You are on page 1of 9

SPE 114127

Reliability of Multiphase Flowmeters and Test Separators at High Water Cut


Kelechi Isaac Ojukwu, Petroleum Development Oman, and John Edwards, Schlumberger

Copyright 2008, Society of Petroleum Engineers

This paper was prepared for presentation at the 2008 SPE Western Regional and Pacific Section AAPG Joint Meeting held in Bakersfield, California, U.S.A., 31 March–2 April 2008.

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 have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper 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 SPE copyright.

Abstract

The use of multiphase flowmeters (MPFM) for well test measurements is increasingly becoming a standard practice replacing
conventional test separators. These MPFMs are usually tested and calibrated in laboratory controlled flow loops using
idealized fluids in steady state conditions. However for high water-cut, high gas-volume-fraction and low pressure unstable
flow these controlled conditions are far from reality which can lead to unforeseen errors in the field. Recent experience shows
that in certain conditions, the various types of multiphase flowmeters react quite differently to the measurement challenges of
transient flows in high water cut and high gas volume fractions (GVF). Some meters can be unreliable in measuring oil rates
in certain conditions which leads to inaccurate estimation of the wells' potential and associated uncertainty in plans for
production optimization. For example the inaccuracies in measured oil rates could be greater than the gain expected from a
stimulation or restoration. An inaccurate measure of oil rate also leads to a poor reconciliation factor and poor estimation of
reserves. By resolving these inaccuracies and allocating oil correctly to wells it is possible to invest in right opportunities
thereby saving unnecessary expenditure. The factors that affect multiphase flowmeter measurement can range from excessive
gas-volume-fraction, low line pressures, solids and unsteady flow.

In 2007 an in-situ comparison was made of two types of multiphase flow meters and two separators to determine the most
appropriate well test device for X Field in the Sultanate of Oman. X Field wells are artificially lifted and have a high water
cut of 80%. The reconciliation factor (also known as the back allocation factor) before this well test campaign was 0.54. The
comparison was made on twelve wells producing to X Field station. Each well was tested for 12 hours in series using an
existing inline MPFM, a low pressure production test separator, a second mobile MPFM, a 1440psi separator, and a
multiphase tracer technique. The objective to determine the most reliable technique(s) of measuring oil production rates was
accomplished, and now the reconciliation factor is 1.03. This paper describes the comparison, how the results were analyzed,
and the comparative error associated with each technique.

Introduction

In 2002 Petroleum Development of Oman (PDO) initiated a change from production separators to multiphase flowmeters.
This decision was taken not only to remedy the deterioration of the reconciliation factor that had reached 0.775 by 20001, but
also for economic reasons. At that time 80% of the wells in North Oman were producing at a GVF of 85% or more with
operating pressures down to 4 bar, and 70% of the wells had a net oil flow of 20 - 100m3/day. The following MPFM
operating envelope specifications were defined to cover the production range from wells in 123 fields:

Operating Envelope for MPFM specification


Gross liquid flow rate 30 to 1200 m3/d Accuracy of the mobile well test multiphase flow meter
GVF 0 to 99% Gas flow rate +/- 10% relative
Water cut 0-100% Water cut +/- 2 % absolute
Piping class of the skid ANSI 600# rated Gross liquid rate +/- 5% relative

Net oil flow rate accuracies are dependent on the actual water cut in the flow and vary for different water cuts. For the
purpose of consistency in the rest of the paper the two types of multiphase flowmeters will be refered as MPFM-A and
MPFM-B.
2 SPE 114127

In 2003 it was believed only MPFM-A had an operational envelope that satisfied this requirement. This was based on three
phase flow loop tests of four different multiphase flowmeter designs using live crude oil and natural gas performed in the
multiphase flow facility located in Daqing Oilfield, Daqing City, China2. Most three phase flow loops used for research,
development and quality control of multiphase measurement devices use known fluids such as diesel, air, nitrogen and water.
The use of live crude and natural gas in the Daqing flow loop used to test the MPFM-A was considered an advantage. This
meter includes a gas conditioning cyclone that reduces the amount of gas in the mixture by separating some of the gas from
the total flow. This reduces the GVF seen by the three phase measurement system in an attempt to achieve the required
specification of uncertainty up to 99% of Gas Volume Fraction.

