Method to save energy for devices with rotating or reciprocating masses

Abstract

A system and method are disclosed for turning off the voltage to a pump jack electric motor during predetermined periods of time to save energy. In the method, the motor's response to closed-loop control may be evaluated over several pump strokes. The periods of the pump stroke when it is feasible to turn off the motor may be identified. The consistency of the measurements over several strokes may be evaluated. The motor may be turned off during predetermined periods on subsequent pump strokes when each pump stroke shows sufficiently similar behavior to that predicted during the closed-loop control process. The system may return to the closed-loop control process after a predetermined period of time to adjust to any changes in the system.

Description

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

N/A

REFERENCE TO MICROFICHE APPENDIX

N/A

BACKGROUND OF THE INVENTION

1. Field of the Invention

This invention relates to electric motors used to operate pump jacks and other devices with rotating or reciprocating masses.

2. Description of the Related Art

A pump jack is an above ground driving device for a reciprocating piston pump installed downhole in an oil well. The pump jack mechanically lifts liquid out of the well when there is not enough bottom hole pressure for the liquid to flow by itself to the surface. The pump jack is often powered by an electric motor that receives electrical power from an electric utility grid. A pump jack converts the rotary mechanism of the motor to a vertical reciprocating motion to drive the downhole pump. There are many different designs of pump jacks, including, but not limited to, conventional, the Lufkin Mark II, beam-balanced, air-balanced, slant hole and conventional portable. Pump jacks are available from many different suppliers, including Lufkin Industries, Inc. of Lufkin, Tex. and Cook Pump Company of Coffeyville, Kans.

The pump jack electric motor usually rotates a set of pulleys to a gear system or transmission, which in turn drives a pair of cranks or crank arms. For a typical conventional pump jack design, the cranks raise and lower an end of a lever or beam, known as a “walking beam,” that is pivoted on a sampson post or A-frame. A curved metal box known as a “horse head” is on the other end of the walking beam from where the crank arms are connected with the beam. A counterweight or reciprocating mass is typically attached to one end of the cranks. A pitman arm usually spans between the counterweight and the end of the walking beam opposite the horse head. A cable connects the horse head to a vertical polished rod, which is connected to the vertical string of tubulars or sucker rods miming to the downhole pump.

The counterweight assists the motor in lifting the string of sucker rods or tubular string. When the motor lifts the counterweight upward, the horse head moves downward, pushing the sucker rods or tubular string downward. After the counterweight reaches the top of its rotation, it swings around and assists the motor to rotate the walking beam in the opposite direction using the counterweight's momentum and mass (kinetic energy). When the counterweight is free-falling downward from its uppermost position, the horse head moves upward, lifting the string of sucker rods upward. U.S. Pat. No. 4,051,736 proposes an improved pump jack for reciprocating an oil well pump.

Although there are different downhole pump designs, downhole pumps have traditionally comprised a plunger or piston reciprocating within a pump barrel located at or near the end of the production tubing. Two independent valves typically accomplish the pumping action. A standing check valve may be secured in the pump barrel beneath the piston, and the piston may include a traveling check valve. The upstroke of the piston opens the standing valve, and draws fluid into the pump barrel as the traveling valve remains closed. The downstroke of the piston opens the traveling valve and forces upward the fluid from the pump barrel as the standing barrel remains closed. U.S. Pat. Nos. 3,578,886; 4,173,451; and 6,904,973 propose downhole pumps.

It is well known that electric motors can enter an energy generation mode of operation. For an electric motor used with a pump jack, an energy generation mode can occur at any time during the rotation of the counterweight, depending on the condition of the balance between the counterweight and the tubular or rod string. The condition of the balance may fluctuate from pumping stroke to stroke, depending on the amount and composition of fluid being lifted by the rod string in each stroke. The polished rod and attached sucker rod or tubular string may be moving upwards or downwards in the energy generation mode.

A well owner must pay his electrical bill based upon the amount of power the pump jack motor consumes. The amount of energy consumed is measured by an energy meter. In the past, the amount of power consumed was measured by an analog electricity meter. Many digital electricity meters are now used. The energy meter, whether of analog or digital design, may be configured, at the discretion of the utility company, to allow or prevent crediting the customer for generated energy that is supplied back to the power grid. A pump jack system is such an inefficient generator that the quantity of consumed energy required to produce any generation significantly exceeds the generated energy. Therefore, regardless of whether the utility company credits generated energy, it is always beneficial to the customer to avoid energy generation.

During periods of generation, a motor will attempt to attain a voltage that exceeds the utility's line voltage, thereby causing current to flow in the opposite direction. The load provided by the utility grid serves as a brake, limiting the acceleration of the motor that would have otherwise occurred. This braking action of the motor prevents the falling weights of the pump jack from developing additional kinetic energy that might have assisted the pumping action. This converted kinetic energy could have served as an alternative to electrical energy from the utility grid.

In the past, engineers have tried unsuccessfully to save significant amounts of energy by turning off the pump jack electric motor during a portion of the pump jack cycle that may have included a period of generation. This has been attempted with various mechanical switches and relays. However, the parameters of the downhole pumps and wells vary over time, so these mechanical solutions have not worked.

Fluid flow in the well may vary as the well fills, and then “pumps off.” In some cases the volume of fluid pumped may change from one stroke to the next. The changing volumes, densities, viscosities, weights, and other properties of materials and/or fluids pumped, such as gas, oil, water, and slurry, may greatly alter the combined weight of the rod string and the column of fluid, thereby affecting the balance of the system and the demand on the motor. In some wells the tubular strings may be thousands of feet in length. The influx of different fluids into the well over time will significantly impact the operation of the motor.

With the introduction of the microprocessor, it became possible to turn off the electric motor by observing the current and voltage. However, the problem was in knowing when to turn the electric motor back on. Various open-loop fixed time delays were attempted in the past, but these attempts failed since the parameters of the downhole pumps and wells vary over time. Failure to turn the motor back on at the appropriate time can cause diminished energy savings and/or other undesirable effects.

When an AC induction motor is lightly loaded, reducing the voltage supplied will cause the motor to operate more efficiently, thereby saving energy. This is particularly evident in the case of single-phase motors, and to a lesser extent, three-phase motors. A large three-phase motor is usually naturally efficient at any load greater than approximately one-third of the motor's rated load.

The variation in the phase angle between the voltage applied to a motor and the current it draws bears an inverse relationship to the power drawn by the motor. A high phase angle indicates a lightly loaded motor, and a low phase angle indicates a heavily loaded motor.

The majority of pump jacks use three-phase motors as small as 5 horsepower (HP), but typically greater than 20 HP. These motors may be subjected to a periodically varying load, with the periods typically ranging from 5 seconds to 12 seconds. During a typical pumping stroke, the motor experiences a heavy load once or twice and a light load once or twice. Depending on the geometry and balancing of the pump jack, the falling weights (either the counterweights or the rod string) can force the motor beyond its synchronous speed, thereby causing it to behave as a generator. During such a time, the phase-angle exceeds 90 degrees.

U.S. Pat. No. 6,489,742 proposes a motor controller that includes power conveyance to an induction motor with a digital signal processor that calculates and optimizes supply of current for existent motor loading from a power supply and main voltage through a control element. Pub. No. U.S. 2010/0117588 proposes a motor controller and method for saving energy in an AC induction motor at every load wherein the motor is calibrated at two or more load points to establish a control line, which is then programmed into a non-volatile memory of the motor controller.

Pub. No. U.S. 2009/0046490 proposes an IGBT/FET-based energy savings device, system and method wherein a predetermined amount of voltage below a nominal line voltage and/or below a nominal appliance voltage is saved. Pub. No. U.S. 2009/0051344 proposes a TRIAC/SCR-based energy savings device, system and method wherein a predetermined amount of voltage below a nominal line voltage and/or below a nominal appliance voltage is saved. Pub. No. U.S. 2009/0200981 proposes a system and method for providing constant loading in AC power applications wherein at least one turn-on point of at least one half cycle of a modulating sine wave is determined, at least one turn-off point of the at least one half cycle of the modulating sine wave is determined, and at least one slice located between the at least one turn-on point and the at least one turn-off point in removed. Pub. No. U.S. 2010/0033155 proposes a power supply for IGBT/FET drivers that provides separated, isolated power to each IGBT/FET driver.

