Presented at ISA 2004 Exhibit and Conference – Edmonton Section, April 2004

Application of Sonar-Based, Clamp-on Flow Meter in

Oilsand Processing
Daniel L. Gysling
CiDRA Corporation
50 Barnes Park North
Wallingford, CT 06492
USA

Eb Mueller
Syncrude Canada Ltd.
Edmonton Research Centre
9421-17 Avenue
Edmonton, Alberta T6N 1H4

Abstract
Data is presented demonstrating the applicability of sonar-based, clamp-on flow measurement to several, long-
standing flow measurement challenges within the oilsands industry, including hydrotransport, coarse tailings, and
bitumen froth flow lines.

Sonar-based flow measurement technology was developed and field proven in the oil and gas production industry
over the last five years and provides robust, accurate volumetric flow rate measurement for a broad range of process
fluids, slurries, pipes sizes and flow conditions. Sonar-based flow metering technology utilizes an array of sensors to
listen to, and interpret, unsteady pressure fields within process flow lines. The methodology is implemented using
strain-based sensors which clamp-on to existing process piping. Sonar-based flow monitoring systems determine
volumetric flow rate by measuring the speed at which self-generated, coherent flow structures convect past the
sensor array. Using similar sonar-based array processing techniques as those used for volumetric flow, sonar-based
flow meters can also determine entrained air (or any other gas) levels by measuring the speed at which sound
propagates within the process flow lines. The speed of sound in the process flow lines provides an accurate and
robust, clamp-on method for determining entrained air levels in aerated liquids. Data is presented showing entrained
air levels on a 16-inch diameter froth line exiting a steam-driven deaerator.

Introduction This paper addresses the application of sonar-

The oilsands processing industry relies based flow technology to the flow measurement
extensively on large-diameter-pipe conveyed challenges in the oilsands industry, specifically
slurries to transport materials throughout the to three, broadly representative classes of
process of converting oilsands into synthetic slurries commonly encountered in the oil sand
crude oil (Gray, 2002). Historically, obtaining processing industry.
accurate and reliable measurements of these 1) Hydrotransport lines which
slurries has proven technically difficult and transport and condition the
economically challenging, due primarily to their oil sands for bitumen
abrasive characteristics and difficult to extraction,
characterize physical properties of solid/liquid 2) Tailings lines, which return the
mixtures. tailings for eventual use in
land reclamation

BI0102 Rev. A 1
 Presented at ISA 2004 Exhibit and Conference – Edmonton Section, April 2004

3) Bitumen froth lines, which

transport the bitumen for
additional upgrading.

Sonar-based Flow Meters

Sonar-based flow measurement technology was
first introduced into the oil and gas industry in
1998 for use in down-hole multiphase flow
metering applications (Kragas, 2002), and is
currently being applied to a wide range of other
industries, including chemical processing
Figure 1: Coherent Structures within Turbulent
industries, mineral processing and pulp and Pipe Flows
paper industries (Gysling, 2003).
The Convective Ridge
Sonar-based flow measurements utilize an array As discussed above, sonar-based flow meters
of sensors, aligned axially along the pipe, to use the convection velocity of coherent
characterize and interpret naturally occurring, structures (eddies) inherent within turbulent pipe
unsteady pressure fields within process piping. flows to determine the volumetric flow rate.
Although applicable to single phase flows as The sonar-based algorithms determine the speed
well, sonar-based flow measurement techniques of these structures by characterizing both the
were specifically developed for multiphase temporal and spatially frequency characteristics
flows. of the flow field. These algorithms have, at their
roots, many of the concepts developed to
Figure 1 illustrates the naturally occurring, self- address the beamforming problem in underwater
generated, coherent structures present within sonar processing (Nielsen, 1991). For a series of
turbulent process flow of Newtonian fluids. As coherent eddies convecting past a fixed array of
shown, the time-averaged axial velocity for a sensors, the temporal and spatial frequency
turbulent pipe flow velocity profile is a function content of pressure fluctuations are related
of radial position, from zero at the wall to a through a dispersion relationship, expressed as
maximum at the centerline of the pipe. The flow follows:
near the wall is characterized by steep velocity
gradients and transitions to relatively uniform
  U convect k
core flow near the center of the pipe. Naturally
occurring, self-generating, turbulent eddies are Here k is the wave number, defined as k=2in
superimposed over the time averaged-velocity units of 1/length,  is the temporal frequency in
profiles. These coherent structures contain rad/sec, and Uconvect is the convection velocity or
fluctuations with magnitudes on the order of phase speed of the disturbance, and  is the
10% percent of the mean flow velocity and are spatial wavelength. The dispersion relationship
carried along with the mean flow. Experimental basically states that temporal frequency
investigations have established that eddies observed at a fixed location are proportional to
generated within turbulent boundary layers the convection speed and inversely proportional
remain coherent for several pipe diameters and to the spatial wavelength of the disturbance.
convect at, or near, the volumetric averaged flow
rate in the pipe (Schlichting, 1979). Although In sonar array processing, the spatial / temporal
this description is based on the empirical and frequency content of time stationary sound fields
theoretical understanding of Newtonian fluids, are often displayed using “k- plots”. K- plots
this has proved to be a useful framework with are three-dimensional power spectra in which
which to consider many non-Newtonian flows the power of a sound field is decomposed into
such as pulp suspensions, froth slurries and bins corresponding to specific spatial wave
others. numbers and temporal frequencies, the power
level is represented by a color map shown to the

