Note: Descriptions are shown in the official language in which they were submitted.
CA 02485824 2002-08-08
SELF-CALIBRATED ULTRASONIC METHOD OF IN-SITU
MEASUREMENT OF BOREHOLE FLUID ACOUSTIC PROPERTIES
BACKGROUND OF THE 1NVF;NTION
Field of the Invention
The present invention relates generally to oil well logging and monitoring.
More particularly,
the present invention relates to determining the acoustic properties of a
borehole fluid.
Description of the Related Art
To recover oil and gas from subsurface formations, wellbores or boreholes are
drilled by
rotating a drill bit attached at an end of a drill string. The drill string
includes a drill pipe or a
coiled tubing that has a drill bit at its downhole end and a bottom hole
assembly (BHA) above
the drill bit. The wellbore is drilled by rotating the drill bit by rotating
the tubing and/or by a
mud motor disposed in the BHA. A drilling or wellbore fluid commonly referred
to as the
"mud" is supplied under pressure from a surface source into the tubing during
drilling of the to
wellbore. The drilling fluid operates the mud motor (when used) and discharges
at the drill bit
bottom. The drilling fluid then returns to the surface via the annular space
(annulus) between
the drill string and the wellbore wall or inside. Fluid returning to the
surface carries the rock
bits (cuttings) produced by the drill bit as it disintegrates the rock to
drill the wellbore.
A wellbore is overburdened when the drilling fluid column pressure is greater
than the
formation pressure. In overburdened wellbores, some of the drilling fluid
penetrates into the
formation, thereby causing a loss in the drilling fluid and forming an invaded
zone around the
wellbore. It is desirable to reduce the fluid loss into the formation because
it makes it more
difficult to measure the properties of the virgin formation, which are
required to determine the
presence and retrievability of the trapped hydrocarbons. In underbalanced
drilling, the fluid
column pressure is less than the formation pressure, which causes the
formation fluid to enter
into the wellbore. This invasion may reduce the effectiveness of the drilling
fluid.
_._._ _ ~ _ ,
CA 02485824 2002-08-08
A substantial proportion of the current drilling activity involves directional
boreholes (deviated
and horizontal boreholes) and/or deeper boreholes to recover greater amounts
of hydrocarbons
from the subsurface formations and also to recover previously unrecoverable
hydrocarbons.
Drilling of such boreholes require the drilling fluid to have complex physical
and chemical
characteristics. The drilling fluid is made up of a base such as water or
synthetic material and
may contain a number of additives depending upon the specific application. A
major
component in the success the drilling operation is the performance of the
drilling fluid,
especially for drilling deeper wellbores, horizontal wellbores and wellbores
in hostile
environments (high temperature and pressure). These environments require the
drilling fluid to
excel in many performance categories. The drilling operator and the mud
engineer determine
the type of the drilling fluid most suitable for the particular drilling
operations and then utilize
various additives to obtain the desired performance characteristics such as
viscosity, density,
gelation or thixotropic properties, mechanical stability, chemical stability,
lubricating
characteristics, ability to carry cuttings to the surface duaving drilling,
ability to hold in
suspension such cuttings when fluid circulation is stopped, environmental
harmony, non-
corrosive effect on the drilling components, provision of adequate hydrostatic
pressure and
cooling and lubricating impact on the drill bit and BHA components.
A stable borehole is generally a result of a chemical andlor mechanical
balance of the drilling
fluid. With respect to the mechanical stability, the hydrostatics pressure
exerted by the drilling
fluid in overburdened wells is normally designed to exceed l:he formation
pressures. This is
generally controlled by controlling the fluid density at the surface. To
determine the fluid
density during drilling, the operators take into account prior knowledge, the
behavior of rock
under stress, and their related deformation characteristics, fornaation dip,
fluid velocity, type of
the formation being drilled, etc. However, the actual density of the fluid is
not continuously
measured downhole, which may be different from the density assumed by the
operator.
2
CA 02485824 2002-08-08
Further, the fluid density downhole is dynamic, i.e., it continuously changes
depending upon
the actual drilling and borehole conditions, including the downhole
temperature and pressure.
Thus, it is desirable to determine density of the wellbore fluid downhole
during the drilling
operations and then to alter the drilling fluid composition at the surface to
obtain the desired
density and/or to take other corrective actions based on such measurements.
