U.S. patent application number 11/398503 was filed with the patent office on 2007-10-11 for fluid distribution determination and optimization with real time temperature measurement.
Invention is credited to Gerard Glasbergen, David O. Johnson, Jose Sierra, Diederik van Batenburg, Mary Van Domelen, John Warren.
Application Number | 20070234789 11/398503 |
Document ID | / |
Family ID | 38573687 |
Filed Date | 2007-10-11 |
United States Patent
Application |
20070234789 |
Kind Code |
A1 |
Glasbergen; Gerard ; et
al. |
October 11, 2007 |
Fluid distribution determination and optimization with real time
temperature measurement
Abstract
Fluid distribution determination and optimization using real
time temperature measurements. A method of determining fluid or
flow rate distribution along a wellbore includes the steps of:
monitoring a temperature distribution along the wellbore in real
time; and determining in real time the fluid or flow rate
distribution along the wellbore using the temperature distribution.
A method of optimizing fluid or flow rate distribution includes the
steps of: predicting in real time the fluid or flow rate
distribution along the wellbore; comparing the predicted fluid or
flow rate distribution to a desired fluid or flow rate
distribution; and modifying aspects of a wellbore operation in real
time as needed to minimize any deviations between the predicted and
desired fluid or flow rate distributions.
Inventors: |
Glasbergen; Gerard; (Gouda,
NL) ; van Batenburg; Diederik; (Delft, NL) ;
Van Domelen; Mary; (Katy, TX) ; Johnson; David
O.; (Spring, TX) ; Sierra; Jose; (Katy,
TX) ; Warren; John; (Cypress, TX) |
Correspondence
Address: |
SMITH IP SERVICES, P.C.
P.O. Box 997
Rockwall
TX
75087
US
|
Family ID: |
38573687 |
Appl. No.: |
11/398503 |
Filed: |
April 5, 2006 |
Current U.S.
Class: |
73/152.55 |
Current CPC
Class: |
E21B 47/103 20200501;
E21B 41/0092 20130101; E21B 47/07 20200501; E21B 47/00
20130101 |
Class at
Publication: |
73/152.55 |
International
Class: |
E21B 47/08 20060101
E21B047/08 |
Claims
1. A method of determining fluid distribution along a wellbore, the
method comprising the steps of: monitoring a temperature
distribution along the wellbore in real time; and determining in
real time the fluid distribution along the wellbore using the
temperature distribution.
2. The method of claim 1, further comprising the step of optimizing
the fluid distribution.
3. The method of claim 2, wherein the optimizing step further
comprises comparing a desired fluid distribution with the fluid
distribution determined using the temperature distribution.
4. The method of claim 1, wherein the determining step further
comprises inputting the temperature distribution to a predictive
device, so that the predictive device predicts the fluid
distribution.
5. The method of claim 4, wherein the predictive device includes a
neural network.
6. The method of claim 4, further comprising the step of inputting
the fluid distribution to an optimization device.
7. The method of claim 6, wherein the optimization device modifies
inputs to the predictive device, so that a deviation of the fluid
distribution from a desired fluid distribution is minimized.
8. A method of optimizing fluid distribution along a wellbore, the
method comprising the steps of: predicting in real time the fluid
distribution along the wellbore; comparing the predicted fluid
distribution to a desired fluid distribution; and modifying aspects
of a wellbore operation in real time as needed to minimize any
deviations between the predicted and desired fluid
distributions.
9. The method of claim 8, further comprising the step of monitoring
a temperature distribution along the wellbore in real time, and
wherein the predicting step further comprises predicting the fluid
distribution along the wellbore using the temperature
distribution.
10. The method of claim 8, further comprising the steps of
monitoring a temperature distribution along the wellbore in real
time, and determining a current fluid distribution along the
wellbore using the temperature distribution.
11. The method of claim 8, wherein the predicting step further
comprises inputting a real time temperature distribution along the
wellbore to a predictive device, so that the predictive device
predicts the fluid distribution.
12. The method of claim 11, wherein the predictive device includes
a neural network.
