Handbook of Multi Phase Metering

Handbook of multiphase metering
Produced for
The Norwegian Society for Oil and Gas Measurement
by:
John Amdal, Saga Petroleum

Harald Danielsen, Statoil
Eivind Dykesteen, Fluenta
Dag Flølo, Sandsli Drift
Jens Grendstad, Kongsberg Offshore
Hans Olav Hide, MultiFluid International
Håkon Moestue, Norsk Hydro
Bernt Helge Torkildsen, Framo Engineering
TABLE OF CONTENTS
1. INTRODUCTION 4
2. SCOPE 5
3. DEFINITIONS 6
3.1 Terms related to multiphase flow 6
3.2 Terms related to metrology 9
4. APPLICATIONS OF MULTIPHASE METERING 14
4.1 Well testing 14
4.2 Production metering 18
4.3 Comparison of multiphase meters with conventional methods 20
5. MULTIPHASE FLOW 22
5.1 Flow regimes in vertical flow 24
5.2 Flow regimes in horizontal flow. 25
6. CLASSIFICATION OF MULTIPHASE METERS 27
6.1 Separation type meters. 27
6.2 In-line meters 29
6.3 Other categories of multiphase meters 29
7. IMPLEMENTATION OF MULTIPHASE METERS 31
7.1 Selecting the point of installation 31
2
Handbook of multiphase metering Page 3
7.2 Prediction of flow regime at the design stage 31
7.3 Use of Mixers 32
7.4 Converting flow rates measured at flowing conditions to flow rates at standard conditions 33
7.5 A format for initial evaluation of multiphase meter implementation 33
8 PERFORMANCE SPECIFICATION 37
8.1 Uncertainty description 37
8.2 Establishing the reproducibility of a given meter 41
8.3 Rated conditions of use 41
8.4 Influence quantities and how they influence accuracy 42
8.5 A format for the presentation of a multiphase meter performance specification 42
9. TESTING OF MULTIPHASE METERS 46
9.1 Test facilities 46
9.2 Types of tests 51
9.3 A format for presentation of summary test results. 53
10 QUALIFICATION OF MULTIPHASE METERS 57
11. CALIBRATION OF MULTIPHASE METERS 59
11.1 Factory Calibration 59
11.2 Independent Laboratory Calibration 60
11.3 Field Calibration 60
11.4 In Situ Calibration 60
11.5 Implementation of calibration results 62
3
1. INTRODUCTION
The need for multiphase flow measurement in the oil and gas production industry has
been evident for many years. A number of such meters have been developed during
the last few years by research organisations, meter manufacturers, oil & gas
production companies and others.
These developments employ different technologies, and the prototypes have been
quite dissimilar in design and function. Some lines of development have been
abandoned.
Only during the past year or two have meters been developed and tested to the stage
at which multiphase flow measurement is a realistic option in an industrial
environment. The number of uses and users is now expected to increase.
Multiphase flow measurement has yet to be established as a separate discipline.

Meters from different manufacturers will always differ in their design, function and
capabilities. In order to promote mutual understanding of multiphase flow meters and
their use among users, manufacturers and others, some form of guidelines or user
manual would seem appropriate. This document has been written to serve that
purpose and to help provide a common basis for the field of in-line multiphase
measurement systems.
It is not our intention that this document should be regarded as a final report.
Rather,we hope that it will initiate more international work in which the issues and
topics raised here can be further developed.
4
2. SCOPE
This document is intended to serve as a guide for users and manufacturers of

multiphase flow meters. Its purpose is to provide a common basis for, and assistance
in, the classification of applications and meters, as well as guidance and
recommendations for the use of such meters.
The document may also serve as an introduction to newcomers in the field of

multiphase flow measurement, with definition of terms and description of multiphase
flow in closed conduits being included.
The primary focus is on in-line meters for direct measurement of true multiphase flow
of oil, gas and water. Even if the individual flow rates of each constituent are of
primary interest, fractions of oil, gas and water are sometimes useful as operational
parameters.
Other meters, e.g. separation meters and model/calculation type "meters", do not fall
within the scope of this document, and are only briefly discussed. Other constituents
than oil, gas and water are not dealt with.
The performance of a multiphase meter in terms of accuracy, repeatability, range, etc.

is of great importance, as is the user’s ability to compare different meters in these
respects. One section covers this issue, and proposes standard ways of how
performance can be described.
Related to performance are the testing and qualification of the meters, which are also
covered. Guidance is provided to help optimise the outcome of such activities.
Since meters are in-line, flow rates are measured at process operating conditions.
Conversion of flow rates to standard conditions, which involves multiphase sampling,
knowledge of composition and mass transfer between phases at fluctuating pressures
and temperatures, is only briefly dealt with here.
5
3. DEFINITIONS
The terms defined below are split into two categories. One section defines terms that
are commonly used to characterise multiphase flow. Another section defines
metrological terms that may be useful in characterising the performance of a
multiphase meter.
3.1 Terms related to multiphase flow
Emulsion:
Colloidal mixture of two immiscible fluids, one being dispersed in the other in the
form of fine droplets.
Flow regime:
The physical geometry exhibited by a multiphase flow in a conduit; for example,
liquid occupying the bottom of the conduit with the gas phase flowing above, or a
liquid phase with bubbles of gas.
Fluid:
A substance readily assuming the shape of the container in which it is placed; e.g. oil,
gas, water or mixtures of these.
Gas:
Hydrocarbons in the gaseous state at the prevailing temperature and pressure.
Gas-liquid-ratio (GLR):
The gas volume flow rate, relative to the total liquid volume flow rate (oil and water),
all volumes converted to volumes at standard pressure and temperature.
Gas-oil-ratio (GOR):
The gas volume flow rate, relative to the oil volume flow rate, both converted to
volumes at standard pressure and temperature.
Gas volume fraction (GVF):

The gas volume flow rate, relative to the multiphase volume flow rate, at the pressure
and temperature prevailing in that section. The GVF is normally expressed as a
percentage.
6
Hold-up:
The cross-sectional area locally occupied by one of the phases of a multiphase flow,
relative to the cross-sectional area of the conduit at the same local position.
Homogeneous multiphase flow:

A multiphase flow in which all phases are evenly distributed over the cross-section of
a closed conduit; i.e. the composition is the same at all points.
Mass flow rate:

The mass of fluid flowing through the cross-section of a conduit in unit time.
Multiphase flow:
Two or more phases flowing simultaneously in a conduit; this document deals in
particular with multiphase flows of oil, gas and water.
Multiphase flow rate:

The total amount of the two or three phases of a multiphase flow flowing through the
cross-section of a conduit in unit time. The multiphase flow rate should be specified
as multiphase volume flow rate or multiphase mass flow rate.
Multiphase flow velocity:

The flow velocity of a multiphase flow. It may also be defined by the relationship
(Multiphase volume flow rate / Pipe cross-section).
Multiphase flow rate meter:

A device for measuring the flow rate of a multiphase flow through a cross-section of
a conduit. It is necessary to specify whether the multiphase flow rate meter measures
the multiphase volume or mass flow rate.
Multiphase fraction meter:

A device for measuring the phase area fractions of oil, gas and water of a multiphase
flow through a cross-section of a conduit.
Multiphase meter:
A device for measuring the phase area fractions and flow rates of oil, gas and water
of a multiphase flow through a cross-section of a conduit. It is necessary to specify
whether the multiphase meter measures volume or mass flow rates.
Oil:
7
Hydrocarbons in the liquid state at the prevailing temperature and pressure

conditions.
Oil-continuous multiphase flow:

A multiphase flow of oil/gas/water characterised in that the water is distributed as
water droplets surrounded by oil. Electrically, the mixture acts as an insulator.
Phase:
In this document, “phase” is used in the sense of one constituent in a mixture of
several. In particular, the term refers to either oil, gas or water in a mixture of any
number of the three.
Phase area fraction:

The cross-sectional area locally occupied by one of the phases of a multiphase flow,
Phase flow rate:

The amount of one phase of a multiphase flow flowing through the cross-section of a
conduit in unit time. The phase flow rate may be specified as phase volume flow rate
or as phase mass flow rate.
Phase mass fraction:

The phase mass flow rate of one of the phases of a multiphase flow, relative to the
multiphase mass flow rate.
Phase velocity:
The velocity of one phase of a multiphase flow at a cross-section of a conduit. It may
also be defined by the relationship (Superficial phase velocity * Phase area fraction).
Phase volume fraction:

The phase volume flow rate of one of the phases of a multiphase flow, relative to the
multiphase volume flow rate.
Slip:
Term used to describe the flow conditions that exist when the phases have different
velocities at a cross-section of a conduit. The slip may be quantitatively expressed by
the phase velocity difference between the phases.
Slip ratio:
8
The ratio between two phase velocities.
Slip velocity:
The phase velocity difference between two phases.
Superficial phase velocity:

The flow velocity of one phase of a multiphase flow, assuming that the phase
occupies the whole conduit by itself. It may also be defined by the relationship (Phase
volume flow rate / Pipe cross-section).
Velocity profile:
The mean velocity distribution of a fluid at a cross-section of a conduit. The velocity
profile may be visualised by means of a two- or three-dimensional graph.
Void fraction:
The cross-sectional area locally occupied by the gas phase of a multiphase flow,
Volume flow rate:

The volume of fluid flowing through the cross-section of a conduit in unit time at the
pressure and temperature prevailing in that section.
Water-continuous multiphase flow:

A multiphase flow of oil/gas/water characterised in that the oil is distributed as oil
droplets surrounded by water. Electrically, the mixture acts as a conductor.
Water cut (WC):

The water volume flow rate, relative to the total liquid volume flow rate (oil and
water), both converted to volumes at standard pressure and temperature. The WC is
normally expressed as a percentage.
Water-in-liquid ratio (WLR):

The water volume flow rate, relative to the total liquid volume flow rate (oil and
water), at the pressure and temperature prevailing in that section.
3.2 Terms related to metrology
9
The accuracy of multiphase meters should be specified by terms which are in

conformance with "The international vocabulary of basic and general terms in
metrology" issued by ISO.
Other standards based on the above document may also be used, e.g. BS 5233
(1986): "Glossary of terms used in metrology."
Some of the definitions of BS 5233 which may be particularly relevant to multiphase

flow measurement are quoted below (or form part of the definitions).
Accuracy:
The quality which characterises the ability of a measuring instrument to give
indications equivalent to the true value of the quantity measured.
NOTE: The quantitative expression of this concept should be in terms of uncertainty.
Conditions of use:
The conditions which must be fulfilled in order to use a measuring instrument
correctly, taking account of its design, construction and purpose.
NOTE: The conditions of use can refer, among other things, to the type and condition
of the subject of the measurement, the value of the quantity measured, the values of
the influence quantities, the conditions under which the indications are observed, etc.
Error of measurement:
The discrepancy between the result of the measurement and the true value of the
quantity measured.
NOTE 1: In general, "true value" may be replaced by "conventional true value".
NOTE 2: The discrepancy can be expressed as either: the algebraic difference

between these two values, i.e. (error of measurement) = (result of measurement) -
(true value), or as the quotient of that difference and the value of the quantity
measured.
These two forms of expression are often identified as "absolute error" and "relative
error" respectively.
Effective range, measuring range, working range:
10
The range of values of the measured quantity for which any single measurement,
obtained under specified conditions of use of a measuring instrument, should not be in
error by more than a specified amount.
NOTE: The effective range may be the whole or a specified part of the interval
between the lower and upper range limits.
Influence quantity:
A quantity which is not the subject of the measurement but which influences the value
of the quantity to be measured, or the indications of the measuring instrument, or the
value of the material measure reproducing the quantity.
Limiting conditions of use:

The ensemble of the ranges of values of influence quantities and of the measured
quantity which an instrument can withstand without consequential damage or
degradation of performance when it is subsequently operated under rated conditions
of use.
Lower range limit:

The minimum value of the quantity to be measured, for which the instrument has been
constructed, adjusted or set.
Range:
The interval between the lower and upper range-limits.
Rated conditions of use:

The ensemble of the ranges of influence quantities and of the measured quantity
within which the performance of the instrument is specified.
Reference conditions:
The ensemble of reference values or reference ranges of different influence quantities
affecting a measuring instrument.
NOTE: It is customary to select the reference conditions to correspond to a

conveniently established environment for testing that falls within the range of
conditions expected in the use of the instrument.
Reference range:
A particular range of values of an influence quantity stated in the specification of a
measuring instrument as a basis for determining its intrinsic error.
11
Reference value:
A particular value of an influence quantity stated in the specification of a measuring
instrument as a basis for determining its intrinsic error.
Repeatability:
A quantitative expression of the closeness of the agreement between the results of
successive measurements of the same value of the same quantity carried out by the
same method, by the same observer, with the same measuring instruments, at the
same location at appropriately short intervals of time.
Reproducibility:
A quantitative expression of the closeness of the agreement between the results of
measurements of the same value of the same quantity, where the individual
measurements are made under different defined conditions, e.g.:
by different methods, with different measuring instruments, by different observers, at

different locations, after intervals of time that are appropriately long compared with
the duration of a single measurement, or under different customary conditions of use
of the instruments employed.
Span:
The algebraic difference between the upper and lower values specified as limiting the
range of operation of a measuring instrument.
Example: A thermometer intended to measure over the range -40 0C + 60 0C has a

span of 100 0C.
Supplementary error:
The error of a measuring instrument arising from the fact that the values of the
influence quantities differ from those corresponding to the reference conditions.
Total error, overall error:

The whole error of a measuring instrument under specified conditions of use.
Uncertainty of measurement:
That part of the expression of the result of a measurement which states the range of
values within which the true value or, if appropriate, the conventional true value is
estimated to lie.
12
NOTE: In cases in which there is adequate information based on a statistical

distribution, the estimate may be associated with a specified probability. In other
cases, an alternative form of numerical expression of the degree of confidence to be
attached to the estimate may be given.
Upper range limit:

The maximum value of the quantity to be measured, for which the instrument has
been constructed, adjusted or set.
13
4. APPLICATIONS OF MULTIPHASE METERING
Conventional single-phase meters require the constituents or "phases" of the well

streams to be separated and stabilized upstream of their points of measurement.
For production metering this requirement is met automatically at the outlet of a

conventional process plant. As the main purpose of such a plant is to receive the sum
of well streams in one end and to deliver transportable (i.e. single, stable phases) in
the other end, the stream has been made measurable at the same time as it has been
made transportable.
The need for multiphase metering arises when it is necessary to meter the well
stream(s) before they are fully processed as described above. Some areas in which
multiphase metering may be employed, are described below.
As this document is intended to be a guide for users or potential users of multiphase

meters in the oil and gas production industry, application areas in other industries are
not dealt with here.
4.1 Well testing
Reservoir engineers need to monitor the performance of each single well constantly
in order to optimise production and the lifetime of the field.
The criticality of the uncertainty of the figures obtained from well tests depend on a
number of factors that differ from field to field. For most large fields in the North Sea,
decisions with a very high financial impact are based on well-test results, e.g. shutting
down wells, drilling new wells, reducing production rate from the reservoir, etc.
4.1.1 Conventional well testing.
The conventional means of doing this include using a test separator into which the
production from each well is separetly led. See Figure 4.1.
The stream from the well being tested is separated into two (or three) "phases": high
vapour-pressure oil and gas (and water).
14
Test Gas SM
manifold
Test
separator
Oil SM
Water SM
Gas
1st-stage
separator
Oil
Water
Production
manifold
From wells
Figure 4.1 1st-stage production separator, and test separator (SM = single phase metering
station).
The stream of each of the above phases is metered and sampled at the same time as
certain parameters such as choke opening, wellhead flow pressure, separator pressure
and separator temperature are recorded. (Other parameters may also be recorded, and
samples may not be taken at each well test. On production platforms samples may
only be taken at intervals of some years).
Each well may be tested at one or more settings of the well's choke. For each setting,
recordings of the above measurements are made.
After the test, the analyses of the samples taken and the data recorded at the test are
mathematically combined in order to estimate the well's contribution to the output
streams of the production plant, e.g. sales gas, stabilised oil and water.
These flow rates are the main data from a well test and are determined for one or
more choke settings. The flow rates are plotted against choke settings, wellhead
flowing pressure or some other parameter and are used until the next well test, to
calculate the theoretical contribution made by each well to the output streams of the
process plant.
The metering uncertainty of conventional single-phase meters on a test separator

varies from field to field and in most cases is very difficult to estimate.
15
In the first place this is due to the fact that in most cases, single-phase meters,
normally orifice plates for gas and turbine meters for hydrocarbon liquid, operate
slightly into the two-phase area where their uncertainty cannot be predicted.
Secondly, the oil industry tends to be very relaxed in the calibration and inspection of
these meters.
The test separator measurement can therefore under normal operating conditions not
be expected to give an uncertainty of better than +5 % to +10 % of reading of each
phase volume flow rate. When special precautions are taken, e.g. by use of heaters
and/or pumps to avoid two-phase flow in the metering legs, an uncertainty of ± 2-3
% of reading of each phase volume flow rate may be possible to meet.
The only readily available field experience that can offer an indication of the
uncertainty of test separator meters is the ratio between the actual production of a
process plant and the sum of the theoretically calculated flow rates from all its wells.
For most fields this ratio will deviate by between 2% and 10 % from unity. It should
be noted, however, that these percentages are not the same as the uncertainty of the
test separator's meters. There is also a contribution of unknown magnitude from the
mathematical combination described above, from the method of setting up well
performance curves and from records of choke settings, well-down time, etc., during
regular production.
The above statements on uncertainty are primarily intended to illustrate that it is

difficult to suggest general uncertainty requirements for a multiphase meter if this is
to replace a conventional test separator.
The only way to determine the minimum requirements is via dialogue between
metering specialists and petroleum engineers in each particular case.
4.1.2 Welltesting by multiphase meters.
The multiphase meter may be installed in the same way as the test separator, see
Figures 4.2 and 4.3.
16
Test
manifold
Multiphase
meter
Gas
1st-stage
separator
Oil
Water
Production
manifold
From wells
Figure 4.2 Multiphase metering replacing test separator and its meters
Test Gas SM
manifold 1
Test
separator
Test
manifold 2 Oil SM
Water SM
Multiphase
meter
Gas
1st-stage
separator
Oil
Water
Production
manifold
From wells
Figure 4.3 Multiphase metering can be used to increase overall testing capacity (SM = single
phase metering station)
4.1.3 Continuous monitoring of well performance
Well testing as described in the previous chapter is, in essence, a regular sampling of
well performance data. A change in the performance of one or a few wells may not be
detected during the periods between the tests.
17
Also, as mentioned in the previous chapter, the estimated performance of each