BI0102 Rev. A 2
 Presented at ISA 2004 Exhibit and Conference – Edmonton Section, April 2004

right of the k- plot. On a k- plot, the power volumetrically averaged flow rate as a low-order
associated with a pressure field convecting along function of Reynolds to provides accurate flow
with the flow is distributed in specific regions measurement. Figure 3 illustrates the how the
that satisfy the dispersion relationship developed 6 inch Sonar Meter Perfromance
above. For turbulent boundary layer flows, this Slope of Ridge Determines Flow rate
35
region is termed “the convective ridge”
Slope of
Convective Ridge:
17 fps

(Beranek, 1992) and the slope of this ridge on a 30

k- plot indicates the speed of the turbulent

Measured Velocity, fps

25
eddies. Thus, identifying the slope of the 20
Slope of
Convective Ridge:
5 fps

convective ridge provides a means to determine Slope of

Convective Ridge:

15 27 fps

the convection speed of the turbulent eddies; and

10
with calibration and knowing the cross-sectional
area of the pipe, this slope provides a means to 5

determine the volumetric flow rate. 0

0 5 10 15 20 25 30 35

Reference Velocity, fps

Figure 3: Calibration Data showing

representative k- plots at various volumetrically
averaged flow velocities
Slope of
Convective slope of the convective ridge increases with flow
Ridge =
10.13 ft/sec rate. Using standard calibration / verification
procedures, sonar-based flow meters, with
diameters ranging from 3-30 inches, consistently
demonstrate accuracies of +/-0.5% accuracy for
flow velocities of 3-30 ft/sec. Figure 4 shows
data from a family of sonar-based flow meters
ranging from 4 to 10 inches in diameter.
Figure 2: K- plot generated from an array of Calibration Flow Curve
sensors for water flow in at 10 ft/sec in an 8-inch 8000

10" SONARtrac Meter

pipe
Measured SONARtrac Flow Rate (gpm)

7000
8" SONARtrac Meter
6" SONARtrac Meter
6000
4: SONARtrac Meter

Figure 2 shows a k- plot generated from the 5000

using the 8-inch diameter sonar-based flow 4000

meter, operating at a water-only calibration 3000

facility. As shown, the power contours show a 2000

well-defined convective ridge. A parametric 1000

optimization method was used to determine the 0

“best” line representing the slope of the ridge.

0 1000 2000 3000 4000 5000 6000 7000 8000
NIST Traceable Reference Flow Rate (gpm)

For this case, a slope of 10.13 ft/sec was Figure 4: Volumetric Flow Rate measured using
determined. Sonar-based convective flow meter versus
Reference Flow Rate
Calibration
The k- plot shown in Figure 2 illustrates the Slurries and Non-Newtonian Fluids
fundamental principles behind sonar-based Sonar-based flow metering technology has been
convective flow measurements, namely that evaluated in hundreds of industrial applications
axial array sensors can be used in conjunction in numerous industries. Figure 5 shows a
with sonar processing techniques to determine comparison of a mag meter and a sonar-based
the speed at which naturally occurring turbulent flow meter on 3.5% consistency paper slurry in
eddies convect within a pipe. The slope of the operating in a commercial paper mill. Although
convective ridge can be calibrated to the the sonar-based flow meter is reporting flow

