You are on page 1of 12

SPE 92053

Successful Field Surveillance Using Portable Multi-Phase Flow Meter in a High Gas-
Volume Fraction and High Water-Cut Application in East Kalimantan, Indonesia
S. Wangsa and R. Latief, Total Indonesia, and J. Kaura, D. Finley, and A. Ogilvie, Halliburton Energy Services, Inc.

Copyright 2005, Society of Petroleum Engineers


was designed within a predetermined set of constraints so as to
This paper was prepared for presentation at the SPE Europec/EAGE Annual Conference held in achieve the desired results.
Madrid, Spain, 13-16 June 2005.
The GKO vessel was installed immediately upstream of the
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
multiphase meter. After incorporating the GKO vessel with the
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to MPFM, the new “hybrid” system was used to test the same wells
correction by the author(s). The material, as presented, does not necessarily reflect any position
of the SPE, their officers, or members. Electronic reproduction, distribution, or storage of any part that had been tested earlier. The results using the MPFM and the
of this paper for commercial purposes without the written consent of the Society of Petroleum GKO vessel yielded liquid rates and gas rates that were within
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 ±10% of the test separator.
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.
Introduction
Abstract Accurate well testing is extremely important in all field
Total Indonesia had been finding it increasingly more difficult to development and reservoir management scenarios. The accuracy
obtain direct measurement of produced oil from some of their of the well testing systems is critical to providing the data for
remotely located fields. The production wells that needed testing decision analysis in ongoing drilling and work-over programs.
were in a mature field in a swamp location and had high water- However, well-test systems generally have been cumbersome
cuts (WC) and extremely high gas-volume fractions (GVF). The with large footprints, have numerous mechanical failure
WC and GVFs in these fields range from 80 to 100% and 90 to potentials, and very often, provide less than desirable
100%, respectively. Moreover, most of the wells have low capabilities for repeatability in flow measurements. A MPFM
potential and are very sensitive to back-pressure. All these that can generate accurate flow measurements can provide
characteristics as well as their logistics made accurate significant cost savings and improved well management and
measurement of production from the wells in this area extremely optimization1.
challenging. As a result, the time interval between the well tests MPFM technology has improved significantly in the last few
was ever increasing (more than one month). years and is becoming increasingly accurate and repeatable
In order to achieve more frequent performance data, and when compared to the separator.2 However, the MPFM accuracy
thus, enable more effective work-over decisions, the suggestion is very dependent upon the conditions to which the meter is
was made to evaluate use of a portable multi-phase flow meter subjected since the MPFM performance is affected by the
(MPFM) to resolve the measurement problems. increasing GVF in the flow stream. The MPFM operates within
The initial well tests yielded MPFM total liquid flow-rate acceptable accuracy range for well test purposes with a GVF of
readings that were ±30% of the test separator readings; the up to 85%. Beyond this GVF, the accuracy of the MPFM
MPFM gas rates were ±25% of the separator reading. The reduces and is subject to acceptance by the operating company.
operator’s requirement was for the MPFM to be within ±10% of The high GVF, high WC, and low-potential productivity
the test separator readings. Since these disparities were characteristics make accurate measurement of production from
attributed to the extremely high GVFs that had been encountered the wells in this area extremely challenging. This is due to the
on these wells, modifications to the present MPFM system were inherent nature of the fluid flow and the physics of the sensors
needed to measuring the multiphase flow. This producing field consists of
layered reservoirs stacked on top of each other. The reservoirs
• Reduce the GVF going through the meter
form a fluvial deltaic environment with the surface locations of
• Reduce the uncertainty of the meter measurements
the wells in a swampy area of the island. The reservoir pressure
• Improve the accuracy of the meter has declined as a natural result of production over an extended
• Increase the meter operating envelope. period of time. In order to sustain the production and keep the
Subsequently, a gas knock-out (GKO) vessel was designed field economically viable, various artificial-lift techniques were
and constructed to address the above needs. The GKO vessel deployed. In order to effectively manage the layered reservoirs
2 SPE 92053

