See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/263947103

Robust Aqueous–Nonaqueous Hybrid Process for

Bitumen Extraction from Mineable Athabasca Oil
Sands

Article in Energy & Fuels · April 2012

DOI: 10.1021/ef300270j

CITATIONS READS

17 376

5 authors, including:

Jaroslaw Drelich Zhenghe Xu

Michigan Technological University University of Alberta
217 PUBLICATIONS 3,896 CITATIONS 368 PUBLICATIONS 9,338 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Hydrocyclone Project View project

Zinc Biostent Project View project

All content following this page was uploaded by Jaroslaw Drelich on 24 July 2014.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
 A Robust Aqueous-Nonaqueous Hybrid Process for Bitumen
Extraction from Mineable Athabasca Oil Sands 1
Sanjay Kumar Harjai,1 Chris Flury,1 Jacob Masliyah,1 Jaroslaw Drelich2 and Zhenghe Xu1
1
Department of Chemical and Materials Engineering
University of Alberta
Edmonton, AB T6G 2G6, Canada
2
Department of Materials Science and Engineering
Michigan Technological University
Houghton, MI 49931, USA

ABSTRACT

Mining and processing of Athabasca oil sands in Alberta, Canada, is a great success story of
government-industry collaboration, fulfilling increased domestic and world-wide demands for oil.
However, economic and environmental incentives still exist in the oil sands industry to enhance
oil recovery and reduce energy consumption and water use, while minimizing green-house gas
emissions and tailings ponds. The current industrial bitumen extraction processes, after surface
mining, are exclusively water-based and operate at elevated temperatures, typically between 45
and 50oC. Robust low-temperature processes, with reduced in-take of feed water, that are less
sensitive to ore characteristics are in a strong demand from both environmental and economical
point of views. In response to this demand, we propose a robust aqueous-nonaqueous hybrid
bitumen extraction process, in which diluent such as kerosene and naphtha is added to the oil
sands prior to oil sands slurry preparation to decrease bitumen viscosity and enhance bitumen
liberation. With the proposed hybrid bitumen extraction process, the oil sand processing
temperature can be reduced to ambient temperature. To prove this concept, bitumen recovery
tests were carried out on four Athabasca oil sand ores of good to poor processability, using a
Denver flotation cell operated at ambient temperature. Adding kerosene or naphtha to oil sands
at 4-11 wt% of the bitumen content was found to significantly enhance flotation recovery of
bitumen and bitumen froth quality, especially for poor processing ores. It was found that
kerosene addition not only increased bitumen liberation kinetics determined using our novel in
situ bitumen liberation visualization flow cell (BLVFC), but also improved bitumen aeration
measured by induction time apparatus.

Keywords: Athabasca oil sands, aqueous-nonaqueous hybrid process, bitumen liberation,

bitumen aeration, bitumen flotation

1
Published in Energy & Fuels, Vol. 26, No. 5, 2012, pp. 2920-2927.

1
INTRODUCTION
With great efforts to unlock the vast oil resources in Athabasca oil sands, hot water-based
processes pioneered by Clark have been the principal technology for bitumen extraction from
mineable oil sands (Gray et al. 2009). Although oil production from these oil sands resources is
providing a security for fossil fuel needs into a foreseeable future, the industry is facing great
challenges of water shortage and green-house gas emissions related to bitumen production from
mineable oil sands. To alleviate the challenges associated with water shortage, a great deal of
effort has been made on developing either effective water-recycle technologies (Mikula et al.
2009; Wu et al. 2009) or non-aqueous-based extraction processes (Painter et al. 2010; Hogshead
et al. 2011; Wu and Dabros 2012). The concepts of using supercritical CO2 (Brough et al. 2010;
Roopa and Dowe 2010), ionic liquids (Painter et al. 2010) or solvent (El-Haj et al. 2009) for
bitumen extraction have been proposed and tested at laboratory scales. Although promising, the
technical and economical challenges inherited in these concepts set the roadblocks for them to
move forward to pilot scale tests. With regard to the reduction in green house gas emissions,
effort has been made in developing lower temperature extraction processes. The use of
hydrotransport slurry pipeline for oil sands conditioning to reduce the temperature of bitumen
extraction from the original 80-90°C (Clark 1944) to current limit of 45oC (Long et al. 2007) has
been a successful story of oil sands industry in reducing energy intensity and green house gas
emissions of oil production from oil sands. However a further reduction in operating temperature
with current water-based extraction technology is considered impractical. Furthermore, the
performance of processing at lower temperatures has been found to be more sensitive to ore
characteristics and water chemistry. There is a clear need to develop a robust, low temperature
process for bitumen production from mineable oil sands, which is the subject of this
communication.
In dealing with difficult ores such as Utah oil (tar) sands, Hupka et al. (1983, 1987, 1991, 2004)
developed a diluent-assisted bitumen extraction process at the University of Utah. In this process,
about 5 to 10 wt% of kerosene on the basis of bitumen content was added to Utah oil sands and
water-based bitumen extraction was conducted at 50-55oC and mild alkalinity (pH 8 to 9)
controlled by sodium carbonate addition. By applying the knowledge of solid hydrocarbon (coal)
flotation using kerosene and methyl isobutyl carbinol (MIBC) as process aids, OSLO (Other Six
Lease Owners) developed two distinct extraction processes known as OSLA cold water
extraction at 5-30oC (OCWE) and OSLO hot water extraction at 30-90oC (OHWE), collectively
known as OSLO processes (Sury 1990; Sury et al. 1995). In OSLO process, kerosene (as
collector) and MIBC (as frother) at a 2:1 mass ratio and 0.3 wt% of bitumen content were used
as process aids to improve bitumen recovery. By eliminating the use of caustic as process aids, a
unique feature of these two OSLO processes is the improved settling of extraction tailings (Sury
and Stone 1995). Based on the promising results of OSLO, Syncrude extended the OCWE
process to low energy extraction (LEE) process operated at temperatures between 5oC and 35oC.
In LEE process, MIBC and kerosene at a 1:2 mass ratio were used as the process aids. The
amount of MIBC and kerosene required to achieve an acceptable process performance was found
to vary, depending on the ore grade. Low-grade oil sands need more MIBC/kerosene to attain
desired levels of bitumen recovery (Kasperski 2001).
Further development of LEE process by Sury (1990) led to a patented process of low
temperature bitumen extraction. In this patented process, the oil sand slurry was vigorously
mixed with a conditioning and frothing agent for a sufficient period to release bitumen from oil
sands at temperatures above the freezing point to 35oC, preferably in the range of 2oC to 15oC.

