U.S. patent application number 13/915326 was filed with the patent office on 2013-10-17 for feedback control using a simlator of a subterranean structure.
The applicant listed for this patent is Schlumberger Technology Corporation. Invention is credited to Baris Guyaguler, Jonathan Anthony Holmes, Andreas Theodoros Papadopoulos.
Application Number | 20130275105 13/915326 |
Document ID | / |
Family ID | 40902475 |
Filed Date | 2013-10-17 |
United States Patent
Application |
20130275105 |
Kind Code |
A1 |
Guyaguler; Baris ; et
al. |
October 17, 2013 |
FEEDBACK CONTROL USING A SIMLATOR OF A SUBTERRANEAN STRUCTURE
Abstract
To provide feedback control in a simulation framework, any one
of plural output metrics from a simulator of a subterranean
structure is selected. A value for the selected output metric is
received from the simulator. In response to the received value of
the selected output metric and a target value of the selected
output metric, at least one setting of the simulator is adjusted by
a feedback controller.
Inventors: |
Guyaguler; Baris; (Aberdeen,
GB) ; Papadopoulos; Andreas Theodoros; (Oxford,
GB) ; Holmes; Jonathan Anthony; (Reading,
GB) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Schlumberger Technology Corporation |
Sugar Land |
TX |
US |
|
|
Family ID: |
40902475 |
Appl. No.: |
13/915326 |
Filed: |
June 11, 2013 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12473310 |
May 28, 2009 |
8463457 |
|
|
13915326 |
|
|
|
|
61061540 |
Jun 13, 2008 |
|
|
|
Current U.S.
Class: |
703/10 |
Current CPC
Class: |
E21B 49/00 20130101;
G06F 30/20 20200101; G01V 11/00 20130101 |
Class at
Publication: |
703/10 |
International
Class: |
G06F 17/50 20060101
G06F017/50 |
Claims
1. A method comprising: selecting any one of plural output metrics
from a simulator associated with a subterranean structure;
receiving a value for the selected output metric from the
simulator; and in response to the received value of the selected
output metric and a target value of the selected output metric,
adjusting a setting of the simulator by a fuzzy logic feedback
controller executed on a processor.
2. The method of claim 1 comprising determining an error value as a
difference between the received value of the selected output metric
and the target value, fuzzifying the error value, applying a fuzzy
rule to the fuzzified error value to provide a fuzzified response,
defuzzifying the response to provide a response value and adjusting
the setting of the simulator based at least in part on the response
value.
3. The method of claim 2 wherein the selected output metric
comprises pressure and wherein the response value comprises a value
for adjusting a choke setting.
4. The method of claim 1 wherein the simulator comprises a surface
network model that models a fluid choke for choking fluid flow
associated with the subterranean structure.
5. The method of claim 4 further comprising adjusting a fluid choke
in a surface network associated with the modeled subterranean
structure based at least in part on the adjusting of the setting of
the simulator by the fuzzy logic feedback controller.
6. The method of claim 4 wherein the selected output metric
comprises pressure and wherein the setting comprises a choke
setting for the fluid choke of the surface network model.
7. The method of claim 1 wherein the simulator comprises a surface
network model, wherein the target value comprises a target pressure
and wherein the adjusting the setting acts to maintain the target
pressure.
8. The method of claim 7 wherein the setting comprises a choke
setting for a fluid choke modeled by the surface network model.
9. The method of claim 1 further comprising selecting a plurality
of the plural output metrics, receiving a plurality of
corresponding values for the selected plurality of output metrics,
determining a plurality of error values based on the corresponding
values and at least one target value, fuzzifying the plurality of
error values, applying at least one fuzzy rule to the fuzzified
error values to provide at least one fuzzified response,
defuzzifying the at least one fuzzified response to provide at
least one response value and adjusting at least one setting of the
simulator based at least in part on the at least one response
value.
10. The method of claim 1, wherein the adjusted setting comprises a
controllable variable that is provided as an input to the
simulator, wherein the simulator is associated with multiple
controllable variables, wherein a simulation of the subterranean
structure performed by the simulator is based on values of the
controllable variables, the method further comprising: selectively
coupling any one or more of the plural output metrics from the
simulator to any one or more of the controllable variables.
