Method for autonomous control of a chemical injection system for oil and gas wells

(54) Title: METHOD FOR AUTONOMOUS CONTROL OF A CHEMICAL INJECTION SYSTEM FOR OIL AND GAS WELLS (30) Unionist Priority: 11/02/2007 us 11 / 934,513 (73) Holder (s): National Coupling Company, INC (72) Inventor (s): Maurice Slot, Richard R. Watson, Robert E. Smith IV (57) Abstract: method for autonomous control of a CHEMICAL INJECTION SYSTEM FOR OIL AND GAS WELLS. A control program for a positive displacement measurement system is released which measures the time required for the displacement of a free piston in a cylinder of known volume to determine an average flow during the full stroke of the piston. The system can also measure and record inlet and outlet pressures or differential pressure between fluid inlet and outlet. The control program positions a four-way valve that can function as an adjustable metering orifice in response to the average measured fluid flow and / or changes in inlet and outlet pressures to achieve the desired flow. At the end of each stroke, the four-way valve is repositioned to reverse the flow of fluid through the measuring cylinder. The system can review the valve position adjustments for both the direct stroke and the inverted stroke based on the measured time required for a full stroke at a certain valve position. In this way, the system automatically and iteratively compensates for changes in fluid properties and fluid pressure. A cleaning cycle is provided, which gradually opens the valve gradually and which can culminate in displacement with the valve fully open in both directions of fluid flow, in the event of an obstruction of the valve orifice.

<img file="BRPI0804350A2_D0001.tif" />

ΡΙ0804350-7 "METHOD FOR AUTONOMOUS CONTROL OF A CHEMICAL INJECTION SYSTEM FOR OIL AND GAS WELLS"

Cross Reference to Related Orders: None.

Federally Sponsored Research or Development Statement: None.

BACKGROUND OF THE INVENTION

1. Technical Field

This invention relates to chemical injection systems for oil and gas wells. More particularly, it concerns autonomous control systems for the injection of liquid chemical treatment agents into subsea wells.

2. Description of Related Technology including information disclosed under 37 ^!. CFR 1.97 and 1.98

A variety of chemical agents are injected into hydrocarbon wells to control corrosion, hydrates, asphaltenes, paraffins, fouling and the like. Typically, these chemical agents are in the liquid phase and are pumped into the well at a selected speed using a chemical injection system. For subsea wells, the chemical supply and the pump can be located on a production platform, and are commonly connected to the wellhead via an umbilical cable. If the measurement of the chemical agent is carried out only on the surface, any leak in the umbilical cable or its connectors will give an erroneous indication of the amount of chemical agent that is injected into the well. In addition, each subsea well may require its own injection system on the platform and the umbilical cable connection.

Certain prior art measurement systems employ a variable orifice and an adjustable orifice that allows remote control of the flow in each well. Other systems, measuring the prior art, are based on pressure-compensated flow control ι '&lt;

- an adjustable pressure regulator and a fixed orifice can maintain a constant flow in each well.

Frequently, flow measurement over a wide range is necessary during the life of the well. The orifice measurement is limited in relation to the strip and is subject to film coverage, clogging and differentiated fluid properties.

Contamination by particulates in long chemical injection lines is inevitable and can block the small holes required for measurement and control. In-line filters are a further complication that affects the reliability of the system, increasing capital costs and requiring periodic maintenance (which increases operating costs).

US patent 6,973,936 to Richard R. Watson discloses a fluid injection system that controls the distribution of fluid from a supply line to a selected well at an adjustable rate. A free piston divides a cylinder in the first and second chambers. A multiposition valve comprises a first position for passing fluid from the supply line into the first chamber to move fluid from the second chamber back through the valve to an injection point, and a second position for passing fluid from the supply line. to the second chamber to move fluid from the first chamber back through the valve to the injection point. A control system in communication with a position sensor determines the travel time of the free piston to selected positions and selectively adjusts a variable valve opening to adjust the flow, alternates between the first and second positions and periodically increases the opening valve for cleaning.