As noted above, an important function of the drilling fluid is to transport
cuttings from the
wellbore as the drilling progresses. Once the drill bit has crE:ated a drill
cutting, it should be
removed from under the bit. If the cutting remains under tlae bit it is
redrilled into smaller
pieces, adversely affecting the rate of penetration, bit life and mud
properties. The annular
velocity needs to be greater than the slip velocity for cuttings to move
uphole. The size, shape
and weight of the cuttings determine the viscosity necessary to control the
rate of settling
through the drilling fluid. Low shear rate viscosity controls the carrying
capacity of the drilling
fluid. The density of the suspending fluid has an associated buoyancy effect
on cuttings. An
increase in density usually has an associated favorable affect on the carrying
capacity of the
drilling fluid. In horizontal wellbores, heavier cuttings can settle on the
bottom side of the
wellbore if the fluid properties and fluid speed are not adequate. Cuttings
can also accumulate
in washed-out zones. Determining the density of the fluid downhole provides an
indication of
whether cuttings are settling or accumulating at any place in the wellbore.
In the oil and gas industry, various devices and sensors have been used to
determine a variety
of downhole parameters during drilling of wellbores. Such tools are generally
referred to as the
measurement-while-drilling (MWD) tools. The general emphasis of the industry
has been to
use MWD tools to determine parameters relating to the formations, physical
condition of the
tool and the borehole. Very few measurements are made relating to the drilling
fluid. The
majority of the measurements relating to the drilling fluid are made at the
surface by analyzing
samples collected from the fluid returning to the surface. Corrective actions
are taken based on
3
CA 02485824 2002-08-08
such measurements, which in many cases take a long time and do not represent
the actual fluid
properties downhole.
SUMMARY OF TIDE INVENTION
The problems outlined above are in large part addressed by a self calibrated
ultrasonic method
of in-situ measurement of borehole fluid acoustic properties. In a preferred
embodiment of the
present invention, a method for determining a borehole fluid property includes
(i) generating an
acoustic signal within a borehole fluid, (ii) receiving reflections of the
acoustic signal from the
fluid, and (iii) analyzing a reverberation portion of the acoustic signal to
determine the property.
The analyzing of the reverberation portion may include obt<~ining a
theoretical reverberation
signal and relating the measured reverberation signal with the theoretical
reverberation signal to
determine the borehole fluid property.
In another preferred embodiment of the present invention, a processor adapted
to provide real-
time estimates of a borehole fluid property includes an input terminal and a
processing portion.
The input terminal receives a data signal corresponding to a reflected
acoustic wave. The
processing portion separates the data signal into a first reflection portion
and a resonance
portion and convolves the first reflection portion response to yield a
theoretical reverberation
response.
In yet another preferred embodiment of the present invention, a tool for
measuring borehole
fluid properties includes a body, an acoustic transducer, and a metal disk.
The body houses the
transducer and metal disk. A borehole fluid enters the tool ahrough an opening
in the body,
flows in between the transducer and metal disk where it is measured, and exits
the tool.
Thus, the present invention comprises a combination of featurea and advantages
which enable it
to overcome various problems of prior devices. The various characteristics
described above, as
well as other features, will be readily apparent to those skilled in the art
upon reading the
4
_.. ____ ~"~~ F . ~,,~.-,~~. _~~ ~..__..__ _.._ro.
CA 02485824 2002-08-08
following detailed description of the preferred embodiments of the invention,
and by referring to
the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
For a more detailed description of the preferred embodiment of the present
invention, reference
will now be made to the accompanying drawings, wherein:
Figure lA is a general schematic showing a tool in a preferred embodiment;
Figure 1B is a cut-away view illustrating component parts of Figure lA;
Figure 2 illustrates waveform reflection and reverberation;
Figure 3 is a graph showing a received acoustic waveform;
Figure 4 is a diagram illustrating the component parts of Figw-e 3;
Figure SA is a diagram of a subterranean system built in accord with a
preferred embodiment;
Figure SB is a diagram of the above ground system built in accord with a
preferred
embodiment;
Figure 6 is a general flow diagram of a preferred embodiment;
1 ~ Figure 7A is a flow diagram of a preferred embodiment; and
Figure 7B is a flaw diagram of a preferred embodiment.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
Figure IA illustrates a general overview of a tool submerged downhole. Shown
are tool 10,
fluid vent 20, formation 30, and well fluid 210. Fluid vent 20 provides a
means for well fluid
210 to enter and exit tool 10. 'While in tool i 0, well fluid 210 is measured
for its acoustic,
properties.
Figure 1 B is a cross-sectional view of the tool showing acoustic measurement
components.
Inside tool I0, where fluid vent 20 is located, are acoustic transducer 200
and metal disk 220..
