See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/271014497

New Laboratory Core Flooding Experimental

System

Article in Industrial & Engineering Chemistry Research August 2014

DOI: 10.1021/ie501866e

CITATIONS READS

4 1,113

3 authors, including:

Alexey Baldygin David Nobes

University of Alberta University of Alberta
7 PUBLICATIONS 24 CITATIONS 112 PUBLICATIONS 683 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Alexey Baldygin on 18 January 2015.

The user has requested enhancement of the downloaded file.

 Article

pubs.acs.org/IECR

New Laboratory Core Flooding Experimental System

Aleksey Baldygin, David S. Nobes, and Sushanta K. Mitra*,

Micro and Nanoscale Transport Laboratory, Department of Mechanical Engineering, University of Alberta, Edmonton, Alberta T6G
2G8, Canada

Optical Diagnostic Group, Department of Mechanical Engineering, University of Alberta, Edmonton, Alberta T6G 2G8, Canada

ABSTRACT: This paper focuses on nding ways to improve the traditional core ooding experimental setup that has been used
by the reservoir engineers over the past decades. The new proposed setup can be used in contemporary studies related to
enhanced oil recovery. This setup has a possibility of using dierent ooding agents, e.g., surfactant, polymer, emulsion, oil and
water. It also includes an automated euent analysis, which has been developed to provide estimates on oil recovery eciency.
For validation purposes, the core ooding setup has been tested with an unconsolidated one-dimensional sand pack as a porous
medium. Traditional water ooding experiments with paran oil and water were conducted at rst. Also, two types of emulsion
ooding techniques were tested for the sand packs: the direct emulsion ooding and the water ooding followed by the emulsion
ooding as an example to exploit the capability of the new setup to successfully perform enhanced oil recovery techniques.
Hence, this setup provides a valuable tool for the reservoir engineers to test the dierent ooding strategies in a laboratory scale
experiment, before committing to huge resources in terms of man-power and cost in actual drilling operations in oil reservoirs.

INTRODUCTION
The ever pressing need for exploring new hydrocarbon
automated core ooding platform that can be tested for various
types of reservoir conditions using some of the advancements
resources in some of the unexplored regions of the world1 in uid controls and data processing and automation. For
validation purposes for the developed system, initially experi-
has made oil and gas industry to compete with each other and
ments are conducted with water and paran oil only, viz., water
develop new and ecient processes for extracting this precious
ooding. To anticipate the future performance of core ooding
commodity out of the natural reservoirs. However, this is a
system for potential enhanced oil recovery (EOR) studies,
huge undertaking for the overall industry and, for instance,
applicability of the system for ooding complex uids, like
China alone spends close to 13 billion dollars each year for oil
emulsion, is also demonstrated here.

and gas exploration,2 which involves selection of preferred sites,
drilling operations including the appropriate design of injection
and production wells, etc. In recent years, the cost has also
escalated due to the tendency of the industry to try dierent
ooding uids (chemicals, surfactants, polymers, etc.)35 to
displace the resident oil phase and achieve record level of oil
production (92 mb/d).6 The obvious question that arises is that
whether there can be ways one can predict the oil recovery
process for new geological formations and more importantly
the applicability of a new ooding strategy (i.e., combination of
wells, ooding uids, etc.) in a laboratory scale without
embarking on a costly venture of direct drilling in oil elds.
To this extent, the oil and gas industry heavily relied on
laboratory core ooding experiments, where a core pack, often
unconsolidated sand particles710 or, in some cases, con-
solidated cores (sandstone and carbonate outcrops),1114
mimicking the porous natural reservoir, are traditionally
ooded with an injection uid and the resultant output in
terms of oil recovery is measured. However, this age old core
ooding system and the process need to be made more robust,
automated, and more in tune with the advancements of uid
machineries of the twenty-rst century, rather than relying on a
core ooding system, which was introduced way back in early
1930s.15 This is more of the need of the hour, particularly when
the industry is exploring new options for ooding the existing
oil elds abandoned after the primary recovery process.16 Here,
we provide a critical solution to this pressing need for the oil
and gas industry by providing a robust, versatile, precise, and
TRADITIONAL CORE FLOODING SYSTEMS
The traditional core ooding system, as described in the
existing literature,9,11 has three main components: the up-
stream, the core block, and the downstream. The upstream
supplies relevant uids (saturation and ooding agents) to the
core block using syringe pumps and piston accumulators (PA).
The core block contains the porous media and often simulates
the reservoir conditions, which could include overburden
pressure and a thermal jacket to maintain reservoir temper-
ature. The downstream collects the euent from the core block
by using fraction collectors17 or two-/three-phase separa-
tors,18,19 the analysis of such data often leads to the estimation
of the eciency of tested oil recovery method. All system
components are connected to a data acquisition module
through which the user can communicate, which is referred
here as a control module. Such experimental techniques work
well for single-phase uids (such as water or oil) and certain
restricted two-phase systems (oil and water mixture as euent
from the production wells). New advancements, both in the
eld and experimental techniques, for enhanced oil recovery
dictate new requirements to the age old traditional core
Received: May 7, 2014
Revised: July 2, 2014
Accepted: August 8, 2014
Published: August 8, 2014

2014 American Chemical Society 13497 dx.doi.org/10.1021/ie501866e | Ind. Eng. Chem. Res. 2014, 53, 1349713505
Industrial & Engineering Chemistry Research Article