SUMMARY OF THE INVENTION

The present invention can be incorporated into a control program for a positive fluid displacement measurement system that measures the time required for displacement of a free piston in a cylinder of known volume to determine an average flow during a full stroke of the piston . The system can also measure and record the differential pressure inlet and outlet pressures between the fluid inlet and outlet. The controller can take the flow commands from a client underwater control capsule and adjust the flow by partially opening a four-way valve each time the valve is inverted.

The control program precisely positions a four-way valve that can function as an adjustable measuring orifice in response to the measured average flow and / or changes in the inlet and outlet pressures to achieve the desired flow. At the conclusion of each stroke, the four-way valve is repositioned to reverse the flow of liquid through the measuring cylinder.

The system can review stored valve position adjustments for both direct and inverted travel based on the measured time required for a complete stroke at a currently stored valve position. In this way, the system iteratively compensates for all changes in fluid properties and fluid pressure. Certain embodiments of the invention additionally comprise an optional cleaning cycle which progressively opens the valve gradually and can, if necessary, open the valve completely in both directions of the fluid in order to clear an obstruction of the valve orifice. It has been found that shut-off or door valve construction are the most suitable designs for reliable operation when high pressure fluids are contaminated with rigid particulate material. The standard operating procedure for clearing a blocked valve of this type is to move it to the fully open and completely closed positions. This allows the accumulated particles to pass and the sealing elements to be swept away or to cut all remaining obstructions. The use of this valve construction for the two-position four-way valve and the control of the valve actuator in response to the stroke time of the cylinder results in precise flow control with excellent resistance to contamination.

The actuation of the two-position four-way control valve can be performed with a conventional stepper motor that drives a ball screw to convert rotation into linear motion. This combination has been found to give very high accuracy to the linear position of the valve. This precision allows the valve to be partially opened, thus creating a precision hole every time the valve is moved.

As the valve is moved to allow flow to the first side and then to the other side of the cylinder piston, the flow is regulated by the precision orifice created by the partially opened valve. In certain embodiments, a contact time is included after the piston of the cylinder has completed its displacement and the flow has been stopped. This provides precise control of the volume of chemicals injected into the continuous production stream from the oil or gas well over a period of time.

A system according to the invention can be designed to accommodate all chemicals currently used for the control of corrosion, hydrates, asphaltenes, paraffins and incrustations in hydrocarbon wells. Changes in viscosity or fluid density do not require recalibration of the positive displacement measurement system. Verification data can be transmitted to the client subsea control capsule.

The use of a plurality of systems according to the invention can provide the guarantee of the flow of treatment chemicals to multiple subsea wells from a single umbilical cable.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a schematic diagram of a prior art chemical injection apparatus that can be controlled by the method of the present invention.

Figure 2 is a cross-sectional view of the control valve used in the apparatus of Figure 1 in a first position.

Figure 3 is a cross-sectional view of the control valve shown in Figure 2 in a second position.

Figure 4 is an enlarged cross-sectional view of part of the valve shown in Figure 2.

Figure 5 is a schematic diagram of a modified chemical injection system for use with the present invention.

Figure 6 is a flow chart representing the steps of a method according to an embodiment of the invention.

Figure 7 is a graph of the flow produced by a representative control valve in particular as a function of the number of steps taken by a stepper motor that drives the valve actuator.

Figure 8 is a flow chart representing the steps of a method according to a second embodiment of the invention.

DETAILED DESCRIPTION OF THE INVENTION

Figure 1 schematically illustrates details of a measuring body 12 interconnected with a control system 14 and a multiposition valve 16 actuated by actuator 45 in a chemical injection system 10. The measuring body 12 has a hole 20 to contain chemical fluid a be distributed in a well. An axially movable free piston 22 in the hole 20 divides the measuring body 12 into first and second chambers 24, 26 with variable volume. The free piston 22 seals the measuring body 12 with a sealing element, such as the O-ring 25. Conventionally, the measuring body 12 and the free piston 22 comprise a cylinder and piston assembly, as shown. The first and second inlet - outlet holes 28, 30 are provided to pass fluid into and out of the first and second chambers 24, 26. The supply line 33 supplies chemical fluids at high pressure through the multiposition valve 16 to the measuring body 12.

