Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02445296 2006-03-06
. = .
APPARATUS AND METHOD FOR CONTROLLING THE SPEED OF A PUMP
IN A WELL
The present invention relates to a controller for pumps used in [oil] wells
and more
particularly to a method and apparatus for controlling a pump speed.
Background of the Invention
In recovery of oil from oil wells, pumps are used to draw crude oil from the
well
bore to the surface well head. The crude oil extracted generally consists of a
combination
of oil, natural gas, grit, wax and water. The pumps generally comprise two
types, namely,
continuous flow or on-off pumps, and are powered by either electrical or
natural gas
motors. Upon emerging at the well head, the crude oil is passed via a pipe to
separation
tanks where the oil is removed from the mixture extracted from the well bore.
The oil
may also be temporarily stored in the separation tanks.
The maximum obtainable production rate for a well depends on the rate of
migration of crude oil from its geological formation to the well bore. The
well bore is
unique in having both an inflow and an outflow. The inflow represents the
quantity of
crude oil that a local formation can deliver to the well bore, whereas the
outflow (or rate
capacity) represents the quantity of crude oil that can be delivered to the
surface (or well
head). Typically, the quantity of oil that a pump is able to extract from the
well bore (or
rate capacity) exceeds the rate of flow of the crude oil from the local
formation into the
well bore. This situation is normally exacerbated with age of the well. Also,
the actual
flow rate of crude oil into the well bore can deviate significantly at any
particular point in
time from an average flow rate for that well.
Thus, it may be seen that if the rate capacity of a pump exceeds the rate
capacity
of the well, the pump is then operating below maximum efficiency. As the cost
of
operating the pump is relatively high, this reduced efficiency translates into
a wasted
cost. Furthermore, severe pump degradation may be caused by having a pump
operate
above the well production rate. Conversely, if the pump rate falls below the
wells
production rate, oil accumulates in the well bore resulting in an equilibrium
established
between oil flowing into the well bore from the formation and causing a
resultant drop in
production. Furthermore, for progressive cavity
1
CA 02445296 2006-03-06
type pumps or continuous flow pumps it is necessary to always maintain fluid
in the well
bore. Thus, control of the pump rate is relatively more critical in this case.
Thus, there exists the need for a method and apparatus to control pump rates
in
response to changing rates of oil flow. There has been many attempts in the
prior art to
mitigate some of these problems, and in particular, the reader is referred to
the applicant's
U.S. patent number 5,525,040 which describes prior art attempts.
Summary of the Invention
A method for controlling production flow in a well, said method comprising:
a. receiving a signal indicative of a production flow associated with said
well;
b. providing a speed control signal to a pump operating in said well to
control
said pump speed by increasing the pump speed while monitoring the flow and
continuing to increase the speed if the monitored flow increases.
A production flow controller for a well, comprising:
c. an input for receiving a signal indicative of a production flow associated
with
said well; and
d. a circuit, providing a speed control signal to a pump operating in said
well to
control said pump speed by increasing the pump speed while monitoring the
flow and continuing to increase the speed if the monitored flow increases.
A further aspect of the invention provides for the predetermined parameter
being
the pump speed.
A still further aspect of the invention provides for a processor means
including:
a. means for determining a temperature difference between the first
and second temperature sensing means the temperature difference
being indicative of a flow rate in the well;
b. means for generating the output signal being indicative of a pump
speed;
2
CA 02445296 2003-10-14
(c) means for storing a table of flowrates versus the predetermined pump
speeds;
(d) means for determining a rolling average of the flowrates;
(e) means for comparing the current rolling flow average to a stored flowrate
and
either incrementing the pump speed if the stored flowrate exceeds the average,
or decrementing the pump speed if the flowrate is less than the average;
(f) means for updating the table.
A further aspect of the invention provides for the temperature-sensing means
to be a
linear RTD.
Brief Description of the Drawings
A better understanding of' the invention will be obtained by reference to the
detailed
description below in conjunction with the following drawings in which:
Figure 1 is a block diagram of a controller according to the present
invention;
Figure 2 is a cross-sectional view of a probe according to the present
invention;
Figure 3 is a schematic diagram of the controller unit shown in Figure 1;
Figure 4 is a diagram of an RTD response curve;
Figure 5 is detailed circuit diagram of the controller unit of figure 3;
Figure 6(a) is a flow chart of a variable speed control algorithm;
Figure 6(b) is a detailed flow chart of the set-speed step of Figure 6(a); and
Figure 7 is a flow chart of an on-off speed control algorithm.
