CN112188937A

CN112188937A - Capacitance elimination method of tuning fork for fluid property measurement

Info

Publication number: CN112188937A
Application number: CN201980034564.8A
Authority: CN
Inventors: H.R.塞伦; M.冈扎莱兹; S.丘塔克; M.德芬鲍
Original assignee: Saudi Arabian Oil Co
Current assignee: Saudi Arabian Oil Co
Priority date: 2018-06-20
Filing date: 2019-06-20
Publication date: 2021-01-05
Also published as: JP2021529304A; WO2019246410A1; SG11202011832WA; SA520420536B1; KR20210023978A; EP3810339A1

Abstract

An apparatus for determining an uncharacterized property of a downhole fluid. The apparatus comprises an oscillation driver circuit comprising an amplifier having an output and an input, a feedback loop between the output and the input of the amplifier or logic gate; an electromechanical resonator disposed within the feedback loop such that the resonator is driven by the oscillation driver circuit, wherein a resonant frequency of the resonator defines an oscillation frequency of the oscillator circuit; and switching means for causing the oscillator circuit to cease driving the resonator so that the decay rate of the oscillation of the electromechanical resonator in the uncharacterized fluid can be observed. The electromechanical resonator is encapsulated in a conductive layer to protect the resonator from the capacitive effects of the downhole fluid.

Description

Capacitance elimination method of tuning fork for fluid property measurement

Technical Field

The present invention relates generally to the field of petroleum engineering, and more particularly to a method for eliminating the effects of parasitic capacitance in an electromechanical device used to obtain in situ measurements of viscosity and density of downhole fluids at a subterranean well.

Background

A variety of hydrocarbons, brines, other liquids and gases, and supercritical fluids, slurries, foams and emulsions are produced from, found in, used in the construction of, or injected into subterranean wells. These fluids will be collectively referred to as downhole fluids. Knowledge of the physical properties (e.g., density and viscosity) of these fluids is critical to drilling, completing, operating, and abandoning wells. These wells may be used to recover hydrocarbons from a subterranean reservoir, inject fluids into a subterranean reservoir, and monitor the condition of a subterranean reservoir.

Fluids include substances in liquid, gaseous and supercritical states. Downhole fluids include one or more fluids produced from the earth, such as hydrocarbons, brine, and other fluids occurring in subterranean reservoirs, as well as fluids such as brine, carbon dioxide, and methane that may be injected into the subsurface to increase the production of hydrocarbons or for disposal purposes. Downhole fluids also include slurries containing liquid and solid components (e.g., drilling mud and cement) used to construct wells. One or more downhole fluids may be found simultaneously in a subterranean well, such as in a multiphase flow, and they may interact to form emulsions and foams. Downhole fluids are also understood to include materials that are fluids at reservoir temperatures and pressures, even though they may be solids at cooler temperatures closer to the surface.

Downhole fluid properties include the viscosity and density of each fluid phase, as well as the effective viscosity and density of a polymeric fluid composed of multiple fluid phases. Newtonian fluids are characterized by a single viscosity. In non-newtonian fluids, such as slurries, the viscosity may vary with flow conditions, for example with the stress or shear rate applied to the fluid. The properties of non-newtonian fluids also include rheological parameters that describe this dependence of velocity on flow conditions.

It is known that downhole fluid properties vary with temperature and pressure, and that the characteristics of such variations are important properties of downhole fluids. This variation is described, for example, by the PVT (pressure-volume-temperature) characteristic of the fluid, which describes how the density varies with pressure and temperature, or by the viscosity varying with pressure and temperature. As the pressure and temperature of the fluid change, the fluid may change state, such as condensing from a gas to a liquid (e.g., at the dew point), boiling from a liquid to a gas, or transitioning to a supercritical or non-supercritical state. Other types of downhole fluids include structured fluids or dispersions, such as emulsions, suspensions, and foams, which may undergo structural changes with pressure, temperature, concentration, or other chemical or thermodynamic variables. These changes may be dynamically detected as changes in their viscosity and/or density. For example, one fluid may dissolve in another fluid, and the pressure and temperature conditions (e.g., bubble point) in which the fluid becomes dissolved or no longer dissolves or the pressure and temperature conditions at which solids may precipitate from the fluid are important properties of the fluid. The changes in these conditions and the depth or location in the well where such dissolution and precipitation occurs are critical information for optimal production or injection of fluids into the well. In addition, the density (or API gravity) and viscosity of the oil dictate its type and value, and as a function of depth, can be used to understand reservoir structure and compartmentalization. Asphaltene content can also be inferred from the viscoelastic properties of the hydrocarbons produced. Knowledge of the PVT characteristics of the produced fluid is also important to optimize the surface facility design, including determining the optimal pressure for the surface separator. Changes in state, dissolution and precipitation are usually accompanied by changes in the viscosity and density of the fluid, so measurements of changes in viscosity and density with pressure and temperature can determine the temperature and pressure at which these changes occur.