2
The conditioning agent in this case was kerosene, diesel or kerosene/diesel blend at a
concentration of 100-800 ppm. MIBC at 50-400 ppm was used as frother. Recently, Khraisha
(1999) investigated the water-based extraction of oil sand from Wadi Isal in Jordan. In his study,
bitumen extraction was found to be extremely poor even at temperatures as high as 80oC. With
kerosene addition, bitumen recovery increased up to seven times as compared to hot water
extraction without NaOH addition.
Several investigators concluded that there is a close correlation between bitumen recovery and
bitumen viscosity (Hupka et al. 1983, 1987, 2004; Hupka and Miller 1991; Schramm et al. 2003;
Drelich 2008). Hupka et al. (1983, 1987, 2004) demonstrated that a reduction in bitumen
viscosity allows the processing temperature to be lowered by at least 40 degrees as compared to
the original Clark’s process developed for bitumen extraction with boiling water. Kerosene or
other solvent (Yang et al., 1989) was typical added to the oil sands during their pretreatment step
to reduce bitumen viscosity. The amount of kerosene added depends on the original bitumen
viscosity, oil sand grade and intended processing temperature. The kerosene addition levels can
be predicted by using the bitumen viscosity curves developed as a part of the study. The study
shows that to achieve a satisfactory coefficient of separation, the bitumen viscosity must be
reduced to below 2 Pa.s at the processing temperature for a predetermined conditioning time
(penetration time necessary for the diluent to mix with oil sand prior to digestion in order to
uniformly reduce the bitumen viscosity). Schramm et al. (2003) also obtained a bitumen
viscosity threshold value of similar magnitude, about 3 Pa.s, to achieve a satisfactory bitumen
recovery from the Athabasca oil sands. Most recent analysis performed by Long et al. (2007)
suggests that the viscosity threshold value for the Athabasca oil sands is about 7 Pa.s. Although
the viscosity reduction by diluent addition is a key element in improving bitumen recovery, the
alternation of interfacial properties by solvent addition should not be overlooked as the bitumen
liberation is largely controlled by interfacial properties (Drelich and Miller 1994).

CONCEPT OF AQUEOUS-NONAQUEOUS HYBRID EXTRACTION PROCESS

With the proven improvement in bitumen recovery by solvent addition and difficulties
encountered in nonaqueous-based extraction processes (Wu and Dabros, 2012), it is natural to
look into a water-diluent hybrid process which is robust in response to variability of oil sands
ores and water chemistry, and effective at low (ambient) operating temperatures. This is
particularly attractive if one closely examines the current water-based bitumen production
process as shown by solid-line connected flow chart of Figure 1. In water based extraction
process, mined oil sands ores are mixed with hot water and chemical additives, mainly caustics,
to make slurry. The resultant slurry is conditioned through slurry hydrotransport pipelines in
which bitumen becomes liberated and aerated, an invention of oil sands industry. The
conditioned slurry is then fed to a separation vessel where aerated bitumen floats to the top of the
vessel as bitumen froth, typically containing 60 wt% bitumen, 30 wt% water and 10 wt% solids.
To meet the requirement of upgrading for hydrocracking and/or hydrotreating, the bitumen froth
needs to be cleaned by adding a large amount of solvent, typical at 1:2 of naphtha:bitumen
volume ratio or 2.5:1 of hexane:bitumen volume ratio. After separating the solids and water by
gravity separation, the solvent-diluted bitumen is sent to a solvent recovery unit where the clean
bitumen of less than 2 wt% water and 0.5 wt% solids is produced while the recovered solvent is
recycled back to the process. Since an ample amount of solvent is used anyway in the current
bitumen production process, it would be feasible to distribute the solvent addition up front to

3
mined oil sands as indicated by dashed line in Figure 1. After soaking the oil sands at
crushers/breakers with a desired amount of solvents, the current water-based bitumen production
process can be applied equally, but at much lower operating temperatures. Such a concept opens
the prospect of processing oil sands at reduced temperatures and chemical additives without
compromising bitumen recovery and froth quality. Equally important is an anticipated reduction
in difficulties of treating oil sands tailings of aqueous-nonaqueous hybrid process, as caustic
addition is no longer needed for this process, minimizing dispersion of clays and hence avoiding
the generation of slowing settling clay dispersions of negligible consolidation capacity.

