SPE-198359-MS

Operational Experience with MultiPhase Flow Meters

Anton Skopich, Edward Neubauer, John Clarke, and Chingiz Bopiyev, Tengizchevroil

Copyright 2019, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Annual Caspian Technical Conference held in Baku, Azerbaijan, 16 – 18 October 2019.

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
TCO has been using multiphase flowmeters (MPFMs) since early 2000 and has gained significant
operational experience during this time. MPFM’s have replaced conventional separators because of higher
operating ranges and they provide instantaneous measurements for operating decisions.
MPFMs from several manufactures are used in the Tengiz and Korolev fields and have several
applications such as gas oil ratio monitoring on each well in the Sour Gas Injection Project pilot, at the
metering stations for well testing which is used to calibrate the well performance models and production
allocation, daily water production monitoring and well clean ups.
TCO has been working with vendors to improve flowmeter performance and reliability in a sour
environment and extreme weather conditions. Several modifications were implemented to improve
inline temperature measurements. To ensure MPFMs are operating properly and reliably requires a
good understanding of the principles the manufacturer uses to calculate flowrates, meter components,
operating envelopes, daily monitoring of key parameters, ability to trouble-shoot problems that do occur,
regular maintenance and dedicated Production Engineers to manage the process. Getting accurate surface
measurements is a challenge due to the wide range of well rates (1,600 to 25,000 BOPD), large seasonal
temperature variation (from -40 °C to +50 °C) and high H2S content.
This paper provides an overview of MPFM usage in Tengiz and describes lessons learnt, best practices
and challenges, from the perspective of a Production Engineer who is responsible for ensuring the MPFM’s
are providing accurate and reliable data.

Introduction
Multiphase flowmeters have been used in Tengiz field Kenter et al (2006) for more than 15 years. During
this time, the company has accumulated good working experience. To make multiphase flowmeters provide
accurate and reliable measurements, engineers need to ensure proper PVT input, calibration and a good
working maintenance strategy.
Currently, the Tengiz field has almost ideal conditions for multiphase flowmeter applications: high line
pressure, mild gas volume fraction (GVF) 70-75%, flowing bottom hole pressure (FBHP) above the bubble
point thus a known GOR ~450 m3/m3 and a very low water cut (<5%). Conventional test separators were
initially used throughout the Tengiz field. It was difficult to maintain the separators in the large seasonal
2 SPE-198359-MS

temperature variation experienced in Tengiz. Many wells were also producing above the design rate of the
separator, high line pressure in flowlines resulted in poor separation and inaccurate well tests and poor
allocation factors.
MPFM have tested wells higher than 24,000 BOPD and allow wells to be tested across a wide range rates
to develop accurate IPR curves for each well. Well tests (high and low) are used to develop flowing tubing
pressure (FTP) rates which are used for production allocation. The MPFM rates stabilize in less than one
hour, can be used to quickly test all the wells at the metering station (MS) to find a problem well.

Tengiz MFPM Strategy

Multiphase flowmeters from two vendors are used in Tengiz: Meter A and Meter B. Both types are installed
on wells in the Sour Gas Injection (SGI) area for continuous gas/oil ratio (GOR) monitoring as described by
Darmentaev et al (2010). These meters are sized for SGI producers that are producing in the range of 8,000
- 9,600 BOPD. Larger flowmeters are installed at the meter stations (MS) because the wells have higher
potentials. Locations with Meter A are shown as orange triangles on the map on Fig. 1, where large triangles
indicate MS’s and small – wellsite meters. Meter B are shown as blue circles with size respectively.

Figure 1—Field Map with MPFM locations

Both A and B type meters use total mass flow and volume fractions to determine the mass flow rate of
the oil, water, and gas. A venturi is used to calculate the total mass flow. The vendors use different methods
to calculate the phase fractions of oil, water, and gas. Meter A uses a gamma densitometer to measure the
mixture density to calculate the gas-liquid ratio, permittivity to determine the ratio of hydrocarbons to water,
and gas slip is determined by bubble velocity. Meter B uses dual energy gamma capture to determine oil,
SPE-198359-MS 3

gas and water mass fractions and correlations to estimate gas slip. Tengiz PVT tables are used to convert
measurements to standard conditions in both meters.

