SPE 71478
Field Testing Coriolis Mass Flowmeter in Central Ghawar, Saudi Arabia
Imran Abbasy, SPE, Saudi Arabian Oil Company
Copyright 2001, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the 2001 SPE Annual Technical Conference and
Exhibition held in New Orleans, Louisiana, 30 September–3 October 2001.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300
words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
Abstract
Field test results of a Coriolis mass flowmeter are presented
for a crude oil gathering facility located in a mature part of the
Ghawar field in Saudi Arabia. Due to the maturity of the field,
conventional well test measurements under low flow rate and
high water cut conditions suffered from inaccuracies and gave
erroneous and sometimes incorrect readings; consequently
there was a need to find a replacement mode of measurement.
This paper highlights advantages mass flowmeter technology
has to offer under such difficult conditions. The Coriolis
technology has been around for several years but its
performance has not been adequately documented in oilfield
applications and only recently has become economically
viable and user-friendly as a result of the manufacturer’s
efforts to expand their application base.
In the past, 3-phase testing was necessary due to inability of
conventional sensors to accurately measure flow at low and
high water cut ranges. The instrumentation was expensive to
maintain and large quantity of chemicals were required to
achieve proper separation. Moreover, error was introduced due
to an additional water rate measurement that was necessary to
quantify the total water production.
This field test covered a very wide range of flow rates and
water cuts and gave an insight into the improvements this
technology has to offer. The test results established the
accuracy and repeatability of this measurement. This
technology has therefore been adopted as the primary
measurement for all well testing in Central Ghawar where the
water cut ranges are extreme. The field test also concluded
that 3-phase testing was no longer necessary, which resulted in
significant saving in maintenance and chemical costs.
Introduction
The test separators in Uthmaniyah Gas Oil Separation Plants
(GOSPs) were originally equipped with a turbine flowmeter
for oil/water emulsion flowrate and a capacitance type meter
to measure water cut. In a low water cut environment, these
measurements were reliable and results were representative of
the well behavior. With the maturing of the field, most wells
now flow at high water cuts and consequently inaccuracies
were experienced in the water cut meter measurements. These
inaccuracies were not entirely unexpected, as theses devices
are only suitable for oil-continuous phases. A field test
conducted by Saudi Aramco concluded that the capacitance
water cut meters were unable to accurately measure water cuts
above 40%, and as such an alternative mode of measurement
was required to overcome this shortcoming1 .
Saudi Aramco started field-testing various multi phase
meters in the early 1990s. The objective of this initiative was
to select an appropriate technology to measure phase fractions
under various operating conditions and to take advantage of
new technology that was becoming available. Furthermore,
reliability and maintenance support was also evaluated for
each type of meter tested. The meters were tested under
different operating conditions and performance was carefully
evaluated2 . A Coriolis mass flowmeter was field-tested on an
onshore location in Saudi Arabia and the results were found to
be within 10% of the test separator measurement and thus
deemed acceptable. There were no operational problems
reported during this testing.
The presence of small fraction of gas in the liquid stream
generally has little effect on the mass flow reading3 . In
situations where there is a larger gas fraction in the liquid
stream, the accompanying density change is interpreted as a
decrease in water cut. Furthermore, mass flow readings are
also flawed due to the absorption of vibration energy by the
gas. It is clear therefore, that Coriolis mass flowmeters are
better suited for applications downstream of the separator or
where the liquid is flowing above the bubble point. For the test
separator application, the actual retrofit of the vessel was
relatively straightforward to implement and thus considered an
added advantage.
2 IMRAN ABBASY
The turbine flowmeter, being a mechanical device, requires
a minimum flow to overcome friction. Consequently, there is a
minimum flow threshold below which the flowmeter does not
make any measurement. The turbine flowmeter installed in the
plant has a limited accuracy outside the 3 to 30 thousand
barrels per day (MBD) range and the readings tend to be
unreliable and non-repeatable whenever the flow rates fall
outside this band. As the fields mature, more and more wells
are producing at low flow rates and deterioration in flowmeter
performance becomes very evident.
Theory of Operation
Coriolis mass flowmeter technology has evolved over the
years from a relatively niche application to widespread
commercial use. Chevron took the initiative and worked with
a Coriolis -meter manufacturer to develop a Net Oil Computer