Figure 1 Schematics of current water-based bitumen extraction process (solid connections)

and a hypothetic aqueous-nonaqueous hybrid bitumen extraction process (dotted connections). A
partion of organic solvent currently used in froth cleaning is added upfront in oil sands ores prior
to slurry preparation to reduce bitumen viscosity and hence bitumen extraction processing
temperature.
In this communication, our objective is to prove the concept of aqueous-nonaqueous hybrid
bitumen extraction process for Athabasca oil sands by bench scale laboratory tests. To develop
this concept into a robust, low temperature bitumen production process, bitumen extraction tests
with different types of oil sands ores were conducted at ambient temperatures. It is hoped that the
hybrid bitumen extraction process could further lower the current oil sand processing
temperature to ambient temperature. By considering effective recycling of the process water as
indicated by the dashed line in Figure 1, the concerns on solvent loss in the process water could
be alleviated. To understand the role of bitumen soaking with solvent in bitumen recovery,
bitumen liberation tests were performed using our newly designed in situ bitumen liberation
visualization flow cell, while bitumen aeration was studied by determining the induction time of
bitumen-bubble attachment. The success of this novel hybrid process could reduce not only

4
energy consumption and green gas emissions, but also loss of water due to evaporation by
recycle of a maximum amount of process-affected water from the improved characteristics of the
tailings due to the use of the hybrid process.

EXPERIMENTAL PROCEDURE

Materials

Oil Sand Ores: Four different Athabasca oil sand ores, three poor processing ores (marked as
P1, P2 and P3) and one good processing ore (marked as G) from Syncrude Canada, were used in
this study. Characteristics of these four ores are given in Table 1. Good processing ore G
contained 14.6 wt% of bitumen and had a low content of fines (2.1 wt% of solids). The bitumen
content in poor processing ores was 9.6-10.6 wt%, while the fines content varied from 11.5 wt%
for ore P1 to 39.1 wt% for ore P2 (Table 1). Ore samples were stored in plastic bags and kept in
a freezer to avoid any weathering effects, such as water and light hydrocarbon evaporation and
bitumen oxidation. The samples were removed from the freezer and left on the laboratory bench
for 3-4 hours prior to each test.

Process Water: Process water from Syncrude Canada was used in all the experiments. The pH
and surface tension of this water were 8.5±0.3 and 68±2 mN/m, respectively. The process water
contained 14-20 ppm K+, 503-743 ppm Na+, 19-20 ppm Mg2+, 48-56 ppm Ca2+, 431 ppm Cl-, 63
ppm SO42-, 647 ppm HCO3- and <2 ppm NO3-.

Diluent Soaking of Oil Sands Sample

The bitumen extraction tests were performed using the procedures schematically shown in Figure
2. Precisely weighed 300-g oil sand ore sample was placed in an aluminum pan and mixed with a
desired amount of solvent. For the proof of concept, only kerosene was used as the solvent in this
study. The solvent from 2 to 15 mL was spread evenly over the ore sample using a spray bottle.
The ore sample was frequently mixed between sprays. The quantity of the solvent corresponded
to less than 4 wt% to more than 40 wt% of bitumen in the ore. After solvent addition, the pan
with the ore sample was covered with an aluminum foil and left on the laboratory bench (~25oC)
for 15 minutes prior to bitumen extraction test. During this period of kerosene soaking, the
kerosene diffused into bitumen and the loss of volatiles in kerosene was found to be negligible

5
by determining a negligible change in the weight of oil sands and kerosene. In industry
implementations of this process, the ore serge bin and conveyor may need to be enclosed to
avoid the loss of volatile to the atmosphere during continuous operations. In this study, 15-min
soaking was found to be sufficient to complete saturation of bitumen with the added solvent. No
significant difference in bitumen recovery was noticed for samples soaked with solvent for 15
minutes as compared to those soaked for 2 hours. However, the solid-to-bitumen ratio in bitumen
froth was usually better for samples with prolonged soaking time.

Figure 2 Schematics of hybrid bitumen extraction tests in which oil sands ores were soaked
with a given amount of solvent (kerosene) for 15 minutes prior to oils sands ore slurry
preparation and flotation.

Flotation Experiments
The flotation experiments were carried out in 1-L Denver flotation cell. The flotation cell was
equipped with a jacket for circulation of cooling/heating fluid to control the temperature of
flotation. In each test, 950 mL of the process water was poured into the Denver flotation cell.
Circulating water at constant temperature of either 25±2oC or 50±2oC from a thermal water bath
was pumped through the jacket of the flotation cell while the Denver cell agitator was running at
1500 rpm without airflow until the process water reaches the desired flotation temperature. Oil
sand sample was then added to the cell and the agitator was allowed to run for 5 minutes without
air flow to condition the slurry. The air was then added to the conditioned slurry at 150 mL/min
and the froth generated was continuously scooped off using spatula and collected separately into
three or four separate thimbles over different time intervals. The flotation lasted typically for 20
minutes. The bitumen froth samples collected in each time intervals were analyzed for bitumen,
solids and water content using Dean Stark apparatus and toluene as refluxing solvent. When
calculating bitumen recovery, the solvent (kerosene) added to the ore was considered and hence
the recovery represents both bitumen and kerosene, i.e. total hydrocarbon recovery (Harjai,
2007).