13. The method of claim 8, wherein the comparing step further
comprises inputting the predicted fluid distribution to an
optimization device.
14. The method of claim 13, wherein the modifying step further
comprises the optimization device modifying inputs to the
predictive device, so that the deviation between the predicted and
desired fluid distributions is minimized.
15. A method of determining fluid distribution along a wellbore,
the method comprising the steps of: inputting a fluid distribution
to a model; predicting temperature distribution along the wellbore
using the model; monitoring temperature distribution along the
wellbore in real time; and modifying the fluid distribution based
on a comparison between the predicted temperature distribution and
the monitored temperature distribution.
16. The method of claim 15, wherein the predicting step further
comprises predicting pressure distribution in the wellbore, the
monitoring step further comprises monitoring pressure distribution
in the wellbore, and wherein the modifying step further comprises
modifying the fluid distribution based on a comparison between the
predicted pressure distribution and the monitored pressure
distribution.
17. The method of claim 15, wherein the predicting step further
comprises inputting at least one parameter to a model, so that the
model predicts the temperature distribution.
18. The method of claim 17, further comprising the step of
optimizing the fluid distribution by modifying the parameter input
to the model, then predicting the temperature distribution based on
the modified parameter, then modifying the fluid distribution based
on a comparison between the predicted temperature distribution
based on the modified parameter and the monitored temperature
distribution.
19. The method of claim 18, further comprising the step of
modifying the parameter based on a comparison between the modified
fluid distribution and a desired fluid distribution after the step
of modifying the fluid distribution based on the comparison between
the predicted temperature distribution based on the modified
parameter and the monitored temperature distribution.
20. A method of determining flow rate distribution along a
wellbore, the method comprising the steps of: monitoring a
temperature distribution along the wellbore in real time; and
determining in real time the flow rate distribution along the
wellbore using the temperature distribution.
21. The method of claim 20, further comprising the step of
optimizing the flow rate distribution.
22. The method of claim 21, wherein the optimizing step further
comprises comparing a desired flow rate distribution with the flow
rate distribution determined using the temperature
distribution.
23. The method of claim 20, wherein the determining step further
comprises inputting the temperature distribution to a predictive
device, so that the predictive device predicts the flow rate
distribution.
24. The method of claim 23, wherein the predictive device includes
a neural network.
25. The method of claim 23, further comprising the step of
inputting the flow rate distribution to an optimization device.
26. The method of claim 25, wherein the optimization device
modifies inputs to the predictive device, so that a deviation of
the flow rate distribution from a desired flow rate distribution is
minimized.
Description
BACKGROUND
[0001] The present invention relates generally to equipment
utilized and operations performed in conjunction with subterranean
wells and, in an embodiment described herein, more particularly
provides a method for fluid distribution determination and
optimization using real time temperature measurements.
[0002] Several methods have been used in the past for determining
fluid distribution along a wellbore. Among these are flowmeter
logging, evaluation of pressure response, qualitative evaluation of
temperature profile or distribution and evaluation of temperature
profile after shut-in.
[0003] Unfortunately, each of these methods has its shortcomings.
Flowmeter logging only provides an indication of flow rate at a
single point in the wellbore. Multiple logging runs may be made,
but each logging run still produces only an indication of flow rate
at a single point. Pressure measurements at surface and/or at
downhole locations also provide indications of flow rate at only
discrete points in the wellbore.
[0004] Past evaluations of temperature profiles have only been
qualitative, that is, a determination may be made as to whether or
not fluid flows into certain intervals, but quantitative
measurements of flow rate distribution along the interval are not
provided. Evaluations of temperature profiles after shut-in do not
provide real time determinations of fluid distribution, and
therefore cannot be used to modify or optimize an operation as it
progresses.
[0005] Thus, it will be appreciated that improvements are needed in
the art of fluid distribution determination and optimization. It is
among the objects of the present invention to provide such
improvements.
SUMMARY
[0006] In carrying out the principles of the present invention,
methods are provided which solve at least one problem in the art.