5
CA 02485824 2002-08-08
As can be seen, well fluid 210 enters tool 10, flows between acoustic
transducer 200 and metal
disk 220, and exits tool 10
Figure 2 illustrates the acoustic wave path and metal disk reverberations for
a downhole
acoustic wave. Shown are acoustic transducer 200, well fluid 210 and metal
disk 220. Well
fluid 210 and disk 220 each has its own impedance, labeled Zm and ZS,
respectively. Also
shown is acoustic signal 250, including firsf reflected portion 260, disk
reverberation portions
271-276 and transmitted wave portions 280, 282, 284 and 286 through the disk
in the same
well fluid.
To measure the reflection coefficient of the well fluid, the <~coustic
transducer 200 sends out
14 acoustic signal 250, which is preferably an ultrasonic impulse with a
characteristic frequency
of about 500 kHz, then switches to the receive mode. The; impulse frequency is
preferably
set at the expected resonance frequency of the disk. The acoustic signal 250
travels through
the well fluid 210 and strikes the disk 220. The largest portion of the energy
of the impulse is
reflected back to the transducer as reflected portion 260 while a small amount
of signal enters
the disk as wave 280. When the well fluid 210 is water, the reflected wave
form has an
amplitude of about 93% of the initial impulse. The portion of the signal that
entered the disk
is reflected back and forth between the disk/fluid interface and the disk/tool
interface, as
illustrated by wave reverberations 271-276. At each reflection some energy is
transmitted
through the interface, dependent on the acoustic impedance contrast, and is
either directed
back toward the transducer or out into the tool. The signal inside the disk is
quickly
dissipated in this manner at a rate directly dependent on the acoustic
impedance of the
material outside the disk according to the equation:
Rl - ~ZI - Z2~ ~ ~Z1 '~-' Z2~ (,1~
where Rl is the reflection coefficient, and Z1 and Z2 are the impedances of
the materials at the
interface in question. In a preferred embodiment, the thickness of the metal
disk is set to one
half of the resonant wavelength of the transducer signal.
6
CA 02485824 2002-08-08
The acoustic transducer 200, now acting as a receiver or transducer, sees a
waveform
consisting of a loud initial reflection followed by an exponentially decaying
reverberation
signal. Figure 3 illustrates the measured acoustic waveform received at the
transducer 200.
If time t=0 is the time of generation of the acoustic wave a.t the acoustic
transmitter, then the
time Tt~" represents the transit time (the time for the travel of this
acoustic wave to the disk
and back to the transceiver). Since the distance is fixed, the transit time
Ttran provides a.n
indication of the acoustic velocity of the fluid. Also shown in Figure 3 are
the Time Offset,
Toff, and the Resonance Window, TW;", both of whose significance is explained
below.
Figure 4 illustrates the individual waveforms; both first reflection and
reverberations, that sum
to provide the waveform of Figure 3. The waveform received by the transducer
is the sum of
the initial reflection waveform with each reverberation waveform, where each
reverberation i.s
delayed an amount proportional to the width of the disk. Further, because the
acoustic
transducer is not a perfect transmitter, it "rings" somewhat upon the
transmission of an acoustic
wave. This transducer "ringing" also is included in the detected waveform, and
may be
accounted for by the present invention.
Figure 5 illustrates a device built in accord with a preferred embodiment.
Shown in Figure
5A is acoustic transducer 200, analog-to-digital converter 500, a processor
510 for recording
start time and gain, waveform compression chip 520, and multiplexer 530.
Waveform
compression chip 520 could alternately be part of a processor. Also shown are
downhole
transmitter 540 connected to multiplexer 530 and telemetry cable 545.
Referring now to
Figure 5B, at the surface are located uphole receiver 550, demultiplexer 560,
transmission
line 564 carrying tool information to processor 590 for a data log 595,
transmission line 570
carrying gain and start time information to uphole processor 590, and waveform
decompression chip 580. Attached to decompression chip 580 is processor 590.
Processor
590 generates data suitable for a log 595.
7
..... _... _ ,. ..w .x~~~..w~~ .~~ --. r>n f . _.
CA 02485824 2002-08-08
Referring now to both Figs. SA and SB, acoustic transducer 200 collects data
of metal disk
reflection and reverberation. This acoustic waveform is digitized by analog-to-
digital
converter 500 and sent to processor 510, which detects the first reflection
from the digitized
signal. Processor 510 then computes the relevant start tirr~e and transit
time. Because the
total waveform data may be greater than the bandwidth capacity of transmission
line 545,
digital compression 520 is preferably performed. Suitable compressions include
wavelet and
ADPCM (Adaptive Differential Pulse Code Modulation) techniques, which work
well for
smoothly varying data. The compressed waveform from digital compression chip
520 is then
multiplexed 530 with the other tool information. Dowruhole transmitter 540
sends this
multiplexed data to the surface. Sending the data to the surface allows
processing by faster,
more sophisticated machinery.