Bitumen Liberation Analysis

A novel in situ flow visualization cell equipped with a stereo-optical microscope and a high
resolution digital camera, designed specifically for the bitumen liberation process from oil sands
ore, was used to study the role of kerosene addition on bitumen liberation from sands grains. The
details on the experimental setup were given in a previous report (Srinivasa 2010). Briefly, the

6
flow visualization cell consists of a sample holder with a glass frit on the bottom. A thin layer of
bitumen fed to vacuum distillation unit is spread on a filter paper of the same dimension as the
glass frit. A small lump of oil sand sample with or without kerosene addition/soaking is pressed
onto the bitumen-smeared filter paper placed in the sample holder. The excess oil sands sample
is carefully removed from the top using a knife to obtain a flat surface of oil sands to be exposed
to flowing process fluids. It was noticed that as the bitumen film on the sand grains starts to
recede under the liquid flowing over them to form bitumen droplets on the sand grains, it is
necessary for the oil sand grains to remain intact without the slightest displacement to obtain real
time, high quality images under the stereo-optical microscope. For this purpose, the sample
holder is sealed by a Teflon o-ring under the clamps and connected to a low level vacuum line.
The vacuum is used to hold the oil sands samples in place. The cell is sealed from the top with a
glass slide in a metal cover plate. A set of baffles is provided at the exit end of the fluid to
prevent eddy formation inside the cell. Process water of pH 8.5 from Syncrude Canada Ltd.,
maintained at 35oC with a water bath, was pumped into the flow visualization cell with a
peristaltic pump while the images were recorded by computer at a desired 100-millisecond image
capture speed. The digital camera transfers high quality images to a computer screen. The entire
bitumen liberation process recorded with a high resolution camera is analyzed off-line to
determine degree of bitumen liberation (DBL) as a function of run time. The degree of bitumen
liberation was calculated by determining the percentage of clear sand grains out of total sand
grains visible on each frame (Srinivasa 2010).

Induction Time Measurement

Bitumen aeration was studied by determining the induction time of bubble-bitumen attachment
in process water. A custom-built induction time devise (Gu et al. 2003) was used to measure the
minimum contact time required for an air bubble to attach to a bitumen surface, which is known
as the induction time. The shorter the induction time, the faster the attachment occurs. A
vacuum distillation feed bitumen was used for induction time measurement. The bitumen with
or without kerosene addition/soaking was placed in a semi-spherical bowl cut on a circular
Teflon disk. The excess bitumen was removed by a razor blade to make the bitumen surface
even with the top surface of the Teflon disc. The bitumen surface prepared as such was allowed
to sit in a covered glass dish at room temperature for approximately 30 minutes to make a
surface of near-molecular smooth finish under a slightly fluidic condition. The smooth sample
was transferred into a small rectangular cell that had been filled with the process water from
Syncrude Aurora plant. The solution pH was adjusted using 1 M NaOH and HCl solutions to pH
8.5. The bitumen sample was allowed to equilibrate in the process water for 30 minutes prior to
induction time measurement.
To determine the induction time, an air bubble of 1.5 mm in diameter was generated by a
microsyringe at one end of a glass capillary tube, with the other end of the glass capillary tube
connected to a speaker diaphragm used as a drive for the bubble. The bubble was brought down
into contact with bitumen surface in the test solution for a specified period and then retracted
from it at a constant speed. The attachment was visually observed with the aid of a charge-
coupled device (CCD) camera connected to a computer monitor. This process was carried out 20
times per specified contact time and the number of attachments was recorded to calculate the
percent attachment (number of attachments over the total number of trials). The percent
attachment was plotted as a function of bubble-bitumen contact time. The contact time with a
50% attachment was considered as the induction time. All the measurements were carried out at

7
35°C. The initial separation distance between the air bubble and bitumen surface was kept
constant at 0.25mm, while the bubble displacement was kept at 0.4 mm.

Characterization of Tailings Water

A critical aspect of using solvent for enhancing bitumen recovery is minimizing solvent loss,
which requires analysis of solvent distribution during the extraction process. To estimate the loss
of kerosene in the tailings water, which is a major concern, the solvent analysis in this study is
limited to tailings water from bitumen extraction experiments, conducted for poor processing
ore-P1 soaked in kerosene at 50oC and 25oC for 15 min. The tailings generated during the
bitumen extraction from kerosene-soaked ore using Denver cell were carefully transferred into
500-mL PPCO (Polypropylene copolymer) centrifuge tubes and centrifuged at 33 rev/s for 20
min to separate solid particles from the tailings water. The solids-free tailings water was
carefully transferred into 250-mL glass bottles and sent to a commercial analytical laboratory
(Maxxam Analytics, Edmonton Environmental) for further analysis. The carbon content in the
tailings water samples was determined chromatographically. Since kerosene is a mixture of
C10-C20 hydrocarbons, its concentration was determined by difference method based on the
amount of C10-C20 in the water samples.

RESULTS AND DISCUSSION

Effect of Temperature
Flotation tests were first conducted at two different temperatures (25 and 50oC) to provide
reference data for flotation of bitumen from kerosene-soaked oil sands ores. The tests were done
using one good and two poor processing ores to demonstrate the effect of temperature on
bitumen recovery and froth quality from various oil sands ores. The results in Figure 3 show a
quite high bitumen recovery of 80% from good processing ore G even at room temperature of
25oC. Increasing processing temperature from room temperature to 50oC did increase bitumen
recovery from 80% to above 90%, as anticipated. For both poor processing ores P1 and P2,
bitumen recovery obtained at room temperature is very low at 30 to 48%, respectively.
Increasing temperature to 50oC increased bitumen recovery significantly to above 80%, reaching
the processability of good processing ore G at room temperature. The viscosity of the Athabasca
bitumen at 25oC is about one order of magnitude larger than that at 50oC (Helper and Smith
1994; Hupka et al. 2004; Zhang 2012), which has been widely considered as the cause for lower
bitumen recovery at lower processing temperatures.