production well becomes less accurate than the metering results of the well test
because of such factors as inaccurate well-performance plots, inaccurate recording of
choke settings, etc..
As multiphase meters become available, it is possible to install them on the flow line
of individual production wells (Figure 4.4). This offers the benefit of a continuous
monitoring of each well, as well as eliminating some of the uncertainties in calculating
the theoretical performance of the well.
Gas
1st-stage
separator
Oil
Water
MM MM MM Production
manifold
From wells
Figure 4.4 Multiphase meters on each well's flow line replacing test separator and its meters
(MM = multiphase meter)
4.2 Production metering
When a single process plant or multiphase pipeline is used to process (or transport)
the streams from more than one license area, it is necessary to meter the production
from each license area separately, before it enters the common process plant or
pipeline.
The metering of the production from each license area is used in a procedure to
allocate each field owner's ownership in the streams at the outlet of the common
process plant. Consequently this production metering is governed by national
regulations for fiscal metering.
18
4.2.1 Conventional metering systems
In a conventional production scheme production is metered on single-phase streams

at the outlet of the process plant.
Production from the individual license areas may be calculated as the sum of
theoretical well production rates for the wells of that particular license area. This
requires each well to be routed to a test separator. In such a case additional valve and
manifold arrangements have to be installed on the sea-floor and a dedicated test line
has to be installed (Figure 4.5).
Satellite field Production platform
SM Gas
Test
separator
SM
1st-stage
SM separator
SM
Inlet
separator
SM
2nd-stage
SM separator
Test Oil
manifold Water
Production
manifold
Test
line
Field “B” Field “A”
Figure 4.5 Satellite field “B” with multiphase production line and test line (SM = single phase
metering station)
4.2.2 Production metering by multiphase meters
Alternatively, well testing and production metering from the satellite field “B” may be
done by means of multiphase meters. This removes the need for a separate test line
and manifold system. Assuming that a dedicated inlet separator would still be needed
on the production platform, a typical multiphase production metering concept could
be as shown in Figure 4.6.
19
Satellite field Production platform
1st-stage
separator
Inlet
MM separator
2nd-stage
separator
MM MM MM
Field “B” Field “A”
Figure 4.6 Satellite field "B" with multiphase meters for well testing and production metering
(MM = Multiphase meter)
4.3 Comparison of multiphase meters with conventional methods
A conventional single-phase measurement system for well testing and production has
the following features:
+ the test separator makes sampling of the process fluids readily available
+ the availability of a test separator may be necessary for well cleaning purposes
+ traceable single-phase measurements, according to well-accepted procedures, can
be performed
− it does not lend itself to continuous monitoring or metering
− long transport lines from a satellite require a long period for separator conditions
to stabilise
− test separator, test lines, manifolds and valve systems have high installation and
operational costs
− a large volume is needed on the production platform
A multiphase metering system for well testing and production metering has the
following features:
+ continuous monitoring or metering is possible

20
+ installation and operating costs are low compared to those of a conventional

system
+ test separator, test lines, manifolds and valve systems are eliminated
+ given the possibility of continuous metering, the overall uncertainty may be lower
than in a conventional system
− sampling of the process fluids is not readily available
− an inlet separator may be required as a slug catcher and for well-cleaning purposes
− multiphase metering is not traceable to any accepted standards
21
5. MULTIPHASE FLOW
Multiphase flow is a complex phenomena which is difficult to understand, predict and

model. Common single-phase characteristics such as velocity profile, turbulence and
boundary layer, are thus inappropriate for describing the nature of such flows.
The flow structures are classified in flow regimes, whose precise characteristics
depend on a number of parameters. The distribution of the fluid phases in space and
time differs for the various flow regimes, and are usually not under the control of the
designer or operator.
Flow regimes vary depending on operating conditions, fluid properties, flow rates and
the orientation and geometry of the pipe through which the fluids flow. The
transition between different flow regimes may be a gradual process. The
determination of flow regimes in pipes in operation is not easy. Analysis of
fluctuations of local pressure and/or density by means of Gamma-ray densitometry
has been used in experiments and is described in the literature. In the laboratory, the
flow regime may be studied by direct visual observation using a length of transparent
piping. Description of flow regimes are therefore to some degree arbitrary, and they
depend to a large extent on the observer and his interpretation.
The main mechanisms involved in forming the different flow regimes are transient
effects, geometry/terrain effects, hydrodynamic effects and combinations of these
effects. Transients occur as a result of changes in system boundary conditions. This is
not to be confused with the local unsteadiness associated with intermittent flow.
Opening and closing of valves are examples of operations that cause transient
conditions. Geometry and terrain effects occur as a result of changes in pipe-line
geometry or inclination. Such effects can be particularly important in and downstream
of sea-lines, and some flow regimes generated in this way can prevail for several
kilometres. Severe riser slugging is an example of this effect. In the absence of
transient and geometry/terrain effects, the steady state flow regime is entirely
determined by flow rates, fluid properties, pipe diameter and inclination. Such flow
regimes are seen in purely straight pipes and are referred to as “hydrodynamic” flow
regimes. These are typical flow regimes encountered at a wellhead location.
All flow regimes however, can be grouped into dispersed flow, separated flow,
intermittent flow or a combination of these (Figure 5.1). Dispersed flow (LB = 0) is
characterised by a uniform phase distribution in both the radial and axial directions.
Examples of such flows are bubble flow and mist flow (Figure 5.2). Separated flow
22
(Ls = 0) is characterised by a non-continuous phase distribution in the radial direction

and a continuous phase distribution in the axial direction. Examples of such flows are
stratified and annular (with low droplet entrained fraction), see Figure 5.3.
Intermittent flow is characterised by being non-continuous in the axial direction, and
therefore exhibits locally unsteady behaviour. Examples of such flows are elongated
bubble, churn and slug flow (Figure 5.4). The flow regimes shown in Figures 5.2 - 5.4
are all hydrodynamic two-phase gas-liquid flow regimes.
Flow regimes effects caused by liquid-liquid interactions are normally significantly

less pronounced than those caused by liquid-gas interactions. In this context, the
liquid-liquid portion of the flow can therefore often be considered as a dispersed flow.
However, some properties of the liquid-liquid mixture depend on the volumetric ratio
of the two liquid components.
I n t e r m it t e n t
f lo w
LB LS
S e p a ra te d f lo w D is p e rs e d f lo w
Figure 5.1 Multiphase flow regime
Bubble
Mist
Bubble Mist
Figure 5.2 Dispersed flow
23
S tratifie d sm oo th
S tratifie d w avy
A nn u lar
An n ula r
Figure 5.3 Separated flow
Elongated bubble
Slug
Slug Churn
Figure 5.4 Intermittent flow
5.1 Flow regimes in vertical flow
Most oil wells have multiphase flow in part of their pipework. Although pressure at
the bottom of the well may exceed the bubble point of the oil, the gradual loss of
pressure as oil flows from the bottom of the well to the surface leads to an increasing
amount of gas escaping from the oil.
24
Transitions between flow regimes in

the vertical tubing of an oil well are
illustrated in Figure 5.5, which shows
the different hydrodynamic flow
regimes which may occur in vertical
Mist f lo w
liquid-gas multiphase flows.
It should be noted that Figure 5.5 is

Ann ular flow only a schematic illustration which is
intended to show the transitions
between the flow regimes as the
Ch urn flow
superficial gas velocity increases from
the bottom of the well up to the
wellhead. In real production tubing it
is rare that more than two or three
Slug flo w
flow regimes are present at the same
time.
Bub ble flow

Figure 5.6 is a qualitative illustration
of how flow regime transitions are
dependent on superficial gas and
liquid velocities in vertical multiphase
No ga s
flow. It should be noted that the
transitions are also a function of
tubing diameter, interfacial tension,
density of the phases, etc..
Figure 5.5 Schematic transitions between

flow regimes in oil wells
5.2 Flow regimes in horizontal flow.
In horizontal flows too, the transitions are functions of factors such as pipe diameter,
interfacial tension and density of the phases. Figure 5.7 is a qualitative illustration of
how flow regime transitions are dependent on superficial gas and liquid velocities in
horizontal multiphase flow. A map like Figure 5.7 will only be valid for a specific
pipe, pressure and a specific multiphase fluid.
25
D IS P E R S E D
BU BB LE
S U PE R FIC IA L LIQ U ID V EL O CIT Y
BU BB LE CHUR N
SLUG
A N N U LA R
S U P E R F IC IA L G A S V E L O C I T Y
Figure 5.6 Multiphase flow map, vertical flow. Vsl is superficial liquid velocity, Vsg is superficial gas
velocity.
D I S P ER S E D
S UP ER FIC IA L L IQ UID V ELO CITY
B U B BL E
E L O N G AT E D
B U B B LE
SL U G
A N N U L AR
M IS T
S T R A T IF I E D
S M O O TH
ST R A T IF IE D
W A VY
S U P E R F I C I A L G A S V E LO C I T Y
Figure 5.7 Multiphase flow map, horizontal flow.
26
6. CLASSIFICATION OF MULTIPHASE METERS
Multiphase meters can be classified into different categories:

• Separation meters
• In-line meters
• Others
6.1 Separation type meters.
Separation type multiphase meters are a class of multiphase meters characterised by

performing a complete or partial separation of the multiphase stream, followed by
in-line measurement of each of the three phases. The test separator which is found
on nearly every production platform is basically a three phase meter. It separates the
three phases and carries out flow measurements of the oil, water and gas.
6.1.1 Separation of total flow
This type of meter is characterised by its separation of the total multiphase flow,
usually a partial separation to gas and liquid. The gas flow is then measured using a
single-phase gas-flow meter with good tolerance to liquid carry-over. The liquid flow
rate is measured using a liquid flow rate meter. The water-in-liquid ratio may be
determined by an on-line water fraction meter.
separator gas meter
Liquid meter
water-in-liquid ratio
Figure 6.1 Principle design of full separation type meter.
27
6.1.2 Separation in sample line
This type of meter is characterised by the fact that separation is not performed on the
total multiphase flow, but on a bypassed sample flow. The sample flow is typically
separated into gas and liquid, whereafter the water-in-liquid ratio in the sample stream
can be determined using an on-line water fraction meter. Total multiphase flow rate
and gas liquid ratio must be measured in the main flow line.
gas/liquid sep.
water-in-liquid ratio
gas/liquid ratio
restriction
multiphase flowrate
Figure 6.2 Principle of a multiphase meter with separation in sample line
In this configuration three measurements are required to determine the mass and
volume of the three phases;
Gas / liquid ratio (GLR): - gamma absorption

- vibrating tube
- neutron interrogation
- weighing
Multiphase flow rate: - cross-correlation using radioactive, acoustic or
electrical signals
- differential pressure using Venturi, V-cone or Dall tube
- mechanical, e.g. positive displacement or turbine
Water-in-liq. ratio (WLR): - electrical impedance
- vibrating tube
28
6.2 In-line meters
In-line multiphase meters are characterised in that the complete measurement of

phase fractions and phase flow rates is performed directly the multiphase flow line,
without any separation of the flow.
The volume flow rate of each phase is represented by the area fraction multiplied by
the velocity of each phase. This means that a minimum of six parameters has to be
measured or estimated. Some multiphase meters assume that two, or all three, phases
travel at the same velocity, thus reducing the required number of measurements. In
this case either a mixer must be employed or a set of calibration factors established.
In-line multiphase meters commonly employ a combination of two or more of the
following measurement techniques:
- microwave technology
- capacitance
- gamma absorption
- neutron interrogation
- cross-correlation using radioactive, acoustic or electrical signals
- differential pressure using Venturi, V-cone or other restriction
- positive displacement / turbine meter
Multiphase meter
Mixer (if specified)
Figure 6.3 Principle design of in-line multiphase meter
6.3 Other categories of multiphase meters
Other categories of multiphase meters include advanced signal processing systems,

estimating phase fractions and flow rates from analysis of the time-variant signals
from sensors in the multiphase flow line. Such sensors may be acoustic, pressure or
other types. The signal processing may be a neural network or other pattern-
recognition or statistical signal-processing system, for example.
29
There are also multiphase metering systems which have been developed on the basis
of process simulation programs combined with techniques for parameter estimation.
Instead of predicting the state of the flow in a pipeline at the point of arrival, its
pressure and temperature can be measured at the arrival point and put into the
simulation program. The pressure and temperature of an upstream or downstream
location also have to be measured. When the pipeline configuration is known along
with properties of the fluids, it is possible to make estimates of phase fractions and
flow rates.
30
7. IMPLEMENTATION OF MULTIPHASE METERS
Given the complexity of multiphase flow as described in Section 5, and the variety of
different approaches to the design of multiphase meters, it is likely that for specific
applications, some multiphase meters may turn out to be more suited than others.
Different meters have specific design and installation requirements which make it
important to do a thorough job of establishing the basis for a particular application,
when production facilities are being designed.
Different multiphase meters have different requirements for

- flow conditioning
- upstream and downstream pipe configuration, and
- orientation of the pipe at the point of installation.
While one instrument requires a mixing device, vertical upward flowing fluid and
quite short upstream and downstream straight lengths, another may require upstream
and downstream straight lengths of several pipe diameters, no mixing device and
horizontal pipe orientation.
7.1 Selecting the point of installation
The field installation can only be expected to reproduce the performance obtained in
the reference installations if the pipe configuration is the same as used in the reference
installations. However, if an instrument can be shown to reproduce its performance in
different pipe configurations, less attention need be paid to the pipe configuration in a
given installation.
A multiphase meter normally indicates measured quantities at flowing conditions.

These quantities often need to be recalculated to quantities at standard conditions.
The uncertainties associated with calculating the quantities at standard conditions
increase if the fluid is in a non-equilibrium state. After a choke over which the
pressure drop is significant, the fluid will be in a non-equilibrium state. Attention must
be paid to this point when choosing a point for installation if quantities at standard
conditions are to be calculated.
7.2 Prediction of flow regime at the design stage
When piping arrangements and systems where in-line multiphase meters will be
usedare being designed, it is important to be able to predict the kinds of flow regimes
31
that will occur. Under many flow conditions pipe configuration, flow conditioning
and orientation will significantly affect the flow regime at the cross-section of the
multiphase meter. Since no instrument measures the composition and velocity of an
infinite number of points over the cross-section, a change in flow profile at constant
flow rates will affect instrument readings. This influence may be large at some flow
conditions and insignificant at others, large for some instruments and small for others,
and is normally very difficult to predict from theory.
Most flow regime prediction methods are based on experience which has been from
low-pressure small-scale two-phase flow loops. Reliable scaling methods do not exist.
Detailed knowledge of three-phase flow regimes in high-pressure full-scale
production systems is therefore limited.
One available method consists of using "correlations" to predict the flow regime that
will exist at a given pipe cross-section. However, such correlations are reputed to
have large uncertainties, and must be used with care.
Another method, which is regarded as more reliable than correlations, is laboratory

testing. Laboratory testing should always be considered when implementing
multiphase meters in new applications, in order to verify the ability of the meters to
provide the required performance under a specific set of conditions. It should be
noted that extrapolation from test conditions to field conditions is uncertain, as
scaling and the effects of using test fluids may introduce significant uncertainties.
Laboratory testing is described in more detail in Section 9 of this Handbook.
7.3 Use of Mixers
Static or dynamic mixers located immediately upstream of the multiphase meter are
used to ensure well-mixed flow or to convert another multi-phase flow regime into
homogeneous flow. Different types of mixers exist, ranging from simple static mixers,
similar to a gas orifice plate, to more sophisticated versions that can absorb and
smooth out slugs. Whereas a static mixer consists of only non-moving parts, a
dynamic mixer utilises moving parts. Both type of mixers cause an instant pressure
drop over the device (most of which will be recovered), to mix the different phases.
If the operation of a multiphase flow meter requires the installation of a separate

upstream mixing device, the mixing device must be installed in accordance with the
manufacturer’s specifications. Furthermore, it is necessary to ensure that these
32
specifications are completely fulfilled during the period of measurement. The mixing
device should be counted as part of the instrument.
The application and use of mixers should always be evaluated in close co-operation
with the manufacturer of the specific multiphase meter.
7.4 Converting flow rates measured at flowing conditions to flow rates at

standard conditions
A multiphase meter usually measures phase fractions and flowrates at process

conditions, i.e. at the pressure and temperature prevailing at the installation point.
Test separator measurements are commonly converted to standard conditions before
presentation to the user. In order to compare multiphase meter measurements with
test separator measurements, both must be converted to the same process conditions,
usually standard conditions (15.0 oC, 1.01325 bara).
These calculations are not trivial, especially in the case of high pressure, real
hydrocarbon systems, which involve mass transfer between the gas and oil phases.
Models for “flashing” crude oil exist, but these require detailed knowledge of the
hydrocarbon composition.
The calculations assume thermodynamic equilibrium, which may not be perfectly true
in a multiphase pipeline, in particular at an installation point close to a major pressure
drop such as at a bend, mixer or other restriction. The uncertainty of such a
conversion is therefore high, and must be taken into account when evaluating the
overall uncertainty of a multiphase metering system, as indeed also with the
conventional systems. Additional uncertainty is caused by the solubility of
hydrocarbon gases in both oil and water.
7.5 A format for initial evaluation of multiphase meter implementation
On the following two pages, two fill-in forms suitable for an initial evaluation of
installation of a multiphase meter is proposed. The fill-in form could be used by
users, or by manufacturers, to compare meter performance specifications with actual
process data at an intended installation point. The forms are intended as a first
evaluation only, and more comprehensive investigations will have to follow before
any certain decitions of suitability can be made. When using the fill-in forms, the
following should be included:
• A sketch showing imortant details of the installation point:

33
− upstream / downstream piping and process equipment

− available space
− other relevant information
• Process conditions:
− the list in the form is not exhaustive, and other process parameters may be
required
− an attempt should be made to identify which flow regimes that are to be
expected at the actual installation
• Expected multiphase production profile:
− data to enter the graph must be given at process condition
− superficial velocity axes should be linear, from zero to any required upper
velocity range
− secondary flowrate axes should be filled in according to the actual line size,
or used to select a more suitable metering line size
− the production profile, or multi-well production rates, should be marked on
the graph, using WLR and year / well no. as legend
• Multiphase uncertainty calculations:
− data to enter the table must be given at process condition
− a representative number of expected production points should be filled in
− average multiphase velocitiy, WLR and GVF may easily be calculated for
each of the selected production points
− absolute and relative deviations in measured phase flow rates may be
calculated for each of the selected production points, using manufacturers
performance specifcation on a format suggested in section 8 of this
document
34
Multiphase meter implementation evaluation