Detailed Description of Preferred Embodiments
Referring to figure 1, a block diagram of a pump controller is shown generally
by
numeral 10. A variable speed purnping unit 12 extracts crude oil from a well
bore 14. which
is then pumped via a conduit 16 to a holding tank 18, or the like. The pump
control system
includes a sensor 20 which is placed in the path of the oil flow in the
conduit 16, in a manner
to be described below. The sensor 20 provides an electrical signal indicative
of flow via a
cable 22 to a main control unit 24. The control unit 24 provides a control
signa126 to control
the variable speed pump unit 12. "1'he control
3
CA 02445296 2003-10-14
signal 26 maintains the pump speed at an optimal level, in order to ensure
efficient extraction
of crude oil from the well bore 14. An external computer 28 may be connected
to the
controller unit 24 in order to download or control parameters of the
controller.
Furthermore, the computer 28 includes a graphical display system for
displaying
information on the controller performance. Each of these elements will be
discussed in detail
below.
Referring to figure 2, a cross-section of the sensor 20 in figure 1, is shown.
The
sensor 20 is a passive device in that it must be powered from the controller
24. The sensor
includes a cylindrical body section 30 and a lower threaded section 32 for
installing in a bore
of a T-pipe section 15 in the conduit 16. Generally, the sensor is installed
relatively close to
the well head. A pair of probes 34 and 36 project from one end of the body 30
so that when
the sensor is inserted into the conduit 16, oil can flow over each of the
probes uniformly. The
actual orientation of the probes within the conduit is not critical, however,
the probes should
project generally perpendicularly to the direction of flow in the conduit. The
probes 34 and
36 are each comprised of a hollow polished stainless steel tube and each
contain a heating
element 38,42 and a temperature sensing element 40,44, respectively. A heating
current
derived from the controller 24 is provided to the heating element 38 and 42
via a suitable
electrical conductor 46 and temperature measurement signals are returned from
the
temperature sensing elements to the controller via a pair of conductors 48.
The conductor 46
and 48 are attached to a connector 49 which may be attached to cable 22.
The sensor operates on a thermal dispersion principle based on Newton's law of
cooling. One of the probes is selected and its heating element is supplied
with a constant
energy, which radiates out as heat. We generally refer to this probe as the
energized probe. Its
counterpart probe or unheated probe is generally called the ambient probe.
Both the probes
provide a temperature signal from their respective temperature sensing
elements. Thus, it may
be shown that the heat input rate into a medium may be expressed by the
equation Q = h t,
where h is the convection heat transfer co-efficient and Ot is the temperature
difference
between the heat source and the medium. In this case, At is the temperature
difference
between the heated and ambient probes. The value h is a function of the flow
rate of the
medium. Hence, h is not constant. Thus it may be seen that the
4
CA 02445296 2003-10-14
temperature differential, between the probes is inversely proportional to the
flow rate of the
medium for a given heat input rate Q.
It may be more accurately stated that the velocity of the fluid is a function
of the
inverse of the square of the difference in temperatures between the two
probes. By heating
one of the probe tips at a constant rate, the difference in temperature
between the probe tips
provides a relative temperature measurement independent of the ambient
temperature of the
fluid.
The calculated velocity of the fluid is proportional to the square of the
energy transfer
into the probe. Therefore, it is important that the energy supplied to the
probe is stable over a
wide range of ambient conditions. Furthermore, in situations where high flow
exists, most of
the radiated heat is absorbed by the passing fluid and carried down stream.
The temperature
thus recorded at either of the energized or ambient probe is approximately the
same.
However, with reduced fluid movement across the probes, residual heat builds
up along the
tip of the energized probe thus resulting in a higher temperature measurement
relative to the
ambient probe. By comparing the energized probe temperature to the ambient
probe
temperature, the flow rate can be estimated to produce a value which is
substantially
independent of the temperature of the oil flowing past the probe. Additional
compensation for
the variation of constant fluid properties from well to well with temperature
is implemented
in the controller 24.