8
Figure 3 Effect of processing temperature on bitumen recovery from a good (G) and two
poor (P1 and P2) processing ores at natural pH of 8.5 without caustic addition, showing a
significant increase in bitumen recovery from poor processing ores P1 and P2 and marginal
improvement in bitumen recovery from good processing ore G with increasing processing
temperature. The total flotation time was 10 min.

Although increasing processing temperature increased bitumen recovery, the results in Figure 4
show a slight penalty of decreasing bitumen froth quality (measured by bitumen to solids ratio,
B/S ratio) with increasing bitumen recovery. For good processing ores, for example, bitumen to
solids ratio dropped from around 10 at room temperature to around 8.3 at 50oC. In mineral
processing, the observed decrease in bitumen froth quality (grade) with increasing bitumen
recovery is known as grade-recovery trade off, most likely due to an increased recovery of fine
solids entrained within process water. Similar grade-recovery trade off was observed for poor
processing ore P2, although it was less significant. For poor processing ores, the bitumen froth
quality is much lower with B/S ratio much less than 1. The lower bitumen froth quality is
attributed to poor bitumen liberation from sand grains of high fines ores as observed in a
previous study (Srinivasa 2010) and more severe fines entrainment to bitumen forth with process
water. More interestingly, when the processing temperature increased from 20oC to 50oC, the
bitumen to solids ratio for poor processing ore P1 remained almost the same even though a
significant bitumen recovery was observed. A substantial drop in B/S ratio from 1.5 to 0.4 was
observed for poor processing ore P2 when bitumen recovery increased from 48% to 86% with
increasing processing temperature from 25oC to 50oC. Overall, these results clearly demonstrate
a critical role of processing temperature in bitumen recovery from oil sands ores, in particular for
processing of poor processing ores.

9
Figure 4 Effect of processing temperature on bitumen froth quality (B/S ratio) from a good
(G) and two poor (P1 and P2) processing ores at natural pH of 8.5 without caustic addition,
showing a marginal decrease in bitumen froth quality with increasing processing temperature.
The total flotation time was 10 min.

Effect of Kerosene Soaking (Hybrid Process)

To prove the concept of aqueous-nonaqueous hybrid extraction process illustrated in Figure 1,
bitumen extraction tests were conducted at room temperature using the oil sands ores with or
without soaking with kerosene. The ultimate goal of the hybrid bitumen extraction process is to
reduce thermal energy intensity and GHG emissions by operating bitumen extraction at ambient
temperatures.
In this study, the dosage of kerosene is scaled by the bitumen content in the ore. For an ore of 10
wt% of bitumen, for example, 0.5 g of kerosene is added to 1 kg of oil sands ore to have a
kerosene dosage of 5 wt% of bitumen. The results in Figure 5 shows a substantial increase in
bitumen recovery from 80% to nearly 99% with good processing ore G when it is soaked with
kerosene at 5 wt% of bitumen. Within experimental error, a further increase in kerosene addition
to 11 wt% of bitumen showed no further increase in bitumen recovery. For poor processing ores,
a much more significant increase in bitumen recovery was observed when soaking of the ores by
kerosene. For poor processing ore P1, for example, the addition of kerosene at 5 wt% of bitumen
increased bitumen recovery from 30% to 60%. A further increase in kerosene addition led to
further increase in bitumen recovery, but to a much less extent. Similar improvement was
observed for poor processing ore P2. Compared with the results of bitumen recovery achieved at
50oC in Figure 3, the results in Figure 5 show that adding kerosene to oil sands ores at 5 to 10
wt% of bitumen in the ore prior to their slurry preparation could reduce bitumen extraction
temperature from 50oC to 25oC without sacrificing oil recovery, clearly demonstrating the
benefit of adding solvent (kerosene) in reducing energy intensity and GHG emissions of oil
sands processing.

10
Figure 5 Effect of kerosene addition/soaking on bitumen recovery from a good (G) and two
poor (P1 and P2) processing ores extracted at 25oC and natural pH of 8.5 without caustic
addition, showing a significant increase in bitumen recovery from poor processing ores P1 and
P2 and marginal improvement in bitumen recovery from good processing ore G . The kerosene
content is in reference to bitumen content in the ore and total flotation time was 10 min.

The results in Figure 6 show a negligible effect of kerosene addition/soaking on bitumen froth
quality for all three ores tested. This finding suggests that kerosene addition/soaking is more
beneficial for improving bitumen recovery without sacrificing bitumen froth quality. Considering
both bitumen recovery and froth quality, the addition of kerosene to oil sands makes extraction
process not only feasible at lower operating temperature, but also less sensitive to ore grade, as
two poor processing ores could reach almost identical bitumen recovery for a given kerosene
addition withot sacrificing bitumen froth quality. A kerosene addition at 10 wt% of bitumen in
the ore seems to be optimal. Further kerosene addition showed diminished return in bitumen
recovery, but may run into the risk of significant solvent loss to tailings. Compared with a
solvent to bitumen ratio of 0.7 to 1 (1:1.4) as used in froth cleaning by naphtha addition, the
solvent to bitumen ratio of 1 to 9 added in the aqueous-nonaqueous ambient temperature
extraction process is acceptable. In fact, only a small fraction of solvent is needed to soaking the
oil sands ores and majority of solvent remains to be added to bitumen froth to minimize solvent
loss.