Table 1. Summary of Existing Core Flooding Systems Used Globally

author year upstream downstream used uids
Hornof and Morrow 1987 metering pump n/a isooctane, brine, water, and oil
Chang and Grigg 1994 syringe pumps with PA multiport rotary valve with vials CO2, brine, oil, and surfactant solution
Legowo and Pratomo 1999 displacement pump Quizix with PA fraction collector and separator water, oil, and dierent microbial cultures
Bagci et al. 2001 syringe pump fraction collector water, oil, and dierent microbial cultures
Sedaee 2004 displacement pump with PA euent condenser with collector steam, CH4, and heavy oil
Nobakht et al. 2007 syringe pump with PA oil sample collector with gas ow meter crude oil, and CO2
Hadia et al. 2007 dual piston syringe pump with PA fraction collector oil and water
Mandal et al. 2010 syringe pump with PA fraction collector oil-in-water emulsion, oil, and water
Samanta et al. 2012 syringe pump with PA fraction collector oil, brine, and chemical slug
Veerabhadrappa et al. 2013 syringe pump with PA euent sample collector oil, water, and polymer
Ko et al. 2014 syringe pump with PA fraction collector with centrifuge tubes brine, oil, and surfactant

Figure 1. Schematic of the new core ooding experimental apparatus.

