You are on page 1of 12

MULTIPHASE FLOW MEASUREMENT

Class # 8110

Robert A. Webb, PE
Advisor, Measurement and Allocation
BP Exploration and Production
200 Westlake Park Blvd.
Houston, Texas USA

Abstract
Recognition of various real cases of multiphase flow are illustrated. Multiphase flow dynamics within pipes is
related to various flow regimes encountered. Primary measurement using Venturi meters is highlighted and
various associated or combined sensing technologies are examined. Techniques for both liquid and gas
dominant flows are examined. Attention is given to researchers in the area of wet gas flow measurement using
differential pressure meters. Techniques discussed include gamma ray densitometry, dual energy gamma
spectroscopy, cross correlation, wet gas differential pressure meter over-read, dual differential pressure, tracer
dilution, and partial separation.
Introduction
What is multiphase measurement? Although I know of no official definition, the definition I like to use is:
…measuring a flow that does or will consist of both gas and liquid components when
that flow reaches the end of the associated process...
Based on this definition multiphase flow measurement is a rather comprehensive subject for all oil and gas
production applications. This paper will attempt to introduce the subject of multiphase flow measurement. The
various applications of this technology will be illustrated as well as the flow dynamics and physical flow regimes
presented by multiphase flow. The paper discusses some of the basic engineering parameters used to describe
and qualify this type of flow as well as the measurement performance and uncertainty that is generally present.
Finally the various techniques and sensor arrangements used for both liquid dominant and gas dominant flows
are given. Ultimately this paper serves only as a short introduction to multiphase flow measurement. The subject
matter is rather complex and will require much more in-depth discussion to provide a more firm basis of
understanding.
Applications
Before looking at the technical details of multiphase flow measurement, one should understand the applications
which are conducive to this type of measurement. Generally, these applications are in the far upstream portion of
the oil and gas business. As shown below, multiphase measurement is used all along the flow path until the gas
and liquid streams have been fully processed and are in a suitable stabilized single-phase condition, which is
when reference metering can be performed. Along the way multiphase measurement applications include
wellhead, flow line, and separator outflow applications. Below are more details on these typical applications for
multiphase measurement.

The most obvious application for multiphase metering is at the wellhead. Seldom do oil or gas well flow in single-
phase condition on their own. Oil wells generally have half or even more of their volume in the gaseous state at
the wellhead flow conditions (e.g. below bubble point pressure). And, even the driest of gas wells generally has
enough free liquid in the stream to discount single-phase measurement techniques. Lastly, water is an inevitable
component of the flow which represents a third, and generally the most difficult phase to measure. Measurement
at the wellhead also represents the most economical advantage to the operation. If wellhead measurement is
accomplished, a constant stream of data which is directly related to the most valuable asset of the production
company, the source (i.e. the reservoir), is also made available. This data when transformed into knowledge
about the reservoir is of truly great value since it represents an opportunity to increase the actual recovery of the
total oil or gas in place. Additionally, provided enough wellhead meters are installed (i.e. on all the wells in a field)
there may be a case to eliminate or at least reduce the need for separation equipment and the entire associated
infrastructure. Eliminating separators (especially in high pressure cases) carries a significant cost reduction
opportunity in both capex and opex. Eventually this results in the ability to drill more wells and further recover
more oil and gas.