and optimize the gas-lift injection, a continuous and consistent (γ), water (β) and gas (α) are calculated from the equations
well-performance monitoring technique that used the temporary above.
testing facilities was required. Due to the logistics of the swamp
well locations, it was difficult to obtain direct measurement of The Water-in-Liquid Ratio (WLR) is defined as:
produced oil from each well, and with the time interval between
well tests continuing to increase (more than one month between β
tests), the field performance could not be properly evaluated. WLR = . . . . . . . . . . . . . . . .(2)
If the use of a portable MPFM could achieve quicker turn- β +γ
around time and effective work-over decisions, the MPFM The set of equations for the calculation of flow rates are
measurements could be used to: listed below:
• improve production allocations (calibration of current
procedure) Oil Flow Rate = A ⋅ V ⋅ γ
• identify trends per zone (shallow, main, deep, etc.) to Water Flow Rate = A ⋅ V ⋅ β . . . . . . . . . . (3)
better define priorities for work-overs
• identify work required on surface facilities. Gas Flow Rate = A ⋅ V ⋅ α

Multi-Phase Meter Operating Principle The cross sectional area of the pipe (A) is known, based on
The major parts of this MPFM are the Venturi insert and the the geometry of the Venturi; the velocity (V) is measured by
electrodes incorporated within the throat of the Venturi insert. cross-correlating the electrical signals from the electrodes; and
The flow rates of oil, water, and gas are calculated based on the the fractions of oil, water and gas are found from the set of three
measurements obtained by the electrodes and the measurement equations for the fractions, as indicated above.
of the differential pressure across the Venturi inlet. No
separating devices, mixers, by-pass lines or radioactive sources Operating Envelope
are used in this MPFM concept. Figure 1 shows the cross- The operational range for this multiphase meter is shown in
sectional view of the meter. The principle of operation is Figure 3. To make the diagram independent of piping and meter
summarized in the block diagram shown in Figure 2. The dimensions, the term flow rate divided by the pipe cross section
differential pressure is measured across the inlet of the Venturi area (also known as the superficial velocity) on the axes has
insert. The capacitance or conductance of the mixture flowing been used. This velocity can be explained also as the velocity of
through the Venturi insert is measured by the electrodes inside one phase, based on its flow rate, if the other phases were not
the Venturi throat. The velocity is found from cross-correlating present. The maximum limit (red line) of the operating range is
the high resolution time signals from pairs of electrodes within determined solely by the maximum limit of the differential
the Venturi insert. pressure cell. In this case, it is 2500 bar. The minimum limit
Unlike most other MPFMs, the MPFM used here does not (blue line) is determined by the lower limit of the differential
measure the fluid-mixture density directly, as there is no gamma pressure cell, as shown in the bottom left corner of the diagram
densitometer. However, the density is determined indirectly (solid blue line).
through the momentum equation (i.e., the “Venturi equation”). However, two other important effects have impact on the
The set of equations used in the calculations of the fractions lower limit of the range:
are listed below: A minimum liquid flow rate is needed to create a continuous
flow through the meter. This means that a certain momentum is
Cap . /Cond . = f (α , β , γ , ε oil / σ water ) 1) required to prevent back flow through the main metering section
in which the velocity is measured by cross-correlation. In this
Density = f ( DP,Velocity) = f (α , β , γ , ρ o , ρ w , ρ g ) . . (1) MPFM, this limit is lower than most other multi-phase flow
meters because of the fact that the velocity is measured in the
α + β + ρ = 1.0 Venturi throat, and thereby, creates a higher velocity. The limit
1)
Capacitance and oil permittivity (εoil) apply for oil-continuous is set at a superficial liquid velocity of 0.6 m/s (blue horizontal
mixtures while conductance and water conductivity (σwater) apply for line) in the main pipe line. An upper limit for the GVF is set at
water-continuous mixtures. 97% (dotted and blue line).