11
Figure 6 Effect of kerosene addition/soaking on bitumen froth quality (B/S ratio) of a good
(G) and two poor (P1 and P2) processing ores extracted at 25oC and natural pH of 8.5 without
caustic addition, showing a negligible effect of kerosene addition on bitumen froth quality
measured by B/S ratio. The kerosene content is in reference to bitumen content in the ore and
total flotation time was 10 min.

Kerosene Loss to Tailings Water

Solvent is typically a higher value product than bitumen. A major concern for implementing
aqueous-nonaqueous hybrid bitumen extraction process is therefore the solvent loss in tailings
water. A set of experiments was designed to estimate the loss of kerosene in the tailings water.
The kerosene loss into the tailings water was analyzed chromatographically by an external
agency, Maxxam Analytics (Edmonton Environmental Ltd.). Since kerosene is a mixture
comprising mainly of C10 to C20 hydrocarbons, any kerosene lost to the tailings water from the
kerosene-soaked ore during the extraction can be estimated by the increase in the C10-C20
content in the tailings water. With this analytical approach in mind, the concentration of C10-
C20 in mg/L present in the tailings water was determined by peak area summation of the
chromatogram under retention volume of C10-C20 for each sample. Table 2 shows the

12
concentration difference of kerosene in the tailings water and estimated percentage losses of
kerosene to tailings water during processing of poor ore P1 at 25oC.
As shown in Table 2, the concentration difference of C10-C20 hydrocarbons in tailings water is
9.4 mg/L at processing temperature of 25oC when the initial kerosene addition is 11 wt% of
bitumen. This concentration level accounts for about 0.1% kerosene loss to the tailings water,
which is considered to be acceptable, in particular when taking the recycle of the process-
affected water into consideration. As the kerosene addition increases to 41.2 wt.% of bitumen,
the concentration difference of C10-C20 hydrocarbons in the tailings water increases to about 57
mg/L, equivalent to 0.4% total kerosene loss to the tailings water. It is therefore highly desirable
to minimize solvent addition to the oil sands ores in the aqueous-nonaqueous hybrid bitumen
extraction process. From technological point of view, the losses of kerosene to the tailings water
are therefore small.
Working Mechanism of Kerosene Addition
To further understand the role of kerosene addition on improving bitumen recovery, the bitumen
flotation kinetics was examined using a third poor processing ore, P3. The results obtained at
35oC are shown in Figure 7. The solid lines in the figure represent predicted flotation recovery
using the flotation rate constant (k) and ultimate flotation recovery (R∞) obtained by fitting the
recovery data (R) to the first order flotation kinetics model, (Wills and
Napier-Munn, 2006). Also shown in this figure are bitumen froth quality (B/S ratio) and
flotation rate constant (k). The reason to run this set of tests at 35oC is to see whether a slight
increase in processing temperature with kerosene addition could lead to an improvement in both
bitumen recovery and froth quality.

Figure 7 Effect of kerosene addition on bitumen flotation kinetics (k) and bitumen froth
quality (B/S ratio) of a poor processing ore P3. The extraction was conducted using a Denver
flotation cell operated at 35oC and natural pH of process water (pH 8.5). A significant increase in
both flotation rate constant (k), ultimate flotation recovery (R∞) and bitumen to solids (B/S) ratio
with kerosene addition was seen.

13
A significant increase in bitumen flotation kinetics with an increase in the flotation rate constant
from 0.57 min-1 to 2.08 min-1 was observed with a kerosene addition at 14.5 wt% of bitumen in
the oil sands ore. The ultimate bitumen recovery also increased from 21.2% to 60.8% and 92.2%
with the addition of 4.5 wt% and 14.5 wt% of kerosene addition, respectively. Accompanying
this increase in flotation rate and ultimate bitumen flotation recovery is the increase in bitumen
froth quality of B/S ratio from 0.57 to 0.15 and 2.08, respectively. The results in Figure 7 clearly
demonstrate the benefit of operating aqueous-nonaqueous hybrid bitumen extraction process at
temperatures slightly higher than the ambient temperature is highly beneficial, increasing both
bitumen recovery and froth quality.
It is well established that bitumen recovery involves two essential steps: bitumen liberation and
aeration. It is therefore of great interest to investigate the key step(s) of solvent (kerosene)
addition in improving bitumen recovery and improved bitumen froth quality by analyzing these
two sub-processes. Bitumen liberation kinetics was analyzed using our recently developed on-
line bitumen liberation visualization flow cell (Srinivasa 2010). The results obtained at 35oC as
shown in Figure 8 revel a significant increase in bitumen liberation kinetics and high ultimate
degree of bitumen liberation with increasing addition of kerosene. For example, a kerosene
addition at 14.5 wt% of bitumen could improve ultimate degree of bitumen liberation from 30%
to 90%. The video images in Figure 8 show much cleaner sand grains with kerosene addition. It
is evident that slower liberation and lower ultimate degree of bitumen liberation from poor
processing ore P3 without kerosene addition is responsible for its lower bitumen recovery
kinetics (k) and lower ultimate bitumen recovery (R∞). Increasing kerosene addition was able to
improve both bitumen liberation kinetics and ultimate degree of bitumen liberation. Such an
improvement in bitumen liberation kinetics can be attributed to the observed decrease in bitumen
viscosity with increasing kerosene addition (Zhang 2012), which leads to an improved migration
of natural surfactant from bulk bitumen to bitumen-water interface. The increased accumulation
of natural surfactant at bitumen-water interface would reduce its interfacial tension and also
affect surface wettability of solids by adsorption of these natural surfactant, both contributing to
improved bitumen liberation. Similar improvement could be accomplished by increasing
processing temperature, leading to improved bitumen recovery as shown in Figure 3. Similar
improvement on bitumen liberation from a glass surface with kerosene addition was reported
recently (Ren et al. 2012).