864
Figure 1 - Applications of Multiphase Flow Measurement

However, as shown in Figure 1 the wellhead applications are not the only place for multiphase measurement.
Within a gas and oil gathering system, often due to economics a multiphase flow line is used. In these cases, and
even when a test separator is used for testing individual wells, the combined flow from a well pad or a set of wells
tied into a common flow line, needs to be metered. Subsea tiebacks are good examples of this scenario. Often
several wells in generally the same area are tied into a flow line that transports the combined production to a host
facility on the surface of the ocean. Furthermore, there are multiple flow lines arriving simultaneously at the host.
Therefore, it becomes important to measure the arriving production form each flow line prior to commingling.
These measurement points are obviously in a multiphase flow condition. This scenario also holds true for
onshore operations when a series or network of production wells at well pads sites are all directed to a central
processing facility. The arrangement creates a nodal network where the output of each well pad is a node. In
order to effectively allocation the total production from the central plant, one must understand the input to the
system from each node. While it is also possible to use periodic well test measurement and sum all the
theoretical quantities that make up the well pad total flow, using a continuous multiphase flow measurement
eliminates the inaccuracies associated with the periodic (i.e. one point in time) type measurements.
Lastly, one must consider the case of separator outflow measurement (i.e. well testing, inlet separator, or intra-
process measurement). According to the definition given above for multiphase flow measurement, separator
outflow measurement also qualifies. Separator outflow measurement is generally treated as single-phase
measurement. However, one should consider it a special case of multiphase flow measurement. Since the fluids
are not fully processed, the liquids will eventually flash and turn into a gaseous and liquid component. Likewise,
the gas outflow will be partially free liquids and will under go condensation and re-gasification along the way.
Also, the liquids may have significant water fractions which will be expelled from the flow. Since the separator
outflow measurements are only truly useful in comparison to post-process measurement (e.g. sales points /
pipeline entry points) there is at least the phase behavior component of multiphase flow measurement built into
the procedure. Furthermore, often the separator outflows are truly not in single-phase conditions at the metering
point. This is especially true for partial separation techniques where some two-phasing is expected and
anticipated. Finally, regardless of where the measurement takes place eventually all the multiphase
measurements are judged or allocated against a single-phase reference measurement.
Multiphase Flow
In order to begin to understand multiphase flow measurement, one must first understand multiphase flow in pipes.
The way gas and liquid fluid components interact within the confined space of the pipe itself is a complex subject
but can be broken down into some basic parameters called “flow regimes”. There various flow regimes range
from 100% liquid to 100% gas and are associated with the pipe orientation (i.e. horizontal versus vertical) due to
the ever-present effects of the earth’s gravity. Figure 2, below illustrates this.

865
Figure 2 - Flow Regimes

The most common engineering parameter or tool in multiphase flow is the flow pattern map.

Figure 3 - Flow Maps

Provided with the superficial liquid velocity (X axis) and the superficial gas velocity (Y axis), the anticipated flow
regime can be found on a flow pattern map. Superficial velocity means the velocity of the liquid or gas component
if it were traveling in the pipe by itself. Flow pattern maps help engineers understand the how variable the flow

866
regime will be over the life of the production well, or through all the operating conditions of a flow line. Obviously,
the measurement will be more difficult if the flow trajectory of the flow crosses a flow regime boundary (for
example from wavy flow into annual mist flow). As will be discussed later, most multiphase measurement
techniques are valid only over one or a finite subset of flow regimes. Migration from one flow regime to another or
even worse a frequent cycling across multiple flow regimes (e.g. slugging), are especially difficult measurement
problems.
Two other important parameters in multiphase flow are “slip” and “gas volume fraction”.