ooding systems. Use of complex uids such as emulsions as a
ooding agent requires modications of a traditional core
ooding system.
Table 1 provides a summary of existing core ooding
systems, used around the world both for academic research and
industrial applications, and often been the gold standard for the
oil and gas industry. One can note that most of the systems use
syringe pumps coupled with piston accumulators,8,1925 which
13498
limits the type of liquids that can possibly be injected into the
core, as the inner surface of the pistoncylinder assembly of the
syringe pump is susceptible to corrosion. The piston
accumulator size limits the volume of uid that can be
continuously injected during ooding experiments and one
needs to perform a priori calculations to determine the size and
specication of the piston accumulator to perform a given core
ooding study. Table 1 also suggests that the fraction collector
dx.doi.org/10.1021/ie501866e | Ind. Eng. Chem. Res. 2014, 53, 1349713505
Industrial & Engineering Chemistry Research
was used for the downstream section in most of the core
ooding systems, which resulted in discrete estimation of oil
and water euents for the recovery curves.8,9,25 Also, post
experimental analysis is required to analyze the fraction of each
euent phase collected in each tube8,2628 of the fraction
collector. In some cases, the ultrasound separator has been used
for the euent collection and analysis.19 It is an eective
method for the experiments with three-phase euent fractions,
where phases are distinctly dierent from each other. For
instance, experiments that involve water/oil/gas ooding have
three components that can be separated in a quick manner
inside the column of the three-phase separator using ultra-
sonication and later can be quantied by volume. These
traditional core ooding systems need to be modied to make
them more versatile in terms of using a wide range of ooding
uids like surfactants, polymers, emulsion, etc. Often in such
scenarios, one needs to calculate the various uid volume
fractions on the y, which requires inclusion of modern uid
handling systems to increase the functionalities of traditional
core ooding systems.
NEW CORE FLOODING SYSTEM
A new core ooding apparatus has been designed and
constructed by taking into account the potential challenges
related to the traditional core ooding system, pointed out
above. This is a further modication to the initial conceptual
design reported earlier by the authors.29 A schematic of the
system is shown in Figure 1.
The upstream side of the system provides continuous
injection of three dierent types of liquids. It consists of uid
holding containers (carboys), pumps, and mass ow meters.
Three carboys (E-1,...,E-3) are directly connected to three dual
piston pumps (Smartline 100, Wissenschaftliche Geratebau Dr.
Ing. Herbert KNAUER GmbH) through 6.35 mm (1/4 in.)
stainless steel tubes. The precision and accuracy of the
upstream pumps at 12 MPa and for ow rate of 1 mL/min
are equal to <0.5% and <1.0%, respectively. To avoid cavitation
at the inlet side of the pumps, the carboys are pressurized with
nitrogen at 0.138 bar (g) (2 psig). Mass ow meters (M13,
Bronkhorst Cori-Tech BV) are installed downstream of the
pumps. Two of the mass ow meters are built of stainless steel,
whereas one of them is made of Hastelloy, to allow use of salt
(brine) water. According to calibration certicates of the mass
ow meters, the measurement error is 0.2% of the full-scale
reading +0.5 g/h. During the calibration procedure, density
deviation was found to be +0.11, 0.16, 0.27, and 0.01 kg/m3
for water, oil, emulsion, and downstream mass ow meter,
respectively. The density measured using three mass ow
meters was compared with the one calculated using Bronkhorst
High-Tech FLUIDAT software under specic uid temperature
and pressure conditions. For example, with the water ow rate
of 0.5 cm3/min for 30 min (approximate time to ll a single test
tube at the fraction collector), the error is 0.28 mL. Together
with the pumps and coupled with check valves, they form liquid
dosage systems (LDS-1,...,LDS-3), which can keep a constant
mass/volume ow rate at a preset pressure. There are three
LDS in the current conguration to deliver water, oil, and
emulsion into the core block as needed. The LDS system can
work simultaneously if it is needed for certain experimental
conditions. In addition, the upstream has open access to rell
the carboys, to allow continuous injection of liquids for lengthy
experimental runs. There is no need to stop an ongoing
experimental run for relling purposes, which is a shortcoming
13499
Article
in the case of a traditional system with piston accumulators at
the upstream side.
The core block consists of the core holder (in which the
porous medium is usually packed, which can be sand pack or
outcrops), inline pressure sensors, and dierential pressure
sensors. A biaxial horizontal core holder, shown in Figure 2, is
Figure 2. Exploded view of the core holder (1) cap, (2) end plugs
body, (3) body, (4) internal surface, and (5) tting 1/8 NPT to 3.175
mm (1/8 in.) tube).
used for containing and packing the medium. The core holder,
shown here, is designed to perform one-dimensional core
ooding experiments. However, any other two-dimensional10
or three-dimensional7,8 core holders, as and when available, can
be plugged into the system through appropriate connections to
3.175 mm (1/8 in) stainless steel tube. The dimensions of the
core holder used here as an example are length = 284 mm
(11.18 in) and diameter = 38.1 mm (1.5 in). This design
incorporates minimal uid leakage between the inner surface of
the coreholder and the sand pack. At the same time, it also
reduces the dead volumes at the inlet and outlet sides of the
sand pack due to redesigned entrance region, shown in Figure
3. The entrance region incorporates modications to the
patents published back in 1984 and 1985.30,31 It combines the
advantages of both patents and thereby replaces/modies the
distribution plates as per current needs. A sand-blasted plastic
tube is installed inside the core holder to prevent such uid
leakage. The distribution plugs, located at the two ends of the
core holder, have built-in stainless steel medium with the mesh
size of 40 m and 1.59 mm (1/16 in.) thickness and
distribution plate printed in a rapid prototype machine. This
combination redistributes the liquid across the cross section
and reduces dead volumes. The design for the distribution
plates was optimized by performing separate simulations using
commercial software. Particularly, the size of the holes in the
radial direction30 was optimized to guarantee uniform ow
distribution through the porous medium with known range of
permeability and porosity. For pressure monitoring, there are
three pressure sensors (FP2000 series, Honeywell International
Inc.), TP-2, TP-1, and dP-1, to measure the inlet pressure at the
upstream, the outlet pressure at the downstream, and
dierential pressure across the porous media.
The downstream section has an euent management system
that allows it to quantify dierent fractions of euents removed
from the core block. The mass fraction of each component
needs to be calculated in order to estimate the amount of oil
recovered, which is often the ultimate goal for such a core
ooding system. The euent management system has a mass
ow meter and a fraction collector. The fraction collector uses
graded test tubes with 15 mL volume each. The error in a
measuring volume of liquid collected in a test tube (430055,
Cole-Parmer Canada Inc.), used for checking the euent
