PETROLEUM SOCIETY PAPER 2008-096

Correlation of Natural Surfactant

Partitioning and Bitumen Recovery in
Oil Sands Extraction
R.J. MIKULA, V.A. MUNOZ, O. OMOTOSO
Natural Resources Canada, CANMET Energy Technology Centre,
Advanced Separation Technologies Laboratory

K. PERU, J. HEADLEY
Environment Canada, Northern Waters Research Institute

This paper is accepted for the Proceedings of the Canadian International Petroleum Conference/SPE Gas Technology Symposium
2008 Joint Conference (the Petroleum Society’s 59th Annual Technical Meeting), Calgary, Alberta, Canada, 17-19 June 2008. This
paper will be considered for publication in Petroleum Society journals. Publication rights are reserved. This is a pre-print and
subject to correction.

Abstract surfactant concentration in the water phase and inversely to

The role of natural surfactants in bitumen extraction from oil
sands has been extensively researched, but these studies have
focused primarily on surfactants that are soluble in the water
phase. Recent work at CETC–Devon has shown strong
correlations between bitumen weathering or oxidation and poor
recoveries. The reason for the poor recoveries has been shown
to be an increase of surfactant concentration in the
hydrocarbon phase. With oxidized ores, this hydrocarbon-phase
surfactant concentration increases, resulting in a stronger
association between the mineral and bitumen components in the
ore. This association between the mineral and bitumen
components results in a decrease in bitumen recovery, poorer
froth quality, or both.
Quantification of the surfactant partitioning between the
hydrocarbon and water phases has been related to bitumen
recovery for a variety of naturally occurring oxidized ores, as
well as for ores that were oxidized under controlled laboratory
conditions. Optimal recoveries were correlated directly to
surfactant concentration in the bitumen.
Introduction
The CANMET Energy Technology Centre–Devon (CETC–
Devon) at Devon, Alberta has evaluated a variety of extraction
protocols over the years in order to systematically investigate the
parameters affecting the extraction process. An effective batch
extraction test protocol needs to be reproducible, correlate with
commercial experience, and permit extraction assessments using
core samples that are sometimes small. The CETC–Devon protocol
meets all of these criteria and, although it seems unlikely given the
vast differences in scale, very good correlations have been
established between the industrial experience (at 500,000 tonnes per
day) and the 500-g batch extraction test. In any case, regardless of
the details of the batch extraction procedure, a consistently executed