11. The method of claim 1, wherein the adjusted setting comprises a
controllable variable that is provided as an input to the
simulator, and wherein the controllable variable comprises a
variable that affects a behavior of a controllable flow entity
modeled by a surface network model.
12. The method of claim 1, wherein the output metrics are selected
from among pressure, temperature, flow rate, flow volume, and phase
saturation.
13. A system comprising: a processor; a simulator executable on the
processor to simulate at least a surface network associated with at
least one reservoir using a surface network model, the simulator to
produce output metrics when performing a simulation, and wherein
the simulator has input control variables that control the
simulator; and a fuzzy logic feedback controller executable on the
processor to configurably couple any of plural ones of the output
metrics to any of plural ones of the input control variables,
wherein the feedback controller adjusts a particular one of the
control variables in response to particular one or more output
metrics that have been coupled to the particular control
variable.
14. The system of claim 13, wherein the model models at least one
controllable flow entity associated with the surface network,
wherein the input control variables are to control the at least one
controllable flow entity.
15. The system of claim 14 wherein the at least one controllable
flow entity comprises a fluid choke.
16. The system of claim 13, wherein the simulator further comprises
at least one reservoir model.
17. The system of claim 13, wherein the output metrics are selected
from among pressure, temperature, flow rate, flow volume, and phase
saturation.
18. An article comprising at least one computer-readable storage
medium containing instructions that upon execution by a processor
is to: receive input regarding configuration of a fuzzy logic
feedback controller that is coupled to a simulator to simulate at
least a surface network associated with at least one reservoir
using a surface network model, wherein the received input specifies
a particular one of plural output metrics from the simulator to
couple to a particular one of control variables input to the
simulator; configure the fuzzy logic feedback controller according
to the received input; and operate the fuzzy logic feedback
controller to couple the particular output metric to the particular
control variable, wherein the fuzzy logic feedback controller
receives values of the particular output metric from the simulator
and adjusts the particular control variable in response to the
received values of the particular output metric.
19. The article of claim 18, wherein adjustment of the particular
control variable adjusts a setting of a controllable flow entity
modeled by the surface network model.
20. The article of claim 19, wherein the controllable flow entity
modeled by the surface network model comprises a fluid choke.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of U.S. patent
application Ser. No. 12/473,310 filed May 28, 2009 issuing as U.S.
Pat. No. 8,463,457 on Jun. 11, 2013, which claims the benefit of
U.S. Provisional Patent Application Ser. No. 61/061,540 filed Jun.
13, 2008, both which are incorporated herein by reference in its
entirety.
BACKGROUND
[0002] There are generally three phases that are undertaken to
obtain hydrocarbons or other fluids from a given field of
development (or from a well). The phases are exploration, appraisal
and production. During exploration, one or more subterranean
structures (e.g., formations or reservoirs) are identified that may
include fluids in an economic quantity.
[0003] Following successful exploration, the appraisal phase is
conducted. During the appraisal phase, operations, such as drilling
wells, are performed to determine the size of the field and how to
develop the field. After the appraisal phase is complete, the
production phase is initiated. During the production phase, fluids
are produced from reservoirs in the field.
[0004] In one or more of the phases discussed above, simulation of
subterranean structures (e.g., reservoirs) can be performed by
using models of such subterranean structures. Performing
simulations using models of subterranean structures can assist
operators in better understanding the subterranean structures such
that more effective strategies can be developed to produce fluids
such as hydrocarbons from the field.
[0005] However, conventional techniques of performing simulations
may not offer the accuracy or flexibility that may be desired.
SUMMARY
[0006] In general, feedback control in performing simulation of a
subterranean structure is provided. Any one of plural output
metrics from a simulator of the subterranean structure is selected.
A value for the selected output metric is received from the
simulator. In response to the received value of the selected output
metric and a target value of the selected output metric, at least
one setting of the simulator is adjusted by a feedback
controller.
[0007] Other or alternative features will become apparent from the
following description, from the drawings, and from the claims.