In a first valve position shown in figure 1, conceptually illustrated by the alignment of parallel line segments 18 with lines 31 and 33, the fluid passes from the supply line 33, through the multiposition valve 16, line 29 and the orifice inlet - outlet 30, and into chamber 26. As the fluid passes into the chamber 26, the fluid pressure impels the free piston 22 towards the end 34 of the measuring body 12, decreasing the volume of the first chamber 24 and displacing the fluid out through the inlet - outlet 28. The fluid that exits through orifice 28 passes through line 27, back through valve 16, and out through line 31 to an injection point in the well.

In a second position (not shown), which can be conceptually visualized by sliding the crossed flow lines 15 on the valve 16 to the left to align with lines 31 and 33, the fluid passes from the supply line 33, through the multiposition valve 16, line 27, from the inlet - outlet 28, and into the chamber 24. As the fluid passes into the chamber 24, the fluid pressure impels the free piston 26 towards the end 26 of the measurement body 12, decreasing the volume of the chamber 26 and displacing the fluid out through the inlet port. - outlet 30. The fluid leaving the orifice 30 passes through line 29 back, through valve 16, and out through line 31 to the same injection point in the well. Thus, by repeatedly reversing the direction of the multi-function valve 16 after the free piston 22 has reached a selected position, the fluid can move continuously from line 33 to line 31 to the injection point in the well.

Position sensors 38 and 40 are included to detect the position of the free piston

22. Position sensors 38, 40 are in communication with the control system 13, as represented by the dashed lines 39, 41 by means of a conventional device, such as a wired, fiber optic or wireless signal. When the free piston 22 reaches selected positions, the position sensors 38, 40 signal the control system 14, in response to which the control system 14 can selectively reverse the position of the multi-position valve 16 to reverse the direction of travel of the free piston 22.

Because the selected positions are known, the relative displacement of the free piston 22 is also known, corresponding to a known volumetric displacement of the fluid of the measuring body 12, computed as the product of the displacement of the free piston 22 and the cross sectional area of ​​the hole 20. The control system 14 includes an internal timer to synchronize the displacement of the free piston 22 between the selected positions, as signaled by the position sensors 38, 40. Therefore, a volumetric flow is also known that can be computed as the divided volumetric displacement by the travel time. The multiposition valve 16 includes a variable valve opening discussed below in conjunction with figures 2-4, to control the flow between the supply line 33 and the measuring body 12. The control system 14 selectively adjusts the variable opening of the valve in response to free piston travel time 22. If the travel time is too long, indicating a lower flow than the desired flow, the control system 14 can increase the variable opening of the valve to increase the flow. Conversely, if the travel time is too short, indicating a flow greater than the desired flow, the control system 14 can selectively decrease the valve opening to reduce the flow. In this way, the flow of the fluid distribution in the well is controlled.

As shown in figure 1, preferably, the selected positions of the free piston 22 are the positions of the free piston 22 having reached both ends 34, 36 of the measuring body 12. The selected positions of the free piston 22 can alternatively be any place along the free piston displacement range 22, and need not be at the ends 34, 36 of the measurement body 12. In typical embodiments, as illustrated, position sensors 38, 40 are substantially in the same axial position as the selected positions. Conventional position sensors, such as spring loaded pins, or magnetic or infrared proximity sensors, can be used. In other embodiments, position sensors may conceivably not need to be axially aligned with the selected positions. A position sensor can additionally comprise an optional pressure transducer 49 or a flow transducer 42. Optional orifice valves, such as they can comprise the sealing elements 43, 44 on the free piston 22, can be included to seal the inlet orifices - exit 28, 30 when the free piston reaches the ends 34, 36. This can decrease the pressure on line 31 more dramatically and thereby provide a more distinct indication that the free piston 22 has reached the end of its displacement. An indication like this can provide a replacement to confirm or replace the position sensors 38 and 40.

The terms “first position” and “second position” in relation to valve 16 refer, in general, to the direction resulting from the flow, instead of a fixed position of the components of valve 16, because, in general, there is a degree of adjustability in each of the two positions, such as to adjust the flow. Figure 2 shows a detailed view of the preferred embodiment of the multiposition valve 16 in the first position of the valve, partially open to limit the flow through the valve. Figure 3 shows a detailed view of the preferred valve 16 in the second position of the valve, also partially open. Figure 4 shows in close detail a part of the gate-type valve 16 in the first position of the valve in figure 2.