In the intervening years as MPFM-A replaced production station separators the reconciliation factor failed to improve. In
2007, after three years of using MPFM-A, the X Field reconciliation factor had actually dropped to 0.54. This was the
catalyst for a local investigation; a well test campaign that would compare all available flow measurements to resolve this
long lasting metering problem. The objectives were to determine the actual production from each well, to check the validity
of the multiphase flowmeters, and to improve the reconciliation factor.

The comparison of various flow measurements is only meaningful if a reference rate is available. The X Field production
station 3-phase low pressure test separator, named TSEP-B in this paper, was used for that purpose. A second reference
available for net oil rate is the multiphase tracer technique3. This device has been used to verify oil and water rates. The third
reference is the export meter from the facility. A 1440psi horizontal separator called TSEP-A and a multiphase meter called
MPFM-B were also provided by different vendors for validation of MPFM-A. Therefore the X Field provided an excellent
opportunity to compare MPFM-A, MPFM-B, a high pressure well test separator (TSEP-A), a low pressure well test separator
(TSEP-B) and a multiphase tracer. This comparison was repeated in twelve wells with 12 hour tests of each well.

X Field

The X Field gathering station has two production separators; a sour separator for five wells producing from a shallower
carbonate formation, and a sweet separator for five wells producing from a deeper clastic formation and two wells from
nearby Y field, also producing from a clastic formation. Sour wells could not be produced through the TSEP-B. This is due to
restrictions on flaring any gas containing H2S. Five of the X Field sweet wells are on gas lift; the other seven have electrical
submersible pumps (ESP). The fluid properties are tabulated in Figure 1.

Well Modeling

It is possible to further validate each well’s production independently of the actual testing using Nodal analysis. Recent
wireline formation pressure transient buildups and infill well static pressures confirmed that the carbonate matrix (but not the
fracture system) in X Field was still above bubble point. The clastic layered formations of Y Field have strong aquifer
support, and have been kept above bubble point. Considering little lateral flow from the various areas of the reservoir, it is
fair to assume that the solution GOR from earlier PVT analysis is still applicable; 30 m3/m3 for X-9, 10,12,17,25 and 100
m3/m3 for Y-2, 3. These GOR were used as an input to the Nodal analysis software to model the ESP lifted wells. For gas
lifted wells the measured GOR is the sum of lift gas and produced gas, and flowing gradient surveys were acquired and used
as an input to another Nodal software to obtain estimates of the production rates as shown in Figure 2. Solution GOR for X-
7,14,28,29,30 range from 60 to 128 m3/m3 depending on the producing layers. BSW measurements were acquired
periodically from samples in each well throughout the 12 hour test periods. The GOR in/outflow models and these BSW
measurements were used to reject any unrealistic production test results, and to validate the final choice of the “best”
measurement for each well. For example from the measured gas rate and the GOR the calculated oil rate should be
comparable to the measured oil rate, and in ESP lifted wells the discharge and intake pressures should be consistent with the
measured phase density and nodal analysis model.

Well and Multiphase Flowmeter Preparation

One line connected the two multiphase flowmeters, two separators and multiphase tracer in series, with a return line to the
production station. Each metering technique has a somewhat different favorable operating range, and some wells fell outside
their respective envelopes. However the objective was to determine which meter was most suitable for X Field, not to
determine whether the meters fulfilled their published specifications. Both multiphase flowmeters were calibrated and
verified prior to the comparison. All the wells were left undisturbed to flow for many weeks prior to these well tests. Figures
3 to 8 show the layout of the equipment in the test loop, photographs taken during the testing and description of each device.
SPE 114127 3