BRIEF DESCRIPTION OF THE DRAWINGS
[0008] FIG. 1 is a schematic diagram of an exemplary arrangement
that includes a subterranean structure that can be simulated using
a simulation framework with feedback control, according to an
embodiment;
[0009] FIG. 2 illustrates an exemplary model of a subterranean
structure that can be used by a simulator used in the simulation
framework according to an embodiment;
[0010] FIG. 3 is a block diagram of a system that includes a
simulator and a feedback controller, according to an
embodiment;
[0011] FIG. 4 depicts triangular (H) and trapezoidal (VH) fuzzy set
membership functions to illustrate a fuzzy logic feedback
controller according to an embodiment;
[0012] FIG. 5 is a flow diagram of a process of feedback control in
a simulation framework, according to an embodiment; and
[0013] FIG. 6 is a flow diagram of configuring a feedback
controller according to an embodiment.
DETAILED DESCRIPTION
[0014] In the following description, numerous details are set forth
to provide an understanding of the present disclosure. However, it
will be understood by those skilled in the art that some
embodiments may be practiced without these details and that
numerous variations or modifications from the described embodiments
are possible.
[0015] In accordance with some embodiments, a control system with a
feedback controller is used in a simulation framework that provides
a numerical simulator of a subterranean structure, such as a
reservoir, formation or other structure underneath an earth
surface. The feedback controller used according to some embodiments
can be either a Proportional Integral Derivative (PID) feedback
controller or a fuzzy logic feedback controller. A PID feedback
controller and fuzzy logic feedback controller are discussed
further below.
[0016] Using such feedback controller allows for provision of
generic feedback control where any observable metric from a
simulator (of a subterranean structure) can be coupled to any
controllable variable that is an input to the simulator. More
generally, a control system that includes the feedback controller
and simulator of a subterranean structure is able to select any one
or more observable metrics (from multiple possible observable
metrics) output from the simulator to couple to any of one or more
controllable variables (from among multiple possible controllable
variables) that are input to the simulator. The selectivity of the
output observable metric and the input controllable variable can be
based on predetermined input, such as user selection and/or
predefined criteria. This generic any-to-any feedback control in a
simulation framework allows for greater flexibility.
[0017] Using feedback control in a simulation framework for
subterranean structures can help an operator perform one or more of
the following: maintain desired operating conditions in a field;
formulate a better plan or strategy for development of a field; or
provide a better design of a system (e.g., an automated system)
used to produce hydrocarbons or other fluids from the field. For
example, operating strategies can be developed, and/or feasibility
of installing physical control devices in the field can be
determined. More generally, providing feedback control in a
simulation framework for simulation of subterranean structures
assists operators in better understanding a particular field (or
well) that is to be developed.
[0018] FIG. 1 illustrates an exemplary arrangement for extracting
fluids from a reservoir 100 in the subsurface. Wellbores 102 have
been drilled into the subsurface to intersect the reservoir 100.
Although not shown, completion equipment can be provided in the
wellbores 102 to assist in producing fluids from the reservoir 100
to the earth surface. Surface equipment 104 is deployed to allow
for extraction of the fluids and to carry the extracted fluids to
target locations. The arrangement of FIG. 1 can be a land-based
arrangement or a marine arrangement (in which the reservoir 100 is
located underneath a sea floor or other water bottom surface.
[0019] The reservoir 100 can be modeled by using a reservoir
simulation model. The reservoir simulation model is used by a
simulator to perform simulations, which produces output values for
one or more metrics based on input data values. As shown in FIG. 2,
a reservoir simulation model 200 can be a three-dimensional (3D)
model, which has a grid to partition the simulated reservoir into
multiple cells 202. Each cell 202 represents a portion of the
reservoir and is characterized by the physical properties of the
portion of the reservoir (e.g., temperature, pressure, phase
saturations, oil viscosity, porosity, permeability, etc.). Instead
of a 3D model, a two-dimensional (2D) model or even a
one-dimensional (1D) model can be used instead. As yet another
alternative, the model can be a time-lapse model that considers a
time dimension--for example, the time-lapse model can be a 4D
model.
[0020] FIG. 3 shows a simulation system 300 that is able to use the
model 200 of FIG. 2 (1D, 2D, 3D, or time-lapse model). The
simulation system 300 includes a simulator 304 that performs
simulation of a reservoir using the reservoir model 200. The
simulation system 300 further includes a feedback controller 306,
which can be a PID feedback controller or a fuzzy logic feedback
controller, in accordance with some embodiments. The simulator 304
and feedback controller 306 can be software modules executable on a
processor 310, which can include one CPU (central processing unit)
or multiple CPUs (whether run on one computer or multiple
computers). The simulation system 300 can be implemented with one
or multiple computers, for example.