In particular, in relation to figure 2, the schematic diagram of figure 1 and the closest view to figure 3, the multiposition valve is indicated, in general, at 16, with a body 17. A port 50 is positioned in a cavity 52 in the body 17. The port 50 has a hole 54 which, in the position shown, communicates with an inlet port 32 and a first flow passage 56 which extends through the body 17 to a first exchange port 57. Thus, in this position, chemical fluid supplied by the aforementioned supply line 33 flows into the body 17 through the inlet port 32, through the port hole 54 and through the first flow passage 56, out through the first exchange port 57 to line 29. As shown, the fluid passes through line 29 into the measuring body 12, and another fluid passes from the measuring body 12, through line 27, back to valve 16. Then, the flow passes back into the body 17, through the second exchange orifice 59, into a second flow passage

58, passes around port 50 into an outlet port 53, and out through an outlet port 55. Finally, the flow out through outlet port 55 passes through line 31 to the injection point in the well, as exposed.

In figure 3, the port hole 54 is positioned, in contrast, in communication with the inlet port 32 and the second flow passage 58. Thus, the flow of line 33 passes through the inlet port 32 into the passage inlet 51, through the port hole 54, and through the second flow passage 58, exiting through the second exchange orifice 59 to line 27. As shown, the fluid passes through line 27 into the measuring body 12, and another fluid passes from the measuring body 12, through line 29, back to valve 16. Then, the fluid passes back into the body 17 through the first exchange port 57, into the first flow port 56, into the outlet port 53, and out through the outlet port 55. Finally, the flow out through outlet port 55 passes through line 31 to the injection point in the well. Thus, as described in relation to figure 1, the flow between the valve 16 and the measuring body 12 can be reversed by moving the valve between the first and second valve positions shown, respectively, in figure 2 and figure 3, but in in each case, the liquid flow is from line 33 to line 31 to the injection point in the well.

In the mode shown in figures 2 - 4, the stepper motor 45 drives a ball screw 48 to axially move the port 50 in the cavity 52, adjusting the size of the flow path defined between the port hole 54 and the first flow passage 56, thereby adjusting the flow to a desired flow. The port 50 can move axially to switch between the first position of the valve of figure 2 and the second position of the valve of figure 3. Experienced in the technique, they perceive alternative devices to move the gate, different from the stepper motor 45.

The Hall effect device using motion perception and motion limit switches can offer greater reliability in extreme environments. Since there are no moving parts involved in the sensor or magnet, the typical life expectancy is increased when compared to traditional electromechanical switches. In addition, the sensor and magnet can be encapsulated in a suitable protective material. The Hall effect device, when properly packaged, is immune to dust, dirt, mud and water. These characteristics make the Hall effect device particularly preferred in a system according to the present invention for the perception of the piston position, when compared to alternative devices, such as optical and electromechanical perception.

Controlling flow in a few gallons a day at a pressure drop of several hundred pounds per square inch requires a very small orifice of just a few thousandths of an inch. The valve used in a particularly preferred embodiment of the invention is a gate-type shut-off valve with a rejection rate of 2,880: 1. This valve provides the required small orifice and reverses the flow for each positive displacement load so that clogging is avoided:

Certain prior art systems used filters to prevent small particulate flow control holes from clogging, but typically these filters need to undergo subsea maintenance, which is highly costly. Prior art devices also used capillary holes, which have a larger area, for a given flow restriction, and these can be made to adjust their hole diameter by means of a tapered screw thread so that flow rates can be changed, and a larger temporary opening can be used to provide contamination cleaning. With each of these solutions, measuring a wide range of flow rates is a separate necessary task that requires costly flow instrumentation. The flow cannot be precisely measured by pressure loss through an orifice of unknown restriction, as in the situation with partial particulate obstruction. The present invention can include a device for creating the small orifice matched with a four-way gate valve that is constantly displaced to avoid obstruction, and can also be completely opened to allow the particulate to pass. What's more, the system provides very accurate flow measurement that is immune to partial obstruction or changing the properties of the fluid or covering the orifice with film - all conditions that are present and can destroy conventional meters that are based on a drop in pressure through an orifice.