BI0102 Rev. A 3
 Presented at ISA 2004 Exhibit and Conference – Edmonton Section, April 2004

~5% lower than the mag meter, apart from this further complicates the flow measurement, as
offset, the two measurements are does the possibility of the hydrotransport line
indistinguishable. A representative k- plot containing up to several percent of entrained air
constructed from the sonar-based flow meter by volume.
data shows a clearly defined convective ridge,
similar to that observed in the water calibration Currently, modified venturi (or wedge) meters
data. In addition to the convective ridge, a are the predominant devices used to measure
nearly vertical ridge is also present. As will be flow rates in hydrotransport lines. Mechanical
developed later, the slope of this vertical ridge wear of these meters results in high
contains information regarding the speed of maintenance, calibration, and replacement costs,
sound propagation in the pipe and can be used to providing an incentive for operators to evaluate
determine entrained gas levels within the alternative measurement technologies. In efforts
process mixture. to reduce total lifetime costs, eliminate process
Pulp Suspension interruptions due to flow meter maintenance and
3.5% Consistency Stock-line
S o n arT rac an d M ag F lo w improve accuracy, sonar-based flow meters have
1100.00 1045.00
been evaluated on several Oilsand
1050.00 995.00 hydrotransport lines. The initial technology
1000.00 945.00
evaluation was performed at a pilot facility at
Syncrude Research in Edmonton during June
SonarTrac Flow (gpm)

Mag Flow (gpm)

950.00 895.00
2003.
900.00 845.00

850.00 795.00
The pilot facility was designed to simulate the
commercial hydrotransport and primary
800.00 745.00
separation processes of oil sands on a reduced
10:04

10:33

11:02

11:31

12:00

12:28

12:57

13:26

13:55

14:24

scale. A sonar-based flow meter was evaluated

T im e

S onarT rac F low (G P M) Mag F low (vbfc 4212.pv)

on 4-inch line, using a slurry mixture composed

Figure 5: Comparison of a mag meter and a to simulate hydrotransport operation. The meter
sonar-based flow meter on a 3.5% consistency was evaluated as composition of the process
paper slurry in operating in a commercial paper
mill Volumetric FlowVolumetric
& Entrained Measurement
Flow Rate and Entrained Air

300.0 3.000%

Volumetric Flow Rate

GVF
Oilsands Hydrotransport 250.0 2.500%

Hydrotransport technology is the primary

method for conditioning and transporting mined 200.0 2.000%

oilsand from the mine to the extraction facilities 150.0 1.500%

in which the bitumen is extracted from the water

and solids contained in the slurry (Gray, 2002). 100.0 1.000%

Hydrotransport lines typically operate at

densities of ~1450 kg/m^3 to 1600 kg/m^3 with
50.0 0.500%

flow velocities of 4 to 5 m/sec. These flows 0.0 0.000%

present several challenges to conventional flow

12:14:24 12:28:48 12:43:12 12:57:36 13:12:00 13:26:24

Time

metering technology. Firstly, the flows are Figure 6: Volumetric Flow and entrained air
extremely abrasive, containing >50% solids by measurement on 4-inch pilot hydrotransport line
mass with particle distributions ranging from
several microns up to several inches in diameter. mixture was transitioned from predominately
These flows typically exhibit some level of aerated-water mixture to a non-aerated oilsand
stratification (Crowe, 1998), as evidenced by the slurry. The sonar-based flow meter measured
preferential wear of the lower portion of total volumetric flow and entrained gas volume
hydrotransport lines, requiring the pipes to be fraction. The results of the test are shown in
rotated on a periodic basis. The presence of Figure 6. During the ~1 hour test, the entrained
~10% by volume of non-conductive bitumen air levels are varied. The sonar-based flow meter

BI0102 Rev. A 4
 Presented at ISA 2004 Exhibit and Conference – Edmonton Section, April 2004

indicated that the entrained air transitioned from show the acoustic and convective ridges
~3.5% to less than 0.2%, spanning the levels of simultaneously.
entrained air anticipated in commercial Figure 9 shows a k- plot for the hydrotransport
hydrotransport lines. operating in full solids loading of ~1500 kg/m^3
and relatively low entrained air level of ~0.2%.
A wedge meter and density meter served as As shown, the convective ridge is clearly
reference measurements. Figure 7 shows that evident and, the comparison of the k- plot for
the sonar-based flow meter agrees reasonably hyrdrotansport slurries (Figure 9) with that for
well with the reference meter, reporting ~8% water (Figure 2) is a good indication of
lower than the wedge meter for the low-density applicability of sonar-base flow measurement to
slurry and reporting ~3% higher than the wedge oilsand slurries.
meter for the high-density slurry.