Capacitance or conductance, DP, and velocity are the input Initial Well Test Set-up Description and Results
measurements and are being measured by the electrodes and DP The initial well tests were conducted on a testing barge that
transmitter. The permittivity of the oil (εoil), water conductivity consisted of the standard well-test equipment with a three-phase
(σwater) and densities of oil, water and gas (ρo, ρw, ρg) are input separator and an atmospheric tank. The well-test equipment
parameters entered into the user interface. The fractions of oil occupied the majority of the topside on this 20 meter barge. The
MPFM was installed upstream between the wellhead and the
SPE 92053 3

choke manifold. The first set of well tests from the MPFM was ensure that the other essential parameters such as the superficial
not very promising when compared to the results obtained from liquid velocity (VLS) through the meter and the DP across the
the well test equipment. The MPFM total liquid rate was within MPFM did not fall below the minimum required to obtain
±25% of the separator readings, and the gas rate was within reliable data. As a result, a control valve was placed on the gas
±30%. The following subsections address in detail why the test line in order to assist with the above objective of reducing the
results deviated from those derived from traditional well-testing GVF while ensuring that at the same time, the required
results. conditions were met. Based on the fluid flow rates from the
wells being monitored, it was virtually impossible to achieve
Liquid Rate Comparison complete separation of the flow through the GVF control
Figure 4 shows the liquid rate comparison between the MPFM system. The GVF control system would partially remove some
and the three-phase separator reading after the first set of trials. of the gas before the multiphase flow enters the MPFM. The
It can be seen that the liquid rate is underestimated by the partially separated gas was measured using a separate gas meter,
MPFM. This is primarily due to the relatively high GVF of the and then, combined downstream of the MPFM with the gas
fluid going through the meter. The GVF was so high that it going through the MPFM to allow the total flow to go to the
overshadows the liquid droplets in the flow, and as a result, the pipeline carrying the gas to the gathering station. Figure 7 is a
liquid rate estimated was lower than the liquid actually flowing representation of the gas measurement using the MPFM and the
through the MPFM. GKO vessel. This is solved mathematically with:

Gas Rate Comparison QgT = (Qggko + Qgmpfm) . . . . . . . . . (4)


On the other hand, the gas rate was over estimated. Figure 5
shows the gas-rate response of the MPFM in comparison with Where,
the separator. The deviation of the MPFM gas rates to the high QgT = Total gas-flow rate.
side can be attributed to the nature of flow that is happening with Qggko= Gas flow rate going through the gas knock out
the wells ― essentially, the fluid flowing with high gas-volume separator.
fractions. Qgmpfm= Gas flow rate going through the MPFM.
Had the MPFM and the three-phase separator reading been The amount of gas going through the GKO separator is a
the same, all the points plotted in Figures 4 and 5 would have function of
been on the unit slope straight line going through the origin of 1. the fluid properties
the plot. This was not observed in either of these figures. 2. the respective volumetric flow rates
3. percentage GCV opening.
Solution to the Challenge Qggko = f (Vmix, DP, dρ, percentage opening of the GCV)
Since the wells were producing at a GVF of 90 to 99% and a Where,
WC of 80 to 100%, they were essentially flowing in the water Vmix= mixture velocity through the MPFM.
continuous mode. As a result, the MPFM is measuring in the dρ = density difference between the liquid and gas phase
conductance mode, as opposed to the capacitance mode where In order to keep the project economical, a simple, rugged
the continuous phase would be oil. method to achieve partial gas separation was targeted. And,
When the MPFM is operating in the conductance mode, the since it was also necessary to keep the equipment compact, the
measurement is a function of the following; design engineers decided to base the separation technique on
1. Water cut. centrifugal separation.
2. Salinity of the water In the design, the centrifugal separation is induced by a
3. Gas volume fraction (GVF) going through the meter. cyclonic motion of the fluid owing to the mechanical design of
Since the water cut and water salinity were beyond operator the inlet to the GKO vessel4,5. The centrifugal force causes the
control, the only parameter that could be effectively controlled heavier fluids such as the water and oil to move to the wall of
was the GVF of the fluid going through the MPFM. It was the vessel and slide down. The lighter particles (i.e., the gas)
determined that any reduction achieved in the GVF would assist move up the center of the induced cyclone to the top of the
in achieving more reliable results from the MPFM and that this vessel and exit from the gas outlet. The liquid and some of the
reduction could only be achieved by having a mechanism strip entrained gas pass through the liquid outlet at the bottom of the
some of gas from the flow. While this would lower the GVF vessel. In this case, the gas exiting from the top of the gas
going through the MPFM, it was important that the overall knock-out separator is measured using a simple gas-turbine
measurement of the gas flow also be measured. meter.
Figure 6 is a detailed block-diagram representation of the
conceptual GVF control system. In order to shift the flow Modified Set up and Results
characteristics to the optimum part of the operating envelope, it After the GKO vessel was designed, it was incorporated into the
was important to reduce/modulate the GVF going through the existing setup and was installed on the testing barge upstream of
MPFM.3 However, at the same time, it was very important to the MPFM to achieve the desirable objectives of producing
4 SPE 92053