[0021] An output of the simulator 304 includes an output metric y,
which is fed back to a comparator 308. The comparator 308 also
receives a target (predefined) value of the metric, y.sub.sp, and
compares y to y.sub.sp to produce a comparison value e (which
represents the difference between y and y.sub.sp). The comparison
value e is provided to an input of the feedback controller 306.
[0022] The feedback controller 306 uses the comparison value e to
adjust a value of a controllable variable u that is input to the
simulator 304. The behavior of the simulator 304 is modified by
adjustment of the controllable variable u. Although just one metric
y and one controllable variable u are shown in FIG. 3, it is noted
that there can be multiple metrics y and multiple controllable
variables u. In fact, in accordance with some embodiments, the
control system implemented in the simulation framework provided by
the simulation system 300 allows for any one or more of multiple
metrics y output by the simulator 304 to be selected for coupling
by the feedback controller 306 to any one or more of multiple
controllable variables u that can be input to the simulator
304.
[0023] The selectivity of the output metrics y and input
controllable variables u can be based on selection input 312 stored
in a storage media 302, which can be implemented with one or more
disk-based storage devices, one or more integrated circuit storage
devices, and so forth. The selection input 312 can reflect user
selection, or alternatively, the selection input 312 can be in the
form of selection criteria. The selection input 312 is provided to
the feedback controller 306 to allow the feedback controller 306 to
select one or more of the metrics y and controllable variables u to
couple according to some embodiments.
[0024] Although the comparator 308 is shown outside the feedback
controller 306, it is noted that the comparator 308 can be provided
inside the feedback controller 306. As used here, the term
"feedback controller" is intended to refer to either the feedback
controller with the comparator 308, or the feedback controller
connected to the comparator 308.
[0025] As further shown in FIG. 3, the reservoir model 200 (which
is stored in the storage media 302) can include multiple regions
(region 1, region 2, and region 3 shown) that correspond to
respective portions of a reservoir that can be selected for
simulation by the simulator 304. Alternatively, instead of
separating the reservoir model 200 into multiple regions, the
reservoir model 200 can instead represent the entire reservoir.
[0026] The reservoir model 200 also includes controllable flow
entities 204. A controllable flow entity refers to any component
associated with a reservoir that is able to control fluid flow to
or from the reservoir. For example, a controllable flow entity can
include a well to inject or produce fluids, a component to inject
heat into the reservoir, and so forth.
[0027] Examples of the metric y include any one of the following:
pressure, temperature, flow rate (at bottom-hole, tubing head,
across pipes, across chokes, at surface equipment, etc.), flow
volume, phase saturation, and so forth.
[0028] Examples of the controllable variable u include any one or
more of the following: flow rate at a selected component (downhole
or at the surface); pressure at a selected component (downhole or
at the surface), a component setting, a constraint on fluid flow, a
constraint on temperature, or any other property that affects the
behavior of a controllable flow entity as modeled by the reservoir
model 200. The constraint (u) output by the feedback controller 306
is set such that the difference between the actual value of the
metric y and the target metric value, y.sub.sp, will be reduced (or
minimized).
[0029] A feedback controller is a formal control mechanism that
aims to bring a process operating condition to a target value by
modifying the setting(s) of the process (control variable, u) based
on the discrepancy (error e) between the observed process output
(y) and the desired target value (the set-point y.sub.sp). In the
context of some embodiments, the "process" is the simulator. The
error is simply the difference between the process (simulator)
output and the target value:
e=y.sub.sp-y. (Eq. 1)
[0030] The generic nature of the feedback controller 306 according
to some embodiments allows a flexible system with which the list of
observable metrics (y) and controllable variables (u) may be
combined in order to evaluate many different control scenarios.
[0031] As noted above, the feedback controller 306 can be either a
PID controller or a fuzzy logic controller.