Figure 7: Wedge flow meter comparison with

Sonar-based flow meter on 4-inch pilot Figure 9: Representative k- plot from 4 inch
hydrotransport line hydrotransport slurry flowing at 21 fps with a
density of ~1500 kg/m^3
A sonar-based decomposition of the pressure
field with the 4 inch slurry lines is shown in SONARtrac Installed on
27 inch Line
Figure 8 for the hydrotransport slurry operating
at a relatively low solids content, density ~1.2,

Multiphase Slurry VF & SoS Measurement

Acoustic Ridges
350 fps

Convective Ridge
25 fps
Figure 10: Installation of Sonar-based flow
meters on 27inch commercial hydrotransport line

Commercial Hydrotransport Line

Figure 8: Representative k- plot from 4 inch Following the demonstration at the pilot facility,
hydrotransport slurry flowing at 21 fps with a a sonar-based flow meter was evaluated on a 27
density of ~1200 kg/m^3 and 3.5% entrained air inch Hydrotransport line at the Syncrude
Mildred Lake facility in September 2003. The
but a high aeration level of ~3.5%. The meter was installed near the discharge of a
frequency range of the k- plot was selected to several-mile-long, hydrotransport line initiating

BI0102 Rev. A 5
 Presented at ISA 2004 Exhibit and Conference – Edmonton Section, April 2004

in the North mine. A picture of the installation is coherent structures that can be observed to
shown in Figure 10. The clamp-on meter was convect down the pipe at, or near, the
installed ~10 diameters downstream of a ~30 volumetrically averaged flow.
degree elbow, on a ~30 degree incline.
Tailings Slurry Line
27 inch HydroTransport
Tailings slurries transport the water and solids
that remain after the bitumen has been extracted
1.2
from the oilsand mixtures. Tailings from the
1 extraction process are typically pumped into
tailings ponds for additional processing and
0.8
eventual use in land reclamation. Monitoring
0.6
SONARtrac
Ref -Pump Curve and optimizing the flow and transport of tailings
0.4
is an important aspect of bitumen extraction,
however, monitoring tailings slurries presents
0.2
many similar challenges as those seen in
0 hydrotransport lines.
7:12 AM 8:24 AM 9:36 AM 10:48 AM 12:00 PM 1:12 PM 2:24 PM

A sonar-based flow meter was evaluated on a 24

Time

Figure 11: Volumetric flow rate on 27 inch inch, commercial tailings lines in September
Hydrotransport line- Comparison between Sonar-
2003 at Syncrude’s Mildred Lake facility. Figure
based flow meter and flow derived from pump-
curves 12 shows the location of the meter installed
under the hoarding. As shown, the sonar-based
Data recorded from this installation is shown in flow meter was clamped onto the tailings line
Figure 11. A reference flow rate, provided by ~10 diameters downstream of a modified venturi
the plant DCS and inferred from pump curves, is flow meter. The output of the venturi flow meter
also presented. Over the ~ 4 hours of operation, was provided as a reference flow measurement.
the sonar-based flow meter was in good Here again, the venturi flow rates serve as a
agreement with the flow inferred from the pump representative measure of the flow rate and, due
curves, with both flow measurements tracking to issues associated with erosion and the
the same dynamic changes and reporting the complex nature of solid / liquid flows, contains
same averaged flow rates. It should be noted a significant level of uncertainty regarding
that, while accuracy is important, the reference absolute flow rate accuracy.
measurement available on this line, and most
commercial hydrotransport lines, are typically
not of sufficient accuracy and tracability to draw 24 in
Tailings
definitive conclusions regarding the accuracy of Lines
the sonar-based flow meter. The objective was
to demonstrate that sonar-based flow SONAR Flow
meter Location
measurement principles are applicable to
hydrotransport lines and that the sonar-based
flow measurement provides a flow measurement
consistent with the various other methods Modified
Venturis
currently used to track flow rates in these lines.
To this end, the sonar-based, k- decomposition Figure 12: Location Sonar-based flow meter on
of the pressure field within the hydrotransport is 24inch commercial tailings line
also shown on Figure 11. As shown, the k-
plot exhibits a well-defined convective ridge,
demonstrating that, like numerous other flows in Results for the evaluation are shown in Figure
numerous other applications, oilsand 13. During this period the tailing lines were
hydrotransport slurries contain self-generated, operating at a typical solids loading levels, with
a mixture density of ~1500 kg/m^3. As shown,