(NOC) system that provides accurate data from 0 to 100%
water cut. This work4 is now the basis of most of the
commercial Coriolis meters available in the market today.
Water cut measurement has been the source of large errors
in testing producing wells. Conventional water cut meters are
either operating on the capacitance principle or the differential
head pressure measurements. Capacitance meters provide
reasonable accuracy in oil continuous emulsion range in low
water cut environments. However, large discrepancies are
introduced in these measurements at higher water cut, or
where the oil/water emulsion has a water-continuous phase.
Differential pressure transducers on the other hand perform
poorly due to the fact that friction correction between the two
measuring points is difficult to quantify.
A Coriolis meter consists of a flow tube that vibrates due to
an electro-magnetic drive coil mechanism. Fluid flow through
this tube introduces a twisting motion and the magnitude of
this twisting motion is proportional to the mass flow-rate.
Similarly, the vibrating frequency of this tube is related to the
density of the fluid flowing through it. Instantaneous water cut
(Yw ) is then calculated by the measured emulsion density (ρm)
using the following equation:
Yw = (ρ m - ρ o ) / (ρ o - ρ w ) ………………… (1)
ρ o and ρ w are the densities of crude and produced water
respectively. As is apparent, these parameters have to be
known with certainty and input in the NOC. The Coriolis mass
flowmeter measures the operating temperature of the fluid
stream, which is used to correct the above densities to some
temperature base, usually 60°F. Oil and water phase flowrates
are then calculated as follows:
Qw = Yw . Qt …………………………. (2)
Qo = Qt – Qw …………………………. (3)
SPE 71478
Qt is the flowrate measured by the Coriolis meter and Qw and
Qo are the calculated water and oil rates, respectively. These
rates are instantaneous, which are accumulated and averaged
by the NOC to provide total barrels and average production
rate in barrels/day.
Field Test Methodology
Excessive pressure drop across the sensor can cause gas
breakout, thereby introducing 2-phase flow condition in the
flow tubes. This can result in an improper operation of the
sensor. High fluid velocity through the flow tubes can also
cause erosion. Therefore, care should be taken when sizing the
sensor to ensure that pressure drop and fluid velocity are
within the recommended limit. For this application, ELITE®
CMF-400 sensor was selected, which can handle >60 MBD.
This is a 4” diameter, Stainless Steel 316L device and is the
largest size available in this range. Micro Motion (a division
of Emerson Process Management) manufactures this Coriolis
mass flowmeter. The sensor is shown in Figure 1.
Micro Motion sensor measures the flow rate accurately over
a 100:1 turndown ratio and density in the range of 0 to 5
gm/cc. Consequently, the sensor measures water cut in the
range of 0-100%. Pressure drop through the device is minimal
and is generally less than 5 psi. The Micro Motion is equipped
with a NOC that produces real-time measurements of water
cut, net oil volume flow and net water volume flow. The oil
and water densities at 60°F are manually entered into the NOC
or measured on-line, if desired. The sensor measures the actual
fluid temperature, density and mass flow rate, and then the
NOC extrapolates the standard densities to the operating
temperature. The water cut is calculated at the operating
temperature (using Equation 1), and then referenced back to
60°F.
Oil density is well known for a particular area, however
water density can vary significantly depending on where the
well is located in relation to the flood front. For this test, 8
wells were sampled randomly in the area and it was
determined that the water density was in the range of 1.0435
and 1.0628 gm/cc. Seawater is 1.0484 gm/cc, which
corresponds to 36,000 mg/L Chlorides. The maximum
variation in water cut due to maximum expected error in water
density was calculated to be about 1%. It was therefore
decided to use a fixed density of 1.0547 gm/cc for all testing.
It was however recognized that this approach would be too
simplistic in an environment that was not as mature as ours,
and where injection and formation water mixing had not
occurred to this extent. For oil density, an average value of
0.8371 gm/cc was used and found to be accurate for all the
wells that were tested.
The Coriolis flowmeter was installed upstream of the
Halliburton turbine meter and the capacitance water cut
(BS&W) meter of the test separator. A schematic of test
separator instrumentation is given in Figure 2.
SPE 71478 FIELD TESTING CORIOLIS MASS FLOWMETER IN CENTRAL GHAWAR, SAUDI ARABIA
This interim retrofit was used to check the Micro Motion
readings against the conventional test separator measurements.
Since this instrumentation was on the oil outlet of the test
separator, it was necessary to test the wells in 2-phase mode to
get a representative comparison. In cases where the test
separator was incapable of giving a representative rate, 3-
phase testing was also done and compared against Micro
Motion measurements in 2-phase mode.
Correction for Shrinkage
The conventional oil measurements from the test separator are
used for projecting the oil deliverability from the field,