One example is described below in which fluid and flow rate
distribution along a wellbore are determined in real time. Another
example is described below in which fluid and flow rate
distribution are optimized in real time during an operation.
[0007] In one aspect of the invention, a method of determining
fluid or flow rate distribution along a wellbore is provided. The
method includes the steps of: monitoring a temperature distribution
along the wellbore in real time; and determining in real time the
fluid or flow rate distribution along the wellbore using the
temperature distribution.
[0008] In another aspect of the invention, a method of optimizing
fluid or flow rate distribution along a wellbore includes the steps
of: predicting in real time the fluid or flow rate distribution
along the wellbore; comparing the predicted distribution to a
desired fluid or flow rate distribution; and modifying aspects of a
wellbore operation in real time as needed to minimize any
deviations between the predicted and desired fluid or flow rate
distributions.
[0009] In another aspect of the invention, a method of determining
fluid or flow rate distribution along a wellbore includes the steps
of: inputting a fluid or flow rate distribution to a model;
predicting temperature distribution along the wellbore using the
model; monitoring temperature distribution along the wellbore in
real time; and modifying the fluid or flow rate distribution based
on a comparison between the predicted temperature distribution and
the monitored temperature distribution.
[0010] Among the benefits of the methods described below is the
ability to determine in real time the fluid and flow rate
distributions along a wellbore, so that an evaluation of a wellbore
operation may be conducted as the operation progresses. Another
benefit is that the fluid and flow rate distributions may be
optimized in real time, so that desired fluid and flow rate
distributions may be achieved during the operation.
[0011] These and other features, advantages, benefits and objects
of the present invention will become apparent to one of ordinary
skill in the art upon careful consideration of the detailed
description of representative embodiments of the invention
hereinbelow and the accompanying drawings, in which similar
elements are indicated in the various figures using the same
reference numbers.
BRIEF DESCRIPTION OF THE DRAWINGS
[0012] FIG. 1 is a partially cross-sectional schematic view of a
method embodying principles of the present invention;
[0013] FIG. 2 is a schematic view of a model which may be used in
the method of FIG. 1;
[0014] FIG. 3 is a flowchart of steps in a technique suited for use
in the method of FIG. 1;
[0015] FIGS. 4-8 are exemplary graphs of desired, predicted and
actual fluid distributions during an injection operation in the
method of FIG. 1;
[0016] FIG. 9 is a schematic view of a fluid distribution
determination and optimization technique for use in the method of
FIG. 1; and
[0017] FIG. 10 is a schematic view of a flow rate distribution
determination and optimization technique for use in the method of
FIG. 1.
DETAILED DESCRIPTION
[0018] It is to be understood that the various embodiments of the
present invention described herein may be utilized in various
orientations, such as inclined, inverted, horizontal, vertical,
etc., and in various configurations, without departing from the
principles of the present invention. The embodiments are described
merely as examples of useful applications of the principles of the
invention, which is not limited to any specific details of these
embodiments.
[0019] In the following description of the representative
embodiments of the invention, directional terms, such as "above",
"below", "upper", "lower", etc., are used for convenience in
referring to the accompanying drawings. In general, "above",
"upper", "upward" and similar terms refer to a direction toward the
earth's surface along a wellbore, and "below", "lower", "downward"
and similar terms refer to a direction away from the earth's
surface along the wellbore.
[0020] Representatively illustrated in FIG. 1 is a method 10 which
embodies principles of the present invention. As depicted in FIG.
1, fluid 12 is injected into a wellbore 14 via a production tubing
string 18, and then into an area 20 of the wellbore below a packer
set in a casing string 22. Although the area 20 is depicted as
being cased, in other embodiments of the invention the area could
be uncased.
[0021] Eventually, the fluid 12 flows into a formation, strata or
zone 24 via perforations 26. If desired, the fluid 12 may also be
flowed into another formation, strata or zone 28 via separate
perforations 30. The zones 24, 28 could be isolated from each other
in the wellbore 14 by a packer set in the casing string 22, if
desired.
[0022] In this manner, a portion 34 of the fluid 12 flows into the
upper zone 24, and another portion 36 flows into the lower zone 28.
