Document Sample
Downhole_Fiber-Optic_Multiphase_Flowmeter Powered By Docstoc
					                            Downhole Fiber-Optic Multiphase Flowmeter

This article, written by Technology Editor Dennis Denney, contains highlights of paper SPE 77655, "Downhole Fiber-Optic
Multiphase Flowmeter: Design, Operating Principle, and Testing," by Tor K. Kragas, SPE, F.X. Bostick III, SPE, and Christopher
Mayeu, Weatherford Completion Systems; Daniel L. Gysling, SPE, CiDRA Corp.; and Alex van der Spek, Shell Intl. E&P B.V.,
originally presented at the 2002 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, 29 September-2 October.

Real-time, downhole, multiphase flow-rate data are of significant value for
production optimization, particularly for high-cost deepwater developments and
complex multilateral wells. Continuous measurements of downhole flow rates in
producing wells generally have not been available because of difficulties in
designing and deploying measurement devices. The design, operating principles,
and flow-loop testing of an all-fiber-optic multiphase flowmeter for downhole
deployment are presented.


Traditionally, permanent monitoring has primarily referred to downhole pressure and
temperature measurement. However, development and deployment of accurate and
reliable downhole flowmeters has extended the types of in-well measurements
available to fluid flow rates and phase fractions. Applications of downhole flow data
include the following.

         Zone production allocation.
         Identifying and localizing production anomalies.
         Directly determining productivity index.
         Commingling production.
         Reducing the number of surface tests.
         Reducing the surface facilities.

Fig. 1 shows hydrocarbon flow regimes that may occur in the wellbore. Although the
downhole environment imposes significant design challenges for hardware
manufacturers, the higher pressures and lower gas-volume fractions (GVFs) that
exist downhole typically result in conditions that are more conducive to multiphase
flow measurement than conditions at the surface.
Fig. 1--Influence of pressure on GVF and flow regime for oils with varying gas/oil
ratio (GOR).

Flowmeter Design

A permanent, downhole fiber-optic-based flowmeter was developed to provide real-
time measurements of pressure, temperature, flow rate, and phase fraction
(holdup). It uses an array of Bragg grating sensors that enable multiple all-optical
measurements, multiplexed on a single fiber-optic cable. The meter contains no
downhole electronics, is nonintrusive, and provides full-bore access. It is deployed
during completion as part of the production-tubing string.

The flowmeter system comprises two modules. The upper assembly is a gauge
carrier that houses a fiber-optic pressure and temperature transducer. The lower
assembly contains the optical flow and phase-fraction sensors. The pressure and
temperature transducer and the flowmeter sensors can be interrogated with a single
optical fiber.

The flowmeter assembly consists of an inner sensor tube, an outer sleeve, fiber-optic
flow and phase-fraction sensors, and an optical bulkhead connector. The sensor
contains a smooth inner bore with no intrusions or pressure ports in the flow path.
The optical sensors are on the outside of the sensor tube and encapsulated within a
hermetically sealed annular cavity by welding an outer sleeve to two upsets on the
sensor tube.
Operating Principle

The flowmeter makes two fundamental measurements on the flowing fluid stream:
bulk velocity and speed of sound through the fluid mixture. From these
measurements, together with knowledge of the densities and speeds of sound of the
individual phases at the measured temperature and pressure, the individual-phase
flow rates are determined in two-phase flowing systems.

Speed-of-Sound Measurement. To measure the speed of sound of the fluid mixture,
the meter uses unsteady pressure measurements to “listen” to the propagation of
production-generated noise in the production tubing across an array of optical
sensors. The noise can be from any production-associated source, including flow
through perforations or downhole choke valves, gas-bubble breakout, electrical
submersible pumps, and gas-lift valves. No artificial noise source is needed. The
unsteady pressure measurements are obtained at multiple locations in the meter and
provide sufficient distance and time resolution to determine the speed of sound
through the produced fluid.