1
procedure will allow for the identification of extraction trends in
terms of the impacts of clays, water chemistry, temperature, or other
variables(1-3). Typically, bitumen recovery and bitumen froth quality
are the most important processability parameters investigated,
although tailings properties (both fluid fine tailings and sand tailings)
also vary and can have important operational consequences
depending upon the ore feed and the extraction conditions. The
focus of this paper is the partitioning of natural surfactants between
the bitumen and water phases in the extraction process. Many
factors affect surfactant partitioning, including temperature and
water chemistry. Over the range of extraction temperatures that is
typically evaluated (20 to 80ºC), surfactant partitioning is more-or-
less unaffected by temperature variations. Water chemistry
variations, however, can be extremely important in the release of
naphthenic acids (the major surfactant component in bitumen) at
elevated pH, or in the precipitation of naphthenic acids from solution
in hard water (high calcium or magnesium concentrations)(2-7).
Experimental
Batch Extraction Testing
Relationships have been established between the fines content,
(commonly defined by the industry as the minus 44-m mineral
fraction) and bitumen content in the Athabasca oil sands ore.
Positive correlations between the bitumen content and recovery are
also well established, as is the negative correlation between fines
content and bitumen recovery. Figure 1 shows a compilation of
extraction recoveries collected over the years for extraction tests with
varying degrees of severity. The low-energy process represents data
collected at 50ºC, but using the extraction protocol developed by
Syncrude Canada Ltd. to emulate the Clark hot water process, which
involved tumbler conditioning, nominally at 80ºC, and the addition
of caustic (NaOH) as a process aid. The medium-energy process is
the CETC–Devon protocol for stirring rate, stirring time, etc., but at
50ºC, designed to emulate the hydrotransport conditioning process in
use by the industry today. Finally, the high-energy process is data
collected using the CETC–Devon protocol, but at 80ºC. The black
bars represent one standard deviation uncertainty in the data sets or
an estimate of the scatter in the data. The poorly processing ore
labeled as oxidized is significantly outside the relationship for
bitumen content and bitumen recovery, and for this ore type, a
process aid such as NaOH might be added in order to improve
bitumen recovery.
Batch extraction testing was carried out on a 500-g sample
basis following a procedure extensively described elsewhere(1).
Sodium hydroxide was used as a process aid in order to
improve the recovery of the oxidized ore to bring it in line with
the expected grade-recovery relationship.
Microscopic Assessment of Degradation or
Surfactant Level in the Hydrocarbon
Bitumen degradation is assessed by determining the presence
of structures in a field of view, measuring them directly using
the cross-hair reticule in the eyepiece, and focusing at different
depths in the structures. In addition to the formation of
structures, the bitumen froths from some problem ores show
variation in the bitumen fluorescence and/or show areas with
very high concentrations of emulsified water droplets less than
10 µm in diameter. The measurements are performed by
2
applying a point-count linear analysis, commonly used in
petrographic studies of coal samples using a single cross-line
eyepiece micrometer. Sample movement is controlled by an
automatic point counter. As each field comes into view, the
sample structure immediately under the cross line is identified
until the surface of the sample has been covered (usually 100 to
200 fields of view).
The microscopic examination was performed in the
fluorescence mode using a Nikon Eclipse E-800 light
microscope attached to a Bio-Rad MRC-1024 confocal laser
scanning microscope. The froth samples were prepared for
microscopic examination by transferring part of the froth from
the batch extraction test into a flat-bottom, 25 x 25 x 2 mm (l w
d) metallic cell and gently covering it with a cover slip.
Mass Spectroscopy for Assessment of
Naphthenic Acids in the Process Water
All aqueous samples are analyzed directly without prior
extraction using loop injection mass spectrometry. Calibration
is established based on a set of five standards prepared from
stock solution using appropriate dilutions with Milli-Q water.
Quantitative analysis is conducted on a Quattro Ultima mass
spectrometer (Micromass, UK) equipped with an electrospray
ionization interface operating in the negative ion mode. Typical
mass spectrometer conditions are as follows: source
temperature 90°C, desolvation temperature 220°C, cone voltage
setting 62 V, capillary voltage setting 2.63 kV, cone gas N2 158
L/h, desolvation gas N2 489 L/h. Typical value for low mass
resolution is 14.1 and high mass resolution 14.3; ion energy
1.7. Entrance voltage is set to 95 V, collision energy 13 eV, and
exit voltage 55 V. The multiplier is set at 450 V. Full-scan MS
is employed for extracts of the mixtures. Loop injections are