reliable results; i.e., reducing the uncertainty related to the liquid GVF of the fluid flowing through the MPFM, the wells were in
flow rates and improve the accuracy. Figure 8 shows the the optimum part of the operating envelope. The uncertainty of
production and ID of the section where the GKO vessel change liquid measurement was reduced. Hence, results obtained for all
was implemented. Figure 9 shows the picture of the actual the fluid rates through the system were within the tolerance level
system after assembly. Upon deploying the system back into the of ±10% as specified by the operating company.
field on the same wells, favorable results were observed. These
are described below. Conclusions
Upon deploying the system back into the field, the same The following results have been obtained from the modifications
wells were tested again in order to verify the accuracy of the applied to the MPFM
modified MPFM system. Again, the results were favorable. • A simple and rugged gas-knock-out vessel design was
successfully applied in a field application and achieved high
Liquid Rate Comparison efficiency with the MPFM in this welltest application.
It can be seen in Figure 10 that the majority of the data points • The liquid-rate measurement uncertainty in a high-gas-
for the liquid flow rate are within the allowable error margin of volume fraction (GVF) environment can be reduced to
±10%. The results are shown in the comparison of the flow-rate provide better results by the use of a GVF control
readings between the MPFM and the standard well-test mechanism. This can be seen by the reduction in deviation
separator. These improved results are owing to the reduced GVF of the liquid-rate measurement from ±30% to ±10%. For the
going through the MPFM, which provides optimum conditions purposes of definition, deviation here is the difference
for the meter. between the flowing rates measured from the MPFM
It can be seen that there are two data points in Figure 10 that against that measured from the test separator.
are outside the prescribed limit, and these are attributed to the • The reduction in GVF was between 6% and 12%. This was
following conditions: dependent upon the varying flowing conditions in the wells
1. The point in the upper half of the diagonal in the graph is included in this case history.
due to the initial fine tuning problems with the gas • The MPFM with the GKO vessel is cost effective and
knock-out separator. inherently safe when compared to the traditional ways of
2. The second point in the lower half of the diagonal in the field surveillance, which employ a test separator, calibration
graph is due to low liquid rates going through the meter, vessels, burner booms, and additional piping. This cost
and as a result of the low liquid rate, the MPFM is at a efficiency results from the reduction in equipment needed as
lower operating limit with regards to the DP and VLS. well as the approximately 60% reduction in manpower
This shows that when the operating limits of the meter required to manage the equipment. In addition to providing
are not maintained, it cannot achieve reliable liquid greater cost efficiency, reducing manpower exposure to
readings. hazardous operating conditions and areas enhances
personnel safety.
Gas Rate Comparison • This system has preferred application in areas with strict
The gas-rate comparison between the MPFM and the separator environmental codes relating to flaring of the hydrocarbons.
is shown Figure 11. It is observed that majority of the gas This is due to the inherent measurement technique, which
measurement points are within the required ±10% deviation does not require separation of the liquid and gas for
margin compared to the standard-condition flow rate derived metering. Traditional field surveillance equipment requires
from the separator calculations. the separated gas to be flared unless an expensive
compression system is available to compress the gas and
GVF Reduction inject it back in to the pipeline.
The tests performed with the GKO vessel resulted in the desired • As it stands today, MPFM technology is applicable to oil-
lower GVF going through the MPFM. The reduction in the GVF bearing fields with gas-volume fractions (GVF) of up to
was between 6 to 12 percent, depending upon which well was 85% at flowing wellhead conditions. However, beyond this
tested. Figure 12 shows the GVF before and after using the GVF, the uncertainty associated with the MPFM reading
GVF control system during a well test on a particular well. It can increases. Use of the gas knock-out vessel has reduced this
be seen that the reduction in GVF is achieved as soon as the uncertainty, and as a result, has shifted the operating
GKO vessel is operated to control the GVF. The change in the envelope towards the “Optimum” region of the MPFM
GVF was closely monitored on all the wells and recorded during operating envelope. This has resulted in increased operator
the entire test period. The gas line on the GKO separator was confidence in the system.
intermittently checked for any liquid carryover. No signs of • The footprint of the MPFM is very small, and with the
liquid carryover or traces of liquid were seen in the gas line. inclusion of gas knock-out vessel, there is no significant
Figure 13 is a bar-graph illustrating GVF comparison for increase in the systems footprint. In addition, the physics of
different wells on which the system was tested before and after measurement of this particular MPFM; i.e., no radioactive
the GVF control skid was used. Upon achieving the reduction in source for density determination, makes this system
SPE 92053 5