BI0102 Rev. A 6
 Presented at ISA 2004 Exhibit and Conference – Edmonton Section, April 2004

the sonar-based flow meter is reporting flow rate annular flow regimes, in which the motion of the
~15% below those reported by the venturi, fluid through the pipe set ups an essentially
however, the dynamic behavior of the meters is rigid, core region in the center of the pipe,
quite similar. Note the sonar-based flow data surrounded by a highly-sheared, less-viscous
has a noticeably faster update rate. A annular region which effectively lubricates the
representative sonar-based k- plot is also core flow (Joseph, 1999). From an operational
shown in Figure 13, indicating that, like perspective, the primary result of this self-
hydrotransport flow lines, tailings lines generate lubrication of bitumen froth (often termed
naturally occurring, coherent disturbances that Natural Froth Lubricity or NFL) is a significant
convect at, or near, the volumetrically averaged reduction (several orders of magnitude) in the
velocity in the pipe. As mentioned above, no pressure drop required to pump a given amount
definitive conclusions regarding accuracy can be of bitumen froth compared to that predicted by a
drawn from this test. model using the static viscosity and Newtonian
24 inch Tailings Line
model of the fluid rheology.
1.4
Another consequence of the shear-thinning
1.2
behavior is that the core-annular flow regime
1 establishes undulations in the core/annular
0.8
Ref - Wedge
interface, often termed “tiger waves” due to the
0.6
SONARtrac
visual patterns observed in laboratory studies
0.4 using clear wall pipe sections for bitumen froth
0.2 operating in NFL. These tiger waves have an
0 associated pressure field that is coherent over
12:00 PM 2:24 PM 4:48 PM 7:12 PM 9:36 PM 12:00 AM 2:24 AM 4:48 AM
several pipe diameters and moves with the fluid.
time
Thus, although bitumen froth flows are far from
Figure 13: Volumetric flow rate on 24 inch turbulent in the classical sense, they do exhibit a
Hydrotransport line- Comparison between Sonar- self-generated, coherent pressure field that can
based flow meter and venturi meter be tracked with sonar-based flow meters to
determine the volumetric flow.
Bitumen Froth Applications
The end product of the separation processes is Sonar-based flow meters were first evaluated on
bitumen. Bitumen is the primary component in bitumen froth in July 2003 in a test at the
bitumen froth, which typically contains 60% Saskatchewan Research Council (SRC),
bitumen, 30% water, and 10% fine solids and sponsored by Syncrude. The test demonstrated
clays. The transport of bitumen through that sonar-based flow methods could be
pipelines has received a significant amount of applicable to bitumen froth exhibiting NFL,
study in recent years. The challenge in definite conclusions about the suitability of this
transporting bitumen through pipelines is technology in a commercial setting could not be
derived from its high static viscosity, on the drawn due to limitations in the experimental set-
order of 105 times that of water at standard up.
conditions. If bitumen froth behaved as a
Newtonian fluid, i.e. the shear stress was strictly Figure 14 shows a convective ridge measured on
proportional to shear rate, pumping bitumen a 6 inch diameter, schedule 160, pipe for a
froth would require several orders of magnitude bitumen froth flowing at ~ 9 fps. As shown, the
more power than that required to pump water, convective ridge for the bitumen froth is similar
rendering pipeline transfer impractical. to those presented for the other Newtonian and
non-newtonian fluids. The sonar-based flow
Fortunately, bitumen froth is a shear thinning meter reported flow rates in reasonable
fluid. The viscosity of these types of non- agreement with available flow estimates used as
newtonian fluids decreases as the fluid shears. references. The primary reference available
Shear thinning fluids tend to exhibit core- was the rotational speed of the positive

BI0102 Rev. A 7
 Presented at ISA 2004 Exhibit and Conference – Edmonton Section, April 2004

displacement pump, but unfortunately, not

enough data points could be collected to draw
any definitive, quantitative conclusions
regarding the accuracy of the sonar-based flow
meter. However, the similarity of the convective
ridge identified for the froth flows with those for
other applications is a strong indicator of the
applicability of sonar-based methods to bitumen
froth.