[0032] A PID controller has the following form in a continuous time
domain:
u ( t ) = K P [ e ( t ) + 1 T I .intg. 0 t e ( t ) t + T D e ( t )
t ] , ( Eq . 2 ) ##EQU00001##
where:
[0033] t is time
[0034] u(t) is the feedback controller output (process input) at
time t,
[0035] e(t) is the observed error (y.sub.sp-y) at time t,
[0036] K.sub.P is the proportional gain constant,
[0037] T.sub.I is the integral time constant, and
[0038] T.sub.D is the derivative time constant.
[0039] Since the process according to some embodiments is a
numerical simulator, which is discrete in time, the discrete form
of Eq. 2 is:
u ( k ) = K P [ e ( k ) + 1 T I i = 0 k e ( i ) .DELTA. t ( i ) + T
D .DELTA. e ( k ) .DELTA. t ( k ) ] , ( Eq . 3 ) ##EQU00002##
where:
[0040] k is the current time-step index,
[0041] e(i) is the observed error at time-step I,
[0042] .DELTA.e(i) is the change in error over the previous
time-step (e(i)-e(i-1)), and
[0043] .DELTA.t(i) is the i.sup.th time-step size
(t.sub.i-t.sub.i-1).
[0044] It can be seen that PID output is determined by three
components (proportional component, integral component, and
derivative component), as described in
TABLE-US-00001 TABLE 1 COMPONENTS CONTRIBUTING TO A PID CONTROLLER
OUTPUT Tuning Term Parameter Purpose Proportional e(k) K.sub.P The
proportional term makes a change to the output that is proportional
to the current error value. Higher proportional gain helps the
process output reach the set-point quicker. Integral i = 0 k e ( i
) .DELTA.t ( i ) ##EQU00003## K.sub.P/T.sub.I The integral term
takes all previous errors and their durations into consideration
and accelerates the change of process output towards the set-
point. However if the integral term's contribution is too high, it
may cause the present process output to overshoot the set-point
value. Derivative .DELTA.e ( k ) .DELTA.t ( k ) ##EQU00004##
K.sub.P .times. T.sub.D The derivative term takes the current
change in error into consideration and slows the rate of change of
the controller output in order to suppress the overshoot that
results from the integral term. Table 1. Controllers with no
derivative component are referred to as PI controllers, and
controllers with only the proportional term are referred to as P
controllers.
[0045] The performance of a PID controller depends on its
parameters K.sub.P, T.sub.I and T.sub.D. Thus, parameter tuning is
performed to improve the performance of the control system
according to some embodiments. One technique of performing
parameter tuning involves using a set of empirical values resulting
from a series of tuning experiments, such as that provided by the
Ziegler-Nichols technique, as described in K. .ANG.strom et al.,
"PID Controllers: Theory, Design, and Tuning," Instrument Society
of America" (1995).
[0046] The Ziegler-Nichols technique is carried out as follows:
[0047] 1. Disable the integral and derivative terms. [0048] 2.
Perform runs, varying K.sub.P, until the observed value (y) starts
oscillating with constant amplitude and frequency. [0049] 3. Record
the value for K.sub.P, referred to as K.sub.U (ultimate gain).
[0050] 4. Record the period length of the oscillation, referred to
as T.sub.U (ultimate period). [0051] 5. Calculate the
Ziegler-Nichols parameters from empirical formulas.
[0052] It is often not possible to achieve constant amplitude and
frequency, because the process response generally changes with
time. However, it is generally sufficient to use an averaged period
length when oscillations are observed.
[0053] In those cases where the derivative term's value as given by
the Ziegler-Nichols technique is too high, which can result in
oscillatory behavior, the derivative term's value is decreased by 1
or even 2 orders of magnitude. In this way, stable behavior can be
recovered while preserving the benefit from the derivative term's
action in terms of sharper convergence to the set-point.