This multiplexed data is received by uphole receiver 550 and is separated into
component
parts by demultiplexer 560. Waveform decompression chip 580 provides the
reconstructed
waveform to processor 590, which also receives start: time information. Upon
the
determination of the reflection coefficient of the well fluid, processor 590
combines with
position information and creates a log 595.
Figure 6 illustrates a general method for the present invention. In block 600,
an observed
waveform is provided uphole far processing. In some embodiments, it may be
desirable to
stack waveforms (block 610). The waveform's transit time I,T,ran) is obtained
in block 620, as
well as the time windows To ff and TW;n. The definition of transit time was
explained above
with reference to Figure 3 and may be easily measured by a first reflection
detector portion of
processor 510. Tuff and Tv,,;n are then selected to obtain a time window TW;"
that contains
reliable reverberation information. Tuff, measured from tile time of receipt
for the initial
reflection, is a time window that encompasses the initial reflection. As such,
its duration is
dependent upon the duration of the acoustic impulse transmitted by acoustic
transducer 200
8
CA 02485824 2002-08-08
and the nature of the drilling fluid. Tuff also preferably accounts for error
introduced because
of the real-world shortcomings of the acoustic transducer (transducer
"ringing"), and thus Toy
may be slightly longer than if chosen theoretically. Nonetheless, To ff is
about 15
microseconds. TW;n is juxtaposed with Tuff and is a time 'window of interest
because TW",
contains reverberation information uncontaminated by the first reflection. The
duration of.
TW;~ should be brief enough so that noise and reverberations occurring in the
tool 10 do not
make unreliable the received disk reverberation waveforms. Nonetheless, so
that a reliable
wave train containing sufficient data is obtained, TW;" preferably includes at
least four
reverberations. Thus, TW;" is about 12.8 microseconds.
The tool calibration may be obtained as follows. First, the reflection
waveforrn defined by
Toff iS transformed to the frequency domain by use of DFT (Discrete Fourier
Transform).
Referring back to Figure 6, proper modeling applied to the first reflection
portion 260, as
defined by Tuff, gives a theoretical prediction of what the reverberation
waveform contained
in TW", should look like. To accomplish this, in block 630 the first
reflection signal is
transformed by Fast Fourier Transform (FFT) into its frequency domain
equivalent. This
yields S(cn). Because the modeling is done in the frequency domain, amplitude
and phase
errors are eliminated. This error elimination simplifies mathematical
processing (and hence
faster processing is obtained):
Alternately, instead of transforming each first reflection individually, to
enhance accuracy, the
first reflections from multiple firings may first be averaged and the result
transformed in block
630 by FFT processing into the frequency domain to yield S(tn). A most
reliable first reflection
average may be obtained by discarding f rst reflections that have amplitudes
above or below a
preset deviation from a moving average of preceding first reflections.
In block 640, a theoretical prediction of the reverberation waves is obtained
by multiplying
(convolution in time domain) the frequency-domain first reflection signal
S(cn) with a
frequency-domain theoretical response equation R(cn) to obtain a frequency
domain version
9
CA 02485824 2002-08-08
X(cu) of the reverberation signal x(t). Assuming a flat mel;al disk, the
theoretical frequency
domain response may be modeled by the following:
4Z",Z~ (ZS - Zm)
R~~)- Zm -ZS + (Zm +ZS)3 e-~2m vs (2)
Zm +zs, Z -i2t~Cr
1- ZS Zm
Z", + ZS
Where
R(o.~) = the reflection coefficient for angular frequency c~
Zm, ZS, = impedances for mud and metal disk, respectively
VS = the speed of sound in the metal disk., and
CT = the thickness of the metal disk.
The above equation assumes that the transducer generates waves having normal
(i.e.,
perpendicular) incidence on the disk. VS, Z5, and CT can be measured very
precisely as basic
physical properties of the metal disk.
In block 640 the frequency domain signal X(co) is transformed back into the
time domain by
1 ~ use of an Inverse Fast Fourier Transform (IFFT). As such, block 640
provides the theoretical
reverberation response x(t) for the observed initial reflection waveform(s) in
the time domain.
This theoretical reverberation response is also a function of the borehole
fluid impedance Zm.
Once the results are converted to the time domain, a relationship is
established between the
theoretical response and the received response. Next, a method is used to
determine the
borehole fluid properties in block 650.