Proportional-integral-derivative (PID) control is a widely used technique applied to control algorithms and feedback mechanisms. A PID controller, as it is generally referred to, calculates a value based upon an “error.” Typically, the “error” is calculated as the difference between a measured process variable and a desired set point or target value. The PID controller attempts to minimize the error by adjusting the process control variables. In essence, the PID controller is a digital filter that has proportional, integral, and derivative parameters. The proportional value determines the reaction to the current error, the integral value determines the reaction based on the sum of the recent errors, and the derivative value determines the reaction based on the rate at which the error has been changing.

The above discussed U.S. Pat. Nos. 3,578,886; 4,051,736; 4,173,451; 6,489,742; and 6,904,973; and Pub. Nos. U.S. 2009/0046490; 2009/0051344; 2009/0200981; 2010/0033155; and 2010/0117588 are incorporated herein by reference for all purposes in their entirety.

A need exists to efficiently manage the energy usage of a pump jack electric motor, particularly during the energy generation mode. It would be desirable to substantially eliminate the energy generation mode if possible.

BRIEF SUMMARY OF THE INVENTION

A pump jack electric motor may be supplied with the full unmodified line voltage, and a motor controller may monitor the phase angle. When the phase angle is observed to exceed a predetermined threshold phase angle, the motor may be considered to be extremely lightly loaded and on the verge of entering the energy generation mode. This is considered a potential turn-off time. Upon crossing the threshold phase angle, a closed loop motor controller system may be activated to control the supply voltage to the pump jack electric motor. By controlling the supply voltage to the motor, the observed phase angle may be maintained at a value no greater than a target phase angle.

By allowing some current flow, primarily of a reactive nature, the closed loop controller system remains enabled and an observable feedback parameter may be used as an indication of the motor load condition, to which the motor controller may react, allowing power to be supplied when an increase in load necessitates an increase in supply voltage. The indication of increased motor load condition may be detected from a decrease below a predetermined threshold in the firing angle used to control the motor controller's switching devices. This is considered a potential turn on time. The above process may be repeated over several pumping cycles to insure consistency. If a close correlation is observed, then the system may enter an open loop on-off state.

In the on-off state, the phase angle may be monitored during every pumping stroke to insure that it crosses the established turn-off threshold at the time predicted during closed loop control. If the phase angle does not meet the expected timing, the system may be returned to closed loop control to repeat the measurements. If the expected timing is met, the voltage supplied to the motor may be completely turned off at the established turn off time from the closed loop control. Immediately after being turned back on at the established turn on time, the observed phase angle may be evaluated and compared with a predetermined ideal phase angle value. If the observed phase angle exceeds the ideal value, it is an indication that the voltage could have remained off for longer. If the observed phase angle is less than the ideal value, then the voltage should have been turned on sooner. Adjustments may be made to the off duration to optimize the behavior on subsequent strokes. The voltage to the motor may remain fully on until the next predicted turn off time.

The system may be periodically forced back into the closed loop control to re-evaluate the timing. If there is little or no change observed in the timing, small or no adjustments may be made, and the system may return to the open loop on-off condition. If a substantial change is found in the closed loop control timing, then the system may remain in closed loop control until a consistent or repetitive timing pattern is observed, at which point the system may be returned to the open loop on-off state with new turn off and turn on times.

The system and method take advantage of the highly periodic variation of the load on the motor in a pump jack application. The system and method accurately predict those times during a pumping stroke when the power to the motor may be completely turned off, as well as the appropriate time at which the power should be reapplied. Completely turning off the power to the motor results in superior energy savings.

The system and method provide a means by which to monitor and characterize the periodic variation in the load on a pump jack motor based on the observed phase angle between the applied voltage and the consumed current. Once the load is characterized, a determination is made whether the opportunity exists to save energy by turning off the motor during portions of each stroke. The system and method predict when this opportunity should occur in subsequent pumping strokes and define criteria to be met in subsequent pumping strokes as conditions for turning the motor off and for turning it on again. An algorithm may be implemented that adaptively adjusts to gradual changes in the behavior of the pump jack, and safeguards may be implemented to recognize and react to sudden, large changes in this behavior.

BRIEF DESCRIPTION OF THE DRAWINGS

A better and further understanding of the present invention can be obtained with the following detailed descriptions of the various disclosed embodiments in the drawings in which like parts are given like reference numerals, and wherein:

FIG. 1 is a block diagram of a digital signal processor (DSP) with hardware inputs and outputs.

FIG. 2 is a block diagram of a DSP-based motor controller.

FIG. 3 is a diagram showing a phase rotation detection method.

FIG. 4 is a flow chart showing a phase rotation detection method.

FIG. 5 is a graph showing power control device outputs for positive phase rotation.

FIG. 6 is a graph showing power control device outputs for negative phase rotation.

FIG. 7 is a block diagram of a window comparator.

FIG. 8 is a schematic of the window comparator.

FIG. 9 is a graph of a current waveform and zero-cross signals.

FIG. 10 is a schematic of a virtual neutral circuit.

FIG. 11 is a graph showing power control device outputs for single phase applications.

FIG. 12 is a three-dimensional graph showing a three-dimensional control line.

FIG. 13 is a three-dimensional graph showing a control line projected onto one plane.

FIG. 14 is a graph showing a two-dimensional plotted control line.

FIG. 15 is a graph showing a sweeping firing angle/duty cycle in a semi-automatic calibration.

FIG. 16 is a graph showing a directed sweep of a firing angle/duty cycle.

FIG. 17 is a graph showing plotted semi-automatic calibration data.

FIG. 18 is a graph showing plotted semi-automatic calibration data.

FIG. 19 is a graph showing plotted semi-automatic calibration data.

FIG. 20 is a flow chart of a semi-automatic high level calibration.

FIG. 21 is a flow chart of a semi-automatic high level calibration.

FIG. 22 is a flow chart of a manual calibration.

FIG. 23 is a flow chart of a fixed voltage clamp.

FIG. 24 is a graph showing a RMS motor voltage clamp.

FIG. 25 is a graph showing a RMS motor voltage clamp.

FIG. 26 is a flow chart of a stall mitigation technique.

FIG. 27 is a graph showing the stall mitigation technique.

FIG. 28 is an elevational view of one embodiment of a pump jack positioned with a tubular string in a well.

FIG. 29 is a plot of observed phase angle versus time for a pump jack motor in an open loop mode.

FIG. 30 is the system block diagram connected to the motor.

FIG. 31 is a plot of observed phase angle versus time for a pump jack motor in a closed loop control mode with a reduction of motor voltage to achieve a target phase angle within one complete pumping cycle.

FIG. 32 is a single phase waveform plot of incoming line voltage.

FIG. 32A is a heavily chopped single phase waveform plot of the voltage supplied to the motor after the application of pulse width modulation (PWM) techniques.

FIG. 32B is a lightly chopped single phase waveform plot of the voltage supplied to the motor after the application of PWM techniques.

FIG. 32C is a variably chopped single phase waveform plot of the voltage supplied to the motor after the application of PWM techniques.

FIG. 32D is the plot of FIG. 31 illustrating the periods when heavy chopping, light chopping, and no chopping may occur.

FIG. 33 is a block diagram of a motor controller disposed between an existing control panel and a pump jack electric motor.

FIG. 34A is a plot of observed phase angle versus time for an exemplary pump jack motor in an open loop mode over several pumping cycles showing exemplary variations in phase angle.

FIG. 34B is a plot of observed phase angle versus time for the pump jack motor of FIG. 34A but in a closed loop control mode over several pumping cycles with reductions of motor voltage to achieve a target phase angle.

FIG. 34C is a plot of observed phase angle versus time for the pump jack motor of FIG. 34B in an open loop mode with the motor turned off during predetermined periods.

FIG. 35 is a flow chart of a summary of a method for turning off a pump jack electric motor during predetermined periods.