Referring now to figure 3, the controller 24 is shown in greater detail. The
sensor
electronics is shown schematically by block 20. The controller 24, includes a
heater coristant
current source supply 51 which provides a constant current to the heater
elements 38 and 42
located in the sensor 20. Each of the heater elements 38 and 42 are connected
to a respective
switch 54 and 56. These switches 54 and 56 are selectively controlled via a
micro-
controller 58 for selecting either one of the heater elements 38 or 42 to be
heated.
As described earlier, each of the heater elements has in close proximity
thereto a
temperature sensing element 40 and 44. The temperature sensors in this case
are platinum
RTDs (resistance-to-temperature devices). As may be seen in figure 3, each of
the RTDs 40
and 44 have one of their inputs 59 connected via a switching multiplexer 60 to
an RTD
constant current source 66. The output of the temperature sensor resistors and
40 are
connected via the multiplexer 60 to the analog input of an analog-to-digital
converter 64
through a buffer amplifier 65. The analog-to-digital converter 64 provides a
digital input to
5
CA 02445296 2003-10-14
the micro-controller 58 which is indicative of the temperature measured by a
respective
RTD 40 or 44. As seen in figure 4, the RTD devices are linear devices and are
capable of
exhibiting a linear resistance change over an approximate temperature range of
-19 C to
150 C. The micro-controller 58 then processes this input data described with
reference to
figures 6(a), 6(b) and figure 7. A digital-to-analog converter 67 has its
digital inputs driven
by an output of the micro-controller 58 to produce an output analog signal
indicative of a
speed control signa126 for control of the pumping unit 12 shown in figure 1.
In addition, an RS232 interface and driver support circuitry 72 is provided
for
communication with the micro-controller 58 by the external computer 28.
Additional E 2
PROM 73 is provided for storage of constants and additional parameters.
Referring to figure 4, a resistance-to-temperature graph 74 illustrating the
relationship
between the resistance and temperature of the RTD is shown generally by
numeral 80. It may
be seen that the relationship is relatively linear over a large temperature
range. This has the
advantage in that over a period of time, the temperature of the resistor may
be sampled by the
analog-to-digital converter 64 and an integer interpolation routine may be
used to determine
values of resistance between the sampled points. Thus, it is not required that
a large amount
of memory be utilized in the micro-controller in order to store a lookup
table, as for example,
when a non-linear thermistor is used as temperature sensing element.
By providing heating elements in each of the probes of the sensor 20, allows
for each
of the probes to be periodically made the energized probe. In the case of oil
wells with high
paraffin wax content, if only one of the probes is heated, then over a long
period of time, wax
would tend to accumulate on the unheated probe. This would result in skewed
temperature
readings. However, by providing heaters in both probes and providing a means
for switching
between the heaters in the probes reduces wax build up on the probes.
Furthermore, the
lifespan of the sensor is extended by switching the heating elements between
the probes since
constant heating of only one of the probes results in severe degradation of
the lifespan of that
probe.
Figure 5 is a detailed circuit diagram of the controller 24, wherein the micro-
controller is a type 68HC705.
Referring now to figures 6a and 6b, an algorithm implemented by the micro-
controller 58 for controlling the output signal 26 to the pump, is indicated
generally by
numeral 90. The micro-controller switches the constant power source 51 to one
of the
6
CA 02445296 2003-10-14
heaters 38 or 42 by activating one of the switches 54 or 56. The micro-
controller then obtains
a first Ti and second T2 digitalized temperature measurement from the input
signal received
from the analog-to-digital converter 64 by sending a signal to the multiplexer
60 to select in
sequence the temperature probe 40 or 44. The difference between these
temperatures T is
calculated and is indicative of a flow measurement. These flow measurements or
temperature
differentials are combined into an average of most recent samples called a
rolling flow
average. The micro-controller samples the temperature approximately once ever
second. The
controller stores a sixteen element rolling window of samples. Once sixteen
samples have
been included in a rolling window, the newest sample replaces the older sample
prior to the
latest average being calculated. That is, a rolling average is calculated over
a sample of
sixteen elements every second with each element being discarded after 16
seconds. The
process of obtaining flow measurements is continuous and proceeds in parallel
with other
processing by the micro-controller.