Pressure transducers 49 and 49 'can provide the controller with more information with which to establish the degree of valve opening, but they cannot measure or check the flow. Flow measurement and verification are provided by sync circuits and position sensors on the positive displacement cylinder.

If the displacement cylinder fails to travel in the expected time, a condition that indicates an obstruction, the controller can operate the four-way valve to the fully open position to allow debris to pass.

An orifice cannot be used as a reliable underwater flow measurement device because it is subject to clogging and covering with film (coating) by chemicals passing through it. Chemicals that are measured in a chemical injection system for a hydrocarbon well may have film-coating characteristics as a desired trace. Common flow measurement devices in the prior art use a pressure loss measurement through an orifice to indicate flow. If a cheap pressure drop measurement system cannot be used, conventional alternatives are costly. In addition, no other measuring device, such as turbine, ultrasonic, tourbillon or thermal mass type, can meet the range of a displacement cylinder according to the present invention. All of them are limited to the total range of 100.Ί to 200: 1. This means that they can measure precisely 1 gallon per day (GPD) (3.79 L per day) up to about 200 GPD (757.09 L per day). A system according to the present invention can measure 1 GPD (3.79 L per day) up to more than 3,000 GPD (11,356.24 L per day). Also, many chemicals for the treatment of wells are non-Newtonian fluids, that is, their viscosities change with pressure in a non-linear manner, a feature that makes accurate flow measurement more challenging for most measurement technologies. previous technology, but which has no effect on a system that employs a positive displacement cylinder.

Now, with reference to figure 6, a particular preferred embodiment of the invention is disclosed in the form of a flow chart representing certain steps of a method for controlling a valve in a chemical injection system. The chemical injection system can comprise a processor, and a method can be implemented as instructions for the processor that can be stored on a storage medium.

As shown in figure 6A, the process can start with manual entry

100 with an operator inserted the desired flow of the chemical to be injected. The flow may have the dimensions of a unit of volume per unit of time. Flow rates for chemical injection systems used in conjunction with oil and gas wells in the domestic energy industry are often expressed in gallons per day (GPD). In certain modalities, the insertion of the desired flow can be carried out by an operator located on a production platform on the high seas, and the command can be transmitted to the controller at or near the head of the underwater well, through an umbilical cable. . The command can also be transmitted via a telemetry system from an offshore installation or from another unit on the high seas.

In a particularly preferred embodiment, system initialization includes driving the valve actuator to a mechanical limit by controlling a stepper motor that drives the actuator to walk numerous steps in a direction that exceeds the number of steps previously determined to correspond to the entire displacement of the actuator. One or more inversions of the actuator, followed by the attempt to move “direct” in excess of the inverted displacement, can be used to ensure that the actuator remains firm in relation to the mechanical limit. Thus, although the starting position of the valve may be unknown upon system startup, an initialization routine can be used to move the valve to a known position. In block 105, the system can determine initial valve settings (number of steps) for both the direct and inverted valve position from the desired flow inserted in 100, a stored flow curve 115 and valve actuation position data 120 - that is, the number of steps from the closed position of the valve to the point at which the valve orifice begins to open in a certain direction. In a particularly preferred embodiment, initialization includes moving the valve actuator from the mechanical limit position to a “central” closed position defined to be the midpoint between the “direct” valve actuation position and the valve actuation position “Inverted”. As an example, using the flow curve of figure 7, if the desired flow rate is 40 GPD (151.42 L per day) and the actuation position of the direct valve is 33 steps from the “central” closed position, then the initial direct valve adjustment will be 171 steps from the center (138 + 33). It has been found that the actuation position of the valve depends on the valve and can vary from valve to valve and / or change after maintenance on the valve or valve actuator. Direct and inverted valve settings can be loaded into registers designated for this function.

The flow curve data 115 can be in the form of a digitized flow curve, such as the curve shown in figure 7. In a particularly preferred embodiment, the flow data is tabulated for each step of a valve actuated by a motor step. In other embodiments, the flow curve data may be in the form of a mathematical representation - for example, slope and intercept values ​​for a substantially linear flow curve. In the case of modalities that use digitized curves, the system can comprise a device for interpolating data points using conventional curve adaptation techniques.