DETAILED DESCRIPTION OF THE INVENTION

With reference to FIG. 1, a block diagram of a digital signal processor (DSP) 1 and hardware inputs and outputs is shown. The DSP 1 can observe the operational characteristics of a motor and make corrections to root mean square (RMS) voltage for the motor that is running and under closed loop control. Hardware inputs 2 capture phase zero crossing inputs 36, phase line voltage 37, phase motor voltage 38 and current 9 and passed through the DSP 1 for processing and then onto power control devices through the power control device outputs 14.

Referring now to FIG. 2, a block diagram of a system and method of the DSP-based motor controller 4 is shown. First, the motor controller 4 reads the voltages 37 of each phase A, B and C and current 9 to capture the zero-crossing inputs 36. At this point voltage 13 and current 9 may be converted from analog to digital using converters 62. Next, computations 63 of motor phase angle for each phase are calculated to yield an observed phase angle 5. Next, a target phase angle 10 which has been derived from a preprogrammed control line 6 is compared to the observed phase angle 5. The difference between the target phase angle 10 and observed phase angle 5 yields a resulting phase error signal (11, 28) which is processed by a PID controller 12 which has proportional, integral and differential components. The output from the PID controller 12 is the new control voltage (13, 29) to the motor 3, which can be obtained through the use of power control devices 33, such as TRIACs, SCRs, IGBTs or MOSFETS, to yield power control device outputs 14 of RMS motor voltage 13 supplied with line voltages 50 for each phase for maximum energy savings.

In this closed loop system, the voltage 13 of each phase of the motor 3 and the current are continually monitored. The motor controller 4 will drive the observed phase angle 5 to the point on the calibrated control line 6 corresponding to the load that is on the motor. At this point, maximum energy savings will be realized because the control line 6 is based on known calibration data from the motor 3. The motor controller 4 can control the motor 3 just as if a technician set the voltage 13 by hand. The difference is that the DSP 1 can dynamically respond to changes in the load in real-time and make these adjustments on a cycle by cycle basis.

Referring now to FIG. 3, in a three-phase system, the motor controller 4 is used to automatically determine the phase rotation. Zero-crossing detectors on the line voltages provide an accurate measurement of the angle between the phase A line voltage zero crossings 15 and the phase B line voltage zero crossings 16. For positive phase rotation 18, the angle is nominally 120° and for negative phase rotation 19, the angle is nominally 60°.

Referring to FIG. 4, a flow chart for phase rotation detection is shown. After a power-on-reset (POR) 20, it is easy for the motor controller 4 to determine positive phase rotation 18 and the negative phase rotation 19. First, the time is measured from phase A line voltage zero crossings to phase B line voltage zero crossings 39. Next it is determined if the time is greater than or less than 90 degrees 40. If greater than 90 degrees, than it is an ACB rotation 42. If the time is less than 90 degrees, than it is an ABC rotation 41. The motor controller 4 can control three-phase or single-phase motors with the same basic software and hardware architecture. For the three-phase case, depending on the phase rotation, the motor controller 4 can drive power control device outputs 14.

Referring now to FIG. 5 which shows power control device outputs for positive drive rotation, the motor controller drives phase A power control device outputs 14 and phase B power control device outputs 14 together during the phase A line voltage zero crossings 15 turn-on time as indicated by the oval 22a. Similarly, the motor controller drives power control devices which drive phase B 16 and phase C power control device outputs 14 together during the phase B turn-on time as indicated by the oval 22b. Finally, the motor controller 4 drives phase C 17 and phase A power control device outputs 14 together during the phase C power control device outputs 14 turn-on time as indicated by the oval 22c. Note that the example shown in FIGS. 5 and 6 depicts a firing angle/duty cycle 23 of 90°.

Referring now to FIG. 6 which shows the TRIAC drive outputs for negative phase rotation, the motor controller 4 drives phase A power control device outputs 14 and phase C power control device outputs 14 together during the phase A line voltage zero crossings 15 turn-on time as indicated by the oval 22c. Similarly, the motor controller 4 drives phase B 16 and phase A power control device outputs 14 together during the phase B line voltage zero crossings 16 turn-on time, as indicated by oval 22a. Finally, the motor controller drives phase C power control device outputs 14 and phase B power control device outputs 14 together during the phase C line voltage zero crossings 17 turn-on time, as indicated by oval 22b.

Now referring to FIG. 7, a block diagram of a window comparator is shown. The DSP based motor controller uses the window comparator 88 to detect zero-crossings of both positive and negative halves of a current wave form. When RMS motor voltage is reduced by the motor controller, it is difficult to detect zero crossings of current waveform because the current is zero for a significant portion of both half cycles. First, motor current is provided 89, a positive voltage is provided 90 as a reference for a positive half cycle and a negative voltage is provided 91 as a reference. Next, the current, positive voltage and negative voltage are presented to two comparators 92 and are then passed through an operation (OR) gate 93 to create a composite zero-cross digital signal 94.

As further illustrated in FIG. 8, a schematic of the window comparator 88 is shown. The motor current is provided 89, a positive voltage is provided 90 as a reference for a positive half cycle and a negative voltage is provided 91 as a reference. Next, the current, represented as a positive voltage and negative voltage, is processed by two comparators 92 and are then passed to an OR gate 93 to create a composite zero-cross digital signal 94.

Further, FIG. 9 shows graphs of a current waveform 95, a positive voltage half cycle 96, a negative voltage half cycle 97 and an OR function 98.

Now referring to FIG. 10, a schematic of a virtual neutral circuit is shown. A virtual neutral circuit may be used as a reference in situations where three phase power is available only in delta mode and there is no neutral present for use as a reference. The virtual neutral circuit comprises three differential-to-single-ended amplifiers 77. Because phase to phase voltages are high, input resistors 78 are used to form a suitable attenuator 79 together with feedback resistors 80 and ground reference resistors 81. Because the danger exists of a loss of phase, protection diodes 82 are used to protect the differential-to-single-ended amplifiers 77. The differential-to-single-ended amplifiers 77 are coupled to a summing amplifier 83 through DC blocking capacitors 84 and summing resistors 85 together with the feedback resistor 80. The output of the summing amplifier 83 is boosted by amplifier 27 thereby providing a low impedance output which is at neutral potential. Additional resistors divide a supply rail thereby allowing the summing amplifier 83 to handle alternating positive and negative signals. An alternate connection is available in the event that a neutral 86 is available along with a jumper block for alternate neutral connection 87.

Referring now to FIG. 11 showing a power control device output 14 for a single-phase application, the output 14 for phase A is turned on each half-cycle based on a power control device output 14 derived from the voltage zero-crossing input 15. The power control device output 14 for phase B line voltage zero crossings and phase C line voltage zero crossings are disabled in the DSP 1 and the hardware may not be present. The power control device outputs 14 are not paired as they were in the three-phase case.

Referring now to FIG. 12 which illustrates a three-dimensional control line for the motor operating space of a motor bounded by an observed phase angle 5 on the y-axis. A controlled firing angle/duty cycle 23 showing the decrease in voltage is shown on the x-axis and the percent load 24 on a motor is shown on the z-axis.

Every motor operates along a parametrical control line 25 within its operating space. For example, when a given motor is 50% loaded and the firing angle/duty cycle 23 is set to 100°, a phase angle 5 of approximately 55° is observed.

The parametrical control line 25 shown in FIG. 12 is defined by five parametric operating points 26 ranging from a loaded case 44 in the upper left corner, to an unloaded case 45 in the lower right corner. Furthermore, the parametrical control line 25 has special meaning because it is the line where a motor is using the least energy possible. If the firing angle/duty cycle 23 is increased and the motor voltage 13 decreased then a motor would slow down and possibly stall. Similar results would be seen if the load on the motor 3 is increased.

As illustrated in FIG. 13, the parametric control line 25 may be parameterized and projected onto one plane described by phase angle 5 in the vertical direction and the firing angle/duty cycle 23 in the horizontal direction.