Slope of
Convective
Ridge =
9.1 ft/sec

Figure 15: Installation of Sonar-based flow meter

on 16 inch commercial bitumen froth line

16in Froth from Deaerater

10000 50

Figure 14: Representative K- plot from 6inch, 9000 45

schedule 160 bitumen froth flowing at 9 fps 8000 40

7000 35

Following the proof-of-concept testing of sonar- 6000

5000
30

25
Flow rate
GVF

based flow measurement at SRC, sonar-based 4000 20

PumpAmps

flow measurement was evaluated on a 16-inch 3000 15

froth line, downstream of froth deaerator at 2000 10

Syncrude’s Mildred Lake facility. In this 1000 5

0 0

application, in addition to reporting volumetric 5:02 PM 5:09 PM 5:16 PM 5:24 PM 5:31 PM 5:38 PM 5:45 PM 5:52 PM

flow, the sonar-based flow meter also provided a time

measure of the entrained air levels of the froth. Figure 16: Volumetric flow rate, Entrained Air,
The meter was installed in a vertical section of and Pump Motor Current for a 16-inch froth line
pipe with upward flow, approximately 30 ft exiting Deaerator
downstream of a variable speed, centrifugal
pump, used to pump the froth from the steam A representative sonar-based k- decomposition
deaerator to a froth header. A picture of the of the pressure field within the froth line is
installation is shown in Figure 15. A time shown in Figure 17. The flow rate is at the
history of the output of the sonar-based flow lower end of the range of flow rates observed at
meter is shown in Figure 16. As indicated by ~6 fps. The convective ridge is clearly evident,
motor current, the pump is being cycled on a ~6- as is a nearly vertical ridge in the center of the k-
8 minute period, resulting in variations in the  plot. As mentioned for the aerated pulp
flow rate and aeration levels of the froth on the slurries presented above, this nearly vertical
same time scale. The flow rate is cycling ridge is termed the acoustic ridge. The slope of
between ~3000 gpm and ~9000 gpm the acoustic ridge is a measure of the speed at
corresponding to volumetrically averaged flow which sound propagates within the process fluid,
velocity of ~5 fps to ~15 fps. Again, no which, in turn, provides a measure of the
quantitative reference was available, however, entrained air within the process mixture.
the time history of the flow rate seems consistent
with the time history of the pump motor current.

BI0102 Rev. A 8
 Presented at ISA 2004 Exhibit and Conference – Edmonton Section, April 2004

efforts to accurately measure flow rates and

density (and therefore percent solids) of bitumen
froth streams.

Future Work
The data presented in this paper demonstrates
the applicability of sonar-based flow meter to
overcome many of the long-standing flow
Slope of
Convective measurement challenges in oilsands processing
Ridge = industry. Long-term commercial trials and
5.95 fps
installations are currently underway to examine
the performance of sonar-based flow meters in
several applications including hydrotransport
Figure 17: Representative K- plot from 16inch, lines, underflow tailings lines, froth lines and
bitumen froth flowing at 6 fps final tailings lines. In most cases, the
performance of the sonar-based meters is being
Figure 18 shows a k- plot of the same data compared directly to that of mag meters, wedge
series, however, it is looking at a higher meters, and venturi meters installed inline,
frequency range, from ~10-60 Hz, instead of ~1- forming a basis to assess the repeatability and
10Hz in Figure 17. At the higher frequency reliability of sonar-based flow meters.
range, the slope of the acoustic ridge can be Additionally, several laboratory tests are
precisely determined and used to calculate the planned to assess the accuracy of sonar-based
amount of aeration in the froth. In this example, flow meters on slurries over a range of flow
sound is propagating at 109 fps, corresponding regimes typically encountered in oilsand
to ~15% gas volume fraction at a line pressure processing, including varying degrees of
of 20 psia. stratification. The authors intend to report the
results of these investigations when the results
become available.