Two embodiments for relating theoretical and measured responses in block 640
include 1) a
curve fitting method and 2) a non-linear waveform inversion method. Both
methods calculate
.n _~.~»~~r .,,,~....~ .".,m.
CA 02485824 2002-08-08
theoretical waveform response based on Equation 2. However, the curve fitting
method uses
fewer theoretical modeling steps than the inversion method.
Figure 7A illustrates the curve fitting method, where a measurement equation
is determined.
As an initial matter, for a reverberation window of interest, TW;", the
natural log of the sum of
the reverberation waveform amplitude (SW) varies linearly with well fluid
impedance. That
is, a linear relationship between well fluid impedance and SW may be expressed
as:
Zm = A + B In (SW) (3)
where Su. is the sum of the reverberation waveform amplitudes and has the
form:
(4)
SW = ~~x(t)I
t
the lower case x(t) being the amplitude at any given point in the
reverberation waveform
contained in T",,in.
For the curve-fitting method, block 640 includes blocks 700-760. In block 700,
an initial
theoretical fluid impedance Zm is chosen. In block 710, the theoretical
response R(co) is
calculated in accordance with Equation 2. In block 720, the first reflection
is convolved with
the theoretical response obtained in block 710. In block 730, the Inverse Fast
Fourier
Transform (IFFT) is performed to obtain a theoretical reverberation waveform.
hText, the
summed amplitudes of the theoretical reverberation waveform SW is determined
in block 740.
In block 750, the theoretical response R(co) and reverberation waveforrn
amplitude sum S~"
are stored. In block 760, it is decided whether or not additional data is
needed. If additional
data is necessary, another theoretical fluid impedance Zm may be chosen in
block 700. To
determine the coefficients in this linear relationship, steps 700-760 are
repeated at least twice
for different assumed fluid impedances Zm. Each time, the resulting sum SW is
calculated.
From these multiple points, (5~,,, Zm), the coefficients A, B, can be
determined using the least
squares curve fitting in block 770. With the relationship, the measured
impedance Zm can be
determined from the observed SW using Equation 4 in block i'80.
11
CA 02485824 2002-08-08
Lastly, in block 650 (Figure 6), SW is substituted into Equation 3, and well
fluid impedance
Zp, is determined. The acoustic velocity of the fluid may also be calculated
in block 650.
Because the separation between the transducer and disk is known, the velocity
is calculable
from the measured transit time TtTan. From the impedance (p) and velocity (v),
the fluid
density (Zm) can be calculated due to the relationship: Zm = pv.
As mentioned above, in a second embodiment, non-linear w.aveform inversion may
be used in
block 640 to determine the relationship between theoretical and measured
reverberation. While
the waveform inversion method is slower than the curve fitting method
described above; it
produces more accurate results because it matches entire reverberation
waveform window
using both amplitude and phase. As a result, many fluid acoustic properties
including density
and attenuation can be calculated simultaneously. A preferred method employs
the Levenberg-
Marquardt method. See generally W. Press et al., Levenberg ll~Iarquardt
Method, p. 542
(Numerical Recipes in C, 1988).
In the non-linear waveform inversion embodiment shown in Figure 7B, fluid
properties such
as velocity, density, and attenuation are initially estimated in block 800. In
block 810, the
theoretical response R(co) is calculated in accordance with Equation 2. In
block 820, the first
reflection is convolved with the theoretical response obtained in block 710.
In block 830, the
Inverse Fast Fourier Transform (IFFT) is performed to obtain an estimated
reverberation
waveform. In block 840, the error between the estimated and measured waveforms
is
determined. The error is calculated according to Equation 5.
Error = ~ I observed - theoretical ~ I (5)
In block 850, the error calculated in block 840 is compared to a predetermined
tolerance. If the
calculated error is greater that the predetermined tolerance, another estimate
is performed in
block 800 using the Levenl~rg-Marquardt method. This cycle is repeated until
the calculated
error is less than the predetermined tolerance. When the calculated error is
less than the
12
CA 02485824 2002-08-08
predetermined tolerance, the estimated fluid velocity, density, and
attenuation are accepted as
the measured properties in block 860.
While preferred embodiments of this invention have been shown and described,
modifications
thereof can be made by one skilled in the art without departing from the
spirit or teaching of this
invention. The embodiments described' herein are exemplary only and are not
limiting. For
example, while the present invention has been described for use while drilling
a well, it rnay also
be used during completing and producing. Many variations and modifications of
the system and
apparatus are possible and are within the scope of the invention. Accordingly,
the scope of
protection is not limited to the embodiments described herein, but is only
limited by the claims
that follow, the scope of which shall include all equivalents of the subject
matter of the claims.
I3