Further, as shown in FIG. 14, the parametrical control line 25 may be displayed on a two-dimensional graph. On the x-axis, increasing firing angle/duty cycle 23 may be equated with a decreasing motor voltage. This is because small firing angle/duty cycles result in high voltage and large firing angle/duty cycles result in low voltage. The motor controller will drive the observed phase angle 5 to the point on the control line 25 that corresponds to the load presently on a motor. To accomplish this, a DSP computes the phase angle 5 between the voltage and current.

Referring back to the block diagram of FIG. 2, the DSP 1 then computes the next target phase angle 5 based on the present value of the RMS voltage 13, or equivalently the present value of the firing angle/duty cycle. The difference between the observed phase angle and the target phase angle 10 results in a phase angle error, which is processed through a PID controller 12 or similar device to generate a new control target. This control target changes the voltage in such a way as to minimize the phase angle error. The target phase angle 10 is dynamic and it changes as a function of the firing angle/duty cycle.

As stated above, the motor controller 4 will drive the observed phase angle 5 to the point on the control line 25 that corresponds to the load presently on the motor 3. This operating point 26 provides the maximum energy savings possible because the control line 25 is calibrated directly from the motor 3 that is being controlled.

This method for calibration is called semi-automatic calibration. The semi-automatic calibration is based on the DSP 1 sweeping the control space of the motor. As shown in FIG. 15, sweeping the control space means that the DSP increases the firing angle/duty cycle 23 and records the current 9 and firing angle/duty cycle 23 of each phase at discrete points along the way. Thus, in this manner it is possible to see the beginning of the stall point 21 of the motor. A well-defined linear portion of observed calibration data curve obtained from sweeping the control space 7, which is used to determine points on the control line 6, has a constant negative slope at lower firing angle/duty cycles 23. Then, as the firing angle/duty cycle 23 continues to increase, the current 9 begins to flatten out and actually begins to increase as the motor 3 begins to slip and starts to stall, called the “knee” 31.

As shown in FIG. 16, subsequent sweeps can be directed at smaller ranges of motor voltages to “zoom in” on the knee. The motor controller 4 requires multiple sweeps in order to get data that is statistically accurate. There is a tradeoff between the number of sweeps and the time required to calibrate the control line 25. A measure of the quality of the calibration can be maintained by the DSP 1 using well known statistical processes and additional sweeps can be made if necessary. This is true because the DSP 1 has learned the approximate location of knee 31 from the first sweep.

There is little danger of stalling during the semi-automatic sweep because of the controlled environment of the setup. A technician or operator helps to insure that no sudden loads are applied to the motor 3 under test while a semi-automatic calibration is in progress.

The process of sweeping the control space can be performed at any fixed load. For example, it can be performed once with the motor 3 fully loaded and once with the motor 3 unloaded. These two points become the two points that define the control line 25. It is not necessary to perform the calibration at exactly these two points. The DSP 1 will extend the control line 25 beyond both these two points if required.

There are many numerical methods that can be applied to find the stall point 21 in the plot of the current motor voltage 23. As shown in FIG. 17, a method is to use the “least squares” method to calculate a straight line that best fits the accumulated data tabulated from the first five motor voltages 23.

The continuation of this method is shown in FIG. 18. Using the previous data points the value of the current 9 may be predicted. Graphically, the DSP 1 is checking for one or more points that deviate in the positive direction from the predicted straight line.

As shown in FIG. 19, the DSP 1 is looking for the beginning of the knee in the curve. The first point that deviates from the predicted control line may or may not be the beginning of the knee 31. The first point with a positive error may simply be a noisy data point. The only way to verify that the observed calibration data curve obtained from sweeping the control space 7 is turning is to observe data obtained from additional sweeps.

Semi-automatic calibration may be performed in the field. Referring now to FIG. 20, a flow chart showing how semi-automatic calibration is performed is shown. First the motor 3 is placed in a heavily loaded configuration 44. Ideally this configuration is greater than 50% of the fully rated load. Next a calibration button 32 on the motor controller 4 is pressed to tell the DSP 1 to perform a fully-loaded measurement. The DSP 1 runs a calibration 46 which requires several seconds to explore the operating space of the motor 3 to determine the fully-loaded point. The motor controller 4 indicates that it has finished this step by turning on an LED.

Next the motor 3 is placed in an unloaded configuration 45. Ideally this configuration is less than 25% of the rated load. Then a calibration button 32 on the motor controller 4 is pressed 47 to tell the DSP 1 to perform an unloaded measurement. The DSP 1 runs the calibration 46 to determine the unloaded point. The motor controller 4 indicates that it has finished calibrating both ends 47 of the control line 25 by turning on a light emitting diode (LED). The DSP 1 then determines the control line 48 using the two measurements and applies this control line when it is managing the motor 3. The values of the control line 25 are stored in non-volatile memory 49.

FIG. 21 shows a more detailed flow chart of the semi-automatic calibration. First a first calibration sweep is run 46 with the motor voltage set at a certain degree 51, depending on if it is a first sweep or previous sweeps have been run 106, in which the motor controller measures the motor 52 until the motor controller detects a knee 53. If a knee 53 is detected the firing angle/duty cycle is decreased by two degrees 54 and the phase angle and the motor voltage are recorded to the memory 55. This process is repeated to obtain at least four sweeps 56 to get a computed average value 57 of the phase angle and the firing angle/duty cycle. If during any step along the calibration sweep, the knee is not detected, then the firing angle/duty cycle is increased by at least one degree 58 and the nest step is measured 59.

An alternative method for calibration is called manual calibration. FIG. 22 shows a flow chart of manual calibration. First a motor is placed on a dynamometer 70. Next the motor is connected to a computer for manual control 71 which allows the motor to be run in a open-loop mode and the firing angle/duty cycle of the AC induction motor to be manually set to any operating point. Then the motor is placed in a fully unloaded configuration 45. Next the firing angle/duty cycle is increased and the RMS motor voltage is reduced 72 until the motor is just about to stall. The firing angle/duty cycle and phase angle are recorded and this becomes a calibrated point which is recorded 73. Then the motor is started with drive elements fully on 74. Then the motor is placed in a fully loaded configuration 44. Next the firing angle/duty cycle is increased or decreased until the RMS motor voltage is chopped by the motor controller 75 until the motor is just about to stall. The firing angle/duty cycle are recorded and this becomes another calibrated point which is recorded 73. Finally the two calibrated points are used to form a control line 76.

When the RMS line voltage is greater than a programmed fixed-voltage, the DSP controller clamps the RMS motor voltage at that fixed voltage so energy savings are possible even at full load. For example, if the mains voltage is above the motor nameplate voltage of 115V in the case of a single phase motor then the motor voltage is clamped at 115V. This operation of clamping the motor voltage, allows the motor controller to save energy even when the motor is fully loaded in single-phase or three-phase applications.

FIG. 23 shows a flow chart of the fixed voltage clamp. First a phase error is computed 64. Next a voltage error is computed 65. Then the RMS motor voltage of the AC induction motor is determined and compared to a fixed voltage threshold 66. If the RMS motor voltage is greater than the fixed voltage threshold then it is determined whether or not control target is positive 67. If the control target is positive then a voltage control loop is run 68. If the RMS motor voltage of the AC induction motor is less than a fixed-voltage threshold, then the control line closed loop is run 69 and the entire process is repeated. If the control target is determined not to be positive then a control line loop is run 69 and the entire process is repeated again.

In some cases, it may not be possible to fully load the motor 3 during the calibration process. Perhaps 50% is the greatest load that can be achieved while the motor is installed in the field. Conversely, it may not be possible to fully unload the motor; it may be that only 40% is the lightest load that can be achieved.

FIG. 24 shows an example of both load points being near the middle of the operating range. On the unloaded end 45 at the right of the control line 25, the DSP 1 will set the fixed voltage clamp 60 of the voltage at minimum voltage 35. When the load on the motor increases, the DSP 1 will follow the control line moving to the left and up the control segment 61. This implementation is a conservative approach and protects the motor 3 from running in un-calibrated space.

As further shown in FIG. 25, on the fully loaded end 44 at the left, the DSP 1 will synthesize a control segment 61 with a large negative slope. This implementation is a conservative approach and drives the voltage to full-on.