Results

The results of the investigation are tabulated and cross plotted in Figures 9 to 14. TSEP-B, multiphase tracer and MPFM-B
gave comparable measurements of net oil. However the 30 minute hydrophobic and hydrophilic tracer measurements are only
30 minute averages; not always representative of a 12 hour average in fluctuating or slug flow. This could have been
improved by periodically repeating the multiphase tracer measurement during the 12 hour test period. In unstable wells the
instantaneous match of multiphase tracer net oil and MPFM-B was better. The multiphase tracer technique only gave
comparable water rates in 50% of the wells for reasons yet unknown. The best estimates of the actual net oil production from
each well came from TSEP-B in the gas lift X Field wells and ESP Y Field wells, while the MPFM-B was best in the ESP X
Field wells. TSEP-B and MPFM-B were not compared in those X Field wells producing H2S because TSEP-B was not
configured to test sour wells due to the flaring restriction. MPFM-A always amplified the net oil rates with errors up to
1550%; this uncertainty reduced when the net oil rate exceeded 50m3/day. Only one MPFM-A production rate was used in
the final tabulation of best results, and this was for the highest net oil producer of 114 m3/d. TSEP-A also performed poorly
due to instability at low line pressures, errors in applying global BSW when using the separator in two-phase mode,
inappropriate high-rated velocity meters on the oil line, unstable liquid level from manual valve control, and lack of
shrinkage/meter factor corrections. TSEP-A could have been improved by configuring it for 3-phase flow under low
pressures and high water cut, including a shrinkage tester, and making insitu meter factor calibrations.

For all cases two or three net oil measurements out of five were in close agreement, and were close to the well model
production estimate. The final choice of the “best”measurement from this smaller group followed these rules: for high GVF
use the TSEP-B low pressure separator, otherwise use the MPFM that most closely matches the multiphase tracer. The x-axis
reference in the two cross plots Figures 11 to 12 is multiphase tracer and MPFM-B. The best oil rates were used with the
export meter to calculate a new reconciliation factor of 1.03, a significant improvement from 0.54 based on MPFM-A. The
monthly reconciliation factor of the X Field gathering station has been consistently around 1.0 since April 2007 when TSEP-
B was reinstated. There have also been additional immediate benefits: Well X-25 was closed after verifying nearly zero net
oil to save electricity and the proactive pump replacement workover replaced with water shut-off activity. New zones were
perforated yielding a gain of 77m3/d net oil. Well X-10, also on the schedule for a proactive pump replacement, was worked
over to add new perforations which improved its net production from 15 to 43m3/d. The almost immediate increase in oil
production from this workover activity based on the best net oil rate measurements exceeded the cost of the entire meter
verification exercise.

Conclusions

TSEP-B is recommended for routine well tests of net oil rates in X Field because it provides the most accurate and consistent
data which has resolved the reconciliation problem at the X Field gathering station. It will be configured to test sour wells.
The TSEP-B oil and water meters should be calibrated every six months. The multiphase tracer technique is recommended
for biannual well test net oil rate validations and accurate reconciliations, and some further investigations is required to
explain the non performance of the tracer technique for water measurement in 50% of the cases. The MPFM-B gave the
closest match to TSEP-B and the multiphase tracer technique, despite errors associated with low pressures and high GVF.
The MPFM-A was only reliable in one of the 12 wells. The utilization of this type of meter is not recommended for this
environment of low oil, high GVF and high water cut.

Such a comprehensive verification of the quality of well test measurements is rare, and provides powerfull insights on the
different methodologies of measurements and allows significant improvement of the reconciliation factor. Operational
recommendations can be drawn on the selection of the optimum testing equipment according to the well behavior and type of
fluid, and on the periodicity of calibration and verification of the meters, thereby ensuring an adequate metrological
performance of the well tests.