typically conducted using a Waters 2695 (Milford, MA)
separation module. The mobile phase consists of 50:50
acetonitrile/water containing 0.1% NH4OH at a flow rate of 200
µL/min. Injection volumes are typically 10 µL.
The results are presented as a distribution of the various "-z"
series of the naphthenic acids, or the cumulative amount can be
expressed as the total naphthenic acid concentration.
Results and Discussion
Oxidized or Degraded Ore Identification
The identification of this particular ore as oxidized ore was based
on a microscopy method developed at CETC–Devon, which
indirectly quantifies the amount of surfactant on the hydrocarbon
phase(8). Under the microscope, hydrocarbon morphology and
interactions with mineral and water components can be related to the
surfactant loading. High surfactant concentrations in the
hydrocarbon enhance association of the hydrocarbon with both the
mineral and water phases. As a result, significant emulsification of
water and/or associations with mineral components can be observed.
Quantification of these microscopic features and, subsequently, the
surfactant concentration in the hydrocarbon can be achieved by a
point counting technique whereby a number of fields of view are
randomly assessed for evidence of water emulsification and mineral
associations. At appropriate magnification, the ratio of the fields of
view having a particular morphology to the total number of fields of
view assessed indicates the degree of oxidation/degradation or, by
inference, the surfactant loading in the hydrocarbon.
Figure 2 shows an image of typical bitumen froth compared to
the degraded or oxidized bitumen froth. Emulsification of water and
association with solids has been related to surfactant concentration in
a variety of heavy oil systems and the same is true for bitumen
froths. It has been shown that this surfactant loading, represented by
the microscopic morphology, is related to oxidative reactions that
can be followed using NIR spectroscopy(9,10). Weathering and
oxidation can dessicate the ore and result in changes to the mineral
components. These more extreme effects also have negative
impacts on extraction while, at the same, time greatly increasing the
surfactant loading on the bitumen phase. In the more severe cases of
weathering, oxidation of the pyrite present in oil sand minerals
produces an acid environment, dissolving calcium carbonate; this
combination of lower pH and increased divalent ion concentration
can severely hamper bitumen flotation processes(11-13). Schramm et
al. related oxidation and weathering to depletion of surfactants and
showed that, by adding caustic to the batch extraction process, the
surfactant concentration could be increased to a maximum value,
that corresponds to a maximum in bitumen recovery(4-6). In the
present study, the microscopic morphology of the bitumen froth is
clearly indicative of an increase in surfactant concentration and a
greater propensity to emulsify water and carry solids with the froth(8).
The increased surfactant concentration in the bitumen increases the
association of bitumen with mineral and water components. This, in
turn, results in either a loss of bitumen with the tailings (mineral
components pulling bitumen out of the flotation process) or an
increase in the mineral and water contents in the bitumen froth
product. The greater the degree of oxidation in the ore, the poorer
the froth quality, and the total bitumen recovery can ultimately be
reduced. It is important to note that oxidation of the bitumen or
hydrocarbon component to increase surfactant concentration in the
bitumen occurs long before dessication or mineral oxidation
becomes apparent(9-11). Most often, the ores with the greatest degree
of oxidation are those that have the greatest connate water content.
Significant operational impacts due to surfactant increases in the
bitumen are apparent long before any of the chemical markers for
oxidation (lowered pH, elevated iron, sulphate, and calcium) can be
measured. By using the microscopic assessment of bitumen
oxidation it is possible to quantify the relative surfactant loading in
the bitumen. In certain cases, it is also possible to determine whether
a particular froth sample is sourced from a single ore or from a blend
of ores with varying degrees of oxidation or surfactant loadings.
Figure 3 shows the variety of froth morphologies observed in
various bitumen froths. Note that the bitumen is imaged in
fluorescence mode and the brightness of the fluorescence is also
related to the surfactant loading or degree of degradation. These
factors have been extensively discussed elsewhere(8), but it is
important to note that there are differences in the detailed
morphological differences or degree of degradation, even for what
would nominally be identified as “degraded” froth for the purposes
of the quantification. This is an important distinction because it
allows one to identify ore blends. For instance, a froth sample in
which only 20% of the fields of view show morphology typical of
surfactant loading might be expected to have relatively mild
indicators. If, however, 20% of the fields of view show extreme