Figure 4 - Constant GVF Example

Since the liquid and gas phases of the flow possess greatly different physical properties (most notably, density),
they travel within the pipe at different rates and the liquid almost always travels slower than the gas (there are
rare occasions in vertical downward flow where the reverse may be true at least momentarily). Figure 4 illustrates
two different cases with the same gas volume fraction, known as “GVF”, but two different slip factors. The result
is two very different liquid/gas combinations within the pipe section. GVF, which is a primary parameter in
multiphase flow, is the fraction of the flow quantity in gas phase as compared to the total. It is also the reciprocal
of the fraction of the flow quantity in liquid phase. The illustration shows two cases, one with slip and the other
with no slip. Slip is the relative velocity of the liquid phase as compared to the gas phase. No slip means both
phases are traveling at the same velocity. In the case of no slip, the pipe cross-section is two-thirds filled with gas
(i.e. in Figure 4 the void fraction of 67% equals the GVF). However, in the case with slip, since the liquid fraction
is traveling slower than the gas, and the liquid occupies a greater fraction of the cross section than would be
expected (i.e. the void fraction is less than the GVF). With no slip (i.e. 100% gas flow or a 100% liquid flow) the
measurement capabilities are very good. However, as the flow becomes multiphase, slip enters and the flow
becomes more and more chaotic.
In the multiphase flow universe (100% liquid to 100% gas) the uncertainty of the measurement increases the
middle of the spectrum when there is a great deal of interaction between the liquid and gas phases and there is a
great deal of slip. Furthermore, the water fraction may be relatively high or low which also greatly influences the
uncertainty of the results. Ultimately the plus or minus uncertainty of multiphase metering can be illustrated over
the GVF range. As shown below in Figure 5, the uncertainty is best at in single phase conditions at either zero or
100 percent GVF for liquid and gas flows respectfully. As there is a mixture of gas and liquid (i.e. GVF values
between 0 and 100%) the uncertainty increases. In the illustration it shows a point in the middle range between
85 and 90% where the uncertainty for either liquid dominant or wet gas techniques is not practical. This
illustration does not provide actual values for the uncertainty, which would likely be a point of debate between
manufactures and users. And, while the illustration is a generality based on the author’s opinion, it is logical to
assume that somewhere in the middle of the GVF range the uncertainty will increase. Whether it becomes
impractical may be the question. Another important point is that water represents a phase that may or may not be
present at any given GVF. Therefore the uncertainty is not a line as much as an area or envelop. Lower water
fraction normally relates to lower uncertainty. Finally, and another important point, is that the illustration shows
that liquid dominant techniques and wet gas techniques do not have uncertainty envelops that overlap within the
range of practicality. There exists a gap between the two techniques. Effectively this indicates that there are

867
regions where there may be no practical multiphase metering solutions (e.g. 85%-95% GVF). This is where most
of the R&D efforts are currently concentrated today, which indicates the recognition of a gap. Last, please note
the illustration is this author’s opinion and others may depict the relative uncertainty of multiphase metering
differently. Also, various flow abnormalities (which are not depicted here) can cause gross errors in the
measurement and high uncertainty anywhere in the flow spectrum and even where one would expect reasonable
performance.

Figure 5 - Uncertainty Over GVF Range

Multiphase (Liquid Dominant) Flow Metering Techniques


Liquid dominant multiphase flow metering is normally called, simply “multiphase” metering and is distinct from the
gas dominant techniques, which are normally called “wet gas” metering. Therefore, the term multiphase metering
is used to refer to liquid dominant flow. Liquid dominant does not mean however that there is always a
predominance of liquid over gas in the flow. It means that the flow techniques are designed to operate from the
liquid end of the spectrum where there is 100% liquid flow. As shown in Figure 5, the actual gas fraction of the
flow reaches into the high 80% range before the techniques fall short from an uncertainty perspective.
The basic premise of most (if not all) the currently available multiphase metering techniques is basically three
equations and three unknowns. The unknowns are gas, oil and water flow. The equations take various forms
and normally involve one or two very basic equations involving the fraction of each phase (α) and the density (ρ)
as follows:

α oil + α water + α gas = 1 and, ρ mix = α oil ρ oil + α gas ρ gas + α water ρ water
These two basic equations plus other various sensing techniques allow the three separate phases (oil, gas, and
water) to be worked out. However, all current commercial meters in this area still rely heavily on modeling of flow
behavior. Since ultimately these models, in turn, heavily depend on flow lab testing, they are only as good as flow
lab’s creation of realistic flows that are encountered in actual operation.
DP Meters
Most multiphase meters use a Venturi meter or other DP device as a basic means of determining the total flow.
This measurement combined with densitometry, cross-correlation, and watercut measurements provides the
basis of the three phase measurement. The Venturi meter is a very well described device and while its reaction
or basic discharge coefficient is not totally understood when it experiences multiphase conditions and thus a
rapidly changing Reynolds Number, it still serves well as a total flow measuring tool.