consequently it is critical that the volumetric rates measured
are corrected to a condition which are commensurate with
what is being delivered downstream of the crude stabilization
plant.
In the past, a set of shrinkage factors were used to correct
the test separator oil rates to the stock tank barrels downstream
of the crude stabilization plant. These factors compensate for
temperature as well as pressure. The Micro Motion readings
are already corrected to 60°F, therefore a new set of shrinkage
values had to be determined to account for pressure shrinkage
alone.
A commercial process simulator was used to develop a set
of shrinkage factors to take into account the various test
separator operating pressures and temperatures. The
simulation data was curve fit using a 3-dimensional curve-
fitting program, to generate a 3-D shrinkage curve, shown in
Figure 3. The simulation used a calibrated model for the
plant. The test separator oil inlet was flashed at a pressure
ranging from 100 psig to 160 psig and a temperature ranging
from 70°F to 180°F. Then the test separator oil was flashed to
50 psig to simulate the second stage separation conditions and
flashed again to a Reid Vapor Pressure (RVP) of 10 psia to
simulate the stock oil conditions at the crude stabilization
plant. These factors are a function of the test separator
temperature and pressure.
Test Results
The field test was carried out over the period February to May
2000. A total of 81 tests were done in 2-phase mode through
the test separator. Some wells were also tested in 3-phase
mode so that all possible operating scenarios were compared.
Total number of wells tested was 41. The wells were chosen
with the objective of subjecting the Micro Motion to the entire
spectrum of flow rates and water cuts that we encounter in the
plant. The liquid from the test separator oil outlet was
measured by the Micro Motion and compared against the
conventional test trap measurements. The validity criteria
chosen was,
Difference in Rate < 500 bbls/day or < 10%
3
Tests were considered to be good when the test separator
and the Micro Motion measurements met one of the above two
conditions. This incidentally is also the standard that is used
for validating routine well tests within Aramco. The results for
the wells tested are plotted in Figure 4 and 5 along with a
10% error band.
The performance of the Micro Motion against conventional
test separator measurements for all tests was as follows:
Oil Rate: 85% (69 tests were valid)
Water Rate; 75% (61 tests were valid)
As can be seen from Figures 4 and 5, the error in oil and
water rates were essentially due to errors in water cut
measurements. This was not entirely unexpected as sited
earlier on in this paper. Reasons underlying these
discrepancies were investigated and the findings are given
below:
Water Cut. It was observed that water cut reading did not
match the conventional water cut meter at very low and high
water cuts, hence the water rates were falling outside the
acceptable range (Figure 5). The conventional water cut
measurement works on the capacitance principle and one
would expect that there would be inaccuracies at the low and
the high end due to the inherent limitation of this gauge.
In high water cut situations, water-continuous phase flowing
through the BS&W meter was electrically shorting out the
capacitance probe, thereby giving a water cut reading that was
significantly higher than actual. This pattern was very
consistent at high water cuts. Micro Motion read a relatively
lower water cut, whereas the BS&W meter constantly gave
artificially high water cut readings. As a result, water rate was
over-estimated and oil rate underestimated by the conventional
well test system. This phenomenon was apparent at water cuts
in excess of 75%.
At water cuts below 5%, oil-continuous phase dominates the
flow through the BS&W meter. The insulating property of oil
forces the capacitance probe to underestimate water and hence
show predominantly oil. Micro Motion on the other hand
consistently saw small quantities of water. Salt analysis of the
samples confirmed the presence of water in these supposedly
dry oil producers. Although water cut from Micro Motion is
an average measured over a certain period of time, the values
were generally closer to the sampled water cuts in comparison
to the BS&W meter measured water cuts. The BS&W meter is
therefore limited in accuracy to the intermediate range.
Flow Rate. During the test it was observed that at low flow
rates (<2 MBD), Micro Motion generally read higher than the
Halliburton turbine flowmeter (Figure 4). To fully understand
this phenomenon, refer to the test separator schematic (Figure
2). The objective of the Level Control Valve (LCV) on the 4”
line is to hold level in the vessel. Consequently, this LCV
4 IMRAN ABBASY
tends to be either barely cracked open or fully closed for low
flow rates. At higher rates however, this LCV remains wide
open most of the time. During the installation of the Micro
Motion it was observed that the 4” LCV 4003-2, and possibly
the 2” LCV 4003-1, were leaking. This was discovered as the
turbine read zero flow with these LCVs closed, whereas the
Micro Motion read a small rate. Once the upstream block
valve was closed, both these meters read zero flow.