dx.doi.org/10.1021/ie501866e | Ind. Eng. Chem. Res. 2014, 53, 1349713505
Industrial & Engineering Chemistry Research
Figure 3. End plugs of the biaxial core holder. (a) Isometric view of end plugs, (b) end plug assembled, (c) cross section view of single end plug (1)
sandblasted internal surface, (2) stainless steel lter, (3) 3D-printed plate, (4) strain ring, (5) end plugs body, and (6) core holders body.
Figure 4. Schematics of the further updates to the developed system.
fraction data, is 0.25 mL. Because this error would be
accumulated from sample to sample, the cumulative error in the
oil production is around 6.5 mL. The error bars incorporated
for each data point in the recovery curves represent the error in
measurements. The mass ow meter (M13, Bronkhorst Cori-
Tech BV) has the same specications as that used for water
ow, installed in LDS-1. The mass ow meter provides the
exibility to deal with euents having dierent densities and it
allows automatic evaluation of volumes of dierent euent
phases, the details of which are described later.
The advantage of the developed system, as it was pointed
earlier, that it is designed as a plug-and-play modular system.
There is a possibility for further modication of the core block,
which is only limited by the specication of the upstream
pumps. The region highlighted with red dashed square in
Figure 1 can be replaced with the subsystem, described in
Figure 4a, quite readily. It has a Hassler core holder,32,33 which
allows it to maintain a conning pressure up to 10 000 psi, to
use actual cores previously extracted from reservoirs (outcrops)
and to measure pressure drop across the core using eight tap-
points. The conning pressure in the present conguration can
be maintained using high pressure N2 line. This further can be
replaced with either high pressure hand pump or a syringe
pump. The temperature inside the core (or along its outer
13500
Article
surface) can be measured using any of the eight tap-points of
cores sleeve via thermocouples installed through the ports
located on the end plug with appropriate pressure sealed
feedthroughs (Series WF, Spectite Inc.). To maintain the
temperature of the core block similar to reservoir conditions,
the core holder can be placed inside an environmental chamber
or ultrathin heat sheets can be mounted at the outer surface of
the core holder covered with berglass pipe insulation. The
chamber allows it to maintain a temperature up to 200 C,
whereas the heat sheets allow it to maintain a temperature up to
100 C. The temperature can be controlled via a programmable
temperature PID controller with feedback from one or more
thermocouples.
Similarly, the downstream section can be also modied to
satisfy the requirements for the inline pressure. For example, a
digital back pressure regulator (EL-PRESS P-712CV, Bronk-
horst Cori-Tech BV) can be installed between the downstream
mass ow meter and fraction collector, as shown in Figure 4b.
It would allow us to set the back pressure up to 103.4 bar (g)
(1500 psig). These modications to the core block and
downstream section of the core ooding developed allow end-
users to obtain the required parameters to mimic desired
reservoir conditions.
dx.doi.org/10.1021/ie501866e | Ind. Eng. Chem. Res. 2014, 53, 1349713505
Industrial & Engineering Chemistry Research
Figure 5. Graphical user interface (GUI) to perform automation for the core ooding system.
One of the limitations of the current system is that the
produced water cannot be separated from the emulsion and
reinjected back to the core block as a ooding uid for further
oil recovery. To utilize such produced water back into the
ooding system, it is required to install additional components
at the downstream to separate emulsion, paran oil, and water.
Such separation of oil/water emulsion can be done using
traditional membranes34 or, more recently, through the use of
microuidics devices.35
The experimental setup can be pressurized with nitrogen, E-
6, and tested for leakage using soap-water (without the porous
sand pack in place). To minimize human intervention and have
robust control on the experiment, a software with a graphical
user interface (GUI) was created for the entire system using
custom design software (LabWindows CVI, National Instru-
ments Inc.). This software allows the user to control and track
dierent components during the experiment, save data, and
analyze various experimental conditions. A commercial data
acquisition (DAQ) system (cDAQ-9172, National Instruments
Corp.) with on board control modules for pressure sensors and
a fraction collector is used for data acquisition. In addition, the
mass ow meters and pumps are connected to the universal
eight port device server (PS 810, Sena Technologies, Inc.)
which transfers and receives signals from personal computer
(PC) through an Ethernet network. Processing of data for
euent measurements, based on fractions of euents, is also
incorporated, the details of which are available in the Euent
Management System section. The developed GUI interface is
shown in Figure 5. Similar to the experimental setup, explained
in details earlier, it can be divided into three subsections: the
upstream, the core block, and the downstream. In addition,
there are components for controlling dierent processes of
experiment, load/save conguration parameters, saving data,
monitor euent properties, pressure, and other parameters
involved in each run.
13501
Article
MATERIALS AND SPECIFICATIONS
Using the newly developed core ooding system, we have
performed both traditional core ooding using water as a
ooding agent (secondary recovery process) and emulsion
ooding using ex situ produced emulsion (enhanced oil
recovery). In this section, we provide the details of dierent
materials used during this experimental validation process for
the new core ooding system along with standard data
extraction in the form of porosity and permeability of the
core pack used for each ooding experiment.
During core ooding experiments, the injection rate was set
at 0.5 cm3/min and the core ooding temperature was at the
room temperature of 22 C. Distilled water (DI) was used for
the water ooding to obtain and to maintain initial water
saturation. Viscous paran oil (CAS Number 8012-95-1,
Sigma-Aldrich Co. LLC) with viscosity of 130 mPas at 25
C and density of 868 kg/m3, measured by the mass ow
meter, was used for oil saturation processes. The ex situ
emulsion used in this study is oil-in-water (70.75 wt % oil and
29.25 wt % water with an average oil drop size of 2.46 m,
provided by Quadrise Canada36) stable emulsion made of
heavy-oil, which is distinctly dierent than the resident oil
(paran oil) in the sand pack. The density and viscosity of
emulsion are 1014 kg/m3 and 122 mPas at 100 s1 shear rate
at 25 C,36 respectively. The chosen emulsion, being heavier
than water and oil, allows the downstream mass ow meter to
detect each phase separately. By setting reference points for the
density of each phase and using the in-house fraction-tracking
software, the mass ow meter can provide online measurements
of dierent euent fractions. In the future, the system can be
further modied to eliminate the need for test tubes.
For each experiment, the sand (silica sand with US sieve
size# 10 (100140), 53251 m) was weighted prior to the
packing inside the core holder. The mass of the sand used for
each pack was recorded to be 557 g and was consistent from
experiment to experiment with 1.5% of deviation. The
procedure for packing the core was as follows. The sand with
dx.doi.org/10.1021/ie501866e | Ind. Eng. Chem. Res. 2014, 53, 1349713505
Industrial & Engineering Chemistry Research Article