Determining the viscosity and density of fluids in a subsurface reservoir can provide important data for optimizing production and reservoir models. Typically, the produced fluid is sampled at the surface. Downhole temperature and pressure conditions are then applied to the sample in the laboratory and its viscosity, density and other properties are measured. However, when hydrocarbon liquids from a reservoir reach surface temperatures and pressures (e.g., as they travel up the well), dissolved gases are released and asphaltenes may precipitate. These changes can be difficult to accurately reverse in the laboratory, and so even if laboratory measurements are made at laboratory temperatures and pressures, the viscosity measured in the laboratory may be different from the viscosity of the fluid in the reservoir. Furthermore, the process of collecting a sample in a well, transporting it to a laboratory and taking measurements there is expensive and time consuming. In addition, the need to transport samples to a laboratory to obtain fluid property data prevents such data from being used in real time in response to changing conditions in the well. Therefore, there is a need for a sensor that can measure downhole fluid viscosity and density in situ under downhole or oilfield conditions.

Viscosity and PVT properties (or phase diagrams) of downhole fluids are typically measured in the laboratory, and these measurements are used to infer the viscosity and density of the reservoir and fluids along the wellbore, and to infer the location of significant transitions such as state changes, bubble points, and dew points. However, these inferences may be inaccurate due to irreversible changes that may occur in the fluid as it is brought to the surface, and uncertainties around the exact location in the actual well that meets certain pressure and temperature conditions. For a discussion of viscosity and PVT characteristics of downhole fluids related to hydrocarbon recovery, see Freyss, Henri et al, "PVT Analysis for Oil resvors," reserve Oil ENGINEERING, The Technical Review, volume 37, phase 1, pages 4-15, publication date: 1/1989, which is incorporated herein by reference in its entirety. Therefore, there is a need for a small, fast and accurate sensor that is capable of measuring the viscosity and density of hydrocarbons along a production well, as these data, in combination with temperature, pressure and depth/location along the well, can be used to determine the true location and condition at which significant transitions occur.

The downhole fluid stream is typically a two-phase or multi-phase fluid stream, composed of two or more different or immiscible fluids. The flow regime (e.g. bulk flow, laminar flow, bubble flow) depends on the flow rate of the different phases as well as the viscosity and density of the phases. Flow conditions can severely impact the effectiveness and durability of downhole equipment, such as manual lift systems. In some flow states, the flow rates of different phases may be correlated, while in other flow states, the flow rates may not be correlated. Knowledge of the flow rate of each phase is important to optimize production and surface facilities as well as to detect production problems (e.g., water infiltration). The simplest flow monitoring sensor can measure the total flow rate (without distinguishing between phases) and measure the volume percentage of the different phases. The flow rates of the phases are determined by multiplying the total flow by the volume percent of each phase. This measurement is accurate only when all phases move at the same speed. In some flow conditions, the different phases move at different speeds, which may lead to inaccurate measurement results. Therefore, there is a need for a small, inexpensive sensor that can measure the instantaneous viscosity and density of fluids it contacts to help determine the flow state, the relative abundance of each phase, the shape and size of the flow structure of each phase, and the degree of velocity correlation between the fluid phases. Because of the limited space in the well, smaller equipment sizes are also required, especially where permanent or cordless sensing is required without significant interference with hydrocarbon production. The present invention addresses these and other needs in the art.

Disclosure of Invention

Embodiments of the present invention provide an apparatus for determining an uncharacterized property of a downhole fluid. The apparatus comprises an oscillation driver circuit comprising an amplifier having an output and an input, a feedback loop between the output and the input of the amplifier or logic gate; an electromechanical resonator disposed within the feedback loop such that the resonator is driven by the oscillation driver circuit, wherein a resonant frequency of the resonator defines an oscillation frequency of the oscillator circuit; and switching means for causing the oscillator circuit to cease driving the resonator so that the decay rate of the oscillation of the electromechanical resonator in the uncharacterized fluid can be observed. The electromechanical resonator is encapsulated in a conductive layer to protect the resonator from the capacitive effects of the downhole fluid.