GOR monitoring in SGI Pilot Area

The Sour Gas Injection pilot was implemented at Tengiz Urazgaliyeva et al (2014) to maintain the pressure
and increase the recovery in the central area of the field. Part of the surveillance program described by
Tursinbayeva et al (2012) was continuous GOR monitoring to determine when gas breakthrough occurs.
MPFM’s are ideal for this purpose. The injected gas has a higher H2S concentration and this will impact the
MPM measurements if not corrected. The problem is the H2S moving from the oil phase to the gas phase as
pressure changes. A PVT study was completed in collaboration with the vendors and made the following
recommendations.
Meter A measurements are impacted by H2S, but to a lesser degree than Meter B. Meter A uses the
mixture density to determine the gas liquid ratio (GLR) (gas density is very different from oil and water)
and uses the permittivity/capacitance to determine the water cut (water conductance much higher than oil
and gas). It iterates back and forth until it converges on the three-phase fractions. The permittivity of gas is
1, the oil is ~2.2, and water is ~50-100 for Tengiz. However, the H2S causes the oil permittivity to increase.
In the past, the oil permittivity was increased by trial and error until the driest well at the meter station had a
water cut close to 0%. After understanding the H2S impact on the permittivity, the adjustment now appears
to be both reasonable and justified. The A meters are typically calibrated with fresh water, the permittivity
is set to ~2.55, and the meter works fine.
For type B meters - additional oil samples were sent to the lab to get refined PVT results at the typical
meter station line conditions (~100 bar and 60 °C) because it was difficult getting the meter to work with
Tengiz fluids. The problem was with the H2S moving from the oil phase to the gas phase as pressure changed.
Meter B uses a solution triangle. The three end points of the solution triangle are the 100% gas, oil and
water points. For most fields, the solution triangle is static, but as the H2S is heavier than most hydrocarbon
molecules, it has a large impact on the solution triangle. Meter B needed a dynamic solution triangle to
compensate for the H2S moving from the oil to the gas phase. The results of the complete study are provided
by Lindsell et al (2015).

Well testing at the Metering Station

Wells connected to a given metering station are tested one by one with 2-12 hour test durations. Based on
the well test results, a Prosper model is built for each well and updated constantly by production engineers.
These well models are used in Petroleum Experts Integrated Production Modeling (IPM) toolkit in Tengiz
Field for short and long-term production forecasting described by Onbergenov et al (2017).

Water production monitoring

The Tengiz field has a weak aquifer, with water also occuring in perched zones, and in tight zones with
poor reservoir quality. The Field has been producing oil with minimal water for an extended period ~ 17
years. The Plants were designed to handle a limited amount of water. Since 2008, isolated water production
has occurred in a few wells located in the highly fractured region. Accurate water measurements became
important to allow timely identification and shut-in of culprit water wells to avoid Plant process upsets. Type
A meters have a capacitive/conductive sensor. If oil dominates in the flow- the capacitive regime is activated,
if water dominates – the conductive regime is activated. Type B meter uses dual energy gamma attenuation
to determine oil, gas, and water mass phase fractions, where water measurement is highly affected by H2S.
To better understand the MPFM’s capacity to measure water production, they were tested by injecting
fresh water into the flowlines upstream of the meters. Type A meter driest well (26,400 BOPD) is ~2.2 km
away to allow proper mixing. Pump rates were varied from 0.3 bbl/min (1.2% WC) to 1.0 bbl/min (3.85%
WC). The Type A meter, shown on Fig. 2, detected the water, showed good repeatability of results. The
4 SPE-198359-MS

water cut was underestimated when compared to theoretically pumped water. The increase in oil rates during
water pumping periods suggested that some water was counted as oil.

Figure 2—Meter A during water pumping test

Type B meter: water was pumped at the driest well (18,600 bbl/d) which is ~1.4 km away from the MS to
allow proper fluid mixing. Pump rates of 0.25 bbl/min (1.9% WC) to 1.5 bbl/min (9.2% WC). Type B meter
response is shown on Fig. 3. Higher oil rates were observed suggesting that pumped water was counted as
oil. The water measurements were more variable and more challenging to identify a water content increase
in the flow.