One problem solved by the method 10, as described more fully below,
is how to determine in real time the flow rate of the fluid 12 as
it flows through the wellbore 14 and into each of the zones 24,
28.
[0023] Another problem solved by the method 10 and described more
fully below is how to optimize the distribution of the fluid 12 in
the zones 24, 28 in real time during the operation. Fluid
distribution is the extent to which fluid penetrates a formation or
zone versus depth along a wellbore. Graphic examples of desired,
predicted and actual fluid distributions are depicted in FIGS. 4-8,
and are described more fully below.
[0024] In the past, DTS systems utilizing an optical conductor 38
(such as an optical fiber in a small diameter tube, or incorporated
into a cable, etc.) have been used to produce a temperature profile
along the wellbore 14. After the injection operation, the
temperature profile from before the operation would be compared to
the temperature profile from during the operation, in order to
determine where the fluid 12 entered the various zones 24, 28 and
how much of the fluid entered each zone. However, these past
methods do not allow the distribution of the fluid 12 to be
determined in real time, so that the injection operation can be
evaluated and optimized during the operation.
[0025] At this point it should be pointed out that the invention is
not limited in any way by the details of the method 10 described
herein or the configuration of the well as illustrated in FIG. 1.
For example, the invention is not necessarily used only in
injection operations, since it may also be used in other types of
operations (such as production, stimulation, completion,
conformance, etc. operations).
[0026] The invention may be used to monitor conditions in a
wellbore prior to a treatment, for example, to determine where
water is being produced and where a treatment gel should be placed.
The invention may be used to place resins for sand control, to
repair gravel packing screens, etc.
[0027] The invention is not necessarily used only in cased
wellbores, since it may also be used in uncased wellbores. The
invention is not necessarily used only where multiple zones have
fluid transfer with a wellbore. A coiled tubing string could be
used to transfer fluid to or from a wellbore. It is not necessary
for an optical conductor to be used to monitor temperature along a
wellbore.
[0028] Therefore, it should be clearly understood that the method
10 is described and illustrated herein as merely one example of an
application of the principles of the invention, which is not
limited at all to the details of the described method.
[0029] Referring additionally now to FIG. 2, a wellbore model 40
which may be used in the method 10 is representatively illustrated.
The model 40 is used to design stimulation treatments or more
general fluid placement/injection. The model 40 predicts pressure,
fluid, injectivity and temperature distribution versus time.
[0030] Actual treatment parameters, such as injection rate, fluid
type and schedules for these, well geometry, reservoir properties,
etc. may be input to the model 40, so that the predicted pressure,
fluid, injectivity and temperature distributions are based on the
actual parameters. Initial fluid distribution (and reaction
parameters, if desired) and pressure and temperature distributions
input to the model 40 may be manually adjusted to obtain a match
between measured and predicted responses versus time. Examples of
models are described in "Field Validation of Acidizing Wormhole
Models," SPE 94695 (2005), the entire disclosure of which is
incorporated herein by this reference.
[0031] Calibration of the model can be conducted based on measured
temperature distribution and one or more measured pressures by
adjusting the reservoir or other relevant properties. This may
require several iterations, and can be automated.
[0032] The downhole pressures may be measured using any type of
pressure sensor, such as optical pressure sensors coupled to the
optical conductor 38. The sensors may be temporary sensors (e.g.,
installed only for the term of the operation) or permanent sensors
(e.g., installed for long term use over the life of the well).
[0033] Note that the optical conductor 38 may be retrievably
deployed, for example, in fracturing or injection operations,
without strapping the optical conductor to the tubing string 18.
However, the optical conductor 38 could be permanently deployed or
strapped to the tubing string 18, if desired.
[0034] Periodically (for example, approximately each minute), a
current measured temperature distribution is available from the DTS
system using the optical conductor 38. An acceptable DTS system for
use in providing the measured temperature distribution is the
OPTOLOG.RTM. DTS system available from Halliburton Energy Services
of Houston, Tex. USA.