868
Figure 6 - Classical Venturi Tube

Measurements with Nuclear Devices


There are two basic uses of nuclear devices in multiphase meters; high energy and low energy gamma ray
nuclear devices. The high energy style is a straightforward technique. With this technique a stream of gamma
rays is passed through flow (normally perpendicular to the flow) and is detected with a photon counter on the
opposite side. The material in the flow absorbs some of the gamma rays and blocks them from passing through
to the other side thus reducing the photon count. The denser a material is the greater the attenuation of the
gamma ray beam will be. When the photon count is compared to the count at a reference point such as an empty
pipe (i.e. air) the average density of the passing stream is determined. Typically, a cesium (137Cs) source is used
and is of high enough energy to pass through the pipe wall thus resulting in a true non-intrusive measurement.
The low energy nuclear devices work at a lower energy level as well as on two separate energy levels. This is
called multiple energy gamma ray spectroscopy. By using a source that emits gamma rays with two or more
different energies a meter can use the attenuation of these distinct spectral lines as input to a model of the
multiphase fluid and sense the fractions (oil, gas, and water) or, essentially, composition. Composition
measurement by gamma ray absorption is a well-established technique in multiphase metering.
Gas
I (EHIGH)

20% GVF
Mix

Oil

Wat 50% WC
I (ELOW)

Figure 7 - Solution Triangle

Like the first case (high energy) the lower/dual energy gamma rays are attenuated as they pass through the flow.
However, the level of attenuation of the two energy levels differs, depending upon the fractions of water, oil and
gas present. Therefore, one can calculate fractions of oil, water and gas based on the attenuation of the two
different gamma energy levels. In this technique the gamma-ray absorption must be measured at a minimum of
two distinct energies. When shown graphically on an X-Y plot for the two distinct energy levels, three distinct
points are determined; namely 100% oil, 100% gas, and 100% water respectively. In operation the meter will
yield a result within the bounds of these three points (sometime called the solution triangle). Provided that all
three fractions must also add to unity, one can determine a single solution for the three fractions.

869
A final point regarding this technology is that it works best with a source that emits low energy X-rays. However
this limits the diameter of the meter because of their greater absorption, and requires a special low-density
window in the meter to allow the X-rays to reach the detector. Thus, this technique is intrusive to a point. Any
adherence of material to the windows will result in errors. Also, this technique depends on understanding the
composition (primarily density) of each component of the flow (i.e. oil, gas, and water). The oil and gas properties
and changes in the properties over time (e.g. shifting GOR as reservoir pressure decreases) is fairly well
understood. However, water is a different story. Changes in the salinity of the produced water will shift the water
point on the solution triangle. Thus, and especially in water flood operations, this may be a real problem since the
injected water may (probably will) eventually find its way to the well bore and shift the water salinity. Without prior
knowledge of exactly when this will happen, the multiphase meter will simply indicate a change in apparent water
fraction. Or, in other words, the water point on the solution triangle will shift unbeknownst to the operators.
Here is an example of a multiphase meter which uses a Venturi meter in combination with a dual energy (low
energy) gamma spectroscopy device.

Figure 8 – Multiphase Meter

Cross-Correlation
Cross correlation is a technique of using two sensors that detect a particular fluid parameter such as electrical
conductivity. When the sensors are placed a known distance apart the time of travel can be used to determine a
stream velocity. The signal signature from the sensors, as opposed to the measurement value, is what is
important in this technique. For example, using a high frequency sensor a shift in conductivity may be sensed at
the first sensor. This same shift (or signal signature) is sensed a fraction of a second later at the second sensor.
This graph illustrates the technique.

Figure 9 - Cross Correlation Example

The time difference (Δt) over a known distance relates to the velocity. However, depending on the flow regime
and the sensing technology used this velocity probably will not represent the average stream velocity. Thus, slip
assumptions and equations will likely also be involved to find a mixture velocity and the resulting phase velocities.