In certain embodiments, at least one property of the uncharacterized fluid may be determined from an observed decay rate of oscillations of the electromechanical resonator. The electromechanical resonator may be implemented as a piezoelectric device having at least two electrodes and a dielectric layer covering the at least two electrodes, the conductive layer surrounding the dielectric layer. The electromechanical resonator may include an anti-corrosion layer over the conductive layer.

In another aspect of the invention, a method and apparatus for determining the PVT properties or phase diagram of a downhole fluid, or dispersed fluid-fluid (emulsion), solid-fluid (suspension), or gas-fluid (foam) system, is disclosed. In one embodiment, density and viscosity measurements are made as the device occupies different depths in the well, so that downhole fluid properties can be measured at different pressures and temperatures encountered at each depth. Based on these measurements at discrete pressure and temperature points, a complete PVT characteristic or phase diagram can be reconstructed by interpolation. This interpolation is typically done by selecting from a series of theoretical PVT characteristics the characteristic that most closely matches the property measured along the well. Alternatively, the series of PVT properties may be determined empirically based on PVT properties measured in a laboratory for similar fluids; the PVT properties that best match the finite data set acquired in the well are selected from the series and assumed to describe the fluid in the well.

In some embodiments, the apparatus further comprises a microprocessor coupled to the switching device and adapted to adjust a duty cycle and a gain rate of the switching device.

Embodiments of the present invention also provide a piezoelectric device for determining an uncharacterized property of a downhole fluid when driven by an alternating voltage. The apparatus includes at least two piezoelectric electrodes formed in a tuning fork shape, the piezoelectric electrodes of the fork shaped portion being operable to mechanically vibrate at a characteristic resonant frequency in response to electrical excitation by an alternating voltage; a dielectric layer covering the at least two piezoelectric electrodes; and a conductive layer covering the dielectric layer and protecting the piezoelectric device from capacitance of downhole fluids.

In certain embodiments, the piezoelectric device includes an anti-corrosion layer over the conductive layer. The conductive layer may be formed of a metal thin film.

Other embodiments of the present invention provide a method for determining a property of a fluid, comprising the steps of: a) exposing an oscillator circuit to an uncharacterized fluid, the oscillator circuit comprising an amplifier having an output and an input, a feedback loop between the output and the input of the amplifier or logic gate, and an electromechanical resonator disposed within the feedback loop such that a resonant frequency of the electromechanical resonator defines an oscillation of the oscillator circuit, the electromechanical resonator being encased in a conductive layer; b) activating the oscillator circuit such that the electromechanical resonator reaches a resonant frequency in an uncharacterized fluid; c) determining a damping ratio of the electromechanical resonator in the uncharacterized fluid while the oscillator circuit is continuously activated; and d) calculating at least one property of the uncharacterized fluid by reference to the gain or negative resistance required to maintain a constant oscillation amplitude by the automatic gain or negative resistance control system.

In certain embodiments, the conductive layer is comprised of a thin metal film. The electromechanical resonator may be implemented using a piezoelectric device having at least two electrodes and a dielectric layer covering the at least two electrodes, the conductive layer surrounding the dielectric layer. The electromechanical resonator may include an anti-corrosion layer over the conductive layer.

These and other aspects, features and advantages will be apparent from the following description of certain embodiments of the invention, the accompanying drawings and the claims.

Drawings

Fig. 1 shows an oscillator circuit according to a first arrangement;

FIG. 2 shows electrical waveforms recorded within the oscillator of the first arrangement;

fig. 3 shows an oscillator circuit according to a second arrangement;

FIG. 4 shows a tuning fork resonator;

FIG. 5 shows an example of measured viscosity data;

FIG. 6 shows an example of measured density data;

FIG. 7 shows a Butterworth-Van Dyke model of a piezoelectric resonator;

FIG. 8 shows an oscillator circuit in which a variable negative resistance is simulated and controlled in a feedback loop to maintain a constant oscillation amplitude;

FIG. 9 illustrates the addition of parallel and series electrical impedances to the resonator that may be required to oscillate the circuit of the present invention when the resonator is in a liquid or otherwise has large damping;

10A and 10B show two vibration modes of a tuning fork oscillator, FIGS. 10C and 10D show tuning fork responses under in-plane shear actuation with an external piezoelectric transducer, with the arrows pointing in the actuation direction;

11A-C illustrate a three electrode (input, ground, output) tuning fork configuration and response to decoupling drive and sense signals;

12A-C illustrate a two-sided (back and front) three-electrode (input, ground, output) tuning fork configuration and response to decoupling drive and sense signals;

FIG. 13 is a differential circuit diagram in which the tuning fork box represents the electromechanical model of the tuning fork with parasitic capacitance C1.