[0035] Referring additionally now to FIG. 3, a technique 42 which
may be used in the method 10 is representatively illustrated in
flowchart form. Of course, the technique 42 may be used in other
methods without departing from the principles of the invention.
[0036] In an initial step 44, the well geometry and planned
treatment schedule with fluid types/properties and other data are
input to the model 40. Possible inputs include reservoir
properties, such as permeability, porosity, mineralogy, acid
reactivity, skin damage, and permeability contrast. Well geometry
may include height of the layers, wellbore tubulars, friction
pressures, etc.
[0037] In step 46, the model 40 is initialized with an initial
fluid distribution versus depth. This initial fluid distribution
may be based on well logs and/or core data or other relevant
data.
[0038] In step 48, the model 40 is initialized with initial data,
such as pressure and temperature versus depth. The DTS system may
be used to supply this data.
[0039] In step 50, the model 40 is used to predict pressure and
fluid distribution versus time. Alternatively, these parameters may
be predicted for a certain future time.
[0040] In step 52, the resulting temperature distribution is
predicted. In step 54, the actual temperature distribution is
determined in real time, for example, using the DTS system.
[0041] As described in the copending patent application entitled
TRACKING FLUID DISPLACEMENT ALONG A WELLBORE USING REAL TIME
TEMPERATURE MEASUREMENTS, attorney docket no. 2005-IP-019088 U1
USA, the fluid properties and injection rate may be modified and/or
chemical reactions may be initiated to enhance detection of
temperature gradient differences in the wellbore 14. This technique
can enable more accurate determinations of fluid distribution along
the wellbore. The entire disclosure of this copending patent
application is incorporated herein by this reference.
[0042] In step 56, the actual pressure at one or more known
locations is determined. An optical conductor with optical sensors,
or any other type of pressure sensors may be used in this step for
measuring the actual pressure(s) in real time, either as part of
the DTS system or separate therefrom.
[0043] In step 58, the fluid distribution input to the model 40 is
modified, based on the actual temperature and pressure
distributions from steps 54 & 56.
[0044] In step 60, the pressure distribution and temperature
distribution versus time are again predicted using the model 40. In
step 62, the predicted pressure and temperature distribution are
compared to the actual pressure and temperature distribution to
determine whether a match is obtained.
[0045] If a match is obtained, then a solution is indicated in step
64, i.e., the fluid distribution input to the model 40 is correct.
If a match is not obtained in step 62, then steps 58 & 60 are
repeated until a match is obtained.
[0046] When additional data becomes available (such as when updated
temperature distribution data is provided by the DTS system and/or
when pressure measurements become available), this process is
performed again. In this manner, the fluid distribution predicted
by the model 40 is periodically updated or "calibrated" as the
additional data becomes available. In order to optimize the fluid
distribution, the planned treatment schedule may be modified based
on the calibrated fluid distribution predicted by the model 40.
[0047] It should be clearly understood that, although certain
inputs have been described above for the model 40, the invention is
not limited to only these inputs. Other inputs, and other
combinations of inputs, could be used for the model in keeping with
the principles of the invention. Thus, it will be appreciated that
the model 40 and technique 42 described above may be modified in
any manner without departing from the principles of the
invention.
[0048] Furthermore, although fluid distribution is described above
as being predicted and optimized using the model 40 and technique
42, it is not necessary for fluids to be injected, for example, the
fluids could instead be produced. Flow rate or injectivity
distribution integrated over time yields fluid distribution, and so
the above described steps wherein fluid distribution is predicted,
determined, etc. may be considered to include prediction,
determination, etc. of flow rate or injectivity distribution, as
well.
[0049] Referring additionally now to FIGS. 4-8, an example of how
the principles of the invention may be beneficially used to
optimize fluid distribution in an acidizing operation is
representatively illustrated. FIGS. 4-8 are schematic graphs of
fluid penetration (on the horizontal scale in units of inches
radially outward from the wellbore) vs. depth (on the vertical
scale in units of feet along the wellbore).