870
Water Fraction
Water fraction is one of the most difficult problems in multiphase measurement. First, water fraction ranges from
nothing to nearly 100% at times. Water possesses much different properties than hydrocarbon. While these
different properties sometimes help they also sometimes hurt the chances of effective measurements. For
example, the dielectric constant for water is many times higher that that of oil or gas. This large difference allows
for certain permittivity measurements to be translated into a water fraction measurement. Conversely, salt is
highly soluble in water and the salinity level of the produced water may change. This, in turn, changes the
permittivity characteristics of the water and degrades the water fraction measurement. However, one water
fraction (or watercut) measurement technique appears not to be influenced by the water chemistry itself. That
technique is optical in nature and uses the relative absorbance of near infrared light. Following is a short
synopsis of the various water measurement techniques in addition to the multiple energy gamma ray
spectroscopy technique described above:
Permittivity
The measurement of permittivity (relative dielectric constant) is a means of estimating the aqueous phase(s) of a
multiphase stream. In particular, permittivity measurement using capacitance or microwave sensors is a common
means of estimating water-cut or water fraction in oil-continuous or wet gas flows.
One type of permittivity measurement is capacitance. Capacitance meters can offer value to a wet gas or
multiphase metering system, as they respond to differences in permittivity of the fluid, primarily an indicator of
water. They can alert a systems operator when a gas flow has become a wet gas flow. A further use of the
capacitance meter is to give watercut estimations for a wet gas flow with an oil / water liquid component.
However, this technique can be compromised by a water-continuous flow regime, in which case some meters
switch modes where they measure conductivity rather than permittivity.
Microwave propagation is another common approach. The technology of microwave propagation has been used
principally for watercut measurement. Like capacitance, it makes use of the fact that the RF permittivity is very
different for oil (2.5) and water (80). Manufacturers have utilized this difference in a variety of implementations to
measure the watercut of a water/oil mixture. Unlike capacitance, microwave approaches do work in water-
continuous as well as oil continuous flow, albeit a challenging measurement, especially in the region of the
inversion point. Also, at least one wet gas meter uses a microwave approach where it combines the microwave
technology for water measurement with an inverted Venturi or a classical Venturi for flow measurement.
Conductivity
In some cases of multiphase flow, the amount of water is great enough that it is the dominant liquid phase. In
these instances, permittivity sensors such as those mentioned above may have difficulty dealing with a
conductive medium in the space where the measurement is to be made. Some meters therefore employ
inductive methods to measure the bulk conductivity of the fluid rather than trying to estimate its permittivity.
Wet Gas Flow Metering Techniques
Wet gas flows are flows that generally stay within the annular mist flow regime. However, there are cases where
wet gas flows stratify and waves occur or even a degree of slugging. The most popular means of measuring a
wet gas flow is to use a differential pressure (DP) meter and adjust the DP reading for the inherent over-read that
occurs due to the liquid. Since a DP meter is a momentum meter and liquid is much denser than gas, just a small
amount of liquid creates significantly more momentum. Thus, the DP device when viewed from a single phase
gas perspective, “over-reads” the DP pressure. By adjusting for the over read, the gas measurement can be
made correct. However, this technique requires information about the liquid amount and density (normally called
“liquid loading”) traveling in the gas flow. Therefore, there are various techniques used to supply information about
the liquid loading. The easiest and often acceptable means is by simply sampling the stream and assuming the
liquid loading present at that point in time is constant. Therefore with this simple technique, the over read
correction is all that is needed.
Wet Gas Flow Parameters
In order to understand wet gas flow one needs to understand two parameters; the Lockhart Martinelli parameter,
(XLM), and the densimetric Froude number (Fr). XLM is a dimensionless parameter that compares the inertia of the
liquid to the inertia to the gas. The Froude number is also dimensionless and relates gas or liquid inertia to the
liquid gravity force. Thus, there is a liquid Froude number and a gas Froude number. The following formulae
apply:

871
A
m ρg Usg ρg UsA ρA
XLM = , Frg = ; FrA = ,
g
m ρl gD ρA − ρg gD ρA − ρg

g
m mA
where the superficial gas and liquid velocities are : Usg = ; and UsA = . Also in the above equations, g
ρg A ρA A
is the gravitational constant, D is the meter inlet diameter and A is the meter inlet cross sectional area.
The wet gas flow regime is generally considered to extend up to a XLM of 0.3. For liquid loading above that point
the flow regime can generally not support an annular mist flow.
DP Pressure (DP) Device Over-Read
Over the years four distinct sets of work in this area have been accomplished by, Murdock (1963), James (1965),
and Chisholm (1967-1977). These researchers worked with orifice meters. DeLeeuw (1997) furthered their work
by extending the work to Venturi meters. DeLeeuw’s work is especially significant given that many wet gas
metering systems (especially subsea systems) use Venturi meters. Additionally, the industry has been collecting
a vast database from wet gas flow loop testing globally over the last several years. The research of these
individuals and the continuing industry work has generally culminated in an understanding of the DP device over-
read which is depicted as follows:

50
Frg 1.59 Uncorrected
Frg 3.24 Uncorrected
40 Frg 4.06 Uncorrected Homogeneous model
Frg 4.89 Uncorrected
Frg 5.82 Uncorrected
30 Homogeneous Model
% Error

20

10

-10
0.00 0.05 0.10 0.15 0.20 0.25 0.30

Lockhart-Martinelli Parameter

Figure 10 - DP Wet Gas Over-Read

This shows the typical over-read for a DP device (in this case an orifice meter but the same general trends are
true of all DP devices). At a Lockhart-Martinelli value of only about 0.03, the gas error is already at 5%. Note that
many DP devices have been tested and all show the over-read of the same general trend as the orifice meter.
However, from all of these tests the orifice meter seems to be the most repeatable. Of course when attempting to
correct for the over-read, repeatability is very critical.
The mechanics behind the over-read is straightforward. A DP type meter responds to momentum. Liquid is much
more dense than gas. Thus, even a small amount of liquid causes a significant increase in the momentum of the
flow. In the example above the 5% gas error shown in the orifice meter at the XLM of 0.03, is equivalent to only
about 36 barrels of condensate in 1000 MSCF of gas. However, keep in mind this is only an example. The
Lockhart-Martinelli parameter is a comparison of gas and liquid densities which change with pressure. The point
here is that ignoring the effect is not a good option given that most gas wells will experience level of liquid loading
that provides a significant bias error.
As mentioned above four individuals have characterized the over-read within their research over the last 40 plus
years, starting with Murdock. Murdock’s work resulted in the relationship:

872
 g,apparent
m
g =
m . Murdock Equation.
1 + 1 .26 XLM
Applying this as a correction brings the apparent gas measurement back in agreement with the actual amount of
gas flowing. The other researchers refined this approach over the years to yield a more comprehensive equation,
which I will not go into in this paper. However, below is an illustration of the orifice meter data shown in Figure 10
with the various orifice meter correlations (i.e. Murdock, James, and Chisholm). Also shown is a homogeneous
model where the wet gas flow is assumed to be truly a homogeneous mixture of the gas and liquid.

40
Uncorrected Data Homogeneous model
35
Homogeneous Model
30 Murdock Correlation
James Correlation
25 Chisholm Correlation

20

15
% Error

10

5 +2%

-5 -2%

-10

-15

-20
0.00 0.05 0.10 0.15 0.20 0.25 0.30

Lockhart-Martinelli Parameter

Figure 11 - Orifice Wet Gas Over-Read Corrections

Dual DP Technique
As stated earlier unless the liquid loading is known, one cannot use the over-read to make a correction to the gas
measurement. Therefore a means to also measure the liquid is needed for a totally self-contained metering
system. One approach to this has been to use two DP devices in sequence where each responds to the liquid
loading differently and demonstrate an over-read curve of differing slopes. Even though both devices are
responding to the same phenomenon and thus to the same equation(s) since the slope of the over-read lines is
greatly different they are treated as a case of two equations and two unknowns.