[0054] Fuzzy logic feedback controllers represent a different
approach to control problems as compared to PID controllers. Fuzzy
logic feedback controllers are based on the foundations of fuzzy
logic. Fuzzy logic is designed to quantitatively represent
qualitative expert knowledge about a process. In traditional set
theory, an element is either in, or out, of a defined set. In other
words, the membership value of an element for a traditional set is
either 1 or 0. For instance, considering the error (e) which is the
difference in the target value (y.sub.sp)) and the actual value (y)
of a metric (e.g., average pressure) in a reservoir region, a
particular error value is either in the set of high error values,
or it is out. However, in the context of analyzing a reservoir, the
sets of pressure (or other metric) errors is not discrete. There
are different degrees of high and low errors in the context of
reservoir analysis. Fuzzy sets aim to capture this by defining
fuzzy set membership functions, whose membership may be defined by
a continuous number between 0 and 1. Examples of fuzzy set
membership functions are illustrated in FIG. 4. FIG. 4 illustrates
triangular (H) and trapezoidal (VH) fuzzy set membership functions.
Other possible types of fuzzy set membership functions include
Gaussian, bell-shaped and singletons.
[0055] Similarly, in the context of reservoir analysis, the
appropriate response to high pressure errors is also not discrete.
There are a range of responses that should be considered, depending
on the range of observations that are made. Hence, similar to fuzzy
set membership functions as in FIG. 4, there would be fuzzy sets to
capture the response (e.g., a high choke setting or a low choke
setting).
[0056] The relations between input and output fuzzy sets are set up
through fuzzy rules. These rules aim to capture an expected
response given the outcome of an observation, for instance:
[0057] If Error is High then Choke Setting is High.
[0058] This constitutes a fuzzy rule. There would be similar rules
for other fuzzy sets of error values (e.g. low, very low, etc.).
There could also be additional fuzzy sets that correspond to other
observations (e.g., the accumulation of errors).
[0059] Once the input/output fuzzy membership functions and fuzzy
rules are determined, the fuzzy logic feedback controller setup is
complete. Given inputs, a fuzzy logic controller processes the
inputs to come up with membership values of its output. This
process is called fuzzy inference (e.g., fuzzy inference using a
max-min technique).
[0060] After fuzzy inference has been completed, the fuzzy logic
feedback controller has to generate a real number that would be the
value for the controllable variable (u) to the simulator. One
technique for obtaining this real number is called defuzzification,
since it involves converting (defuzzifying) fuzzy set memberships
into a real value. There are many different ways to carry out
defuzzification. A center average technique for defuzzification can
be used in some implementations. For the center average technique,
the fuzzy logic feedback controller output is given by:
u = i = 1 n s h i u i i = 1 n s h i , ( Eq . 4 ) ##EQU00005##
where:
[0061] h.sub.i is the membership level of a particular fuzzy set
resulting from a particular input,
[0062] u.sub.i is the output value corresponding to input h.sub.i
that results from a fuzzy rule, and
[0063] n.sub.s is the number of active input fuzzy sets.
[0064] Thus, it is apparent that fuzzy sets tied together with
fuzzy rules may be used to control a process (simulator 304 in FIG.
3). In some cases, if constructed appropriately, the response of
the fuzzy logic feedback controller will be similar to a PID
controller. In fact, it is possible to come up with a fuzzy logic
feedback controller that has three inputs (P, I and D), and a
single output, and one which performs exactly the same as the PID
controller.
[0065] FIG. 5 illustrates a process performed by a control system
that employs a feedback controller and a simulator of a
subterranean structure. The feedback controller (306 shown in FIG.
3) receives (at 402) one or more observable metrics from the
simulator 304 during simulation of a subterranean structure (e.g.,
reservoir) using a model of the subterranean structure. The one or
more observable metrics can be associated with a particular region
(from among multiple regions) of the simulated subterranean
structure. Alternatively, the one or more observable metrics can be
associated with the entirety of the simulated subterranean
structure. Examples of observable metrics include pressure,
temperature, flow rate, flow volume, phase saturation, and so
forth. For a particular region, the observable metric can be an
average value of multiple values of the metric across the
region.
[0066] The feedback controller 306 then calculates (at 404) one or
more controllable variables that are input to the simulator 304 to
control settings associated with the simulator 304. Controllable
variables can include one or more of the following: flow rate at a
selected component; pressure at a selected component, a component
setting, a constraint on fluid flow, a constraint on temperature,
or any other property that affects the behavior of a controllable
flow entity as modeled by the model of the subterranean
structure.
[0067] The controllable variable(s) is (are) provided (at 406) from
the feedback controller 306 to the simulator 304. The controllable
variable(s) causes settings of the model to change, including
settings of controllable flow entities. Examples of controllable
flow entities include one or more of the following: a well to
inject or produce fluids, a component to inject heat into the
reservoir, and so forth.