Referring now to FIG. 26, the DSP-based motor controller uses a special technique to protect a motor from stalling. First, the DSP actively monitors for a significant increase in current 99 which indicates that load on the motor has increased. Next, if a significant increase is observed 100 then the DSP turns motor voltage to full on 101. Next, the DSP will attempt to reduce motor voltage to return to the control 102 and the DSP returns to actively monitoring for a significant increase in current 99. This technique is a conservative and safe alternative to the DSP attempting to track power requirements that are unknown at that time.

As further shown in FIG. 27, a graph of the stall mitigation technique, the load on the motor is represented on an x-axis and time is represented on a y-axis. The bottom line represents the load on the motor 103 and the top line represents the power applied to the motor by the DSP 104. Prior to point a 105, the DSP is dynamically controlling the motor at a fixed load. In between point a 105 and point b 30, the load on the motor is suddenly increased and the DSP turns the motor voltage to full on. At point c 34, the DSP reduces the motor voltage to point d 43.

In FIG. 28, a pump jack 30′ is positioned on the ground adjacent well W. Prime mover or motor 6′ drives gear system or transmission 8′ with drive belt 18′. Motor 6′ may be connected with a electric utility grid for the supply of power. One end of counterweight arm or crank arm 10′ is disposed with gear system 8′, and the other end of counterweight arm 10′ is disposed with counterweight or rotating mass 12′. There are preferably two counterweight arms 10′, with counterweight 12′ disposed between them. Lever or walking beam 2′ pivots on sampson post or A-frame 14′. One end of pitman arm or beam arm 16′ is rotationally attached with one end of beam 2′, and the other end of beam arm 16′ is rotationally attached with rotating mass 12′ and an end of counterweight arm 10′. Beam protrusion or head 4′ is disposed on the end of beam 2′ adjacent to well W. As can now be understood, pump jack 30′ has a conventional design.

One end of cable 20′ is attached with beam head 4′, and the other end of cable 20′ is attached with polished rod or rod 22′. Rod 22′ is disposed with the substantially vertical tubular string or sucker rods 26′ extending in the well W through the production tubing to the downhole pump 28′. Tubular string may comprise sucker rods, pipe, tubulars, or other components used with a pump jack or other similar device to assist in pumping or lifting fluids from a well. The motor 6′ may drive the pump jack 30′ by rotating an end of the counterweight arm 10′ about a horizontal axis. As the counterweight 12′ moves upward, beam 2′ pivots about a horizontal axis on A-frame 14′ and moves the beam head 4′ downward. As the counterweight 12′ moves past its uppermost position, it free-falls downward due to gravity and its momentum, and beam 2′ pivots about A-frame 14′ and moves beam head 4′ upward. The pushing and pulling of the string of tubulars 26′ by the beam head 4′ operates the piston in the downhole pump 28′. The tubular string 26′ moves and reciprocates substantially vertically in the well W.

The motor 6′ is normally in energy consumption mode. However, the motor 6′ may be in the energy generation mode when the falling masses (either the counterweight 12′ or the rod or tubular string 26′) are free-falling, thereby accelerating the motor 6′ beyond its synchronous speed, where the speed is limited by the generated current. Although an exemplary conventional pump jack 30′ is shown in FIG. 28, it is contemplated that all pump jack designs, including, but not limited to, different conventional designs, the Lufkin Mark II design, beam-balanced design, and conventional portable design may be used with the embodiments of the invention. Although the embodiments are shown with pump jacks, it is also contemplated that all of the embodiments may be used with any device having a rotating or reciprocating mass.

Turning to FIG. 29, plot 36′ with observed phase angle on the vertical axis 32′ and time on the horizontal axis 34′ is shown for an electric motor attached to a pump jack, such as motor 6′ and pump jack 30′ in FIG. 28, in open loop mode. The embodiments of the invention described below with FIGS. 30-32D have not been attached to the electric motor; therefore, the motor is in the open loop mode. Second horizontal line 40′ is drawn at an observed phase angle of 90 degrees on the vertical axis 32′. When the plot 36′ exceeds an observed phase angle of 90 degrees, which it does in plot first segment 42′ above second horizontal line 40′, then the motor is in the energy generation mode. At those times when the motor is generating, rather than consuming energy, the current lags the voltage by a phase angle in excess of 90 degrees. The greater the phase angle during generation, the greater the power being generated. The motor is in the heavy energy consumption mode in plot second segment 44′ below first horizontal line 38′. First horizontal line 38′ is drawn at a target phase angle less than 90 degrees on the vertical axis 32′. The target phase angle is discussed in detail below with FIGS. 30 and 31.

In FIG. 30, closed loop motor controller 50′ is schematically shown connected to an electric motor 62′, such as motor 6′ in FIG. 28, which may be connected with a pump jack, such as pump jack 30′ in FIG. 28. Other pump jack designs are also contemplated for use with FIG. 30. Motor controller 50′ may be a PID controller. However, other closed loop motor controllers are also contemplated. A digital signal processor (DSP) based motor controller is contemplated, such as the DSP based motor controller in FIGS. 1 and 2, although other types of DSP based motor controllers are also contemplated. Closed loop motor controller 50′ may be connected with motor (6′, 62′) in the same manner as shown in FIGS. 1 and 2. A microprocessor based controller is also contemplated. In one embodiment, the closed loop controller system may have a PID controller as a component. In the closed loop control system or servo system 48′, controller 50′ may compute 52′ the observed phase angles from the voltage and current supplied to the motor 62′.

Advantageously, no sensors need to be positioned with the motor (6′, 62′), the pump jack 30′ or the downhole pump 28′. Further, the closed loop system 48′ may be adaptive to each individual downhole pump 28′ and to changing parameters and requirements of the pump 28′ and well W over time, including, but not limited to, changing volumes, densities, viscosities, weights, and other properties of materials and/or fluids pumped, such as gas, oil, water, and slurry. Voltage and current monitored by the system 48′ serve as an indicator of the well condition, allowing the system to be adaptive to the changing well parameters. Monitoring the voltage and current on a substantially continuous basis allows for a substantially continuous reading of well conditions. The closed loop system 48′ also adapts when the existing components of the pump jack system are replaced with other components having different characteristics, such as for example replacing the tubular string with a different tubular string having a different weight, or replacing the counterweight with a different sized counterweight, provided that the mechanical system is rebalanced after the components are substituted. After rebalancing of the mechanical system, the embodiments of the invention allow the energy savings to resume.

A target phase angle 58′ input into the controller 50′ may be compared with the computed observed phase angle 52′, and the error 60′ or difference between the two values determined by the controller 50′. It is contemplated that the target phase angle 58′ may be substantially 90 degrees, or the target phase angle 58′ may be greater or less than 90 degrees. At the time of installation, a target phase angle 58′ may be selected that produces optimum results for the motor in use. The target phase angle 58′ may be constant for all motor loads, such as 65 degrees, although other constant target phase angles 58′ are also contemplated. The target phase angle 58′ may also be a variable function of the motor load at any instant. The setting for the target phase angle 58′ may be the lowest possible target phase angle that maintains a sufficiently observable current flow at all times while still supplying sufficient power to meet the motor's requirements at all loads.

The motor controller 50′ may control the supply voltage 54′ applied to motor 62′ based upon the error signal 60′. When the error 60′ is significant because the observed phase angle is too large, such as during the period of open loop energy generation mode, controller 50′ may reduce the supply voltage to the motor 62′ to a lower value, such as to reduce the observed phase angle 52′ to the target phase angle 58′. When the error 60′ is significant because the observed phase angle 52′ is too small, such as during the heavy energy consumption mode, controller 50′ may increase the supply voltage 54′ to the motor 62′ to a higher value to move the observed phase angle 52′ to the target phase angle 58′. In this closed loop system 48′, the voltage and current may be continuously monitored and controlled by the motor controller 50′. It is also contemplated that the supply voltage 54′ may be controlled through the use of power control devices, such as TRIACs, SCRs, IGBTs, or MOSFETs, as shown in FIG. 2. Also, controller 50′ uses timers and pulse width modulation (PWM) techniques to control the supply voltage, which are discussed in detail below with FIGS. 32-32D. Other techniques are also contemplated.