emulsification and association with solids, that would be an
indication that the froth is from a blend of ores, one very seriously
degraded and another relatively fresh.
3
Surfactant Partitioning and the Extraction
Process
It has been shown that caustic addition can improve the
processability (both bitumen recovery and bitumen froth
quality) for oxidized/degraded (or surfactant-rich bitumen) ores.
This was successfully demonstrated not only in laboratory tests,
but also on a commercial scale at the Suncor Energy oil sands
extraction plant. On the laboratory scale, caustic addition has
been shown to move the surfactant from the hydrocarbon
(bitumen) into the water phase. The correlation between
naphthenic acid concentration in the water phase and bitumen
recovery was extensively investigated by Schramm et al.(4-7).
The Schramm work studied poorly processing ores and noted
that the optimum in recovery corresponded to a maximum in the
surfactant concentration in the process water. This study
proposes an alternative explanation, which is that the optimum
in recovery is not due to the surfactant in the water phase, but to
the removal of surfactant from the bitumen phase. In other
words, the maximum in water-phase surfactant concentration is
a symptom of the fact that the bitumen surfactant concentration,
and therefore bitumen-mineral interactions, are minimized.
More work is required to better define the conditions under
which recovery is optimal in order to determine the relative
importance of bitumen chemistry and water chemistry in
determining the processability of various problem ores. In any
case, more attention needs to be paid to the more difficult-to-
characterize bitumen chemistry rather than simply to the clay
and water chemistry as the major factors controlling the
extraction process.
Figure 4 shows the relationship between surfactant
concentration in the hydrocarbon (bitumen) as determined by
microscopy, and the naphthenic acid concentration in the water
phase. The maximum in recovery for this very poorly
processing ore corresponds to the removal of surfactants from
the bitumen and into the water phase. This change in
partitioning from the bitumen to the water phase is relatively
straightforward to monitor using the microscopy methods
developed at CETC–Devon. Figures 5 and 6 show the change
in naphthenic acid speciation in the water phase as the NaOH
concentration increases. The –z family noted in the figures is
related to the number of rings in the naphthenic acid structure
by the formula
CnH2n+zO2 .......................................................................................................... (1)
The factor is a negative even integer where z = 0 denotes no rings,
z = -2 one-ring structures, z = -4 two-ring structures, etc. It has been
suggested that certain z families have greater impacts on toxicity
than others, and the same is probably true for impacts on the
extraction process.
Commercial Implications of Processing
Oxidized or Degraded Ore
Figure 7 shows a froth sample collected from a commercial
operation which clearly shows a difference in fluorescence
behaviour, indicating that a blend of ores produced this froth. Figure
8 shows an image from the same operation, but following passage
through additional piping and a pump. The poorly fluorescing or
degraded (surfactant-rich) froth has a distinctly greater propensity to
emulsify water than the brighter froth. This propensity for
emulsification is related to the surfactant loading in the froth
hydrocarbon, as discussed in the previous section. The greater the
proportion of surfactant-loaded bitumen in the process, the greater
the problems in creating a suitable clean bitumen feed to the
upgrader or refinery(14,15).
In many commercial extraction operations, a fraction of the
bitumen production is from froth collected in the secondary flotation
stage (often flotation units). This secondary flotation is designed to
improve total recovery by collecting bitumen from the primary
separation middlings. The secondary flotation froth is usually
combined with the feed to the primary separation cell where,
presumably, it ultimately reports as the bitumen froth product.
Table 1 shows an assessment of the bitumen degradation (or
surfactant loading) for a variety of ores in batch extraction tests
carried out at 50ºC with no process aids. There is a clear trend for
the secondary froth bitumen to have a significantly higher surfactant
loading. This implies that recirculating this surfactant-rich bitumen
into the primary separation cell does not improve the overall
recovery. A more efficient way to deal with the more surfactant-rich
secondary froth would be to utilize a separate froth treatment process
specially designed to deal with the surfactants and the concomitant
propensity for emulsification with this particular bitumen chemistry.
Summary
Bitumen recovery from poorly processing ores has been
related to the degree of degradation of the bitumen—or the
surfactant loading on the bitumen. Surfactant loading is related
to the degree to which the bitumen has been oxidized. With
caustic addition, minimizing the surfactant loading on the