Acknowledgements

The authors wish to acknowledge the good collaboration spirit between the involved parties in the performance and analysis
of the well testing compaign, and to thank the Ministry of Oil and Gas from the Sultanate of Oman and the management of
Petroleum Development of Oman for giving permission to present some of the field results necessary to illustrate the
methodology presented in this paper.
4 SPE 114127

References

1. K. Busaidi et al, Petroleum Development of Oman, “New Development in Water Cut Meter with Salinity
Compensation”, SPE 77894, SPE Asia Pacific Oil and Gas Conference, Melbourne, Australia, 8-10 October 2002
2. H. Bhaskaran et al, Petroleum Development of Oman, “Multiphase Flow Meter: Experience and Assessment at
PDO”, SPE 84505, SPE Annual Technical Conference, Denver, USA, 5-8 October 2003
3. M. Bohari et al, Petronas Carigali “Multiphase Flow Measurement Using Tracer Technology at Dulang Oil Field,
Malaysia”, SPE 80501, SPE Asia Pacific Oil and Gas Conference, Jakarta, Indonesia, 15-17 April 2003

Definitions

Multi-phase flow A liquid and a gaseous substance flowing simultaneously in a conduit; oil and gas, or water, oil and gas
MPFM Multiphase flow meter, device for measuring the individual oil, water and gas rates in a multi-phase
flow without need for separation
GVF Gas volume fraction is the volumetric flowrate of gas at line condition / volumetric flowrate of total
fluid (gas and liquid) at line condition
WC Water cut is the ratio of the volume flow rate of water and the total liquid volume flow rate, both
volumes converted to the same pressure and temperature
multiphase tracer Measurement of oil, water and gas rates by injecting hydrophobic (for oil), hydrophilic (for water) and
gas tracers in a pipeline while sampling at a designated mixing distance. The samples are analyzed to
estimate inline oil, water and gas rates using the equation below3

Q = qi {Fr − Fb }/ β o {Fs − Fb } Q= Oil Flowrate


qi= Tracer injection rate
Fr= reference signal
Fb= Background signal
Fs= Analysis signal from sample
SPE 114127 5

Description X Field carbonate X Field clastic Y Field clastic

Gross rate, m3/d 150 – 900 100 – 800 100 – 400


Net oil rates, m3/d 0 – 150 0 – 50 0 – 150
Maximum GVF, % 60-97 90-99 90-98
Solution GOR, m3/m3 30 60-128 29-57
Producing GOR, m3/m3 22 100 120
API Grade 27 40 29
Gas Gravity 0.7 0.75 0.7
Average water cut, % 81 90 15
Water cut range, % 20-95 80-98 0-30
Artificial lift system ESP Gas lift ESP
H2S, ppm 20-40,000 0 0

Figure 1. – Relevant fluid and artificial lift properties of X and Y Fields

P r es s ur e ( k P ag)

0 200 0 4 000 600 0 8 000 10 000 120 00 1 400 0


0

200

400

600
D ep th (m eters )

800

100 0

120 0

140 0

160 0

180 0

200 0

Production Pressure Model Design Injection Pressure Design Open / Closing Points (Closing)

Flowing Pressure Survey Design Open / Closing Points (Open)

Data inputs to model

Lift gas 10,000 m3/d


GOR from PVT 125 m3/m3
Flowing tubing pressures from wireline survey As on graph above

Modelled liquid production rates Measured liquid production rates

Gross liquid m3/d 131 131


Oil m3/d 5 5
Water m3/d 126 126
Gas m3/d 10,750 (10,000 lift gas plus 750 formation gas) 14,208 (10,000 lift gas plus 4208 formation gas)
GOR m3/m3 125 842

Figure 2. – Model of the tubing pressures in well X-28 gas lift well, thereby creating an estimate of the flow rate to test the validity of
the production test. On the pressure versus depth plot the black triangles are the model data, which are simulated to match the red
triangles representing the measured pressure data. This match is achieved by changing the production and gas injection rates. In
this example it was possible to achieve a close match of the liquid rates using the measured production rates from the “best” test in
well X-28. The errors in gas rates are larger, partly due to inaccuracy in lift gas measurement; however the objective of the
production test was to determine the best measurements of net oil rate. The deepest gas lift entry at 600m is not optimum, rather
limited by the available gas lift pressure.
6 SPE 114127