In certain embodiments (not shown), the initial system inputs may include the selection of a particular flow curve that may be associated with a particular chemical or mixture of chemicals to be injected or with a certain property of the fluid to be injected, for example, the specific gravity of the fluid, the viscosity of the fluid, the concentration of an active ingredient (s) in a solvent, or the like. In still other embodiments, the initial input may include a correction factor that the system can use to modify a general purpose flow curve previously stored for use with a specific chemical or chemical property, that is, the flow curve stored can be to dilute aqueous solutions, and a correction factor provided allows the system to adapt the curve to a fluid with substantially different rheological properties. However, it is realized that the system according to the present invention will automatically compensate fluids with different properties, and adapting the flow curve to a specific fluid provides an advantage only in the initial adjustments of the valve position and in the first few computations of the corrections of valve adjustment.

Upon system initialization, the position of the free piston 22 in hole 20 can be. unknown. In this way, upon initialization, the system can configure the valve to activate the piston in a known location. In decision diamond 125, the system can first test the actuation of the direct limit switch (indicating that piston 22 is at the end of the direct travel offset). If the actuation of the switch is detected, the process can proceed in block 130 for the inverted stroke sequence. If the direct limit switch is not activated (branch N in diamond 125), the system proceeds to block 140 with a straight stroke sequence (see figure 6B).

After system initialization, the normal forward and inverted alternating cycle flow process can begin. A representative straight stroke sequence is illustrated in the flowchart of figure 6B, and a corresponding inverted stroke sequence is represented in figure 6C.

Now, in relation to figure 6B, the direct stroke sequence starts at block 200 with the valve setting of the current direct stroke (which can be in steps of the actuator limit, the central (closed) position, or the lower valve position recent) being loaded from register 202. At block 204, the control valve is driven by the stepper motor to the most current direct valve setting, and a timer is started (block 205). In this condition, the system is now measuring fluid through control valve 16 from supply line 33 to chamber 26 via line 29. As fluid is pumped into chamber 26, piston 22 moves (to the left in figure 5) displacing fluid in chamber 24, which flows through line 27, valve 16 and line 31 to the injection point of the well. The pressure of the fluid in the injection line 31 can be measured by the pressure transducer 49, while that in the supply line 33 is measured by the pressure transducer 49 '.

The program can include one or more routines that test the piston movement. For example, a straight stroke sequence (figure 6B) is usually inserted from the completion of an inverted stroke sequence signaled by the actuation of the reverse limit switch 40. The movement of piston 22 away from the inverted stroke limit should disable the reverse limit switch 40. This condition can be tested on diamond 208. If the switch remains activated (that is, the piston is still in the actuation range of the limit switch), the system can wait for a selected time interval before taking corrective action. In the illustrated mode, the system expects (at diamond 210) an interval equal to 50% of the expected stroke time (volume of cylinder displacement divided by the selected flow rate) and, if the inverted limit switch remains activated, the valve can be opened 20 additional steps (in block 212). Similarly, the system can now expect (in diamond 216) an additional time interval that, in the illustrated mode, is equal to the expected stroke time (now, cumulatively, 150% of the expected stroke time) for the key- reverse limit to disable (diamond 214). As before, if the piston does not move sufficiently to disable the reverse limit switch, the valve is opened in an additional 20 steps (in block 218). In the illustrated mode, the progressive opening of the valve in case of no piston movement can be repeated in lozenges 220 and 222 with additional opening of the valve in block 224. If the reverse limit switch remains activated (branch N in lozenge 226) and the cumulative time since the valve opening reached 350% of the expected stroke time (branch S in diamond 228), a discharge cycle (described more fully below) can be started (in block 230). However, if the reverse limit switch is deactivated (S branches in diamonds 208, 214, 220 or 226), the System proceeds to the normal straight course sequence block 232.

While the fluid is flowing, the outputs of pressure transducers 49 and 49 'can be periodically sampled and a differential pressure (ΔΡ) can be stored by controller 14. In a particularly preferred embodiment, a running average ΔΡ is stored by controller 14 along with the three most recent ΔΡ values ​​in a PEPS stack. Additional filtering algorithms can be applied to eliminate or reduce the influence of pressure increases that can be encountered during a stroke. This process can be implemented as shown in figure 6B in block 232, whose subroutine is executed at a pre-selected interval measured in diamond 238.