known mass was packed inside the void volume of the core
holder (324 cm3) using a vibration table (VP-181, FMC
Technologies) with the vibrator controller (Syntron Power
Pulse AC, FMC Technologies) set at maximum level. The core
holder with one opened end, containing the inner plastic tube,
was placed vertically on the vibration table. The sand was
poured inside the core holder continuously up to the top of the
internal tube. The core holder was kept for 5 min on a turned
table to let the sand settle and form a dense medium. A rod
with diameter equal to the internal diameter of the plastic tube
was used to compress the sand to produce a dense pack.
Each pack was characterized for porosity and permeability,
which were obtained prior to each ooding experiment. To
obtain a porosity value (pore volume) for each pack, a vacuum
pump (117, Labconco Corp.) was connected to one side and a
vacuum gauge was connected to another side of the core
holder. The core holder was evacuated to 4 kPa absolute (28 in.
mercury gauge) over 30 min. Then the vacuum pump was
disconnected and a tank lled with a known volume of water
was connected to the core holder to allow ow of water into the
core holder. The remaining volume of water in the tank was
recorded using double metric scale glass cylinder (250 mL
Brand 3025, Pyrex). From this measurements, the volume of
water occupied by the pore space can be calculated, which
For each pack, the permeability was obtained using three
dierent methods: with nitrogen, water, and nally, with
residual water saturation during oil ooding. First, the absolute
permeability was found using nitrogen. The mass ow
controller with a built-in pressure sensor was connected to
the inlet and a pressure sensor was installed at the outlet of the
core holder. At dierent ow rates of nitrogen, the pressure
dierence was recorded. Using the Klinkenberg equation,33,38
the absolute permeability was found. Second, the absolute
permeability, with water as the working uid, was calculated for
water saturated pack. Prior to obtaining permeability measure-
ments with residual water saturation, the oil saturation process
was carried out using paran oil. The oil was injected into the
water saturated pack with a constant injection rate of 0.5 cm3/
min for 1 PV of injection; thereafter, the injection rate was
increased to 1.0 cm3/min for an extra 0.5 PV of injection, until
the residual water saturation, Swi, was reached (see Table 2). At
the end of the process, the ow rate was reduced back to the
0.5 cm3/min and the permeability was calculated. The euent
stream was collected in batches in test tubes. The volume of
water and oil collected were compared to the injected one and
used to obtain values for residual water saturation, shown in
Table 2. Porosity and permeability measurements showed that
sand packing process was consistent for all experiments.