bitumen corresponds to a maximum in the bitumen recovery.
This also corresponds to a maximum in the naphthenic acid
concentration in the process water.
Schramm et al. concluded that a minimum surfactant
concentration was required in the water phase in order to
maximize recovery, and that weathered (or oxidized) ores were
deficient in surfactants. The present study elaborates on these
conclusions and shows that poorly processing ores, including
oxidized, weathered, or biologically degraded ores, are, in fact,
surfactant-rich. The surfactant species are on the bitumen phase
and contribute both to recovery losses and to poor froth quality
by promoting bitumen-mineral and bitumen-water interactions.
Caustic addition improves recovery by promoting partitioning
of these surfactants from the bitumen into the water phase.
Further work is ongoing in order to better define the
importance of surfactant partitioning between the water and
bitumen phases and bitumen recovery. There is clearly a direct
link between surfactant concentration in the bitumen and
bitumen-mineral-water interactions, distinct from the more
symptomatic association between water-phase surfactant
loading and bitumen recovery and froth quality. Field studies
have noted that the degraded or oxidized oil sands are most
often associated with ores that have relatively high water
contents, and it is believed that association with underground
aquifers provides either oxygen or bacterial routes to bitumen
degradation.
4
Acknowledgements
The authors would like to acknowledge the contribution of
PERD, the federal Panel for Energy Research and
Development, in supporting this research.
NOMENCLATURE
l.w.d length, width, depth
L/min microliter per minute
m micrometer
NaOH sodium hydroxide
NH4OH ammonium hydroxide
L/h liter per hour
MS mass spectrometry
NIR near infra red
PERD Panel for Energy Research and Development
z two times the number of rings in the naphtenic acid
structure
REFERENCES
1. Omotoso, O.E., Mikula, R.J., Slurry tank conditioning of
oil sands: Evaluation of extraction recovery for oxidized
and problem ores; Division Report WRC 98-27 (CF),
CANMET, Natural Resources Canada, 1998.
2. Clark, K.A., Hot-water separation of Alberta bituminous
sand; Transactions of the Canadian Institute of Mining and
Metallurgy, 47, 257-274, 1944.
3. Kasperski, K.L., Review of research on aqueous extraction
of bitumen from mined oil sands; Division Report CWRC
01-17 (CF) CANMET, Natural Resources Canada, 2001.
4. Schramm, L.L., Smith, R.G., The influence of natural
surfactants on interfacial charges in the hot-water process
for recovering bitumen from the Athabasca oil sands;
Colloids and Surfaces, 14:67-85, 1985.
5. Schramm, L.L., Smith, R.G., Stone, J.A., The influence of
natural surfactant concentration on the hot water process
for recovering bitumen from the Athabasca oil sands;
AOSTRA Journal of Research, 1: 5-13, 1984.
6. Schramm, L.L., Smith, R.G., Some observations on the
aging phenomenon in the hot water processing of
Athabasca oil sands. Part 1. The nature of the
phenomenon; AOSTRA Journal of Research, 3:4:195-214,
1987.
7. Schramm, L.L., Smith, R.G., Some observations on the
aging phenomenon in the hot water processing of
Athabasca oil sands. Part 2. The mechanism of aging;
AOSTRA Journal of Research, 3:4: 215-224, 1987.
8. Munoz, V.A., Kasperski, K.L., Omotoso, O., Mikula, R.J.,
The use of microscopic bitumen froth morphology for the
identification of problem oil sand ores; Petroleum Science
and Technology 21 (9-10): 1509-1529, 2003.
9. Mikula, R.J., Munoz, V.A., Wang, N., Bjornson, B., Cox,
D., Moisan, B., Wiwichar, K., Characterization of bitumen
properties using microscopy and near infrared
spectroscopy: the processability of oxidized or degraded
ores; Journal of Canadian Petroleum Technology, Vol 42,
No. 8, August 2002.
10. Wang, N., Mikula, R.J., Small scale simulation of pipeline
or stirred tank conditioning of oil sands: temperature and
 mechanical energy; Journal of Canadian Petroleum
Technology, Vol 41, No. 1, January 2002.
11. Wiwichar, K., Bjornson, B., Mikula, R.J., Identification
and treatment of weathered ores at Suncor’s Steepbank
mine; CIM Annual General Meeting, Minespace 2001,
Quebec City, QC, May 2001.
12. Mikula, R.J., Munoz, V.A., Zrobok, R., Omotoso, O.,
Kasperski, K.L., A water-based extraction process for
Asphalt Ridge tar sands; Division Report CWRC 00-46
(CF), CANMET, Natural Resources Canada, 2000.
13. Mikula, R.J, Munoz, V.A., Omotoso, O., Laboratory and
pilot experience in the development of a conventional
water based extraction process for the Utah Asphalt Ridge
tar sands; Proceedings of the Canadian International
Petroleum Society, Calgary, June 2006, paper 06-131.
14. Mikula, R.J., Munoz, V.A., Kasperski, K.L., Omotoso,
O.E. (CANMET), Burns, R. (Suncor Energy), Tipman, R.,
Firmin, K. (Shell Canada), Bitumen release mechanisms
and new process development; 7th UNITAR International
Conference on Heavy Crude And Tar Sands, No. 1998.266,
pp. 2091-2103, 1998.
15. Omotoso, O., Mikula, R.J., Urquhart, S, Sulimma, H.,
Stephens, P.W., Characterization of clays from poorly
processing oil sands using synchrotron techniques; Clay
Science, Vol 12 (2) 88-93, 2006.