14
Figure 8 Effect of kerosene addition on bitumen liberation kinetics from poor processing
o
ore P3 at 35 C and pH 8.5. The insets are snapshot video images of visualizing bitumen
liberation at 600 s. A clear increase in bitumen liberation kinetics and ultimate degree of
bitumen liberation (DBL) with increasing kerosene addition is seen.

To investigate the effect of kerosene addition on bitumen aeration, the induction time of bitumen
attachment to air bubble was measured using bitumen samples with and without kerosene
addition/soaking. The results in Figure 9 were obtained in process water of pH 8.5 at 35oC.
From the probability profile of bitumen-air bubble attachment, the induction was found to
decrease from more than 1000 ms to 600 and 510 ms with 4.5 and 14.5 wt% kerosene addition,
respectively. The shortened induction time would greatly enhance bitumen attachment to air
bubbles in flotation and hence increase bitumen recovery kinetics with increasing kerosene
addition as shown in Figure 7. Su et al. (2006) reported a similar improvement in bitumen
recovery and shortened induction time when kerosene covered air bubble was used in the tests.
The increased mobility of bitumen-water interface with kerosene addition appears to be
responsible for the shortened induction as the natural surfactant accumulated at bitumen-water
interface can be pushed away from the contact area under hydrodynamic forces of approaching
bubbles, facilitating the thinning of intervening aqueous film between the air bubble and bitumen
and reducing the induction time.
It is clear that kerosene addition provides a double benefit in enhancing bitumen recovery:
increasing bitumen liberation kinetics and reducing induction time of bitumen-bubble
attachment. The results from our laboratory study clearly prove the concept of ambient
temperature aqueous-nonaqueous hybrid bitumen extraction process to be feasible and attractive.
The implementation of such concept as illustrated in the concept flow diagram of Figure 1 would

15
reduce significantly the energy intensity of bitumen extraction by operating the extraction at
ambient temperature without incur additional operating cost. It should be noted that although
our study is mainly on kerosene addition, with fine tuning, other solvents that is industrially
available, such as naphatha or other heavy residues from bitumen upgrading could be used in
place of kerosene. Assessment in terms of environmental impact, economic viability and process
engineering would be indispensable to compare, aqueous, non-aqueous and hybrid methods for
recovery of bitumen from mineable Athabasca oil sands.

Figure 9 Effect of kerosene addition on bitumen-air bubble attachment in process water of

o
pH 8.5 at 35 C. The addition of kerosene significantly reduces induction time of bubble-bitumen
attachment, measured at 50% attachment.

Conclusion
This study demonstrated the feasibility of aquous-nonaqueous hybrid bitumen extraction process
to reduce the oil sands extraction temperature to ambient temperature without sacrificing
bitumen recovery and froth quality while reducing the energy intension of the process.
Addition/soaking of kerosene (as well as naphtha) to the oil sands prior to slurry reparation
proved to be beneficial to processing the Athabasca ores at 25oC. The proposed novel aqueous-
nonaqueous hybrid process was found to be less sensitive to the characteristics of poor oil sands
ores, maintaining high bitumen recovery. For the Athabasca oil sands ores studied, the optimum
amount of kerosene appear to be in the range of 4 to 11 wt% of the bitumen in the ore. The
kerosene addition to the oil sands ores was found to enhance bitumen liberation kinetics and
ultimate degree of bitumen liberation and reduce induction time of bitumen-bubble attachment.
The solvent loss to the tailing water at 0.1% is considered manageable.

16
Acknowledgement
The financial support for this work from the NSERC Industrial Research Chair in Oil Sands
Engineering is gratefully acknowledged.

References

Brough, S.A.; Riley, S.H.; McGrady, G.S.; Tanhawiriyakul, S.; Romero-Zeron, L.; Willson, C.D.
Low Temperature Extraction and Upgrading of Oil Sands and Bitumen in Supercritical
Fluid Mixtures. Chem. Comm. 2010, 46, 4923-4925.

Clark, K. A. Hot-Water Separation of Alberta Bituminous Sand. Trans. Can. Inst. Min. Metall.
1944, 47, 257–274.

Drelich, J. Wetting Phenomena in Oil Sand Systems and Their Impact on the Water-Based
Bitumen Extraction Process. Miner. Metal. Proces. 2008, 25, 1-12 .

Drelich, J.; Miller, J.D. Surface and Interfacial Tension of the Whiterocks Bitumen and Its
Relationship to Bitumen Release from Tar Sands During Hot Water Processing. Fuel 1994,
73, 1504-1510.

El-Haj, R.; Lohi, A.; Upreti, S.R. Experimental Determination of Butane Dispersion in Vapor
Extraction of Heavy Oil and Bitumen. J. Petrol. Sci. Eng. 2009, 67, 41-47.

Gray, M., Xu, Z.; Masliyah, J. Physics in the Oil Sands of Alberta. Physics Today. 2009, 62, 31-
35.

Gu, G.; Xu, Z.; Nandakumar, K.; Masliyah, J. Effect of Physical Environment on Induction Time
of Air-Bitumen Attachment. Int. J. Miner. Process. 2003, 69, 235-250.