applicable to many global areas as it does not require References


compliance to local radioactive legislations of various 1. Falcone, G., Hewitt, G.F., Alimonti, C., Harrison, B., “Multi-Phase
countries. Flow Metering: Current Trends and Future Developments”,
Journal of Petroleum Technology,” June 2002, pp. 77-84.
2. Hasabe, B., etal., “Field Qualification of Four Multiphase
Future Recommendations Flowmeters on North Slope, Alaska”, SPE 90037 presented at the
Future recommendations include automating the existing system 2004 SPE ATCE held in Houston, TX September 26-29, 2004.
so that the GVF control is automatic, based on signals from the 3. Kaura, Jiten D., “Stretching the Operating envelope of the
computer. This would require taking the GVF signal coming Multiphase Meter Using the Gas Volume Fraction Modulation” M
from the meter combined with the DP across the venturi as well Sc Thesis at Heriot Watt University, 2004.
as the superficial liquid velocity through it. This could be 4. Kurokawa, J., “Gas-Liquid Flow Characteristics and Gas-Separation
accomplished by using a programmable logic controller where Efficiency in a Cyclone Separator,” ASME FED-Vol. 225, Gas
the incoming reading is compared to the set point, and the Liquid Flows, 1995, pp. 51-57.
automated valve adjusts its position to the point where the GVF 5. Chang F, And Dhir, V. K.: “Turbulent Flow Field in Tangentially
Injected Swirl Flows in Tubes”, Int. J. Heat and Fluid Flow, 1994,
control can be exerted by maintaining the other parameters; i.e., 15, pp. 346-356.
the DP and the superficial liquid velocity within the operating
envelope. However, the economic feasibility of this plan must be
SI Metric Conversion Factors
further evaluated based on such factors as the length of the ft x 3.048* E - 01 =m
project and the type of application under consideration. in x 2.54* E + 00 = cm
psi x 6.894 757 E + 00 = kPa
bar x 1.0* E + 05 = Pa

*Conversion factor is exact.

Figure 1 ― Cross-Sectional view of the MPFM


6 SPE 92053

WC Flow
Fraction Oil
Rate
Models Models

Density GVF

Water

Capacitance and
Conductance
Electrodes
X-Correlation (Velocity) Gas

Venturi for
Differential
Pressure

Flow Direction

Figure 2 ― Measurement Principle


SPE 92053 7

10,0

Max limit
Liquid Flow Rate / Pipe area [ m/s ]

8,0
Min limit
DP Min limit
VLS = 0,6 m/s
6,0
GVF = 97 %

4,0

Operational Range
TopFlow Multiphase
Multi-Phase Meter
Flow Meter
2,0

0,0
0 5 10 15 20 25
Actual Gas Flow Rate / Pipe area [ m/s ]
Figure 3 ― Operating Envelope of the MPFM

MPFM Liquid Rate Response to the Flow Conditions

OUTPUT
3000

FastQ
3" FastQ meter
Liquid Flow Rate
Balikpapan
2500
TFE
INPUT
2000
Flow With
Extremely
FastQ (bpd)

High GVF (90


– 99%) & Very 1500

High Water
Cut (80 –
100%) 1000

TFE 9-17 March 2003

500

0
0 500 1000 1500 2000 2500 3000

MPFM Separator (bpd)