Application: ____________________
Installation location: ____________________
Date: ____________________
Reference: ____________________
Installation point piping configuration:
Process conditions:
Pressure: Temperature:
Oil density: Oil viscosity:
Gas density: Gas viscosity:
Water density: Water salinity:
Expected flow
regimes:
Expected multiphase production profile (at process conditions):
15
14 10% gas
25% gas
Superficial liquid velocity (m/s)
13
12
Liquid flowrate (m3/h)
11
50% gas
10
8 Legend:
7
A = Year 0, WLR
6
aa%
5
4
75% gas B = Year xx,
3 WLR bb%
2
C = Year yy,
1 90% gas
WLR cc%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Superficial gas velocity (m/s)
Gas flowrate (m3/h)
35
M U L T IP H A S E M E T E R - U N C E R T A IN T Y C A L C U L A T IO N S
A P P L I C A T IO N : P R O C E S S C O N D IT IO N S : M E T E R D A T A :
F ield: Pressure: M e ter type:
P latform : Temperature: M o d e l:
Installation: No of Meters:
Case: M e ter Size:
Inner Diam :
F L O W R A T E S - A t Process Conditions F L O W C H A R A C T E R IZ A T IO N E S T IM A T E D U N C E R T A IN T Y IN E A C H P H A S E ( 1 )
M U L T IP H A S E
Y E A R W A T E R F L O W G A S F L O W O IL F L O W F L O W V E L . W L R G V F W A T E R G A S O IL
m 3 /h m 3 /h m 3 /h m /s % % relative absolute relative absolute relative absolute
% m 3 /h % m 3 /h r % m 3 /h r
Note: (1) A s s u m ing calibration param eters given within uncertainty lim its specified by supplier, e.g.;
- O il density
- Gas density
- W a ter density
- W a ter conductivity
36
8 PERFORMANCE SPECIFICATION
This chapter will deal with performance statements for multiphase flow meters. Some
of the criteria to be considered are:
• uncertainty
• reproducibility
• influence quantities
8.1 Uncertainty description
Measures of flow element performance represent the difference between how an ideal
flow meter would perform and how the real flow meter actually performs.
The preferred performance specification of a multiphase flow rate meter is given in

terms of the percentage uncertainty of actual oil, gas and water volume flow rates.
However, in many cases this specification will be impractical, and not suited for a best
possible description of the actual meter's performance. Other frequently used
specifications are therefore,
• as a percentage of the actual total multiphase flow rate, or

• as uncertainties in actual liquid and gas flow rates, and with an absolute
uncertainty specification of water-in-liquid ratio.
Tables 8.1 to 8.3 provides an overview of three different methods commonly used to
describe the uncertainty of multiphase meters, applied to typical flow conditions. The
resulting uncertainty numbers are presented in terms of relative (% of phase volume
flow rate) and absolute (m3/hr) uncertainty.
As is seen in these tables, even though the different uncertainty specifications at a first
may look seem similar, they differ significantly.
Other methods for specifying multiphase meters exists, and the purpose of Tables 8.1
to 8.3 is to give some guidance to how different specifications can be compared,
rather than providing a complete overview of the methods. Due to the significant
difference in performance between these methods of expression, the manufacturer
should clearly state his method of specification.
37
These three methods of performance specification will in the following be explained

through examples:
Method 1: Uncertainties relative to actual phase flow rates.
By this method the uncertainty of the meter is described as a fixed per cent (relative)
of the actual volumetric flow rate of each phase.
In the examples shown in Table 8.1, each phase volume flow rate has a +10% relative
uncertainty. The resulting absolute uncertainties in terms of m3/h are calculated as
+10% of the actual phase volume flow rate.
Table 8.1
Specification: +/- 10 % of phase volume flow rate (WLR = 20 %)
Fluid Flow rate Uncertainty Uncertainty

3 3
[m /h] [m /h] [%]
Multiphase 125 +/- 12.5 +/- 10 %
Gas 100 +/- 10.0 +/- 10 %
Oil 20 +/- 2.00 +/- 10 %
Water 5 +/- 0.50 +/- 10 %
Specification: +/- 10 % of phase volume flow rate (WLR = 4.76 %)
Multiphase 125 +/- 12.5 +/- 10 %

Gas 20 +/- 2.0 +/- 10 %
Oil 100 +/- 10.0 +/- 10 %
Water 5 +/- 0.50 +/- 10 %
Method 2: Percentage of total multiphase flow rate.
By this method assumes the uncertainty of the multiphase meter is described as a

fixed per cent of the total multiphase volume flow. The absolute uncertainty in terms
of m3/hr is thus equal for all three components for a given total flow.
In the examples of Table 8.2, an uncertainty level of +5% is used. Of a total flow of
125 m3/hr, this equals +6.3 m3/hr, which is then the absolute uncertainty for all three
phases, independent of the composition (since the total flow is the same for all four
cases).
38
The relative uncertainty numbers can be derived simply by calculating the ratio
between the absolute uncertainty and the actual flow rate for that component:
δQP = (QM * δX)/ QP
where δQP = the relative uncertainty in phase flow rate

δX = the specified uncertainty
QM = the multiphase flow rate
QP = the phase flow rate at QM
Table 8.2
Specification: +/- 5 % of total multiphase flow rate (WLR = 20 %)

3 3
[m /h] [m /h] [%]
Multiphase 125 +/- 6.25 +/- 5.00
Gas 100 +/- 6.25 +/- 6.25
Oil 20 +/- 6.25 +/- 31.25
Water 5 +/- 6.25 +/- 125.00
Specification: +/- 5 % of total multiph. flow rate (WLR = 4.76 %)
Total flow 125 +/- 6.25 +/- 5.00

Gas 20 +/- 6.25 +/- 31.25
Oil 100 +/- 6.25 +/- 6.25
Water 5 +/- 6.25 +/- 125.00
Method 3: Percentage of gas and liquid flow rates, combined with absolute
uncertainty in WLR.
The relative uncertainty in gas flow rate is given specificly, whereas the uncertainties
of oil and water rates results out of uncertainty in two levels. First, there is a relative
uncertainty in the liquid flow rate, and this one must be combined with a second
absolute uncertainty regarding the determination of the WLR of the fluid. It is
assumed that these two uncertainties can be regarded as independant of each other.
39
Table 8.3
Specification: +/- 10 % of gas flow rate

+/- 10 % of liquid flow rate (WLR = 20 %)
+/- 3 % absolute uncertainty in WLR
[m3/h] [m3/h] [%]
Multiphase 125 +/- 12.5 +/- 10.0
Liquid 25 +/- 2.5 +/- 10.0
Gas 100 +/- 10 +/- 10.0
Oil 20 +/- 2.14 +/- 10.7
Water 5 +/- 0.90 +/- 18.0
Specification: +/- 10 % of gas flow rate
+/- 10 % of liquid flow rate (WC = 4.76 %)
+/- 3 % absolute uncertainty in WC
Multiphase 125 +/- 12.5 +/- 10
Liquid 105 +/- 10.5 +/- 10
Gas 20 +/- 2 +/- 10
Oil 100 +/- 10.5 +/- 10.5
Water 5 +/- 3.19 +/- 63.8
The examples of Table 8.3 displays results for a case where the uncertainty of the gas
volume flow rate is +10%. The liquid volume flow rate uncertainty is +10%, which
is combined with an uncertainty of determining the WLR of +3 % absolute.
The absolute uncertainty in the water volume flow rate, ∆VW, is then given by
∆VW = SQRT {(∆WLR * VL)2 + (δVL * VL * WLR)2 }
where WLR = the actual water-in-liquid ratio

∆WLR = the absolute uncertainty in WLR
VL = the actual liquid volume flowrate
δVL = the relative uncertainty in the liquid volume flowrate
The relative uncertainty in water volume flow rate is then simply given by the relation
between the absolute uncertainty in water volume flow rate, ∆VW, and the actual
water volume flowrate.
Accordingly, the absolute uncertainty in oil volume flow rate, ∆Vo, is given by
40
∆Vo = SQRT {(∆WLR * VL)2 + (δVL * VL * (1-WLR))2 }
8.2 Establishing the reproducibility of a given meter
The reproducibility of a meter is a quantitative expression of the agreement between

the results of measurements of the same value of the same quantity, where the
individual measurements are made under different defined conditions.
One significant difference between multiphase meters and single-phase meters is that
most of the uncertainty of a multiphase meter is caused by variations in process
conditions and fluid properties, rather than the uncertainty of the primary
measurement elements. Therefore, the meter’s ability to reproduce its performance
under different process conditions, installation set-ups and flow regimes becomes a
very important parameter.
The reproducibility of a multiphase meter for a set of flow rates may be established by
recording the deviation between values measured by the meter and reference values
obtained from different test facilities. Particular emphasis should be placed on to
establishing the reproducibility from laboratory tests to field test conditions, and to
identify the influence quantities and their effects.
8.3 Rated conditions of use
In a performance description it should be specified for what process conditions the

accuracy of the measurement is given. The following information should at least be
given, together with acceptable boundaries for which the uncertainty specification is
valid:
• temperature range
• pressure range
• oil density and viscosity ranges
• gas density and viscosity ranges
• water density and salinity ranges
41
8.4 Influence quantities and how they influence accuracy
Influence quantities are to be understood as quantities that alter the specified

accuracy of the measurement, such as
• changes from one flow regime to another

• additives, e.g. emulsifiers, wax inhibitors, corrosion inhibitors
• installation effects, upstream straight lengths, bends, etc.
• water salinity, phase densities, etc.
It is important to specify what is influenced by the quantity in question and, if

possible, to quantify the effect on the uncertainty of the measurement.
8.5 A format for the presentation of a multiphase meter performance

specification
On the following two pages, a fill-in form for summarizing the performance
specification of a multiphase meter is proposed. The fill-in form could be used by
users, or by manufacturers, to assemble essensial information from different
manufacturers product information packages, to a common format. When using the
fill-in form, the following should be included:
• A sketch showing imortant details of installation requirements:

− horizontal / vertical upwards / vertical downwards flow
− mixer / not mixer
− straight upstream / downstream lengthts
• Rated conditions of use:
− the list in the form is not exhaustive, and other parameters important for the
particular meter should be included
− flow regimes that the meter is designed to handle should be listed
− the interval in which the influence parameters are allowed to vary, still
maintaining the uncertainty specifications, should be specified
• Influence quantities:
− the list may, or may not, include the same parameters as listed “Rated
conditions of use”; all influence parameters important for the particular
meter should be included
• Operating range:
velocity range
42
− secondary flowrate axes may be used to denote or select a suitable meter

size
− operating range should be marked on the graph, and may be divided into as
many sub-areas as required
• Uncertainty specification:
− uncertainties should be given for each sub-range of the operating range
− uncertainties in phase flowrates should preferably be given relative to actual
phase flowrates
− absolute deviations in WLR and GVF may be given as indicated
− the uncertaity specification may be quoted for as many WLR-ranges as
required
• Additional information:
− method of meter calibration / special calibration requirements must be
identified
− reference to more comprehensive product information should be indicated
43
−
Multiphase meter performance specification
Manufacturer: ____________________
Meter type: ____________________
Date : ____________________
Reference: ____________________
Required Installation Configuration Schematic:
Rated conditions of use:

Flow regimes:
Influence quantities:
Quantity Influencing Effect
Oil density:
Gas density:
Water density:
------
------
Flow regime:
44
Operating range:
15
14 10% gas Operating range

25% gas

13
12
11
50% gas
10
4
75% gas
1 90% gas
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Gas flowrate (m3/h)
Uncertainty specification:
Sub - WLR Uncertainties;
Range Range according to Method 1, 2 or 3
Š%‹‹ Oil Water Gas Liquid WLR

A 0-x
B 0-x
C 0-x
D 0-x
E 0-x
A x-100
B x-100
C x-100
D x-100
E x-100
Calibration requirements:
Reference:
45
9. TESTING OF MULTIPHASE METERS
The term “testing” is understood to mean that the measured results are compared to
data from reference measurements in order to quantify and analyse any deviations
from true values. A test is also a check of the functionality of the instrument,
including determination of the effective range of the instrument. Testing would
normally be performed with flowing hydrocarbons and water.
Several types of testing can be envisaged, and this chapter provides an overview of
those most often referred to, including test facilities, test fluids and the test matrix.
Measurement uncertainty, reproducibility and influence quantities are important
descriptive aspects of the performance of a multiphase meter. Hence, these
parameters are important when test results are evaluated.
9.1 Test facilities
At first, most test facilities were small, purpose-built for a specific design of meter,
and with limited capabilities. Since these early days, test facilities for more general
use, and with larger operating envelopes, have been built.
Whereas the smaller test rigs in many cases are owned by meter manufacturers, the
larger facilities are owned by agencies or research organisations.
In this section aspects of test facilities for multiphase flow meters are discussed. Small
test rigs for special purposes are mentioned only briefly.
9.1.1 Design of laboratory test facilities
A facility for testing multiphase flow meters must have an envelope of operation that
matches the meters to be tested. Otherwise only partial testing is possible.
Flow parameters that must be checked in the course of testing of multiphase flow
meters include:
• individual flow rates of oil, gas and water in order to determine relevant gas/liquid
ratios, as well as water-in-liquid fractions
• flow regimes, by controlling pipework configuration (straight pipe sections,
horizontal/vertical up-down/sloping pipes), fluid viscosities and phase flow rates
• fluid pressure and temperature.
46
The fluid constituents oil, gas and water should be similar to those of the application
fluid.
In any given test facility one or more of the flow parameters listed above may be
impossible to control, limiting the test capabilities.
In a test facility, the fluids are normally circulated in a closed-loop system. There are
at least two options:
• Single phases of oil, gas and water are pumped and measured before being mixed
and passed through the test section. Downstream of the test section, the
multiphase fluid flow is again separated into single phases. Reference
measurements of each single phase are made before mixing, even if a multiphase
reference flow meter downstream of the mixing point can be used.
• Oil, gas and water are first mixed and then pumped continuously as a multiphase
fluid in a closed loop. Gas and/or water fractions can be varied by injecting or
withdrawing fluid into/from the circulating mix. Phase flow rates or fractions are
determined by the mixing procedure and are assumed to be constant until pumping
or composition are changed by adding or withdrawing fluid(s).
In certain cases, the water and oil are circulated as a single phase. Within certain
ranges of water in the oil, the two components will not separate rather remain as one
single phase. Concentration ranges within which water and oil will not separate
depend on the properties of the oil and water as well as on concentrations of surface-
active chemicals.
In principle, a multiphase flow meter could also be used for reference measurement in
such cases. This type of circulation loop lends itself better to functional testing than to
accurate determination of instrument performance.
Multiphase flow meters measure rates at the operating conditions of the fluid as it
passes through the meter. If the reference meters operate at conditions different from
those of the multiphase flow meter, flow rates must be calculated for the conditions
of the multiphase flow meter. This would include calculation of mass transfer between
the phases. Special care must be taken when testing in low pressure loops; small
deviations in pressure will have significant impact on volumetric gas flow rates.
47
9.1.2 Test facilities in a live process
Some oil companies have set up test facilities in their production plants. The test
medium used in such facilities is the ideal one: live well fluids at process operating
conditions.
Various options are available for setting up the test bed in the process. Reference
measurements are normally carried out on single-phase outlets from a separator, e.g.
the test separator.
With this set-up, the selection of test points is limited to wells or fluids which can be
routed via the separator. Fluid properties and phase fractions can be changed only by
changing the well being tested. In principle, the flow rates are selectable. In practice,
however, wells or flow rates may be available for testing only if plant operation is not
hampered.
Some live process test facilities have been modified to offer the option of injecting,
withdrawing or recirculating fluids. In such facilities fluid properties, flow rates and
phase fractions may be selected within a much wider range. Interference with normal
plant operation is reduced. Such test facilities may be complex, and direct reference
measurements may be more difficult to obtain.
In some cases multiphase flow meters are installed in the process for functionality test
purposes. Reference measurements may be limited or non-existent. Even if tests are
very useful, such facilities are not really considered test facilities for the purposes of
this handbook.
9.1.3 Test fluids
The test fluid is an important feature of any test facility. Selection of test fluid is not a
trivial matter and the following considerations must be taken into account:
• type of test facility (hazardous area)

• suitability of test fluid for the test purpose
• working environment, hygienic aspects.
The test fluid is either:
• a model system, using some sort of model oil, water and air or nitrogen, or
48
• a system with live crude, formation water and hydrocarbon gas, with mass transfer
between the oil phase and the gas phase.
Most test facilities use a model system, for reasons of cost, working environment, etc.
In many cases a model system is the only option available. Even operating a model
system may be subject to stringent conditions for use, and the model oil may not
always have been selected for meter-testing purposes only.
Model test fluids serve their purpose, but there are obvious limitations to their
suitability:
• The fluid is not representative of the fluid to be measured, in terms of density,

viscosity (and thus generation of flow regime), dielectric constants, salinity, mass
transfer between the phases, phase-surface active components, etc.
• It is not self-evident that meter calibration carried out using a model fluid is valid
for use on a crude oil well stream.
The problems of using oil products for test fluid are related to the availability of
suitable plant (the cost aspect) and the fact that such plants are built and operated
under a hazardous-area regime. Since the properties of well streams differ, a specific
product used as test fluid may not be representative of any other product or well
stream.
It is possible to synthesise a product-type test fluid from stabilised crude oil, water
with salts added and gas synthesised from methane, ethane etc. Using synthesised
product for the test fluid is practical only for test facilities that employ closed-loop
circulation.
9.1.4 Test matrix
A calibration or test matrix must be defined for each meter to be tested. In principle,
this is no different from other test situations, but with multiphase meters the test
matrix will have a large number of points, due to the very nature of the meter. With
four flow rates per phase, 64 points are needed to cover every possible combination
of pressure, temperature, water salinity, etc. The test matrix soon runs into hundreds
of points.
For this reason it is usually necessary to reduce the number of points from "the full
set", to one or more subsets. With multiphase flow meters, such a reduction is more
difficult and more important due to the very large number of possible variations. The
49
test points which can be omitted with the smallest loss of information of meter
performancemust be identified. It is likely that the "most redundant" points are
different for different types of meters, due to their different working principles.
9.1.5 Reference measurements
Test results are only as accurate as the reference measurements of the test facility.
When the results of multiphase meter tests are being evaluated, the measurement
uncertainty of the reference measurements must also be taken into consideration.
It is very important that reference measurements are carefully controlled. Reference