Figure 3—Meter B during water pumping test

In conclusion, identifying low water cuts (<5%) is challenging in multiphase flow. Both meters identified
water, but the performance of Meter A was better in this trial. This is due to using a capacitance/conductance
type meter.

Workover well Flowbacks

New and workover wells were cleaned-up directly to the Plant Facilities in Tengiz. These wells can flowback
drilling mud, spent acid and stabilized Tengiz crude. It is important to understand the MPFM is set to
measure reservoir fluid and not other fluids pumped into the well. At the beginning of a flowback MPFM’s
measurements will not have high accuracy but can be used to understand relative changes.
SPE-198359-MS 5

MPFM Reliability Performance

Temperature measurements effect on GOR
Initially a temperature sensor was installed on the body of Meter A and took a long time to stabilize
when switching wells through the meter. In addition, the temperature sensor was affected by cold ambient
winter weather, and in some cases varied with ambient temperature during the day and night. Temperature
fluctuations resulted in unstable GOR calculations as it uses the PVT table to convert to standard conditions.
Several attempts were made in collaboration with the vendor to improve flow temperature measurements.
Clamp-on temperature sensors were used in an attempt to get representative temperature of the flow, but
still resulted in unstable GOR readings. Finally, extended intrusive temperature probes were used, and this
resolved the problem. GOR behavior prior and post extended temperature probe installation is shown on
Fig.4. The next generation models of the meter have an improved design with a multivariable transmitter
with intrusive temperature measurements.

Figure 4—Changing LC temperature impacts GOR estimation

H2S migration/intrusion behind electrodes/electronic elements

On Meter A, H2S migration occurred behind the capacitance electrodes and resulted in an electrode failure
and loss of the water cut measurement. However, the integrity of the meter was not compromised. The
inability to fix electrodes on site is a downside in the design of this model. In the next generation model this
issue was addressed and the electrodes configuration was modified. In addition, the design was improved
by making it more modular and allowing the meter to be repaired on-site without the need of taking the
meter to manufacturer’s machine shop.

Inadequate heat tracing

Some MPFM’s have a differential pressure (DP) transmitter with impulse lines tubes that need to be heat
traced. It turned out that the initial heat tracing design was not enough, the impulse lines tubes cooled down
during some interval of the winter. Even though the impulse line tubes did not freeze, the temperature would
be low enough for wax to appear. The impulse line tubes would plug with wax and the DP transmitter would
stop working, shown on Fig. 5. Heat tracing was changed to a higher rating and the impulse line blockage
was eliminated.
6 SPE-198359-MS

Figure 5—Changing LC temperature impacts GOR estimation

Furthermore, issues were found, when the ambient temperature increased, and the heat tracing was left
on. The meter’s computer received temperature higher than the temperature of the flow and resulted in a
GOR estimation error.
Other mechanical issues were experienced with MPFM’s like gamma detectors failures, leaking isolation
valves, computer and power units malfunctioning and burning out. All these issues during the operation
could be addressed timely with parts in stock.

MPFM Maintainability
To keep meters working and providing reliable measurements a regular maintenance strategy has to be
in place. All meters have annual calibration and inspections. All possible upgrades and updates are being
installed and applied. A meter’s gamma system is also calibrated annually. Meter A is calibrated with air
and water gamma calibrations. Oil permittivity and water conductivity has been measured through life
of the meters. Meter B typically has an empty pipe calibration. Pressure and temperature transmitters are
calibrated in a metrology lab every three years, or in case malfunctioning is suspected. Differential pressure
transmitters can be checked when there is no flow through the meter and calibrated on a three years basis.

MPFM Operating Guidelines

It is important to understand the quality control parameters to understand if the meter is not performing or it
is a well behavior. When the MPFM reports questionable results, the first step is to compare measurements
with known values. Most of the wells in Tengiz produce at a stable GOR~ 450 m3/sm3with the reservoir
being above bubble point and water cut close to 0-1%. The oil rate in standard conditions should be ~ 60%
of the total mass flow. Since the MPFM are designed for certain operating ranges, the raw measurements
should be within the expected limits. If the venturi DP is less than ~50 mbar the flowrate is too small to
be accurate. For Meter A the mix velocity range is 2.5 - 35 m/s, above these limit accuracy decreases. The
density range should also be understood as the presence of an unusual density reading could be an indicator
of a meter problem. Mixture density should be within 220-300 kg/m3 depending on line conditions. Mixture
permittivity should be in the range 1.2 -1.5. It is recommended to test other known wells before starting
detailed trouble shooting. DP monitoring – in cases when the oil rate drops by 30 or 50% first check valves
on the test manifold and DP reading on the well choke, then if possible remove the well from test and check
the DP reading to ensure the transmitter is working properly.
SPE-198359-MS 7