Conclusions
Sonar-based flow meters represent a new class
of industrial flow meters well-suited to address
Slope of many of the long-standing flow measurement
Acoustic
Ridge = challenges in the oilsands industry. Laboratory
109 fps
and field data demonstrate the applicability of
this clamp-on flow metering technology to three
broad classes of slurries common in the oilsand
Figure 18: Representative k- plot from 16inch, processing industry – Hydrotransport, Tailings,
bitumen froth flowing at 6 fps showing Acoustic and bitumen froth. Sonar-based flow meter
Ridge indicating a Sound Speed of 109 fps calibrate to within +/- 0.5% on water in
calibration labs and the flow rates reported by
During the test, no reference for gas volume sonar-based flow meter are, in general, are
fraction was available, however, the clearly consistent with available field references. Sonar-
evident acoustic ridge and the well-established, based decompositions of the pressure fields,
first principles, link between speed of sound and termed k- plots, recorded during commercial
entrained air levels indicate that the sonar-based field trails on these non-Newtonian,
technology is well-suited for characterizing the inhomogeneous, abrasive, often-aerated slurries
entrained air levels in bitumen froth. Entrained are shown to exhibit similar characteristics to
air plays an important role in liberating bitumen those recorded for water and other types of
from oilsands, enabling the flotation of the industrial process fluids.
bitumen. However, the entrained air impairs

BI0102 Rev. A 9
 Presented at ISA 2004 Exhibit and Conference – Edmonton Section, April 2004

In addition to providing flow rates, sonar-based ISA Process Control Conference, Philadelphia,
flow meters provide a real-time measurement of PA, and June 2003.
the entrained gases in process lines. It is
believed that this device is the first device to Gysling, D. L. and Loose, D. H., “Sonar-based
monitor entrained air levels on a real-time, full- Volumetric Flow and Entrained Air
bore basis and the industry is currently Measurement for Pulp and Paper Applications”,
evaluating how best to leverage this new presented at the TAPPI Spring Technical
capability. Conference, Paper 58-1, Chicago, IL, May, 2003

Acknowledgements: Schlichting, H., Boundary Layer Theory,

The results presented in this paper have McGraw-Hill, New York, ISBN 0-07-055334-3,
benefited from several years of development 1979.
efforts in sonar-based flow measurement
technology. The authors gratefully appreciate Nielsen, R. O., Sonar Signal Processing, Artech
the efforts of many colleagues and co-workers House, Inc., Norwood, MA, ISBN 0-89006-
and that have contributed to results presented 453-9, 1991
herein.
Beranek, L.L. and Ver, I.S., Noise and Vibration
The authors gratefully acknowledge the efforts Control Engineering, Section 14.6, John Wiley
of Clarence Deroucher, Ian Verhappen, Sean and Sons, New York, ISBN 0-471-61751-2,
Sanders, Scott Hutton and their colleagues for 1992
sponsoring the field trial of the sonar-based flow
measurement at their facilities Crowe, C. et al, Multiphase Flows with Droplets
and Particles, CRC Press, Boca Raton, FL,
References ISBN 0-8493-9469-4, 1998.
Gray, M.R. and Masliyah, J. H., Extraction and
Upgrading of Oilsands Bitumen. – Intensive Joseph, D. D., et al, “Self-lubricated Transport
Short Course,” Department of Chemical and of Bitumen Froth,” J. Fluid Mech. (1999), vol
Materials Engineering, University of Alberta, 386, pp.127-148.
February, 2002.

Kragas, T. K., et al., “Downhole, Fiber Optic

Multiphase Flow Meter: Field Installation”
Presented at the Society of Petroleum Engineers
Annual Conference, SPE Paper #77654,
September 2002.

Kragas, T. K., et al., “Downhole Fiber-Optic

Multiphase Flowmeter: Design, Operating
Principle, and Testing,” Presented at Society of
Petroleum Engineers Annual Conference, SPE
Paper #77655, September 2002.

Gysling, D. L. and Loose, D. H., “Sonar-based

Clamp-on Flow meter for Gas and Liquid
Applications”, presented at the ISA Expo 2003,
Houston TX, October, 2003.

Gysling, D. L., and Loose, D. H., “Sonar-based

Clamp-on Flow meter for Chemical and
Petrochemical Applications,” presented at the

BI0102 Rev. A 10