measurement uncertainty involves meter functionality and reliability, calibration
accuracy, and the transfer of calibration to reference meter use in the test facility
concerned.
In most test facilities, the reference measurements are dynamic measurements for
each single phase upstream of a phase-mixing point. In these cases the reference
meters are "normal" flow meters for oil, water and gas, for any given starting point.
The selection and sizing of flow meters follows normal evaluations and normal
procedures. Special attention must be paid to the reference meter’s susceptibility to
density, viscosity, etc., parameters that will vary in any normal test matrix.
It should be noted that if the test facility has a wide operating envelope in terms of
flow rate, temperature and pressure, suitable "normal" flow meters may be difficult to
find, and two or more parallel meter runs may be required to cover all flow rates.
In test facilities with recirculating flow, one or more phases may not be measured
directly. In such cases the reference measurement uncertainty is increased. Additional
measurements or calculations may be required, such as a water-in-oil meter in the oil
line to measure water carry-over.
The reference flow meters must be subject to periodic calibration, traceable to

national or international standards.
50
9.2 Types of tests
9.2.1 Factory test
By “factory test” is meant a test performed by the manufacturer of the instrument.

This test is usually carried out using facilities owned or controlled by the
manufacturer.
Factory testing may be carried out for several reasons:
• investigation of the performance of a new type of meter during the development

phase
• verification and/or calibration of meters before delivery to customer/user.
Factory tests have certain advantages, as well as a number of limitations:
+ easy access to test facilities and fewer limitations on test time, making larger test
matrices possible
+ relatively inexpensive test time
+ test facility may be purpose-built for a specific make/type of meter
+ the range of phase flow rates may be wide
− the test fluid is normally unlike that of an oil/gas well stream
− flow conditions/regime are likely to be different from the real-life application
− tests cannot be regarded as independent, unless the facility is operated as a part of
the organisation which is independent of production, and with its own quality
program
− normally low pressure
9.2.2 Independent laboratory test
By “independent laboratory test” is meant a test performed by an organisation or

company which is independent of the manufacturer of the meter. An independent test
facility must be expected to have a quality programme with formalised procedures
and reference instrumentation traceable to national or international standards. It is
possible for a laboratory to obtain official accreditation.In principle, it is also quite
possible for a manufacturer to establish independent testing according to the
description above.
51
Independent testing is carried out because such tests are regarded as non-biased,
testing may be standardised to permit comparisons of meter performance or the test
envelope is more suitable. Independent testing would normally provide stronger
verification than in-house tests.
At the time of writing, independent test facilities vary significantly in terms of test
capabilities and, probably, in cost levels. Various test fluids and flow conditions are
available, e.g. model systems and real hydrocarbon fluids. Flow rates, flow regimes,
temperature and pressure ranges differ among different test facilities.
Compared to factory testing, features of independent laboratory testing include:
+ testing is independent and results are un-biased

+ a larger test matrix in flow rates, pressure and temperature is normally possible, as
is testing with different fluids
− testing is more expensive
9.2.3 Field test
By “field test” is meant a test performed in facilities located in an oil/gas production

plant, such that hydrocarbons from a well stream make up the test fluid. One or more
of the fluid phases may be re-circulated in order to generate phase fractions that differ
from the well stream(s). It would be possible to add either fresh water, sea-water or
water with salts added. In cases where the water or other phases are synthesised, the
test is not a true field test.
A field test is usually an independent test, carried out either by the operator of the
oil/gas production plant or under his auspices.
Field testing too has advantages and limitations:
+ the test is performed under real-life conditions, which means that any shortcomings
of the meter not apparent in a laboratory are more likely to be discovered
+ the fluid properties are more similar to the fluid properties in a real well stream
than to the properties of the fluids used in a laboratory
+ flow rates, flow regimes, pressure and temperature are more representative than
flow conditions found in a laboratory
+ data obtained can be used to establish the reproducibility of the instrument
− the range of gas volume fractions and oil/water volume fractions may be limited,
reducing the possible size of the test matrix
52
− flow conditions may not be representative of other fields or installations,

− reference measurements may be poor and/or add to the cost of the test.
9.2.4 In-situ test
“In-situ test” refers to a test which is carried out with the meter installed in the actual
location where it is used. Meters installed for a new application may be subject to in-
situ testing as part of commissioning/start-up. In-situ testing is also relevant for the
periodic testing of meters in normal use, i.e. verification of meter performance
without removing the meter from the line.
An in-situ test has the following advantages and drawbacks:
+ the test provides definitive performance data for a specific instrument in a given
application
− installation for test purposes only (not actual installation for use) is very expensive
− reference measurements may be poor and/or add to the cost of the test.
9.3 A format for presentation of summary test results.
On the following two pages, a fill-in form for summarizing test results of a multiphase
meter is proposed. The fill-in form could be used by users, or by manufacturers, to
present summaries from different tests, and of different meters, using a common
format. When using the fill-in form, the following should be included:
• A sketch showing imortant details of the test installation:

− horizontal / vertical upwards / vertical downwards flow
− mixer / not mixer
− straight upstream / downstream lengthts
− phase commingling point / distance to meter under test
− position of reference measurements
• Process conditions:
− the list in the form is not exhaustive, and other parameters important for the
particular test should be included
− flow regimes that the meter under test has been subjected to, and how these
regimes were controlled or observed
• Reference measurements:
− type and quality of reference measurements shold be provided
− reference to installation point on installation sketch should be given
53
• Multiphase meter calibration prior to test:

− a qualitative description of the calibration performed by meter manufacturer,
or by test institution, prior to the test
− reference to a complete calibration report should be provided
• Test matrix:
velocity range
− secondary flowrate axes should be filled in according to the actual meter size
− test-points should be marked on the graph, using WLR as legend
• Test results summary:
− a representative number of test points should be filled in
− deviations in phase flowrates should preferably be given relative to actual
phase flowrates
− absolute deviations in WLR and GVF may be given as indicated
− the uncertaity specification may be quoted for as many WLR-ranges as
required
− any particular observations during test should be identified in the comments
field
− reference to a more comprehensive test report should be indicated
54
Multiphase meter test summary

Meter identification:
Test location:
Test period:
Test responsible:
Test Installation Configuration Schematic:
Process conditions:
Flow regimes
(how observed):
Reference measurements:
Phase Reference Position Uncertainty Comments
meter type (ref. to Piping
Config. Schematic)
Oil Traceable
Water Traceable
Gas Traceable
Multiphase meter calibration prior to test:

Description:
Calibrated by:
Date of calibration:
Reference to calibration
report:
55
Test matrix:
15
14 10% gas Operating range

25% gas

13
12
11
50% gas
10
4
75% gas
3
1 90% gas
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Gas flowrate (m3/h)
Test results summary:

Reference Test results Šm3/h‹‹ Deviations;
measurements Šm3/h‹‹
Oil Water Gas Oil Wat. Gas Liq. WLR Oil Wat. Gas Liq. WLR
Comments:
Test report reference:
56
10 QUALIFICATION OF MULTIPHASE METERS
“Qualification” of a multiphase meter refers to one or several tests that verify whether
the multiphase meter’s performance meets the stated requirements, and in
particular, whether the reproducibility and reliability of the meter lie within pre-
established limits for different applications.
The basis for the qualification procedure should be tests performed by the
manufacturer, by independent laboratories, or under field conditions.
To ensure the reliability of the qualification process for different kinds of multiphase
meters, test results should be presented in a common format. A pre-ordained and
well-defined test scheme acts as a warranty for the user, ensuring that the test report
contains all necessary and important information.
Users of multiphase meters usually need to qualify the meter for each specific
installation. Such a qualification process may be based on analyses of previous test
results from one or more test facilities, if necessary supplemented with new tests.
Qualification will normally be a joint effort by the user and the vendor of the
multiphase meter.
In determining the effective range of a multiphase meter, we should bear in mind not
only the range of flow rates and compositions to be covered, but also the range of
flow regimes expected in a particular installation. Parameters such as pressure,
upstream and downstream piping configuration, etc. may influence the flow regime,
thus modifying the effective range determined by laboratory tests or field tests under
different process conditions.
Laboratory tests might be used to compare different multiphase meters. Each meter
will have its rated conditions of use within which the best results are obtained.
Therefore, if the comparison forms part of a general evaluation of several multiphase
meters, it is recommended that meters should be tested by different laboratories in
order to obtain a wide range of flow conditions. If, on the other hand, the comparison
is required in order to select a suitable meter for a specific installation, a laboratory
that as closely as possible reproduces the flow conditions at the intended application
must be chosen.
Another important objective of a qualification programme will be to prove the long

term reliability of a multiphase meter under operational condition. In this case, the
57
availaility of high quality reference measurements will be less important. A field test
installation should then be considered.
58
11. CALIBRATION OF MULTIPHASE METERS
A flow rate calibration is defined as a comparison of measured flow rate with