Figure 3. - WellTest equipment layout outside X Field flowstation

rd
Figure 4. - MPFM-A with tie in point, permanently installed Figure 5. - MPFM-B, truck mounted 3 party multiphase meter

rd
Figure 6. – TSEP-A, 1440psi 3 party horizontal separator and a Figure 7. - TSEP-B, low pressure production station horizontal
surge tank test separator in X Field
SPE 114127 7

Device Owner Description Limitation Advantage


rd
multiphase 3 party vendor-1 Sampling and time-of-flight of hydrophylic (for water) and 30 minute average, poor Cost, non-
tracer hydrophobic (for oil) tracers water measurement intrusive
rd
MPFM-A 3 party vendor-2 2 radioactive sources, 2 gamma ray attenuation Looses accuracy at high Portable, inline
permanently measurements and a differential pressure measurement GVF >98% measurement
installed across a venturi spool
rd
MPFM-B 3 party vendor-3 Single radioactive source with dual gamma ray Looses accuracy at high Portable, inline
mobile truck attenuation measurements and a differential pressure GVF >98% measurement
mounted measurement across a venturi spool
rd
TSEP-A 3 party vendor-4 3-phase horizontal 1440psi separator configured as a 2- Separator inappropriately Efficient at
phase separator with manual BSW sampling in liquid line. configured for these well high GVF
Level control performed manually without actuator-wizard. tests. Requires 4 operators
TSEP-B In-house 3-phase horizontal low pressure separator, metering of Flaring constraint would Efficient at
permanent test oil, water and gas phases with auto-level control and not allow testing sour wells high GVF
separator BSW meter in oil line

Figure 8. – Summary of devices used in the comparison

multiphase tracer MPFM-B MPFM-A

Well GVF Lift Oil Water Gross BSW GOR Oil Water Gas Gross BSW GOR Oil Water Gas Gross BSW GOR
% m3/d m3/d m3/d % m3dD m3/d m3/d m3/d % m3/d m3/d m3/d m3/d %

X-7 94 GL 15 986 999 98 67 15 563 8994 578 97 900 36 563 11256 599 94 313
X-9 90 ESP 102 451 553 82 5 53 345 4624 398 87 87 114 340 4091 454 75 36
X-10 68 ESP 13 481 494 97 38 15 262 596 277 95 40 121 334 39 455 73 0
X-12 64 ESP 5 543 548 99 110 14 445 819 459 97 59 231 401 155 632 63 1
X-14 97 GL 19 475 494 69 26 6 248 4203 254 89 701 60 232 3534 292 97 59
X-17 97 ESP 51 29 80 36 2 76 17 3402 93 18 45 68 33 846 101 33 12
X-25 60 ESP 0 806 806 100 4.3 323 489 327 99 114 12 672 215 684 98 18
X-28 97 GL 7 226 233 97 33 1 165 1503 166 99 1503 44 141 0 185 76 0
X-29 98 GL 15 228 243 94 16 5 167 540 172 97 108 41 239 16339 280 85 399
X-30 94 GL 16 962 978 98 61 10 470 5568 480 98 557 67 411 10572 478 86 158
Y-2 98 ESP 118 0 118 0 1 113 8 3969 121 7 35 162 2 3275 164 1 20
Y-3 94 ESP 18 171 149 189 90 8 92 525 1 344 73 0

Figure 9. - Summary of well test results for multiphase tracer, MPFM-A and MPFM-B. Red numbers are considered the “best” of the
five measurements

multiphase tracer TSEP-A (1440psi separator) TSEP-B (low pressure separator)

Well GVF Lift Oil Water Gross BSW GOR Oil Water Gas Gross BSW GOR Oil Water Gas Gross BSW GOR
% m3/d m3/d m3/d % m3/d m3/d m3/d m3/d