Returning to FIG. 30, the controller 50′ reads the voltages of each phase and current in the motor 62′ to capture the zero-crossing points. FIGS. 5 and 6 of Pub. No. U.S. 2009/0046490 propose an oscillogram and circuitry diagram, respectively, of a volts zero crossing point determining means that is contemplated. Other types of volts zero crossing point determining means are also contemplated. Voltage and current may be converted from analog to digital using one or more analog to digital converters for monitoring and/or control purposes, as shown in FIG. 2. Controller 50′ may perform computations 52′ of motor phase angle to yield an observed phase angle. Controller 50′ may compare the observed phase angle 52′ with the target phase angle 58′ and control the motor supply voltage 54′ in response. The phase angle may be monitored in one or more phases. Controller 50′ may be used to automatically determine the phase rotation. A circuit diagram of a phase support means and phase rotation determination means that is contemplated is proposed in FIG. 7 of Pub. No. U.S. 2009/0046490, where multiple phase operations are employed.

Further, it is contemplated that the voltages may be monitored from phase-to-phase or from phase-to-neutral. A schematic of a contemplated virtual neutral circuit is in FIG. 10. Other virtual neutral circuits are also contemplated. A virtual neutral circuit may be used as a reference in situations where three phase power is available only in delta mode and there is no neutral present for use as a reference. It is also contemplated that a window comparator may be used to detect zero-crossings of both positive and negative halves of a current wave form. A window comparator is in FIGS. 7 and 8. Other window comparators are also contemplated. FIGS. 8, 9 and 10 of Pub. No. U.S. 2009/0046490 propose a circuit diagram and oscillogram, respectively, of a half cycle identifying means that is contemplated.

Turning to FIG. 31, plot 64′ with observed phase angle on the vertical axis 32′ and time on the horizontal axis 34′ is shown for an electric motor attached with a pump jack, such as motor 6′ and pump jack 30′ in FIG. 28, in closed loop mode. As in FIG. 29, there is a target phase angle of less than 90 degrees at first horizontal line 38′. Unlike in FIG. 29, the electric motor output represented in FIG. 31 is from a closed loop system 48′ disposed with the motor as shown in FIG. 30. Plot first segment 70′ in FIG. 31 is where the observed phase angle would exceed the target phase angle in open loop mode. However, in closed loop mode in plot first segment 70′ the error signal 60′ creates a control effort by the controller 50′ to reduce the supply voltage 54′ to the motor to maintain the target phase angle 38′. When the observed phase angle would exceed 90 degrees in open loop mode, the large values of observed phase angle create large values of the error signal 60′ in FIG. 30.

During plot first segment 70′, the motor is effectively turned off using PWM techniques, but without actually cutting the power to the motor. There is still current flowing in the motor during this time, which allows the controller 50′ to know when to increase the supply voltage to the motor needed during the energy consumption mode. The real component of the current may be reduced virtually to zero, leaving a reactive component greater than zero. By allowing some current flow when it is reducing voltage, mostly of a reactive nature, an observable feedback parameter is provided that is used in the closed loop control system 48′ as an indication of the load condition, to which the controller 50′ may react, supplying power when needed in the energy consumption phase.

Since the current is of reactive nature, the only power remaining is of an apparent nature. The current flow allows the controller to continuously observe the phase angle between the current and the voltage. The maximum motor voltage reduction occurs approximately at plot first location 66′ in FIG. 31 when the observed phase angle in open loop mode as shown in FIG. 29 would otherwise be at its maximum value greater than 90 degrees.

When the observed phase angle exceeds the target phase angle in closed loop mode, the supply voltage may be reduced with PWM techniques until the observed phase angle reaches the target phase angle. At the beginning of plot first segment 70′ in FIG. 31, the motor controller 50′ reduces the observed phase angle from open loop mode down to the target phase angle. The controller 50′ thereafter maintains the observed phase angle substantially at the target phase angle. Any further reduction in observed phase angle below the target phase angle may be interpreted as an increase in load, to which the controller 50′ may respond by increasing the supply voltage 54′ until the target phase angle is once again reached. The maximum increase of supply voltage to the motor occurs at plot second location 68′ when the observed phase angle drops below the target phase angle. When the counterweight or reciprocating mass is driven by the motor, the values of the observed phase angle will typically be smaller than the target phase angle, which will create an error signal that creates a control effort by the controller 50′ to increase the supply voltage to the motor. The motor is in the heavy energy consumption mode in plot second segment 44′ below first horizontal line 38′.

Turning to FIG. 32, waveform plot 200 of incoming line voltage is illustrated in single phase, although three-phase voltage is also contemplated. In FIG. 32A, PWM techniques have been used to chop out or remove voltage waveform plot segments 204 while leaving voltage waveform plot segments 202. FIG. 32A illustrates heavy chopping of the supply voltage in which large segments 204 of the voltage waveform are chopped out. FIG. 32B illustrates light chopping of the voltage waveform with PWM techniques, wherein the voltage waveform plot segments 206 that are chopped out are smaller than the chopped out segments 204 shown in FIG. 32A. In FIG. 32B, the waveform plot segments 208 that are left are larger than the waveform plot segments 202 that are left in FIG. 32A.

The heavy chopping in FIG. 32A occurs during the period that open loop energy generation mode would be occurring, such as in FIG. 31 at plot first location 66′. In FIG. 32D, the period of heavy chopping 210A is illustrated at plot segment 210. The reduction of voltage shown in FIG. 32A reduces the real component of the current virtually to zero, while leaving a reactive component greater than zero. This is the period when the motor is effectively turned off, while still leaving sufficient current to observe the phase angle.

When the motor is in heavy energy consumption mode, such as occurs in FIG. 31 at plot second segment 44′, then substantially no voltage waveform segments are eliminated, and the motor supply voltage is substantially as shown in FIG. 32. In FIG. 32D, the period of substantially no chopping 212A occurs at plot segment 212.

In FIG. 32D, the DSP controller is in control mode at plot locations 226 and 228. During those periods, the motor is not in heavy energy consumption mode and not in the period when open loop energy generation mode would be occurring. In control mode, light chopping 226A as shown in FIG. 32B may occur or variable chopping as shown in FIG. 32C may occur to control the motor voltage. This may happen when the motor is lightly loaded, saving energy while the motor is still consuming energy. Variable chopping in FIG. 32C uses PWM to chop waveform plot segments (214, 216, 218, 220, 222, 224) of varying sizes to control the motor voltage. The size of the voltage waveform plot segments (214, 216, 218, 220, 222, 224) chopped in FIG. 32C may all be different, leaving voltage waveform plot segments that are also all different sizes.

It should be understood that the motor controller may use any combination or permutation of light chopping, heavy chopping, variable chopping or no chopping to control the observed phase angle of the motor supply voltage to the target phase angle. The digital signal processor (DSP) or motor controller attempts to maintain a substantially constant observed phase angle and will chop the amount required to do so. The DSP controls the motor voltage based on observing the phase angle. The amount of the chopping of the supply voltage may vary.

When the electric motor running open loop is in energy generation mode, the load presented by the utility grid effectively acts as a brake on the motor, thereby limiting its speed. This occurs due to the generated voltage attempting to exceed the voltage presented by the utility, thereby causing the current presented to flow in the opposite direction. When the closed loop controller system and method is applied as shown in FIGS. 30-32D, this braking action may be effectively minimized or removed, and the motor and system will typically speed up during this time. This additional kinetic energy stored in the system will be used to perform a portion of the pumping action without consuming energy in the motor. Minimizing or substantially preventing energy generation eliminates the need to consume energy in other parts of the pumping cycle, thereby saving energy.