Bulk Velocity. The fiber-optic flowmeter uses a cross-correlation technique to
determine the bulk fluid velocity. Cross-correlation techniques rely on axially
displaced measurements of a time-varying fluid property that convects with the flow.
Ideally, the signal from the downstream transducer is a time-lagged version of the
signal from the upstream transducer. Determining this time lag between axially
displaced signals determines the flow velocity, from which the volumetric flow rate is
derived. This system cross-correlates convective pressure disturbance and has been
demonstrated to be equally applicable to single-phase and well-mixed multiphase

Phase Fraction. The flowmeter uses the measured sound speed to determine the
phase composition of two-phase mixtures. The speed of sound for the mixture can
be related analytically to the sound speeds and densities of the individual

This relationship, with some restrictions, holds in general for two-phase gas/liquid
and liquid/liquid mixtures and can be generalized for multiphase mixtures. The
mixing rule provides a first-principles, physics-based relationship linking fluid-
mixture sound speed to volumetric phase fraction through the sound speeds and
densities of the individual phases. The meter relies on the contrast in sound speed to
distinguish the two phases.

Component Flow Rate--Well-Mixed Flows.

The phase fraction determined from the mixture’s sound speed is a measure of the
volume fraction in the meter (i.e., measure of phase holdup). The holdup may or
may not be the same as the flowing volume fraction of the particular phase
depending on whether there is appreciable slip of one phase relative to the other. For
oil/water mixtures without appreciable slip, a homogeneous-flow model can be used.

Fig. 2 shows an overview of the flow-rate determination process. Despite its
simplicity, flow-loop testing has shown that the sound-speed and volumetric-flow
measurements are sufficiently robust and representative to apply this model (within
specified accuracy) to most oil/water flows with mixture velocities greater than 1

Fig. 2--Simplfied flow-rate determination process.

For flow regimes with appreciable slip, a more sophisticated slip model must be used
to interpret the quantities measured by the meter and determine the flow rates of
the individual phases. Factors affecting slip include the total flow rate, relative phase
flow rates, inclination, and fluid parameters. Results from flow-loop testing indicate
that the long-wavelength speed-of-sound measurements from the meter are capable
of determining the in-situ volumetric phase fraction (holdup) for oil/water mixtures,
independent of flow nonhomogeneities, such as those in stratified flows.

Flow-Loop Testing

The flowmeter was tested under a variety of conditions with a variety of fluids at
several flow-loop facilities. Shell’s multiphase flow-loop facility has an articulated
boom that allowed testing the flowmeter from vertical to slightly beyond horizontal
positions. A total of 143 data points from a broad range of flowmeter orientations
was analyzed. The mixture velocity was within the ±5% absolute error with respect
to the reference measurements for the majority of the points.

It should be noted that the reference measurement is the water cut determined by
monitoring the input oil and water flow rates. The meter’s water-phase fraction is a
holdup, or in-situ phase-fraction measurement. Slip between the two phases
manifests itself as a discrepancy between input and in-situ phase fraction. The good
agreement between the two measurements over the broad range of parameters
investigated verifies the measurement technique as well as the homogeneous model.
Because these are liquid-only tests, the mixture velocity is equivalent to the
superficial liquid velocity. The flowmeter measured the total volumetric flow rate
within plus or minus 5% for most of the test points.

Flowmeter Installations

The first field installation of the fiber-optic flowmeter was in October 2000, in Shell’s
Mars A-18 well in the Gulf of Mexico. The meter was installed at a total depth of
6442 m in 896 m of water. The installation was performed as scheduled, with no lost
rig time. The meter has provided valuable pressure, temperature, and volumetric
flow-rate data during well startup that was in good agreement with test separator
reference measurements. Performance of the flowmeter was within specifications
and has exceeded expectations.

A field trial of the flowmeter was conducted in June 2001, in two wells at the Nimr
field in Oman. In one well, the meter was installed in the completion string just
below an electrical submersible pump (ESP), and in the other well, it was installed
just below a beam pump. A surface Coriolis mass flowmeter served as a reference
for the 2-week field trial.

In the beam-pumped well, intermittent flow and relatively high acoustical- and
structural-noise environments characteristic of beam pumps were not compatible
with passive listening techniques used to monitor flow rates. A comparison against
the surface-measured rates was not possible. In the ESP well, comparison of the flow
rates and water cut measured by the fiber-optic flowmeter and the Coriolis meter
were better than expected, plus or minus 2% for water cut and plus or minus 3% for
total rate, and well within specifications.

Source :,1513,1648_2300_6464101_0,00.html
Compiled by : Nugroho Wibisono (