provides the value of the porosity of the sand pack. In addition,
the porosity was calculated using the mass of packed sand, void EXPERIMENTAL VALIDATION OF THE CORE
volume inside the core holder, and bulk density of silica sand37 FLOODING SYSTEM
(2.65 g/cm3). Pore volumes (PV), the volume of uid within
the packed core holder, were in the range from 113.4 to 120.0 Euent Management System. The comparison of
cm3. Obtained numbers from the two approaches are shown in dierent ooding strategies in terms of oil recovery depends
Table 2. on the underlying physical processes in terms of the
displacement of the resident oil phase by the invaded ooding
agent, particularly at the pore-level, which can be evaluated
Table 2. Properties of the Sand Packs
using a Reservoir-on-a-Chip (ROC)39 type system. However,
porosity, % by absolute permeability residual water on a macroscale, where we often rely on integrated cumulative
water/by mass permeability, with residual saturation after measurements, it is important to accurately estimate the
pack measurements Darcy water, Darcy oil saturation Swi
volumes of respective phases through a reliable euent
Run 37/35 10.2 6.25 0.02
#1 management system. To start with, the rst task would be to
Run 36/36 19.8 11.2 0.18 calibrate and ensure that the downstream stream is accurately
#2 measuring both the volume of water and oil for the euent
Run 37/35.8 11.3 6.30 0.06 streams. In total, we have performed the following four runs:
#3 two water ooding experiments (Run #1 and Run #2), direct
Run 35.8/33.4 11.3 7.90 0.06 emulsion ooding (Run #3), and nally water ooding
#4
followed by emulsion emulsion ooding (Run #4). For this
purpose, the fractions of the euent were collected into the test

Figure 6. Comparison of data obtained from mass ow meter and test tube fractionation during water ooding. Vertical axis, recovered mixture/oil
recorded by mass ow meter; horizontal axis, recovered mixture/oil measured with test tubes.