5
 Ore number Primary froth Secondary froth
Vol% degraded or oxidized Vol% degraded or oxidized
1 20 70
2 30 90
3 5 35
4 20 95
5 20 90
6 20 80
7 5 65
8 15 50
9 5 90
10 5 95
11 60 90
12 25 80
13 5 65
TABLE 1. Degree of surfactant loading of the bitumen for a variety of ores in the batch extraction process. The secondary,
or more difficult-to-separate bitumen froth has significantly higher surfactant loading.

100

95

90
y
Total Recovery

ver
co
e 8585
R
l Low Energy
a Low EnergyProcess
Process
tT Oxidized ore
o Oxidized ore
T
80 High EnergyProcess
High Energy Process

Medium
Medium Energy
EnergyProcess
Process
75

70
6 7 8 9 10 11 12 13 14
Per
PerCent
Cent Bitumen inOre
Bitumen in Ore

FIGURE 1. Relationship between bitumen content in the ore and bitumen recovery for different batch extraction test
severities

6
FIGURE 2. Microphotograph showing a bitumen froth with relatively little surfactant loading (top) compared to a froth
with a significant surfactant loading (bottom). Note the lower fluorescence of the degraded bitumen in both images.

FIGURE 3. Microphotograph showing froth with significant surfactant loading as evidenced by a complex emulsion
formation carrying water and solids

7
 90 200

Surfactant in the bitumen

80 160
Bitumen Recovery

Surfactant in ppm
70 120
Surfactant in the water

60 80

Bitumen Recovery
50 40

40 0
10 100 1000 10000
NaOH Addition in ppm

FIGURE 4. Relationship between surfactant concentrations in the bitumen and water phases and batch extraction recovery.
Surfactant in the bitumen phase was determined by association with the microscopic morphology and is in
arbitrary units.

8
 50ppm NaOH test 25

2.5

2.0

1.5

1.0

0.5

0.0
5
10
15
20
14
25 12
10
30 8
6
35 4
2
40 0

FIGURE 5. Naphthenic acid speciation in the water phase when recovery was poor, showing that the single-ring-
structure species dominate (123.5 ppm total naphthenic acid)

9
 1600ppm NaOH test 39

2.5

2.0

1.5

1.0

0.5

0.0
5
10
15
20
14
25 12
10
30 8
6
35 4
2
40 0

FIGURE 6. Naphthenic acid speciation in the water phase when recovery was at a maximum, corresponding to the
maximum in total naphthenic acid (167.3 ppm). In this case the z = -4 or two-ring structures dominate.

10
FIGURE 7. Froth sample from the interstage tank in a commercial scale oil sands operation. This froth is clearly
from a blend of ores with different bitumen chemistries as evidenced by the difference in fluorescence behaviours.

FIGURE 8. A commercial froth sample taken in the same time period as the sample shown in Figure 7. In this case
the sample was taken from beyond the interstage tank but prior to diluent addition and froth treatment. This
downstream sample has been through several metres of piping and a pump and the shear has resulted in a
significant increase in the amount of emulsified water in the poorly fluorescing (surfactant-rich) bitumen.

11