[0068] During simulation, the controllable flow entity(ies) of the
model is (are) operated (at 408) according to the controllable
variable(s) set by the feedback controller 306.
[0069] The procedure of FIG. 5 can be iteratively performed
multiple times at discrete time steps. The time-step size should be
a fraction of the response delay of the simulator output to a given
change in the simulator input (feedback controller output). If the
time-step size is too large, the feedback controller will not be
stable since it will not be able to correctly capture simulator
input/output dynamics.
[0070] FIG. 6 is a flow diagram of a procedure to configure a
feedback controller according to an embodiment. A selection input
312 (FIG. 3) is received (at 502), where the selection input 312 is
regarding observable metric(s) of the simulator 304 to couple to
controllable variable(s) input to the simulator 304. The selection
input 312 can be based on user selection, such as user selection
received through a graphical user interface (GUI) or in a file.
Alternatively, the selection input 312 can be in the form of
selection criteria.
[0071] The feedback controller 306 is configured (at 504) according
to the selection input 312. The configured feedback controller 306
couples the selected observable metric(s) to the selected control
variable(s). The feedback controller 306 is then operated (at 506)
according to the selected configuration, where the feedback
controller 306 adjusts the selected control variable(s) according
to values of the selected observable metric(s) output by the
simulator 304.
[0072] Some example use cases are noted below. In one example, the
control system including the feedback controller and simulator is
useful in the simulation of reservoir management strategies
involving pressure maintenance. In pressure maintenance strategies,
average pressure could be maintained by controlling the choke
settings in a surface network model coupled to the reservoir
simulator.
[0073] The control variable for this example (which is manipulated
by the feedback controller 306) can be the production rate at the
well, where one of the following production rates can be selected
for use: oil, water, or gas rates. The simulator output metric that
can be selected to couple to the control variable can be the
average water saturation within a region of the simulated
reservoir. The set-point for the feedback controller 306 is the
target value for the simulator output metric, in this case the
target average water saturation. Another possible input includes
tuning parameters of the feedback controller 306 (PID controller),
such as tuning parameters obtained according to the Ziegler-Nichols
technique discussed above.
[0074] As another example, the feedback controller 306 can be used
to control the average temperature (simulator output metric) within
a region of a simulated reservoir. The control variable input to
the simulator 304 that can be adjusted by the feedback controller
306 in this case can be the steam injection rate such that a
constant average temperature can be maintained.
[0075] By using the control system including the feedback
controller and simulator according to some embodiments,
sophisticated control mechanisms can be implemented. Also, as new
equipment becomes available in the field (e.g., sensors or unmanned
mechanical devices), using a feedback controller within a numerical
simulation can provide a way of assessing the economical
feasibility of installing such new equipment into existing
facilities.
[0076] The processes depicted in FIGS. 5 and 6 can be performed by
software, hardware, firmware, various logic, and so forth (or any
combination thereof).
[0077] Instructions of software described above (including the
feedback controller 306 and simulator 304 of FIG. 3) are loaded for
execution on a processor (such as processor 310 in FIG. 3). The
processor includes microprocessors, microcontrollers, processor
modules or subsystems (including one or more microprocessors or
microcontrollers), or other control or computing devices. A
"processor" can refer to a single component or to plural components
(e.g., one CPU or multiple CPUs in one computer or multiple
computers).
[0078] Data and instructions (of the software) are stored in
respective storage devices, which are implemented as one or more
computer-readable or computer-usable storage media. The storage
media include different forms of memory including semiconductor
memory devices such as dynamic or static random access memories
(DRAMs or SRAMs), erasable and programmable read-only memories
(EPROMs), electrically erasable and programmable read-only memories
(EEPROMs) and flash memories; magnetic disks such as fixed, floppy
and removable disks; other magnetic media including tape; and
optical media such as compact disks (CDs) or digital video disks
(DVDs).
[0079] While some embodiments have been disclosed, those skilled in
the art, having the benefit of this disclosure, will appreciate
numerous modifications and variations therefrom. It is intended
that the appended claims cover such modifications and
variations.
* * * * *