13502 dx.doi.org/10.1021/ie501866e | Ind. Eng. Chem. Res. 2014, 53, 1349713505

Industrial & Engineering Chemistry Research
Figure 7. Comparison of data obtained from mass ow meter and test tube fractionation during emulsion ooding. Vertical axis, recovered mixture/
oil recorded by mass ow meter; horizontal axis, recovered mixture/oil measured with test tubes.
tubes and compared to the recorded volumes from the mass
ow meter data, as shown in Figures 6 and 7. The error bars
correspond to the error in mass ow meter and fraction
measurements. The vertical axis here shows the total volume of
each liquid recorded by the mass ow meter and the horizontal
axis represents the total volume collected in test tubes. Each
test tube could be either lled with single liquid or can have
multiple liquids, e.g., oil and water. Consequently, points on
graphs could be in a range from 0 to 15 mL. The diagonal line
in these gures provides the measure of the data accuracy
between two measurement techniques.
To record the volume of each phase, the euent ow
through the mass ow meter was analyzed. A density reference
point, ref, was set at 935 kg/m3 for measurements during the
four runs, corresponding to dierent ooding experiments. This
reference point was selected on the basis of initial calibration
runs with paran oil and DI water. As the liquid ows through
the mass ow meter during a single time step, tstep (1 s), the
density, record (kg/m3), and the current volume ow rate, Qcur
(cm3/min), were recorded. On the basis of these quantities, the
uid volume for a given phase,Vj,rec (cm3), was calculated as
Q curtstep
Vj,rec = ; j = oil or water
60
If the density was larger than the reference level ref, the
recorded volume (eq 1) was added to the water value Vwater,rec
and the oil value Voil,rec was kept unchanged. Otherwise, it was
vice versa. This approach was also used for the emulsion
ooding experiments, the values above the reference level
represented water and emulsion phases. The volume of liquid,
which ows through the mass ow meter each second at the
ow rate, 0.5 cm3/min, is approximately 8.3 103 mL.
Figure 6 contains data points obtained during the process of
water ooding for dierent runs, including the one where the
emulsion ooding was followed after the water ooding step
(Run #4). Data points for produced water, shown in Figure 6a,
follow the 45 line trend. Figure 6b shows the similar trends for
the recovered oil in the euent stream. This highlights that
measurements from the mass ow meter can be used for online
determination of the euent fraction.
Results obtained for the two emulsion ooding methods are
presented in Figure 7. In Figure 7a, the results of produced
mixture of emulsion and water are depicted and Figure 7b
shows the volume of recovered oil measured by the two
Article
techniques. Even though one observes a strong correlation
between the two measurement techniques for certain measure-
ment ranges, however, this is not true for the entire range of
measured data, as was the case for Figure 6. In both cases, the
data correlation between the two measured techniques is not as
strong compared to the water ooding results shown in Figure
6. This suggests further calibration for dierent regimes are
required to increase the accuracy level. The worst deviation
from the 45 line trend was found the be 22%.
Core Flooding Experiments. Once the euent manage-
ment system is properly characterized and tested, the new core
ooding system is used to perform dierent ooding
experiments. The recovery curves for each set of ooding
experiment is shown in Figure 8. The presented data was
(1)
Figure 8. Recovery comparison for studied cases: Run #1 and Run #2,
water ooding; Run #3, direct emulsion ooding; Run #4, water
ooding followed by emulsion ooding.
obtained using determined volumes from test tube fractionation
that were matched to the data collected from the euent mass
ow meter. The upstream and downstream mass ow meters
supplemented mass balance calculations. It can be observed
that the rst two water ooding runs ended with recovery
around 60% and 75% for Run #1 and Run #2, respectively. The
obtained numbers were similar to the published data for
unconsolidated sand packs with similar properties.79 It was
observed that after 2 PV of injection of water, the ultimate
recovery increased by only 6% and 8% for Run #1 and Run #2,
respectively. The data suggest that the pore volumes used in the
experiments and the corresponding increase in the recovery are
similar to the eld applications.41 Alternate injection strategy
13503 dx.doi.org/10.1021/ie501866e | Ind. Eng. Chem. Res. 2014, 53, 1349713505
Industrial & Engineering Chemistry Research
must be incorporated to have economic recovery rate.42 Hence,
instead of continuing with water ooding beyond a certain PV,
emulsion was introduced to the system at 3.7 PV. The
breakthrough points for Runs #1 and #2 occurred at 0.3 and
0.24 PV of injected displacement uid, respectively.
For Run #3, the emulsion was used as a primary ooding
agent for the oil saturated porous media. As observed from
Figure 8, the emulsion ooding alone produced larger oil
recovery compared to water ooding for the same PV of
injection. During the emulsion ooding, almost 60% OOIP was
recovered after injecting only 0.5 PV of emulsion. The total oil
recovery was 79% OOIP at 1.25 PV of injection. From 1.01 PV
of injection, additional oil recovery was not observed for Run
#3. The breakthrough for this case happened at 0.72 PV of
emulsion injection. Such a high eciency of the emulsion
ooding compare to the water ooding could be due to late
breakthrough and having its viscosity similar to the viscosity of
the paran oil.
For Run # 4, the emulsion was introduced to the porous
media at 3.7 PV post water ooding. Even at this late stage, the
total oil recovery was found to have increased. An additional
15% of OOIP was recovered. The emulsion injection beyond
4.6 PV did not displace additional oil from the sand pack. The
residual oil saturation at the end of the Run #4 is around 18%,
which is similar to that in case of Run #3 with emulsion
ooding only. It is to be noted that even though emulsion
oodings (both Run #3 and Run #4) produced additional oil
recovery (20% OOIP), it required a large amount of emulsion
to achieve such high recovery.
Therefore, through these core ooding experiments, we have
successfully demonstrated that this newly developed versatile
core ooding system is capable of performing both the
traditional water ooding and more advanced enhanced oil
recovery process using ex situ obtained emulsion as a ooding
agent. This system can be used for academic research purposes
using dierent types of core holders with various combinations
of injection and production wells and, at the same time, can be
adopted as an industry benchmark to perform routine core
studies and has huge potential of replacing the existing core
ooding systems in the core laboratories across the world.
CONCLUSION
A new core ooding experimental setup has been constructed
to improve on the existing system, which has been in use for
several decades. The developed setup allows us to test
enhanced oil recovery techniques in a more ecient time
manner. The system was equipped with the up-to-date
techniques and hardware available in uid measurement
systems. An automated euent analysis incorporating proper-
ties (mass ow rate and density) measured by mass ow meters
has been developed. This allowed the replacement of the
traditional fraction collector in the downstream section for
online determination of euent composition. The entire
experimental setup is also monitored and controlled through
a custom built software.
The core ooding experiments with unconsolidated sand
packs as a medium have been performed for validation
purposes. To understand the applicability of the system,
emulsion (oil-in-water) was used as a ooding agent, as an
example of the advanced technique for enhanced oil recovery.
Here, two dierent strategies for emulsion ooding have been
demonstrated, direct injection of emulsion (Run #3) and water
ooding followed by emulsion ooding (Run #4). It has been
13504
Article
shown that the engineered emulsion worked eectively as an
independent ooding uid and Run #3 produced 79% OOIP
recovery at 1.2 PV of injection. Also, emulsion worked
eectively as a substitute uid after the water ooding in Run
#4 and added an extra 15% OOIP to the total oil recovery at
the cost of an additional 0.93 PV of emulsion injection.
Comparison between these two runs suggests that Run #3 was
eective in terms of the time needed for the oil to be recovered
from the porous media, whereas Run #4 was eective in terms
of the emulsion used for the oil recovery process. Such
estimates of dierent ooding strategies can be seamlessly
executed in the newly developed core ooding system, which
provides opportunities for engineers to test their hypotheses
reliably before embarking into expensive drilling operations in
an actual oil reservoir.
AUTHOR INFORMATION
Corresponding Author
*S. K. Mitra. E-mail: sushanta.mitra@ualberta.ca. Phone: +1
(780)492-2200. Fax: +1 (780)492-5017.
Notes
The authors declare no competing nancial interest.
ACKNOWLEDGMENTS
The authors acknowledge Dr. Patrick Brunelle from Quadrise
Canada Inc. for his valuable inputs and providing emulsion for
experiments. Financial assistance from NSERC CRD (Grant
No. CRDPJ 399500-10) is acknowledged here.
REFERENCES
(1) Albertas Oil Sands 2006; Alberta Department of Energy:
Edmonton, AB, 2007; http://env.chass.utoronto.ca/env200y/
ESSAY09/OilSands06.pdf.
(2) China to invest US$13B in oil and gas exploration this year.
http://business.nancialpost.com/2013/09/16/china-to-invest-us13b-
in-oil-and-gas-exploration-this-year/?\_\_lsa=6fe1-f7f9.
(3) Liu, Q.; Dong, M.; Ma, S.; Tu, Y. Surfactant enhanced alkaline
flooding for Western Canadian heavy oil recovery. Colloids Surf., A
2007, 293, 6371.
(4) Asghari, K.; Nakutnyy, P.Experimental results of polymer
ooding of heavy oil reservoirs. In Proceedings of Canadian Interna-
tional Petroleum Conference, Calgary, Alberta, June 1719, 2008;
Petroleum Society of Canada, 2008; pp 17.
(5) Shen, P.; Wang, J.; Yuan, S.; Zhong, T.; Jia, X. Study of enhanced-
oil-recovery mechanism of alkali/surfactant/polymer ooding in
porous media from experiments. SPE J. 2009, 14.
(6) Oil market report. 2014; http://omrpublic.iea.org/.
(7) Hadia, N.; Chaudhari, L.; Mitra, S. K.; Vinjamur, M.; Singh, R.
Waterflood profiles and oil recovery with vertical and horizontal wells.
Energy Sources, Part A 2008, 30, 16041618.
(8) Hadia, N. J.; Chaudhari, L. S.; Mitra, S. K.; Vinjamur, M.; Singh,
R. Experimental investigation of use of horizontal wells in water-
flooding. J. Pet. Sci. Eng. 2007, 56, 303310.
(9) Mandal, A.; Samanta, A.; Bera, A.; Ojha, K. Characterization of
oil-water emulsion and its use in enhanced oil recovery. Ind. Eng.
Chem. Res. 2010, 49, 1275612761.
(10) Santosh, V.; Mitra, S. K.; Vinjamur, M.; Singh, R. Experimental
and numerical investigations of waterflood profiles with different well
configurations. Energy Fuels 2007, 21, 33533359.
(11) Hadia, N. J.; Chaudhari, L. S.; Aggarwal, A.; Mitra, S. K.;
Vinjamur, M.; Singh, R. Experimental and numerical investigation of
one-dimensional waterflood in porous reservoir. Exp. Therm. Fluid Sci.
2007, 32, 355361.
(12) DElia, S. R.; Ferrer, G. J. Emulsion ooding of viscous oil
reservoirs. In Proceedings of Fall Meeting of the Society of Petroleum
dx.doi.org/10.1021/ie501866e | Ind. Eng. Chem. Res. 2014, 53, 1349713505
 Industrial & Engineering Chemistry Research Article