As can now be understood, the electric power supplied to the motor is “effectively” turned off during the energy generation mode that would occur in open loop, while maintaining the feedback signals of voltage and current to determine when to turn the electric motor back on when the observed phase angle is diminishing. This system and method will constantly adapt to changing parameters in the well, which could not be done in the past. For one example, the motor and system are adaptive to pumping two or more fluids at different times having different densities or weights. Voltage and current monitored by the system serve as an indicator of the well condition, allowing the system to be adaptive to the changing well parameters. By not entering the energy generation mode, the braking action that is created by the open loop energy generation mode may be minimized or eliminated, so the benefit of speed up in the system is obtained. By minimizing or eliminating energy that would otherwise be consumed by the system, energy savings may result both from reduction of the supply voltage to the motor and from the minimization or elimination of the braking action of the motor when in generation mode.

All types and designs of electric motors are contemplated for use with the different embodiments of the invention described above, including, but not limited to, AC induction motors and AC synchronous motors. All types and designs of pump jacks are contemplated for use with the different embodiments of the invention described above, including, but not limited to, all conventional designs, the Lufkin Mark II design, beam-balanced design, and conventional portable design. Although the embodiments have been shown with pump jacks, it is also contemplated that all of the embodiments described above may be used with any device having a rotating or reciprocating mass. Although some of the embodiments have been shown with single phase voltage and current, all of the embodiments of the invention are contemplated with single or multiple phase voltage and current.

Turning to FIG. 33, an exemplary installation of a motor controller 304 implementing the turn-off method is shown. Motor controller 304 may be a DSP based PID controller as described previously and shown in FIGS. 1 and 2, although other closed loop controllers are also contemplated as described previously. Motor controller 304 may be inserted in electrical connection between the existing electrical control panel 302 and the electric motor 306 that drives the pump jack 308. The electric power supply may run through a utility meter 300 to the control panel 302.

FIG. 34A shows the typical variation over time of the phase angle measured from an electric motor driving a pump jack. Plot 314 with observed phase angle on the vertical axis 310 and time on the horizontal axis (not shown for clarity) is shown for an electric motor driving a pump jack in open loop mode. First horizontal line 312 is drawn at an observed phase angle of 90 degrees on the vertical axis 310. A first pumping cycle or stroke occurs during first time segment 316, a second pumping cycle or stroke occurs during second time segment 318, and a third pumping cycle or stroke occurs during third time segment 320.

As exemplified in FIG. 34A, during a typical pumping stroke the pump jack motor experiences a heavy load once or twice, and the motor experiences a light load once or twice. Depending on the geometry and balancing of the pump jack, the falling weights (either the counterweights or the rod string) can force the motor beyond its synchronous speed, thereby causing it to behave as a generator. During such periods, the observed phase angle exceeds 90 degrees.

To begin implementing the on-off method, once initialized, the motor controller may monitor the phase angle and operate in one of three states.

State 1 of the Method

In State 1 of the on-off method, the load profile to which the motor is subjected by the pump jack is evaluated to determine whether there is an opportunity to turn off the motor, and, if so, to identify the appropriate turn-off and turn-on times during the pumping stroke. The load profile may be characterized by phase-angle variations. When turning off the motor during times of energy generation, the motor will accelerate, thereby altering the load profile. Hence, it is desirable to characterize this change. This is accomplished by subjecting the motor to the closed loop PID control process (“heavy chopping”) as a close approximation of the effect of the on-off process.

FIG. 34B shows the effect of the closed-loop PID control process on the motor and its load profile. During initial plot segment 322, the motor is supplied with the full, unmodified, supply line voltage, such as shown in FIG. 34A, during which time the phase-angle is monitored. When the phase angle is observed to exceed a predetermined threshold, such as 70 degrees, as shown at plot location 326, the motor is considered to be lightly loaded and on the verge of entering the energy generation condition. This instant is identified as a potential turn-off time in future pumping strokes. Upon crossing this threshold, the closed-loop PID control process is activated, controlling the motor voltage to achieve the target phase angle, as shown in second horizontal line 313, drawn at a constant target phase angle, such as 60 degrees. Other target phase angles are also contemplated. It is also contemplated that the target phase angle may not be constant, but may vary depending upon motor load condition.

During the closed-loop PID control process, sufficient current flow remains to provide observable feedback for the closed-loop control system. The closed-loop PID control process remains enabled until an increase in load necessitates an increase in the supply voltage above a certain threshold, which is shown at plot location 328. This may be indicated by a decrease, below a predetermined threshold, such as 90 degrees, in the firing-angle used to control the switching devices, such as TRIACS, SCRs, IGBTs or MOSFETS. This instant is identified as a potential turn-on time in future pumping strokes. Upon crossing this threshold firing angle, the full, unmodified supply line voltage may be restored to the motor until the lightly loaded condition is once again identified, which is shown at plot location 330.

To ensure that the above measurements are consistent, the above process may be repeated over several pumping strokes, such as for example four strokes. The period of the altered pumping stroke, and the identified turn-off and turn-on times within the stroke, may be compared among the repeated strokes. If a close or repetitive correlation is observed in all measurements over the several strokes monitored, the system may be considered suitable to enter the open loop on-off state. A close correlation or criteria may be considered to be a variation of not more than 200 milliseconds (ms) the periods of the pumping strokes. Another correlation or criteria may be a variation of not more than 150 ms in the turn-on and turn-off times. Other correlations or criteria are also contemplated. If the correlation requirements are not met, the system may remain in State 1, continuing to employ the closed-loop PID control process, until the required correlation is observed.

A typical pump jack pumping stroke normally has one or two periods during which the motor may be turned off. These two periods may be treated as different from each other in their timing and duration. Either one or both of the two periods may meet the criteria for entering the on-off state. While there may be more than two periods of light loading or generation mode in a single pumping stroke, only the two longest such periods may be considered to provide any meaningful energy savings. Additional periods normally only occur briefly, when the motor is in a “pumped-off” condition. A minimum off-time condition or requirement may be maintained to support a maximum of two off-times that exceed this target. An implementation that supports more off-times is also contemplated. In the event that additional generation conditions are observed, the system may employ the closed-loop PID control process during these times.

State 2 of the Method

Once in the on-off state, the phase-angle may be monitored during every pumping stroke to ensure that it crosses the established turn-off threshold at the time predicted during State 1. If the phase-angle fails to meet the expected timing (within a reasonable predetermined margin, as discussed above), the load profile may be considered to have changed since the timing was established during the closed-loop PID control state. The control system may be returned to State 1 to repeat its measurements.

If the expected timing is met, the voltage applied to the motor may be substantially immediately and completely turned off, as shown in FIG. 34C. It may remain off for the duration computed during State 1. Substantially immediately after turning back on, such as at plot location 324, the observed phase-angle may be evaluated and compared to a predetermined ideal phase angle value. The ideal phase angle ensures that the motor neither generates energy, nor consumes excessive energy immediately after turning back on. This ideal phase angle value may be 75 degrees, although other values are also contemplated.

If the observed value exceeds this ideal value, it may be considered an indication that the voltage could have remained off for longer. If the observed phase-angle value is less than this value, then the voltage could have been turned on sooner. Adjustments may then be made to the off-duration to optimize the behavior on subsequent strokes. The voltage applied to the motor may remain fully on until the next predicted turn-off time. The on-off state, as represented by the observable phase-angle, is shown in FIG. 34C. As there is no current flow when the supply is turned off, there is no observable phase-angle during those periods.

State 3 of the Method

To allow for the possibility of a gradual drift in the load profile over time, the control system may be periodically moved back into State 1 to reevaluate the closed-loop PID control process. This may be done every two minutes. Other time periods are also contemplated. If, after one pumping stroke, little or no change is found in the closed-loop PID control method, small adjustments may be made to the timing requirements, and the control system may return to State 2 for further on-off strokes. If a substantial change is found in the timing, the system may remain in the closed-loop PID control state until a consistent timing pattern may be observed over several consecutive pumping strokes, as described above.