Once this flow is obtained by the micro-controller, tire oil flow at the well
head is
controlled in accordance with the sequence of steps illustrated in figures
6(a) and 6(b).
Initially, an auto reset clock 92 is set to count time down from 48 hours or
any other
convenient time. This clock serves to reset the parameters of the controller
in order to
accommodate drops in motor efficiency over time and to switch the heated
probe.
The micro-controller maintains a speed table of entries having rows of
measured flow
rates Mi and pump speed Si. Thus. at a step 94, this table is initialized. An
initial wait time is
then set at step 96. This period is initially set between 8 to 12 minutes.
It may be noted that for variable speed control applications, the digital-to-
analog
converter delivers 4 to 20 milliamps output signal. By convention, 4 milliamps
represents the
lowest speed setting So of the pump, while 20 rnilliamps represents the
highest speed Si.
setting of the pump. An increment or step in speed is generally designated as
1 milliamp
representing the least step up or step down for change in speed.
In implementing the variable speed control, it is assumed that each increase
in speed
corresponds to some increase in the maximum potential delivery rate of the
pump. Thus it is
the goal to operate the pump at the lowest speed with the delivery rate above
the current
production rate measured for the well. Thus, in order to achieve this, the
speed table, as
described earlier, keeps track by way of the rolling flow average of the
maximum delivery
rate obtained thus far for each selected speed of the pump.
7
CA 02445296 2003-10-14
Changes in speed occur on the basis of time intervals. The length of each
interval is
called the settled time T. Its purpose is to allow changes in the pump speed
and the well's
production rate to be reflected in the rolling flow average. By default, the
length of the settle
time is 2 minutes. At the end of each interval, depending on whether the
rolling average has
increased, decreased or stayed the same, a corresponding change in speed is
initiated. These
changes in speed may be made as a single increment or as an arbitrary number
of increments
per interval.
Thus, referring back to step 98 in figure 6a, an initial speed S; of the pump
is set. The
controller waits a predetermined time at step 99. A new speed is then set at
step 100
according to the algorithm of figure 6(b). The table is initially built from
the lowest speed So
upward, first, the speed is set to So and an initial flow Mo is obtained for
speed So. The speed
is then stepped up to S, and a corresponding flow M, is obtained. This is
repeated for
successive values of speed increments. It is assumed, however, that each step
between a
speed S1 and a speed S;_I corresponds to a corresponding step in the maximum
potential flow
rate. Therefore, if upon obtaining M,+i at speed S;+,, it is recognized that
M;+i <M;, then it is
clear that the well's current production rate is below what the pump can
deliver at speed S;+,.
For example, if M;+t is equal to M;, it indicates that the well at this time
is producing at a
constant rate which corresponds to a speed S;, Otherwise, if M;+1 is less than
Mi, it indicates
that during the settle interval at S,+i production from the well has
decreased. In this case, S,,
may represent a greater speed than is required to support the lowered
production rate.
Therefore, a search of the table is performed beginning at Si, down to So
until the lowest
speed having a maximum deliver rate above the current production rate is
found.
30
8
CA 02445296 2003-10-14
It may therefore be seen that building the speed control table occurs in
conjunction
with varying the pump speed. When production levels or flow rates from the
well increase,
the table is refined while the speed is increased. Conversely, when lower flow
rates are
measured from the well, the table is searched for the minimum speed required
to sustain that
flow rate.
To illustrate how the process of building a table is performed after a drop in
flow rate
is detected, let Sp represent the last speed prior to detecting a drop in flow
rate, and let Si be
the current speed. For example, Sp might be 12 mA and Si might be 9 mA. As
flow rate from
the well increases, the production rate at speed S; as measured by the rolling
flow average
will begin to approach Mi, which is the estimated maximum flow rate at Si. At
the end of an
interval, if the production rate is found to be closer to Mi, then the speed
is incremented up to
Si+1. Assuming production levels continue to improve, the speed is
successively increment up
to Sp,. As this point, the table is continued to be built until either flow
rate decreases or the
maximum speed Sõ is reached.