Engineers of AIME; Paper SPE 4674; Las Vegas, NV, September 30, coal: New insights from a core flooding study. RSC Adv. 2014, 4,
1973. 2277922791.
(13) Hadia, N. J.; Mitra, S. K.; Vinjamur, M. Estimation of (34) Chen, W.; Peng, J.; Su, Y.; Zheng, L.; Wang, L.; Jiang, Z.
permeability heterogeneity in limestone outcrop by pressure measure- Separation of oil/water emulsion using Pluronic F127 modified
ments: Experiments and numerical simulation. Exp. Therm. Fluid Sci. polyethersulfone ultrafiltration membranes. Sep. Purif. Technol. 2009,
2012, 40, 177184. 66, 591597.
(14) Hadia, N.; Chaudhari, L.; Mitra, S. K.; Vinjamur, M.; Singh, R. (35) Fidalgo, L. M.; Whyte, G.; Bratton, D.; Kaminski, C. F.; Abell,
Effect of scaling parameters on waterflood performance with C.; Huck, W. T. S. From microdroplets to microfluidics: Selective
horizontal and vertical wells. Energy Fuels 2008, 22, 402409. emulsion separation in microfluidic devices. Angew. Chem., Int. Ed,
(15) Wyckoff, R. D.; Botset, H. G. The flow of gas-liquid mixtures 2008, 47, 20422045.
through unconsolidated sands. Physics 1936, 7, 325. (36) Brunelle, P. Quadrise Canada Corp., Data sheet: UofA E2
(16) Tzimas, E.; Georgakaki, A.; Cortes, C. G.; Peteves, S. Enhanced Concentrate. Personal communication, 2012.
oil recovery using carbon dioxide in the European Energy System; (37) Bulk density chart. http://www.manusabrasive.com/bulk-
Report EUR 21895 EN; European Communities: Petten, The density-chart/.
Netherlands, 2005; p 124. (38) Klinkenberg, L. The permeability of porous media to liquids and
(17) Nasralla, R.; Alotaibi, M.; Nasr-El-Din, H. Eciency of oil gases. Drilling and Production Practice; Paper API-41-200; New York,
recovery by low salinity water ooding in sandstone reservoirs. In NY, January 1, 1941; pp 200213.
Proceedings of SPE Western North American Region Meeting; Paper SPE- (39) Gunda, N. S. K.; Bera, B.; Karadimitriou, N. K.; Mitra, S. K.;
144602-MS; Anchorage, AK, May 711, 2011. Hassanizadeh, S. M. Reservoir-on-a-chip (ROC): A new paradigm in
(18) Petersen, E. B., Jr.; Lohne, A.; Helland, J. O.; Virnovsky, G. reservoir engineering. Lab Chip 2011, 11, 37853792.
Relative permeabilities for two- and three- phase ow processes (40) Dake, L. P. Fundamentals of Reservoir Engineering, seventeent ed.;
relevant to the depressurization of the Statfjord eld. International Elsevier Science B.V.: Amsterdam, 1998; p 437.
Symposium of the Society of Core Analysis; Paper SCA2008-23; Abu (41) See ref 40, p 364.
Dhabi, UAE, October 29November 2, 2008; pp 112. (42) Baldygin, A.; Nobes, D. S.; Mitra, S. K. Water-Alternative-
(19) Legowo, E.; Pratomo, S. Microbial core ooding experiments Emulsion (WAE): A new technique for enhanced oil recovery J. Petrol.
using indigenous microbes. In Proceedings of SPE Asia Pacic Improved Sci. Eng., 2014, in press.
Oil Recovery Conference; Paper SPE-57306-MS; Kuala Lumpur,
Malaysia, October 2526, 1999; pp 111.
(20) Hornof, V.; Morrow, N. Gravity effects in the displacement of
oil by surfactant solutions. SPE Reservoir Eng. 1987, 2, 627633.
(21) Chang, S.-H.; Grigg, R. Laboratory ow tests used to determine
reservoir simulator foam parameters for EVGSAU CO2 foam pilot. In
Proceedings of Permian Basin Oil and Gas Recovery Conference; Paper
SPE-27675-MS; Midland, TX, March 1618, 1994; pp 483492.
(22) Bagci, S.; Kok, M. V.; Turksoy, U. Effect of brine composition
on oil recovery by waterflooding. Petroleum Science and Technology
2001, 19, 359372.
(23) Sola, S. Feasibility experimental investigation of steam/methane
ooding in a heavy oil reservoir. In Proceedings of SPE International
Petroleum Conference in Mexico; Paper SPE-91968-MS; Puebla Pue,
Mexico, November 79, 2004.
(24) Nobakht, M.; Moghadam, S.; Gu, Y. Effects of viscous and
capillary forces on CO2 enhanced oil recovery under reservoir
conditions. Energy Fuels 2007, 21, 34693476.
(25) Samanta, A.; Bera, A.; Ojha, K.; Mandal, A. Comparative studies
on enhanced oil recovery by alkalisurfactant and polymer flooding. J.
Pet. Explor. Prod. Technol. 2012, 2, 6774.
(26) Veerabhadrappa, S. K.; Doda, A.; Trivedi, J. J.; Kuru, E. On the
Effect of Polymer Elasticity on Secondary and Tertiary Oil Recovery.
Ind. Eng. Chem. Res. 2013, 52, 1842118428.
(27) Ko, K. M.; Chon, B. H.; Jang, S. B.; Jang, H. Y. Surfactant
flooding characteristics of dodecyl alkyl sulfate for enhanced oil
recovery. J. Ind. Eng. Chem. 2014, 20, 228233.
(28) Mandal, A.; Samanta, A.; Bera, A.; Ojha, K. Role of oil-water
emulsion in enhanced oil recovery. 2010 International Conference on
Chemistry and Chemical Engineering, Kyoto, Japan, August 13, 2010;
pp 190194.
(29) Baldygin, A.; Nobes, D. S.; Mitra, S. K. Novel technique for core
ooding experiments. The biennial CSME International Congress,
University of Manitoba, Winnipeg, Canada, June 46, 2012.
(30) Jones, S. C. Perforated end plug plate for testing core samples.
U.S. Patent US4561289 A, 1984.
(31) Jones, S. C. Porous end plug disk for testing core samples. U.S.
Patent US4643019 A, 1985.
(32) Core Laboratories: Hassler Type Core Holder - RCH Series.
http://www.corelab.com/cli/core-holders/hassler-type-core-holders-
rch-series.
(33) Stephen, A.; Adebusuyi, A.; Baldygin, A.; Shuster, J.; Southam,
G.; Budwill, K.; Foght, J.; Nobes, D. S.; Mitra, S. K. Bioconversion of

13505 dx.doi.org/10.1021/ie501866e | Ind. Eng. Chem. Res. 2014, 53, 1349713505

View publication stats