[0050] FIG. 4 depicts a desired final fluid distribution at the end
of the operation. The desired fluid distribution is preferably
planned by experienced professionals to achieve optimum results
(e.g., an acceptable level of stimulation, economy, etc.).
Alternatively, or in addition, the desired fluid distribution could
be planned using computational techniques, expert systems, etc.
[0051] In the depicted example, the operation is planned to include
injection of 10,000 gallons of preflush 66, 10,000 gallons of
mainflush 67 and 10,000 gallons of overflush 68. This schedule
should result in a fluid front of the preflush 66 at approximately
135 inches penetration, a fluid front of the mainflush 67 at
approximately 110 inches penetration, a fluid front of the
overflush 68 at approximately 75 inches penetration and a live acid
edge 69 at approximately 45 inches penetration. These should be
fairly consistent along the wellbore between 4900 and 5000 feet as
illustrated in FIG. 4.
[0052] FIG. 5 depicts the predicted fluid distribution after an
initial 2000 gallons of preflush 66 are injected. Note that the
fluid front of the preflush 66 should be at approximately 35 inches
penetration, and the live acid edge 69 should be at approximately
15 inches penetration, and these should be very consistent between
the depths of 4900 and 5000 feet.
[0053] FIG. 6 depicts the actual fluid distribution after 2000
gallons of preflush 66 have been injected. This actual fluid
distribution may be determined using the technique 42 described
above. Note that, between the depths of 4900 and 4950 feet, the
fluid front of the preflush 66 is at approximately 40 inches
penetration (greater than the predicted 35 inches penetration), and
between the depths of 4950 and 5000 feet the fluid front of the
preflush is at approximately 10 inches penetration (less than the
predicted 35 inches penetration).
[0054] Thus, a comparison between the predicted fluid distribution
(as depicted in FIG. 5) and the actual fluid distribution (as
depicted in FIG. 6) indicates that remedial action will need to be
taken in order to achieve the optimal fluid distribution of FIG. 4.
Among the many beneficial features of the methods and techniques
described herein are that the need for remedial action can be
quickly identified and accurately quantified in real time as the
operation progresses, and the remedial action can be taken in a
timely manner so that the optimal fluid distribution can be
achieved.
[0055] Another beneficial feature of the methods and techniques
described herein is that the model used to predict fluid
distribution may be modified as the operation progresses, so that
the model will more accurately predict fluid distribution during
the operation. Thus, in the present example, a comparison between
the actual fluid distribution as depicted in FIG. 6 and the
predicted fluid distribution as depicted in FIG. 5 indicates that
the model should be modified (for example, by adjusting properties
of the reservoir between the depths of 4950 and 5000 feet, etc.),
and the modification can be accomplished so that subsequent
predictions of fluid distribution during the operation will be more
accurate. In this sense, it may be considered that the model is
"calibrated" as the operation progresses.
[0056] In the present example, the remedial action to be taken
includes injection of a diverter midway between two halves of the
originally planned schedule. FIG. 7 depicts the fluid distribution
after 5000 gallons of preflush, 5000 gallons of mainflush and 5000
gallons of overflush have been injected (i.e., one half of the
originally planned schedule).
[0057] FIG. 8 depicts the fluid distribution after injection of a
diverter 65, followed by an additional 5000 gallons of preflush,
5000 gallons of mainflush and 5000 gallons of overflush (i.e., the
remaining half of the originally planned schedule). Note that the
fluid distribution as depicted in FIG. 8 closely approximates the
desired fluid distribution as depicted in FIG. 4. This result was
achieved by modifying the operation as it progressed, and without
the need to inject fluids in addition to those originally
scheduled, other than the diverter 65.
[0058] In the past, the original schedule of fluids would have been
injected and then, after an analysis of temperature distribution
and other data, it may have been determined that remedial action
including injection of a diverter should be taken. The diverter and
an additional schedule of treatment fluids would have then been
injected in an attempt to achieved the desired fluid distribution.
It will be readily appreciated by those skilled in the art that the
new methods and techniques described herein result in a far more
timely, economical and accurate operation being performed.