reference flow rate made in order to establish a calibration factor. A calibration is
performed when calibration factors are established and implemented in the
instrument, either through software or mechanical/electrical adjustments.
One significant difference between multiphase meters and single-phase meters is that
the uncertainty of a multiphase meter is mainly caused by changes in process
conditions and fluid properties, rather than by the uncertainty of the primary
measurement elements.
The primary measurement elements that make up a multiphase meter can usually be
calibrated according to standard procedures, similar to those used for single-phase
measurement. However, the output of the primary measurements of a multiphase
meter is used as the input to the advanced signal-processing stage, giving individual
phase flow rates as the end result.
Flow rate calibration procedures as we know them from single-phase metering can
therefore not be directly transferred to multiphase meters.
This chapter discuss different ways of calibrating multiphase meters. The chapter is
written in general terms, and not all of the information given need be perfectly valid
for all multiphase meters.
11.1 Factory Calibration
The factory calibration performed by the manufacturer may consist of measurements

of geometric dimensions, gamma-meter count rates and static impedance
measurements in calibration fluids, etc., depending on the working principle of the
primary measurement elements. These calibrations are usually done under static
conditions. Calibration of the primary elements is usually independent of the process
conditions under which the instrument is to be used.
In addition, most manufacturers will run their instrument in a flow loop, either to
perform a flow rate calibration, or to establish a theoretical / experimental signal
processing which is to be valid for the specific meter and be independent of different
process conditions and fluid properties.
59
11.2 Independent Laboratory Calibration
The aim of independant laboratory calibration is to establish a set of calibration

factors and thereby increase the confidence of the multiphase meter calibration
compared to a factory calibration.
The value of any laboratory calibration depends on the reproducibility of the

multiphase meter under changing process and flow conditions. If a multiphase meter
gives the same output for identical flow rates under different process conditions and
physical properties of oil, gas and water, i.e. it displays good reproducibility, the
value of laboratory calibration will be high.
If the reproducibility of the multiphase meter is not known, or is not regarded as

adequate, the laboratory must be able to reproduce process conditions and physical
fluid properties as close as possible to those of the actual application.
We thus recommend using laboratory calibration with great care, and carefully
evaluating all the information available on instrument reproducibility, i.e. previous
tests and field applications, before a calibration test programme is performed.
11.3 Field Calibration
From a calibration point of view, the main difference between an independent

laboratory calibration and a field calibration is that representative fluid properties are
more likely to be obtained in a field test facility than in a laboratory. The limitations of
a field calibration however, are similar to those of an independent laboratory
calibration. In addition, reliable references may be difficult to obtain.
11.4 In Situ Calibration
By “in situ calibration” is meant that the multiphase meter is calibrated at the actual
installation point where it is going to be used. The aim of in situ calibration is to
reduce the uncertainty of the multiphase meter as compared with that obtained from a
factory calibration, or a flow calibration performed in laboratory or field test facility.
Whenever possible, implementation, and periodic verification, of this type of
calibration is recommended. Provision is made that reliable reference measurements
are available.
60
11.4.1 Calibration by using Test Separator as reference.
When the multiphase meter is used to measure a well stream which is occasionally
routed through a test separator, the test separator measurements can be used to
calibrate the multiphase meter.
The results obtained from the test separator must be compensated for phase transition
due to changes in pressure and temperature in the well stream between the location of
the test separator and that of the multiphase meter.
With good instrument repeatability, the uncertainty of a multiphase meter being

calibrated by using a
test separator should be close to the uncertainty of the test separator. This is the case
provided that the distance between the multiphase meter and the test separator is
quite short, and that the flow is not dominated by transient conditions i.e. installation
immediately downstream of a choke valve.
Multiphase meters located at a subsea wellhead can in principle be calibrated using a

vessel prepared for well testing. A multi-rate calibration can be performed on each
multiphase meter and the uncertainty after calibration may be close to that of the test
separator on board the well testing vessel.
11.4.2 Calibration done at start-up of the satellite field
A potential use of multiphase meters is to place one multiphase meter on each single
well head in a satellite field. In this way, test line, test manifold and several valves are
not needed. If individual wells are put into production one by one, each meter can be
calibrated at the start up of each well. If a multi-rate test is done for each well at start
up, it should be possible to obtain quite a good calibration for each meter, provided
the production can be measured by an instrumented inlet separator or test separator.
One possibility is to record sets of rates measured during a multi-rate test with the
multiphase meter and the references, and to establish a calibration curve based on
these data sets. With good instrument repeatability, the uncertainty of the calibration
will primarily be determined by the uncertainty of the test separator.
The calibration can also be done when testing by reduction. When testing by
reduction, the first well is opened and measured using the separator and a multiphase
meter. When the first meter is calibrated, the second well is opened. The increase in
flow rate at the separator will now be due to the production of the second well. If the
61
production of the first well changes, this can be measured by the first meter and
compensated for. Testing by reduction will be more accurate with multiphase meters
placed on each well, since the wells that have not been tested can be measured using
previously calibrated multiphase meters. The uncertainty of the most recently
calibrated multiphase meters will be higher than that of the first multiphase meter.
11.4.3 Limitations of in situ calibration
The accuracy of in situ calibration is limited by the accuracy of the reference

measurements made on site.
Unstabilised liquid hydrocarbons contain some light components which will be

transferred from liquid phase to the gas phase when the pressure is reduced. Thus,
the mass flow rate of hydrocarbons in the liquid and gas phases will change when the
pressure is reduced. For this reason the reference flow rates must be compensated for
this phase transition. If the pressure loss between the multiphase meter and the
reference instruments is small, this effect may be neglected.
If the pressure loss between the multiphase meter and the reference meters is large, a
simulation program can be used to compensate for the effect of phase transition.
However, the uncertainty of such a simulation may be large.
11.5 Implementation of calibration results
The calibration for a multiphase meter can be implemented using one of the following
methods or combinations of these methods.
11.5.1 Matrix calibration
The data obtained from the calibration process can be used to establish a matrix of
calibration factors. When a matrix is used, the instrument chooses the calibration
factor valid for the flow conditions that occur in the pipeline.
11.5.2 Curve-fit calibration
A curve-fit calibration is done by recording measured oil, gas and water rates and
reference rates for many points in a matrix. Using these data, a function which gives
the reference rates as an output for measured rates (or some other measured
parameters) as an input can be derived.
62
11.5.3 Factor calibration
If the meter is to be used mainly in a small range of flow conditions, and it is possible
to obtain reference values for the meter when used to measure a point within this
limited range, a single calibration factor can be established for each of the
components for later use as a valid calibration within the given range.
63

Handbook of Multi Phase Metering

Загружено:

Сведения о документе

Исходное описание:

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Handbook of Multi Phase Metering

Загружено:

Авторское право:

Доступные форматы

Handbook of multiphase metering

The Norwegian Society for Oil and Gas Measurement

John Amdal, Saga Petroleum

3.1 Terms related to multiphase flow 6

3.2 Terms related to metrology 9

4. APPLICATIONS OF MULTIPHASE METERING 14

4.1 Well testing 14

4.2 Production metering 18

4.3 Comparison of multiphase meters with conventional methods 20

5.1 Flow regimes in vertical flow 24

5.2 Flow regimes in horizontal flow. 25

6. CLASSIFICATION OF MULTIPHASE METERS 27

6.1 Separation type meters. 27

6.2 In-line meters 29

6.3 Other categories of multiphase meters 29

7. IMPLEMENTATION OF MULTIPHASE METERS 31

7.1 Selecting the point of installation 31

7.2 Prediction of flow regime at the design stage 31

7.3 Use of Mixers 32

7.5 A format for initial evaluation of multiphase meter implementation 33

8.1 Uncertainty description 37

8.2 Establishing the reproducibility of a given meter 41

8.3 Rated conditions of use 41

8.4 Influence quantities and how they influence accuracy 42

8.5 A format for the presentation of a multiphase meter performance specification 42

9. TESTING OF MULTIPHASE METERS 46

9.1 Test facilities 46

9.2 Types of tests 51

9.3 A format for presentation of summary test results. 53

10 QUALIFICATION OF MULTIPHASE METERS 57

11. CALIBRATION OF MULTIPHASE METERS 59

11.1 Factory Calibration 59

11.2 Independent Laboratory Calibration 60

11.3 Field Calibration 60

11.4 In Situ Calibration 60

11.5 Implementation of calibration results 62

Multiphase flow measurement has yet to be established as a separate discipline.

This document is intended to serve as a guide for users and manufacturers of

The document may also serve as an introduction to newcomers in the field of

The performance of a multiphase meter in terms of accuracy, repeatability, range, etc.

3.1 Terms related to multiphase flow

Gas volume fraction (GVF):

Homogeneous multiphase flow:

Mass flow rate:

Multiphase flow rate:

Multiphase flow velocity:

Multiphase flow rate meter:

Multiphase fraction meter:

Hydrocarbons in the liquid state at the prevailing temperature and pressure

Oil-continuous multiphase flow:

Phase area fraction:

Phase flow rate:

Phase mass fraction:

Phase volume fraction:

The ratio between two phase velocities.

Superficial phase velocity:

Volume flow rate:

Water-continuous multiphase flow:

Water cut (WC):

Water-in-liquid ratio (WLR):

3.2 Terms related to metrology

The accuracy of multiphase meters should be specified by terms which are in

Some of the definitions of BS 5233 which may be particularly relevant to multiphase