As discussed earlier, a turbine flowmeter requires a
minimum flow to overcome friction and to start measuring.
This threshold flowrate is not always achieved due to surging
and in wells that have low productivity. In an attempt to
ensure that this minimum flow is maintained through the
turbine meter at all times, an over-lock system was designed
on the test separator LCVs that would close the outlet valves
of the vessel during low flow rate situations. Once a flow that
is sufficient to start the turbine meter is established in the trap,
the valves would open. Therefore, due to leaking LCVs and
friction, the turbine flow rate inherently lagged behind the
Coriolis mass flowmeter rate on a cumulative basis.
During 3-phase testing, some of the water is separated from
the wet crude by injecting demulsifier, and then each stream is
measured separately using turbine flowmeters to give a total
water rate. This summation from two separate devices results
in a higher measurement error. As per the manufacturer, the
turbine flowmeter has a turndown ratio of 1:10 and a range of
30 MBD. Therefore, the device is inaccurate at rates below 3
MBD.
During 2-phase testing, the turbine meter calculates the flow
rate and the water cut meter measures the water cut in the
fluid. The rates are then calculated as in equations 2 and 3. It
is clear that any error in water cut would have a direct effect
on the calculated oil and water rates.
All the above problems were inherent to the instrumentation
and equipment at the GOSP and therefore had no bearing on
the performance of the Micro Motion. Consequently, tests at
low flow rate (<2 MBD), low water cut (<5%) and high water
cut (>75%) applications were not taken as a benchmark and
excluded from the comparison. The performance of the Micro
Motion was then as follows:
Oil Rate: 94% (76 tests were valid)
Water Rate: 85% (69 tests were valid)
It should be kept in mind that some of the tests that
remained invalid were on wells, which have a history of
giving erratic results. Some of these wells were high water cut,
low rate and prone to surging. Testing on these wells over the
past six months had not been very consistent and there were
large variation in rates. Consequently, their invalidity had to
do with well behavior rather than any instrumentation per se.
The issue of Micro Motion repeatability was studied in
detail. The repeatability criteria chosen was 10% of oil and
SPE 71478
water volumes between tests. Out of the 22 tests that were
repeated in 18 wells, only one test did not repeat within 10%
and a vast of majority repeated within 5%.
Material selection is one of the critical points to be
considered while selecting a Coriolis mass flowmeter for
crude oil applications, as produced oil often contains high
levels of Chloride ions and Hydrogen Sulphide. Generalized
corrosion tables are not sufficient for selecting the material of
Coriolis sensors, as these tables do not take into account effect
of cyclic loading to which Coriolis sensors are subjected.
Corrosion in combination with cyclic loading causes corrosion
fatigue that can lead to cracking of sensor flow tubes.
Therefore, for such applications it is generally recommended
to use Hastalloy-C sensor, which is a Nickel-based Stainless
Steel alloy with vastly superior corrosion resistance and higher
strength.
Conclusions
The results of the field test established the accuracy and
repeatability of the Coriolis mass flowmeter over the entire
range of flow. The measurement does not suffer from the
inherent non-linearity problems of the conventional
capacitance water cut meters. The device is robust, reliable
and as there are no moving parts, maintenance requirements
are minimal.
3-phase testing is no longer necessary in a high water cut
environment and therefore all related instrumentation is
redundant. All piping and instrument loops associated with the
turbine flowmeter and the capacitance water cut meter on the
oil outlet can be removed. Moreover, since testing is to be
done in 2-phase only, all water outlet piping and associated
instrumentation can also be decommissioned. Dismantling of
this redundant hardware would result in substantial saving in
terms of maintenance costs.
Future Plans
It has been decided to gradually phase out all 3-phase testing
in plants where water cuts are high. Several plants in Central
Ghawar are now targeted for the installation of Coriolis mass
flowmeters, which would translate to lower maintenance and
demulsifier costs; and above all, provide more accurate and
repeatable data. A plan is also underway to retrofit these test
separators for 2-phase testing and eliminate unnecessary
hardware.
An initiative is also underway to validate the simulated
shrinkage factors against laboratory-derived parameters. Once
this study is complete, these factors will be applied for oil
allocations and projections for all plants.
At this time CMF-400 is only available in Stainless Steel.
CMF-300 is available in Hastalloy-C metallurgy, however it is
restricted to about 24 MBD flow rate. Based on the review of
wells in the area, CMF-300 should be adequate for this
application and thus this metallurgy issue will be resolved.
SPE 71478 FIELD TESTING CORIOLIS MASS FLOWMETER IN CENTRAL GHAWAR, SAUDI ARABIA 5