[0059] Referring additionally now to FIG. 9, a technique 70 is
representatively illustrated for predicting and optimizing fluid
distribution in the method 10. The technique 70 utilizes a model 72
which may be similar to the model 40, but it should be understood
that any other type of model and any combination of models may be
used in place of the model 72, if desired.
[0060] Inputs to the model 72 include (but are not limited to)
pressure and temperature distributions PTD (these may be the same
as or similar to the pressure and temperature distributions
described above as being input in the technique 42 in steps 48 and
54), geothermal gradient GG (this is similar to the initial
temperature distribution described above as being input in the
technique 42 in step 48), injection rate IR, fluid type FT
(including density, specific heat, etc. of the fluid; these may be
the same as or similar to the fluid properties/schedule described
above as being input in the technique 42 in step 44), well geometry
WG (such as diameters and lengths of tubular strings, deviation,
etc.; these may be the same as or similar to the well geometry
parameters described above as being input in the technique 42 in
step 44), reservoir properties RP (such as rock properties,
porosity, permeability, intrinsic fluids, etc.; these may be the
same as or similar to the reservoir properties described above as
being input in the technique 42 in step 44), and control inputs CI
(such as surface pressure, choke position, etc.). The model 72
outputs a predicted fluid distribution PFD along the wellbore 14 at
an incremental future time (t+n).
[0061] An error evaluation 74 compares the predicted fluid
distribution PFD to the current fluid distribution at present time
(t). Note that the current fluid distribution FD(t) may be provided
by the technique 42 described above and depicted in FIG. 3.
[0062] Any error determined in the error evaluation 74 is used to
modify the model 72, so that future predictions of fluid
distribution FD are more accurate. It will be appreciated that this
technique 70 of continuously predicting the fluid distribution FD,
comparing the predicted fluid distribution PFD to the fluid
distribution determined using the real time temperature and
pressure measurements in the technique 42, and modifying the model
72 to minimize errors in the predictions enables highly accurate
determinations of the fluid distribution in the wellbore 14 to be
available in real time during the course of the operation.
[0063] In another feature of the technique 70, the predicted fluid
distribution PFD(t+n) is input to an optimization device 76 for a
determination of how various aspects of the operations should be
modified to achieve a desired fluid distribution. The desired fluid
distribution is determined prior to the operation, for example, to
deliver certain volumes of stimulation fluid to particular zones or
intervals, etc.
[0064] The optimization device 76 compares the predicted fluid
distribution PFD(t+n) to the desired fluid distribution and
determines whether certain aspects of the operation should be
modified in order to achieve the desired fluid distribution. Of
course, if the predicted fluid distribution is the same as the
desired fluid distribution, then no modifications will be
needed.
[0065] As depicted in FIG. 9, the optimization device 76 may be
used to modify the injection rate IR, fluid types FT and control
inputs CI. These modified inputs are used by the model 72 to again
predict the fluid distribution PFD(t+n), which is then input again
to the optimization device 76 for evaluation. In this manner, the
predicted fluid distribution PFD(t+n) is continuously evaluated,
and aspects of the operation (such as injection rate IR, fluid
types FT and control inputs CI) may be continuously modified to
obtain and maintain the desired fluid distribution (e.g., to
minimize any deviation between the predicted fluid distribution and
the desired fluid distribution) in real time, as the operation
progresses.
[0066] Referring additionally now to FIG. 10, a technique 80 is
representatively illustrated for predicting and optimizing flow
rate distribution in the method 10. The technique 80 utilizes a
predictive device 82 in the form of a neural network, but it should
be understood that any other type of predictive device and any
combination of predictive devices may be used in place of the
neural network, if desired.
[0067] For example, the predictive device 82 may include a neural
network, an artificial intelligence device, a floating point
processing device, an adaptive model, a nonlinear function which
generalizes for real systems and/or a genetic algorithm. The
predictive device 82 may perform a regression analysis, perform
regression on a nonlinear function and may utilize granular
computing. An output of a first principle model may be input to the
predictive device 82 and/or a first principle model may be included
in the predictive device.