X-7 94 GL 15 986 999 98 67 39 593 8589 632 94 220 19 448 7348 467 96 387
X-9 90 ESP 102 451 553 82 5 136 270 9120 406 67 67
X-10 68 ESP 13 481 494 97 38 16 281 0 297 95 0
X-12 64 ESP 5 543 548 99 110 35 551 0 586 9 0
X-14 97 GL 19 475 494 69 26 7 286 8589 293 98 1227 19 182 6272 201 31 330
X-17 97 ESP 51 29 80 36 2 79 28 4085 107 26 52
X-25 60 ESP 0 806 806 100 0 551 0 551 100
X-28 97 GL 7 226 233 97 33 4 168 1892 172 98 473 5 126 4208 131 96 842
X-29 98 GL 15 228 243 94 16 31 149 4800 180 83 155 12 112 5757 124 90 480
X-30 94 GL 16 962 978 98 61 22 464 2887 486 95 131 12 302 5319 314 96 443
Y-2 98 ESP 118 0 118 0 1 121 1 3023 122 1 25 102 0 4901 102 0 48
Y-3 94 ESP 6.4 222 0 228 97 0 16 110 1908 126 87 119

Figure 10. - Summary of test results with multiphase tracer repeated, TSEP-A and TSEP-B. Red numbers are considered the “best”
of the five measurements
8 SPE 114127

250
MPFM-B
Equal Line

MPFM-A TSEP-A

TSEP-A
200 TSEP-B

TSEP-B

150

m3/d
MPFM-B

100

50

0
0 50 100 150 200 250

Multiphase tracer, m3/d

Figure 11. - Oil rate comparison of multiphase tracer to all other measurements. MPFM-A is highly erroneous when the oil rate is
less than 50m3/day and hence no trend line

250
TSEP-A
Multiphase tracer Multiphase tracer

MPFM-A

TSEP-A
200
TSEP-B
Equal Line

TSEP-B

150

m3/d

100

50

0
0 50 100 150 200 250

MPFM-B, m3/d

Figure 12. - Oil rate comparison of MPFM-B to all other measurements. MPFM-A is highly erroneous when the oil rate is less than
50m3/day and hence no trend line
SPE 114127 9

Wells H2S Previous MPFM-A New rates from the Source of the best Differences
measurements best of three measurement
measurements

Gross Oil Net Gross Oil Net Gross Oil Net


m3/d m3/d m3/d m3/d m3/d m3/d

X-7 No Gas lift 909 82 467 19 TSEP-B 442 63


X-14 No 368 46 201 19 TSEP-B 167 27
X-28 No 316 54 131 5 TSEP-B 185 49
X-29 No 247 29 124 12 TSEP-B 123 17
X-30 No 638 40 314 12 TSEP-B 324 28
X-9 Yes ESP 485 120 454 114 MPFM-A 31 6
X-10 Yes 631 88 277 15 MPFM-B 354 73
X-12 Yes 751 130 459 14 MPFM-B 292 116
X-17 Yes 134 122 93 76 MPFM-B 41 46
X-25 Yes 821 97 327 4 MPFM-B 494 93
Y-2 No 225 218 102 102 TSEP-B 123 116
Y-3 No 338 112 126 16 TSEP-B 212 96

Figure 13. - Difference in oil and gross production rates after substituting the MPFM-A production rates with the “best” production
rates from the three alternatives. Note the multiphase tracer measurements were not considered as a possible “best” because they
only represent a 30 minute flow period, not the 12 hour test period used for all the other tests. The 1440psi separator (TSEP-A) was
not used at all as a ‘best” production test meter, and the MPFM-A was used only once. The gas lift wells had the highest GVF, and
were therefore most suitable for the low pressure separator (TSEP-B). With these net oil “best” rates the reconciliation factor is 1.03

250

Multiphase tracer

MPFM-A MPFM-B TSEP-B

200
TSEP-B

MPFM-A

TSEP-A
150

Net oil,
m3/d
TSEP-B MPFM-B MPFM-B TSEP-B MPFM-B MPFM-B TSEP-B TSEP-B TSEP-B TSEP-B

100

50

0
X-7 X-9 X-10 X-12 X-14 X-17 X-25 X-28 X-29 X-30 Y-2 Y-3

Figure 14. – Tabulation of all the net oil rate measurements with the “best” one for each well test indicated with a grey arrow

You might also like