Acknowledgments
The author would like to thank F.A. Hajji, D.P. Delavan and HPTT
D-401 0-100% watercut
S.G. Barlow of Saudi Aramco for their input and invaluable meter AT
4001
comments while writing this paper. Thanks are also due to Water
Outlet
Oil LCV-4003-1
Outlet
Asghar Satvilker of Fisher Rosemount for providing product
information. This is used
only during 3- To LPPT
Strainer
phase testing LCV-4003-2
References FM-4003
Turbine
MicroMotion
1. Stephens, J.F., “Report on Flowmeter and Water Cut Tests in
Hawtah,” Saudi Aramco, Internal Correspondence, November 30,
1992.
2. Al-Taweel, A.B., and Barlow, S.G., “Field Testing of Multiphase To WOSEP
Meters,” SPE 56583, presented at the 1999 SPE Annual Technical Strainer
LCV-4001
FM-4002
Conference and Exhibition, Houston, Texas, October 3-6, 1999. Turbine

3. Taylor, R.D.H., and Nuttall, R.C.H., “On-Line Well Monitoring

and its Application in South Oman Oil Field,” SPE 25619,
presented at the SPE Middle East Oil Technical Conference and
Figure 2 - Instrument Schematic – Test Separator.
Exhibition, Bahrain, April 3-6, 1993.
4. Aslesen, K.S., Bocek, R., Canfield, D.R., and Liu, K.T., “Method
and Apparatus for Testing the Outflow from Hydrocarbon Wells
on Site,” U.S. Patent 4,689,989, September 1, 1987. Micro Motion Shrinkage Factors
Rank 4 Eqn 310
r^2=0.99996061 DF Adj r^2=0.99991685 FitStdErr=0.00010091333 Fstat=28209.729
z=a+bx+cy+dx^2+ey^2+fxy+gx^3+hy^3+ixy^2+jx^2y
a=0.97859106 b=0.00041258311 c=-0.000621469 d=-4.4534132e-07 e=1.9128087e-06
f=1.3221692e-06 g=-3.6910966e-09 h=-2.133067e-09 i=-6.0240899e-09 j=3.1248812e-09

0.995
0.99
0.995 0.985

Shrinkage Factor
0.99 0.98
0.985 0.975
Shrinkage Factor

0.98 0.97
0.975 0.965
0.97 0.96
0.965 0.955
0.96 0.95
0.955 170
160
150
0.95 140
130
150 120
140 110
130 100
90 Separator Temp., deg.
120 80
Separator Pressure, PSIG 110 70 F
100

z = a + bx + cy + dx^2 + ey^2 + fxy + gx^3 + hy^3 + ixy^2 + jx^2y

[Where z = Shrinkage Factor, x = Temperature (°F) and y = Pressure
(psig)]

Figure 3 - Shrinkage Factors from the Process Simulator.

Figure 1 - Micro Motion CMF-400 Elite® Meter.

6 IMRAN ABBASY SPE 71478

20000
18000
MIRCO MOTION, bbls/day

16000
14000
12000
10000
8000
6000
4000
2000
0
0 5000 10000 15000 20000
Test Separator, bbls/day

2-Phase 3-Phase

Figure 4 - Total Fluid, Oil + Water - Micro Motion vs.

Turbine Flowmeter.

100%
90%
80%
MICRO MOTION

70%
60%
50%
40%
30%
20%
10%
0%
0% 20% 40% 60% 80% 100%
Test Separator

2-Phase 3-Phase

Figure 5 - Water Cut - Micro Motion vs. BS&W Meter.