Harjai, S. Effect of Solvent Addition to Oil Sand Processing. MSc. Thesis, University of Alberta,
2007.

Helper, L. G.; Smith, R. G. The Alberta Oil Sands: Industrial Procedures for Extraction and
Some Recent Fundamental Research, AOSTRA Technical Publication Series #14,
Edmonton, AB, Canada, 1994.

Hogshead, C.G.; Manias, E.; Williams, P.; Lupinsky, A.; Painter, P. Studies of Bitumen-Silica
and Oil-Silica Interactions in Ionic Liquids. Energy Fuels 2011, 25, 293-299.

Hupka, J.; Miller, J. D. Moderate-Temperature Water-Based Bitumen Recovery from Tar Sand.
Fuel 1991, 70, 1308–1312.

Hupka, J.; Miller, J. D.; Cortes, A. Importance of Bitumen Viscosity in the Hot Water Processing
of Domestic Tar Sands. Mining Eng. 1983, 35, 1635–1641.

17
Hupka, J.; Oblad, A. G.; Miller, J. D. Diluent-Assisted Hot-Water Processing of Tar Sands,
AOSTRA J. 1987, 3, 95–102.

Hupka, J.; Miller, J. D.; Drelich, J. Water-Based Bitumen Recovery from Diluent-Conditioned
Oil Sands. Can. J. Chem. Eng. 2004, 82, 978-985.

Kasongo, T.; Zhou, Z.; Xu, Z. H.; Masliyah, J. Effect of Clays and Calcium Ions on Bitumen
Extraction from Athabasca Oil Sands using Flotation. Can. J. Chem. Eng. 2000, 78, 674-
681.

Kasperski, K.L. A Review of Research on Aqueous Extraction of Bitumen from Mixed Oil
Sands, CETC-Devon 2001, 01-17.

Khraisha, Y. H. Study of Extraction and Pyrolysis of Jordan Tar Sand. Int. J. Energy Res. 1999,
23, 833-839.

Lelinski, D.; Drelich, J.; Miller, J.D.; Hupka, J. Rate of Bitumen Film Transfer from a Quartz
Surface to an Air Bubble as Observed by Optical Microscopy. Can. J. Chem. Eng. 2004, 82,
794-800.

Long, J.; Xu, Z.; Masliyah, J. On the Role of Temperature in Oil Sands Processing. Energy Fuels
2005, 19, 1440-1446.

Long, J.; Drelich, J.; Xu, Z.; Masliyah, J. Effect of Operating Temperature on Water-Based Oil
Sands Processing. Can. J. Chem. Eng. 2007, 85, 726-738.

Mikula, R.J.; Munoz V.A.; Omotoso, O. Centrifugation Options for Production of Dry Stackable
Tailings in Surface-Mined Oil Sands Tailings Management. J. Can. Petrol. Technol. 2009,
48, 19-23.

Painter, P.; Williams P.; Mannebach, E. Recovery of Bitumen from Oil or Tar Sands Using Ionic
Liquids. Energy Fuels 2010, 24, 1094-1098.

Ren, S.; Dang-Vu, T.; Zhao, H.; Long, J.; Masliyah, J.; Xu, Z. Effect of Weathering on Surface
Characteristics of Solids and Bitumen from Oil Sands, Energy Fuels 2008, 23, 334–341.

Roopa, I.; Dawe, R.A. A Laboratory Study of Recovery with Carbon Dioxide around Critical
Conditions of Trinidad’s Heavy Oil and Tar Sands. Petrol. Sci. Technol. 2010, 28, 1544-
1554.

Schramm, L. L.; Stasiuk, E. N.; Yarranton, H.; Maini, B. B.; Shelfantook, B. Temperature
Effects from the Conditioning and Flotation of Bitumen from Oil Sands in Terms of Oil
Recovery and Physical Properties. J. Can. Petrol. Technol. 2003, 42, 55-61.

Srinivasa, S. Study of Bitumen Liberation from Oil Sands Ores. MSc Thesis, University of
Alberta, Edmonton, 2010.

18
 Su, L.; Xu, Z.; Masliyah, J. Role of Oily Bubbles in Enhancing Bitumen Flotation, Minerals
Engineering. 2006, 19, 641-650.

Sury, K.N. Low Temperature Bitumen Recovery Process. U.S. patent 4,946,597, August 7, 1990.

Sury, K.; Stone, J. Non-Clark Extraction Processes and their Tailings Characteristics, in
Advances in Oil Sands Tailings Research (IV). Alberta Department of Energy, Edmonton,
1995.

Wills, B.; Napier-Munn, T.J. Mineral Processing Technology, 7th Edition; BH Elsevier: Oxford,
2006.

Wu, J.; Dabros, T. Process for Solvent Extraction of Bitumen from Oil Sand. Energy Fuels. 2012,
26, 1002-1008.

Wu, M.; Mintz, M.; Wang, M.; Arora, S. Water Consumption in the Production of Ethanol and
Petroleum Gasoline. Environm. Manag. 2009, 44, 981-997.

Yang Y. J.; Bukka, K.; Miller, J. D. Selection and Evaluation of Diluents in the Modified Hot
Water Process. Energy Proces. Canada 1989, 82, 14-21.

Zhang, M. Role of Bitumen Viscosity in Bitumen Recovery from Athabasca Oil Sands. MSc
Thesis, University of Alberta, Edmonton, 2012.

19

View publication stats