Conclusions
Overall MPFM’s work well in the Tengiz environment. Continuous work on performance improvement has
made it a very important piece of equipment in daily operations. Having a dedicated Production Engineer
working in collaboration with vendor has resulted in Tengiz meters being online for close to 100% of the
time:

• MPFM performance tuning continues in the operational phase beyond commissioning

• Work with the vendor’s team to improve reliability, incorporate new and improved MPFM
technology
• Service Contracts must be in-place prior to meter commissioning

• Development of a in-house service group and continuous input from dedicated Field Production
engineer
• Understanding of the meter’s key troubleshooting parameters is critical to improve MPFM
performance on-site
• MPFMs can deliver accurate and reliable measurements over a wide range of operating conditions

• Understand strengths and weakness of a MPFM type in a given production environment takes time

• It is very difficult to measure to accurately measure low WC’s (<2%)

Acknowledgements
The authors would like to thank the many TCO colleagues from Reservoir Management, Facilities
Engineering and Chevron ETC who supported TCO over the years. Authors would like to extend their
gratitude to TCO partners KazMunaiGas, ExxonMobil Kazakhstan Ventures Inc., Lukoil and Chevron
Corporation for permission to publish this work.

References
Darmentaev, S., Yessaliyeva, A., Azhigaliyeva, A., Belanger, D., Sullivan, M., King, G. R., Feyijimi, T., and Bateman,
P.: "Tengiz Sour Gas Injection Project," paper SPE 139851 presented at the SPE Caspian Carbonates Technology
Conference, Atyrau, Kazakhstan, 8–10 November 2010.
Urazgaliyeva, G., King, G. R., Darmentaev, S., Tursinbayeva, D., Dunger, D., Howery, R., Zalan, T., Lindsell, K., Iskakov,
E., Turymova, A., Jenkins, S., Walker, C., Bateman, P., and Aitzhanov, A.: "Tengiz Sour Gas Injection Project: An
Update," paper SPE 172284 presented at the SPE Annual Caspian Technical Conference and Exhibition in Astana,
Kazakhstan, 12–14 November 2014.
Kenter, J. A. M., Harris P. M., Collins J. F., Weber L. J., Kuanysheva G., and Fischer D. J., 2006, Late Visean to Bashkirian
platform cyclicity in the central Tengiz buildup, Pricaspian Basin, Kazakhstan: Depositional evolution and reservoir
development, in Harris P. M. and Weber L. J., eds., Giant hydrocarbon reservoirs of the world: From rocks to reservoir
characterization and modeling: AAPG Memoir 88, p. 1–48.
Lindsell K., Wang S.W., Clarke J., and Jambayev A., Tengizchevroil; Ricquebourg J., Kaipov Y., Zakharov L., and
Hollaender F., Schlumberger SPE-177422 "In-line measurement of H2S content in flow with multiphase flow meter
– case study in a giant field under sour gas re-injection" paper SPE 177422 presented at Abu Dhabi International
Petroleum Conference and Exhibition, Abu Dhabi, UAE, 9-12 November 2015
Tursinbayeva D., Lindsell K., Zalan T., Dunger D., Kassenov B., and Howery R.: "Tengiz Surveillance: Planning Strategy
and Implementation", paper SPE 160957 presented at Abu Dhabi International Petroleum Conference and Exhibition,
Abu Dhabi, UAE, 11-14 November 2012
Onbergenov U., Aitkazin M., Tang Y., Svyatov A., Kairbekov A.: "Best Practices in Calibration of Integrated Production
Modeling for Tengiz Field", paper SPE 188994 presented at the SPE Annual Caspian Technical Conference and
Exhibition held in Baku, Azerbaijan, 1 – 3 November 2017.