheng, X.H. Gao, Appl. Catal. A 257 (2004)
appreciated Dr. Yu Xia and Miss Hilary Brown for their help to test 145–150.
[30] Y. Hu, Study on direct catalytic cracking of high TAN crude oil PhD thesis,
the double bonds position in the feed. Special thanks to Dr. Richard
China University of petroleum (East China), 2011.
Spacek and Miss Margaret Crawford from the Writing & Study Skills [31] H. Wang, Integration of Catalytic Cracking and Hydrotreating Technology for
Centre at the University of New Brunswick due to language editing. Triglyceride Deoxygenation, PhD Thesis, University of New Brunswick, 2015.
[32] D.C. Elliott, T.R. Hart, G.G. Neuenschwander, L.J. Rotness, A.H. Zacher, Environ.
Prog. Sustain. Energy 28 (2009) 441–449.
References [33] C. Dupont, R. Lemeur, A. Daudin, P. Raybaud, J. Catal. 279 (2011) 276–286.
[34] D.S.J. Jones, P.R. Pujadoı́, Handbook of Petroleum Processing, Springer,
[1] C.B.C. News, Canada Sets Carbon Emissions Reduction Target of 30% by 2030, Dordrecht, 2006.
CBC News, 2015, 2015 http://www.cbc.ca/news/politics/canada-sets-carbon- [35] R.J. French, J. Stunkel, R.M. Baldwin, Energy Fuels 25 (2011) 3266–3274.
emissions-reduction-target-of-30-by-2030-1.3075759. [36] P. Grange, E. Laurent, R. Maggi, A. Centeno, B. Delmon, Catal. Today 29 (1996)
[2] http://www.bioenergywiki.net/Renewable fuel targets 297–301.
[3] I. http://www.iea.org, World Energy Outlook 2008, Chapter 11, International
Energy Agency, IEA/OECD, Paris, 2008.

Please cite this article in press as: H. Wang, et al., Integration of catalytic cracking and hydrotreating technology for triglyceride
deoxygenation, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.009