Figure 4 ― Liquid rate comparison between the MPFM and Separator


8 SPE 92053

MPFM Gas Rate Response to the Flow Conditions

OUTPUT
2500

Gas Flow Rate


FastQ
3" TopFlow meter
INPUT 2000 Balikpapan
TFE
Flow With
Extremely
High GVF (90 FastQ (mscfd) 1500

– 99%) &
Very High
Water Cut
(80 to 100%) 1000

500

TFE 9-17 March 2003

0
0 500 1000 1500 2000 2500
Separator (mscfd)

MPFM

Figure 5 ― Gas rate comparison between the MPFM and Separator

Gas rate out


from gas meter
GVF set GVF Gas Control
point Controller Valve (GCV)
Gas rate out
from MPFM

Monitor GVF and VLS


through the MPFM

Figure 6 ― Conceptual GVF Control system block diagram


SPE 92053 9

Gas
meter

Partial Gas
scrubbing MPFM
system

Figure 7― Gas rate addition

2” 2” gas
wellstream
Production Gas
Wing Valve 6” dia X Turbine
Swab PI Meter
ES 10’ height MPFM
Valve
DV
Service
Wing 39 Master
39 Valves 2” liquid
Annulus 39
Valves
39 Knock Out Pot

FROM
RESERVOIR

OIL AND GAS IN PROCESS


OIL IN PROCESS
GAS IN PROCESS
WATER IN PROCESS

Figure 8 ― P&ID of the Modified System


10 SPE 92053

Gas Outlet

Gas KO
pot

Multiphase
MPFM
Fluid Inlet

Multiphase fluid outlet

Figure 9 ― The GKO pot and the MPFM layout.

System Liquid Rate Response to the Flow Conditions

OUTPUT

Liquid Rate Comparison between the FastQ & Separator with the Gas KO in use

3000

INPUT 2700

Flow with 2400

Extremely 2100
High GVF (90
– 99%) & 1800

Very High
Separator

Water Cut (80 1500

– 100%) 1200

900

600

300

0
0 300 600 900 1200 1500 1800 2100 2400 2700 3000
FastQ w ith the gas KO pot

Gas knock out


separator with
MPFM

Fig 10 ― The Liquid rate comparison between MPFM and Separator


SPE 92053 11

System Liquid Rate Response to the Flow Conditions

OUTPUT
Gas Rate
LiquidComparison Between
Rate Comparison between the MPFM
the FastQ & Separator and
with the Gas KO inSeparator
use
With the Gas KO in Use.
3000

INPUT 2700

Flow with
2400

Extremely 2100
High GVF (90
– 99%) & 1800

Very High

Separator
1500
Water Cut (80
– 100%) 1200

900

600

300

0
0 300 600 900 1200 1500 1800 2100 2400 2700 3000
FastQ w ith the gas KO pot

Gas knock out


separator with
MPFM

Figure 11―Gas rate comparison between MPFM and Separator

GVF through the MPFM

0.9

GVF through the MPFM

0.85
GVF through the meter
when the GKO pot is
clos ed.

0.8 Open the GKO for GVF


control.
GVF

0.75
GVF through the m eter
when the GKO pot is
opened.

0.7

0.65
11:45:00
11:53:00
12:01:00
12:09:00

12:33:00
12:41:00
12:49:00
12:57:00

13:13:00

13:29:00
13:37:00
13:45:00
13:53:00
14:30:00

14:46:00

15:02:00
15:10:00
15:18:00
15:26:00

15:50:00
15:58:00
16:06:00
16:14:00
16:22:00
16:30:00

16:54:00
11:37:00

12:17:00
12:25:00

13:05:00

13:21:00

14:38:00

14:54:00

15:34:00
15:42:00

16:38:00
16:46:00
hh.mm

Figure 12 ― GVF History through the MPFM for GVF control


12 SPE 92053

120
GVF before deploying GKO pot
GVF after deploying GKO pot

100

80
GVF

60

40

20

0
1 2 3 4 5 6 7 8 9
Wells

Figure 13 ― GVF comparison before and after use of the GKO vessel.

You might also like