FIG. 35 is a summarized version of the steps of a method for turning off the supply voltage and turning it back on for predetermined periods. More steps, or fewer steps, are also contemplated. Other steps are also contemplated. In first block 330, the system is initialized to monitor the pump jack motor's phase angle. In second block 332, which corresponds with State 1, the motor's response to closed-loop PID control is evaluated over several pump strokes. The periods of the pump stroke when it is feasible to turn off the motor are identified. In third block 334, the consistency of the measurements over several strokes is evaluated. In fourth block 336, which corresponds with State 2, the motor may be turned off during the predetermined periods on subsequent pump strokes if each pump stroke shows sufficiently similar behavior to that predicted during the closed-loop PID control process of State 1. In fifth block 338, which corresponds with State 3, the system may return to State 1 to the closed-loop PID control process after a predetermined period of time to adjust to any changes in the system.

As can now be understood, the on-off method provides a means by which to monitor and characterize the periodic variation in the load on a pump jack motor based on the observed phase angle. Once the load is characterized, the method determines whether the opportunity exists to save energy by turning off the motor during portions of each stroke. The method predicts when this opportunity should occur in subsequent pumping strokes and defines criteria to be met in subsequent pumping strokes as conditions for turning the motor off and for turning it on again. The system and method implement algorithms that adaptively adjust to gradual changes in the behavior of the pump jack, and implement safeguards to recognize and react to sudden, large changes in this behavior.

The system and method take advantage of the highly periodic variation of the load on the motor in a pump jack application. The system and method accurately predict those times during a pumping stroke when the power to the motor may be completely turned off, as well as the appropriate time at which the power should be reapplied. Completely turning off the power to the motor results in superior energy savings.

The system and method provide a means by which to monitor and characterize the periodic variation in the load on a pump jack motor based on the observed phase angle between the applied voltage and the consumed current. Once the load is characterized, a determination may be made whether the opportunity exists to save energy by turning off the motor during portions of each stroke. The system and method predict when this opportunity should occur in subsequent pumping strokes and define criteria to be met in subsequent pumping strokes as conditions for turning the motor off and for turning it on again. The system and method implement algorithms that adaptively adjust to gradual changes in the behavior of the pump jack, and implement safeguards to recognize and react to sudden, large changes in this behavior.

The foregoing disclosure and description of the invention are illustrative and explanatory thereof, and various changes in the details of the illustrated system and method of operation may be made without departing from the spirit of the invention.

Claims

1. A method of saving energy for a pump jack with a counterweight disposed with a well, comprising the steps of: connecting a closed loop motor controller to an electric motor connected with the pump jack, wherein said motor controller having switching devices;supplying a line voltage and current to said motor;rotating the pump jack counterweight with said motor for a plurality of pumping cycles;observing the phase angle between the voltage and current supplied to said motor with said motor controller;comparing the observed phase angle with a threshold phase angle;adjusting the supply voltage upon reaching said threshold phase angle using a firing angle of said switching devices to maintain the observed phase angle at a value no greater than a target phase angle;reinstating the full supply voltage to said motor when the firing angle falls below a threshold firing angle;identifying the time periods during the pumping cycles that the supply voltage is adjusted;comparing said time periods to establish predetermined turn-off time periods for switching off the supply voltage;after the step of comparing, switching off the supply voltage to the motor during said predetermined turn-off time periods; andswitching on the supply voltage during the other times periods, wherein the other time periods being predetermined turn-on time periods.

2. The method of claim 1, further comprising the step of: observing a turn-on phase angle immediately after beginning the predetermined turn-on time periods.

3. The method of claim 2, further comprising the step of: comparing the turn-on phase angle with an ideal phase angle.

4. The method of claim 3, further comprising the step of: increasing said predetermined turn-off time periods if the turn-on phase angle is greater than the ideal phase angle, and decreasing said predetermined turn-off time periods if the turn-on phase angle is less than the ideal phase angle.

5. The method of claim 4, wherein said threshold phase angle is 70 degrees, said target phase angle is 60 degrees, said threshold firing angle is 90 degrees, and said ideal phase angle is 75 degrees.

6. The method of claim 4, further comprising the step of: after the steps of switching off and switching on, repeating the steps of supplying, adjusting, reinstating, and comparing.

7. The method of claim 6, wherein the step of repeating performed at predetermined times.

8. The method of claim 1, wherein the step of reinstating occurring when there is an increase in a load on said motor.

9. The method of claim 1, wherein the step of adjusting still allowing sufficient current to observe the phase angle.

10. The method of claim 1, further comprising the step of: determining whether the identified time periods during the pumping cycles are repetitive to establish said predetermined turn-off time periods.

11. The method of claim 10, wherein said time periods are repetitive when there is no variation of more than 200 milliseconds in the periods of the pumping cycles.

12. The method of claim 10, wherein said time periods are repetitive when there is no variation of more than 150 milliseconds in said time periods.

13. The method of claim 1, further comprising the step of: moving a tubular string connected with the pump jack in a substantially vertical direction in the well during the pumping cycles;wherein the step of switching off causing the tubular string to move faster than the tubular string would move if there were no step of switching off; andwherein the step of switching off causing the pump jack counterweight to rotate faster than the counterweight would rotate if there were no step of switching off.

14. A method of saving energy for a pump jack with a counterweight disposed with a well, comprising the steps of: supplying a line voltage and current to an electric motor connected with the pump jack, wherein the current having a real component and a reactive component;rotating the pump jack counterweight with said electric motor for a plurality of pumping cycles;observing the phase angle between the voltage and current supplied to said motor with a motor controller electrically connected with said motor;comparing the observed phase angle with a threshold phase angle;controlling the supply voltage to said motor upon reaching said threshold phase angle to maintain the observed phase angle at a target phase angle and causing the current real component to be substantially zero and the current reactive component to be greater than zero;reinstating the full supply voltage to said motor when there is an increase in a load on said motor;identifying the time periods during the pumping cycles that the supply voltage is controlled;setting predetermined turn-off time periods for turning off the supply voltage from said identified time periods; andswitching off the supply voltage to said motor during said predetermined turn-off time periods and switching on the supply voltage during the other times.

15. The method of claim 14, further comprising the steps of: comparing a turn-on phase angle observed immediately after beginning the step of switching on with an ideal phase angle; andincreasing said predetermined turn-off time periods if the turn-on phase angle is greater than the ideal phase angle, and decreasing said predetermined turn-off time periods if the turn-on phase angle is less than the ideal phase angle.

16. The method of claim 14, wherein there is an increase in a load on said motor when a firing angle of said motor falls below a threshold firing angle.

17. The method of claim 14, further comprising the step of: determining whether said identified time periods during the pumping cycles are repetitive to establish said predetermined turn-off time periods.

18. A method of saving energy for a pump jack with a counterweight disposed with a well, comprising the steps of: supplying a line voltage and current to an electric motor connected with the pump jack, wherein the current having a real component and a reactive component;rotating the pump jack counterweight with said electric motor for a plurality of pumping cycles;observing the phase angle between the voltage and current supplied to said motor with a closed-loop PID motor controller in electrical connection with said motor;comparing the observed phase angle with a threshold phase angle with said motor controller;adjusting the supply voltage upon reaching said threshold phase angle to maintain the observed phase angle at a value no greater than a target phase angle;reinstating the supply voltage to said motor when a firing angle of said motor falls below a threshold firing angle;identifying the time periods during the pumping cycles occurring between the threshold phase angle and the threshold firing angle;comparing said time periods to establish predetermined turn-off time periods for switching off the supply voltage;switching off the supply voltage to said motor during said predetermined turn-off time periods and switching on the supply voltage during the other time periods, wherein said other time periods being predetermined turn-on time periods;observing a turn-on phase angle immediately after beginning the predetermined turn-on time periods;comparing said turn-on phase angle with an ideal phase angle;increasing said predetermined turn-off time periods if the turn-on phase angle is greater than the ideal phase angle, and decreasing said predetermined turn-off time periods if the turn-on phase angle is less than the ideal phase angle; andafter the steps of switching off and switching on, repeating the steps of supplying, adjusting, reinstating, and comparing.

19. The method of claim 18, further comprising the step of: determining whether said identified time periods during the pumping cycles are repetitive to establish said predetermined turn-off time periods.

20. The method of claim 18, wherein said threshold phase angle is 70 degrees, said target phase angle is 60 degrees, said threshold firing angle is 90 degrees, and said ideal phase angle is 75 degrees.