[0068] Inputs to the neural network 82 include (but are not limited
to) measured temperature distribution or profile MTP (this may be
the same as or similar to the temperature distribution described
above), geothermal gradient GG (this is similar to the initial
temperature distribution described above as being input in the
technique 42 in step 48), injection rate IR, properties of the
fluids PF (such as density, specific heat, etc.; these may be the
same as or similar to the fluid types/schedule described above),
properties of the wellbore PWB (such as diameters and lengths of
tubular strings, deviation, etc.; these may be the same as or
similar to the well geometry parameters described above),
properties of the intersected zones PZ (such as rock properties,
porosity, permeability, intrinsic fluids, etc.; these may be the
same or similar to the reservoir properties described above as
being input in the technique 42 in step 44), and control inputs CI
(such as surface pressure, choke position, etc.). Any of these
inputs may be the same as or similar to the corresponding inputs
described above for the technique 70.
[0069] The neural network 82 outputs a predicted injectivity or
flow rate distribution PFRD along the wellbore 14 at an incremental
future time (t+n). As discussed above, flow rate or injectivity
distribution integrated over time yields fluid distribution, and so
it should be understood that prediction or determination of flow
rate or injectivity distribution over time also provides predicted
or determined fluid distribution, as well.
[0070] An error evaluation 84 compares the predicted flow rate
distribution PFRD to the current flow rate distribution at present
time (t). Note that the current flow rate distribution FRD(t) may
be provided by the technique 42 described above and depicted in
FIG. 3.
[0071] Any error determined in the error evaluation 84 is used to
modify the neural network 82, so that future predictions of flow
rate distribution PFRD are more accurate. It will be appreciated
that this technique 80 of continuously predicting the flow rate
distribution FRD, comparing the predicted flow rate distribution
PFRD to the flow rate distribution determined using the real time
temperature measurements in the technique 42, and modifying the
neural network 82 to minimize errors in the predictions enables
highly accurate determinations of the flow rate distribution in the
wellbore 14 to be available in real time during the course of the
operation.
[0072] In another feature of the technique 80, the predicted flow
rate distribution PFRD(t+n) is input to an optimization device 86
for a determination of how various aspects of the operations should
be modified to achieve a desired flow rate distribution. The
desired flow rate distribution is determined prior to the
operation, for example, to deliver certain volumes of stimulation
fluid to particular zones or intervals over a certain time,
etc.
[0073] The optimization device 86 compares the predicted flow rate
distribution PFRD(t+n) to the desired flow rate distribution and
determines whether certain aspects of the operation should be
modified in order to achieve the desired flow rate distribution. Of
course, if the predicted flow rate distribution is the same as the
desired flow rate distribution, then no modifications will be
needed.
[0074] As depicted in FIG. 10, the optimization device 86 may be
used to modify the injection rate IR, properties of the fluids PF
and control inputs CI. These modified inputs are used by the neural
network 82 to again predict the flow rate distribution PFRD(t+n),
which is then input again to the optimization device 86 for
evaluation. In this manner, the predicted flow rate distribution
PFRD(t+n) is continuously evaluated, and aspects of the operation
(such as injection rate IR, properties of the fluids PF and control
inputs CI) may be continuously modified to obtain and maintain the
desired flow rate distribution (e.g., to minimize any deviation
between the predicted flow rate distribution and the desired flow
rate distribution) in real time, as the operation progresses.
[0075] As discussed above, the principles of the invention are
useful in operations other than injection operations. For example,
in production operations the input injection rate IR in the
techniques 42, 70, 80 could be replaced with production rate.
Similar modifications may be used for other types of operations, as
well.
[0076] Of course, a person skilled in the art would, upon a careful
consideration of the above description of representative
embodiments of the invention, readily appreciate that many
modifications, additions, substitutions, deletions, and other
changes may be made to these specific embodiments, and such changes
are within the scope of the principles of the present invention.
Accordingly, the foregoing detailed description is to be clearly
understood as being given by way of illustration and example only,
the spirit and scope of the present invention being limited solely
by the appended claims and their equivalents.
* * * * *