Alternatively, if at the end of the interval at speed Si, the production rate
may be
greater than Mi. In this case, M; is no longer the best estimate to the
maximum flow rate at Si.
The new flow rate is then substituted for the old value of Mi. The change to
Mi can also
impact M;+1 if the new value for M; is also greater than M;+,. Therefore, the
table is rebuilt for
Si+,. Thus, it may be seen that changes can precipitate through entries in the
table thus
allowing the controller to constantly fine tune its estimates based on better
information over
time. This is illustrated more clearly in figure 6(b). Once the new speed Si
is set at step 100, a
new settle time is set at step 102.
Besides the settled time, there are two other timing intervals involved in
variable
speed control. These are the initial wait and automatic reset time. The
initial wait time is
simply the settling time for the very first interval in building the table. As
such, it only occurs
once just after the instrument is reset or powered on. The initial wait is
typical h longer than
the settled time.
The automatic reset time is not directly related to variable speed control.
Instead. it is
simply a background timer which upon time out at step 104 initiates an
automatic reset of the
controller. This causes the speed table to be rebuilt. The automatic rest
serves several
purposes as described earlier.
9
CA 02445296 2003-10-14
Referring now to figure 7, a process flow for controlling an on/off type pump
is
shown generally by numeral 170. In this case, the micro-controller 58 may send
a signal to
the digital-to-analog converter 67 one of two signals, namely, a value
corresponding to a
pump-off signal or a value corresponding to a pump-on signal. Alternatively, a
relay 67' may
be provided which turns the pump 12 on or off. The process is divided into
four steps,
namely, establish flow 172, regulate flow 174, timing-out 176 and shut-in 178.
It is to be
noted that each step is associated with a single control parameter which
directs the process of
that step. A default setting is assigned to each control parameter. However,
these parameters
may be easily changed via the external computer 20. The parameters associated
with these
steps are establish flow period, regulate flow cutoff point, timing-out period
and shut-in
period. Generally, these parameters are set at a default value of 15 minutes,
25%, 1 minute
and 30 minutes, respectively.
The establish flow step 172 starts the pump and settles into an interval of
time called
the establish flow period 173. This establish flow period is indicative of a
flow of the current
state of the well. For example, this interval generally covers the time
required for oil to make
its way to the surface and past the probes. Although flow samples are obtained
by the
controller during this period, output signals to control the pump are not
provided during the
establish flow period. Once the establish flow period has expired at step 173,
the process
moves onto the regulate flow step 174.
In the regulate flow period 174, an ongoing flow sample is combined into a
rolling
average called the rolling flow average as described earlier. However in this
case. a rolling
flow average is compared against a regulated flow cutoff point 175. If the
rolling flow
average remains above the cutoff point, a process control cycle remains at
this step. However,
should the rolling flow average drop below the regulated flow cutoff point,
this signals a
pumpoff has occurred and the process moves on to the liming-out step 176.
In the time out step 176, a short period called the time out period is
provided to
confirm whether or not the well has actually pumped off. This avoids instances
where trapped
gas pockets are within the line or short segments of dry pumping have
occurred. During
timing out, the ongoing rolling flow average continues to be compared against
the regulated
flow cutoff point 177. If the rolling average moves back above the cutoff
point
CA 02445296 2003-10-14
before timing out period expires, then the process moves back to the regulate
flow step 174.
Otherwise, at the end of the timing out period, the process moves to the next
step which is the
shut-in step 178.
In the shut-in step 178, the pump is stopped and the well enters an idle state
allowing
time for the well bore to be refilled from the surrounding formation. The
length of time the
well remains idle is determined by the shut in period. Once the shut in period
expires, the
process control begins at the establish flow step 172.
While the invention has been described in connection with a specific
embodiment
thereof and in a specific use, various modifications thereof will occur to
those skilled in 10
the art without departing from the spirit of the invention as set out in the
claims.
The terms and expressions which have been employed in the specification are
used as
terms of description and not of limitations, there is no intention in the use
of such terms and
expressions to exclude any equivalents of the features shown and described or
portions
thereof, but it is recognized that various modifications are possible within
the 15 scope of the
invention as set out in the claims.
11