In the normal course of events, fluid flow continues until piston 22 reaches the end of its direct stroke (left wall of cavity 24 in figure 5) which activates limit switch 38, whose activation is detected in diamond 234. As shown in figure 6C, the stopwatch is stopped at block 270, and the accumulated time in the stopwatch counter is the total time that piston 22 takes to travel in a complete stroke. Since a full stroke displaces a known volume of fluid (as determined by the physical dimensions of cylinder 12 and piston 22), this volume divided by the accumulated time produces the average flow rate of the fluid during that particular direct stroke. In block 272, the average flow measured for the stroke is compared with the desired flow that was entered by the operator at 100.

Corrections (if any) in the direct stroke valve adjustment are computed in block 274. In a particularly preferred embodiment, the difference between the measured flow and the desired flow is equated in a number of steps from the stored flow curve. at 115. The correction can be taken directly from the curve or computed from the first derivative of the curve. As explained, in certain modalities, the system can interpolate the data points in order to determine the correction.

As shown in block 276, the adjustment of the valve adjustment can be further refined by a factor related to a change in the mean ΔΡ of the previous direct stroke. In certain modalities, the correction factor ΔΡ may be a function (in whole or in part) of the selected ΔΡ values, for example, the three most recent ΔΡ values ​​in the PEPS stack of the illustrated modality. Especially at relatively slow flows, a change in ΔΡ just before the end of the course may be more indicative of the ΔΡ likely to be encountered during the next direct course.

The correction of ΔΡ can be derived from empirically determined flow values ​​at various differential pressures. In other modalities, the correction of ΔΡ can be calculated from a function that relates the flow (or motor steps of the valve actuator 45) to ΔΡ.

It should be noted that the process of the present invention will work without ΔΡ data, that is, the absence or failure of a pressure sensor 49 will not disable the system. The corrections computed in block 274 will compensate for changes in ΔΡ. The use of ΔΡ information (in block 276) enables the system to make better predictions of the valve adjustment required to produce the desired flow. However, the iterative process will "zero" in the correct setting even without this data.

In block 278, the revised valve setting to be used for the next direct stroke is stored in the register (or other storage device) designed for this purpose, and the process proceeds to the inverted stroke sequence, as shown in block 282 (and in figures 6D and 6E). Optionally, in block 280, data related to the newly completed course sequence can be recorded before proceeding to the inverted course sequence. Examples of log data include actual stroke time, and time and number of additional valve openings (for example, blocks 212, 218, 224, 240 and / or 246), and whether a discharge stroke (blocks 230 or 256) was required. Any other parameters detected by the system can also be recorded at this stage of the process.

Since the complete displacement of the free piston 22 displaces a known volume of fluid, the time that must be required for a complete stroke of the piston 22 at the desired flow rate can be calculated to produce an expected stroke time. As shown in block 236, the elapsed stroke time can be compared to the expected stroke time and, if the elapsed stroke time exceeds the expected stroke time by a selected margin (100% in the example shown), the system can initiate corrective action - progressive opening of the valve in increments of 20 steps in 200% of the expected travel time (diamond 236) and, again, in 300% of the expected travel time (diamond 244). If the time exceeds 400% of the expected travel time (branch S in diamond 254), a discharge cycle (block 256) is started in the illustrated mode. In each of these routines, ΔΡ readings can be taken and stored (blocks 250 and 260) in a selected and repetitive time interval (diamonds 248 and 258).

Q control of an “inverted” course of motion - that is, a course in which the control valve is positioned in such a way that the flow path 15 is active and the fluid flows into the chamber 24 via line 27 and is expelled by chamber 26 and inside line 29 as piston 22 moves from left to right in figure 5 - is illustrated in figures 6D and 6E. The process is analogous to that illustrated for a “direct” course in figures 6B and 6C and discussed earlier. Reference numbers for corresponding elements in figures 6B and 6C differ by a value of 100 from those in figures 6D and 6E. The current inverted stroke valve setting can be stored in register 302 and loaded into the controller in block 300. Corrections computed for the inverted stroke valve setting in block 374 and (optionally) in block 376 can be stored in register 305 in block 378 and used for the next inverted stroke. Upon completion of an “inverted” stroke, the process returns to the straight stroke sequence (in block 382). In this way, the system continuously iterates the direct and inverted valve settings to provide the requested fluid flow.