Figure 12 - Dual DP Technique

One problem with this technique is that at higher pressures when the gas and liquid densities get closer together,
the slope of the two over-read curves gets closer. This causes great difficulty in discerning the liquid portion

873
(especially at low liquid loading) which in turn influences the gas measurement. Once the liquid loading increases
there is a better ability to discern liquids.

Tracer Dilution Technique


Another approach is to introduce a substance that when collected later in a sample will reveal through calculation
and measuring the concentration levels, the liquid loading in the stream.

Figure 13 - Tracer Dilution Technique

Of course this technique is manpower intensive and may not be conducive to all applications such as very remote
locations or subsea.

Compact – Partial Separation Techniques

Figure 14 - GLCC Compact Separation

Another area that qualifies as a special type of multiphase meters are systems that utilize compact or partial
separation. Normally these include a cyclonic style separation vessel coupled with multiphase or (more often)
two-phase tolerant gas and liquid meters. Since the stream after separation is divided into a most gas stream (i.e.
wet gas) and a mostly liquid stream (i.e. bubbly liquid), more traditional metering solutions may be effective. Due

874
to their nature the styles of meters may be either for wet gas or multiphase, but often there are limitations in flow
regime range due the effectiveness of the separation.
These systems may utilize Coriolis meters for gas and liquid, PD meters for liquid, typical DP type or vortex
meters for gas, and other combinations of common meters. As shown below, some are as simple as a GLCC
(Gas Liquid Compact Cyclonic) separator and the appropriate metering, provided that the gas and liquid metering
can handle the two-phase flows that they will be subjected to. Other designs are much more compact and
marketed directly as a multiphase meter.

Concluding Statement
What has been discussed in this paper is an introduction to multiphase measurement. It, admittedly, has
concentrated on multiphase and wet gas meters and of those devices, only what is commercially available today.
One must realize that multiphase measurement is more than just the meter. Sampling and phase behavior is a
big part of the science.
Multiphase flows can be highly dynamic with wide dynamic range. Many of the measurement techniques used
today are founded in single phase flow metering with stable flow. Applying these measurement techniques in the
dynamic (unsteady) world of multiphase flow is no trivial engineering problem. Furthermore, multiphase metering
applications are inherently closer to the well head than traditional measurement applications. Therefore, a variety
of other complications such as emulsion forming and associated high and difficult to predict viscosity shifts,
foaming situations, scale or Asphaltenes deposits, etc. However, even with the variety of difficulties in multiphase
measurement, it is an essential part of hydrocarbon metrology. The operational and capex advantages that are
made available through multiphase measurement are often very significant and ultimately results in higher
reservoir recovery factors. In fact, various forms of multiphase measurement, including separator outflow
measurement of non-stabilized fluids, such as in well testing, is the first measurement in arguably the most critical
asset of the oil and gas industry; the actual oil or gas reservoir.
Today the multiphase measurement and metering technology is still a rapidly changing area. There are normally
multiple development projects in progress in this area at any one time. Thus, the techniques of today may be
quickly replaced with new innovations. Therefore, when a multiphase measurement issues arises it is always
best to get multiple (unbiased) technical opinions and firmly understand the measurement uncertainty before
proceeding.
Finally, due to the introductory nature of this paper and in order to be brief, no references have been given.
However, several papers on this subject are in the public domain and those interested should seek them out.

Acknowledgements
The following individuals contributed materials to this paper and/or provided technical review:

• David Bromley – BP North America Gas

• Andrew Hall – BP E&P Technology Group

• Bill Priddy – BP E&P Technology Group

• Robbie Lansangan – BP Gulf of Mexico

• Chip Letton – Letton-Hall Group

• Richard Steven – Colorado Experimental Engineering Station, Inc. (CEESI)

875

You might also like