As shown in blocks 230 and 256 (figure 6B) and blocks 330 and 356 (figure 6D), the system can initiate a discharge cycle in the event that the elapsed stroke time exceeds the expected stroke time at a selected margin . A possible cause of less than expected fluid flow is debris that obstructs or partially obstructs a hole in the control valve 16. The system can act to discharge debris from the control valve (or an associated fluid conduit). A discharge cycle according to a preferred embodiment of the invention is illustrated in the flow chart of figure 6F. In this cycle, first, the control valve is activated to its completely open position in one direction (block 460) and then to its completely open position in the opposite direction (block 465). In practice, it has been found that this cycle is almost always successful in unloading obstructive remains from the chemical injection system. Following a discharge cycle, the system can perform an initialization sequence (block 470).

Figure 8 is a flow chart illustrating an alternative embodiment of the invention in which a known fixed volume of chemicals is injected over a predetermined time interval. This modality has a particular advantage in those applications in which the control of the total volume of chemicals injected in a certain period of time is more important than injecting the chemical at a constant speed.

The process can start at manual input 500 with the selection of an average flow by the operator. Since a full stroke of piston 22 displaces a known volume of chemicals, in block 510, the system can compute the time required to shift the volume of chemicals injected during a complete stroke at the selected flow rate. In block 520, the system can be initialized as previously described in order to position piston 22 at the beginning of a stroke. Using the stored flow curve, the setting of the valve that should produce the selected flow is determined in block 530 from the flow curve stored in 535. In block 540, the valve is opened to provide an orifice somewhat larger than that required to achieve the selected flow. The absolute value of the surplus can be a selected percentage increase in the selected flow (for example, X + 10% GPD), a selected incremental increase in the flow (for example, X + 5 GPD) or a pre-selected number of additional steps of the stepper motor that positions the valve (for example, position computed from the flow curve + 15 steps).

A timer can be started in block 545 and then the system can test the piston movement (in diamond 550) by perceiving the previously activated limit switch deactivation. If the piston movement is not detected (branch N at 550) in the illustrated mode, the valve is opened in an additional 20 steps. This process can be repeated (branch A at diamond 557) at selected time intervals and, if no piston movement is detected after a selected cumulative time (branch B at 557), a discharge cycle can be started at block 559 to clear any obstructions in the valve orifice.

Once the piston movement has been detected (S branch at 550), the system can expect (at 560) that the limit switch signals that the piston has reached the end of a stroke and that the known volume of a complete stroke has been injected in the well. The valve position set at 540 should result in a complete stroke that is completed before the 510 computed time interval has elapsed - that is, the system must need to wait for a “contact time” to elapse before starting another stroke. . In diamond 565, the system tests the end of the computed time interval before the limit switch is activated. If true (S branch at 565), there is an error condition (block 570) and the system can take corrective action by correcting the valve setting used in block 530. If the limit switch is not yet activated after a selected interval (diamond 572), the corrective action may include a discharge cycle (block 573), as shown in relation to figure 6F, and / or an upward adjustment of the position the valve.

However, in the normal course of events, the piston will reach the end of a stroke (thus activating the limit switch) before the end of the time interval computed at 510 (S branch at 560). The system can store the time when the limit switch is activated at 575 and then wait (at 580) for the end of the time period at diamond 580.

The time it takes the piston to take a full stroke (recorded at 575) can be used to compute and store a revised valve setting in block 585. Then, this revised setting can be used by the system for the next stroke in the same direction. In this way, the system continually refines the valve setting to compensate for all changes in parameters that can affect flow - for example, supply pressure, viscosity, density, etc.

In block 590, the system orders a corresponding process for a stroke in the opposite direction (which can start in block 530) and the system alternates between direct and inverted cycles, repeating the required valve adjustments.

Although the invention has been described in detail with respect to certain preferred embodiments, there are variations and modifications according to the scope and spirit of the invention described and defined in the following claims.