Fig. 14 shows a similar resonator circuit for a piezoelectric electromechanical resonator, where Lt is the effective inductance representing the mechanical inertia of the resonator.

Fig. 15A and 15B are current amplitude versus frequency curves illustrating ideal resonance in terms of amplitude (fig. 15A) and phase shift (fig. 15B) of the output current when the additional capacitance is not considered.

Fig. 15C and 15D are graphs of current amplitude versus frequency showing parasitic capacitances (Cp, Cm, Cd in fig. 14) at low and high damping conditions (i.e., small and large Rt).

Fig. 16A and 16B are graphs of current amplitude versus frequency response for resonators with varying parasitic capacitance and damping in the case of differential output.

FIG. 17 shows an exemplary embodiment of a piezoelectric tuning fork resonator 1700 according to the present invention.

Fig. 18 is another embodiment of a differential circuit providing automatic gain control.

Detailed Description

In accordance with a broad aspect of the present invention, the inventors have recognized that small, fast and accurate sensors capable of measuring viscosity and density and mounted on a platform that enables them to obtain measurements at various depths along a production well may be used, for example: 1) plotting the PVT properties of the downhole fluid and determining the locations along the wellbore where dew point, bubble point and/or other significant state changes and fluid property transitions occur, 2) plotting the PVT properties of dispersed fluid-fluid (emulsion), solid-fluid (suspension), gas-fluid (foam) systems by rapidly measuring the changes in density and viscosity, 3) more accurately determining the true viscosity and density of the reservoir fluid at reservoir conditions, and 4) determining the individual phase densities and viscosities of the transient phases present at the sensors in the multiphase flow and the time series from which the flow state, shape and size of the flow structure of each phase and the volumetric flow rate of each phase can be more accurately inferred.

According to one embodiment of the invention, an apparatus includes a platform for positioning a sensor at a desired depth in a subterranean well; an electromechanical resonator; an oscillator circuit incorporating an electromechanical resonator as its frequency defining element; and a microcontroller that measures the frequency (or period) of the oscillation of the oscillator circuit and measures the damping of the oscillation. Such an embodiment differs from the prior art approach in that the electromechanical resonator defines the resonant frequency of the oscillator circuit, as opposed to the conventional arrangement of a resonator circuit which is different from the resonator and which has to be tuned until the resonant frequency of the electromechanical resonator is found. This embodiment also differs in that the damping of the oscillation or other energy loss parameter is measured directly to determine the energy loss or damping caused by the fluid contacting the resonator.

Oscillator circuits typically include an amplifier having at least one feedback path from an output of the amplifier to an input of the amplifier. A frequency defining element (e.g. a quartz crystal) is typically included along one such feedback path in such a way as to induce a sustained oscillation at the resonant frequency of the frequency defining element. Such continuous oscillation occurs when the total phase shift around the loop including the frequency defining element and the amplifier is 360 degrees or a multiple thereof and the gain around the loop including the frequency defining element and the amplifier is not less than 1.

The inventors have realised that a circuit arrangement comprising an electromechanical resonator as the main element in the oscillator circuit for determining the oscillation frequency has substantial benefits over prior art approaches, including the first being that there is no need to search through a plurality of frequencies to find the resonance frequency of the resonator when the oscillator circuit starts to oscillate at the desired resonance frequency. This makes measurements using the method of the present invention much faster than conventional methods. Second, the amount of circuitry and complexity required to complete the measurement is greatly reduced. Third, there is no need for a precision oscillator capable of generating a precisely controlled frequency with fine frequency resolution; the circuit of the present invention oscillates only at the required frequency and the frequency can be measured using a simple crystal-based timer, a device that is already available on many small microcontrollers. Fourth, incorporating the resonator as a frequency-defining element into the oscillator circuit means that a large amount of energy is stored in the resonator, which provides a powerful signal for measuring damping due to surrounding fluid. In contrast, driving a resonator with a pulse or step function to observe its resonant frequency and damping results in a much smaller signal because only a small fraction of the energy in the pulse or step function is within the resonant frequency band of the resonator. Fifth, among other advantages, incorporating the resonator into the feedback loop of the amplifier in the oscillator circuit provides an efficient circuit design opportunity to provide amplification of the sensor signal during damping measurements with the same amplifier.

There are various ways to measure damping. In one embodiment, the damping is determined by briefly stopping the electrically driven resonator so that the amplitude decay of the oscillation and the rate of amplitude decay can be measured. This may be accomplished, for example, using an oscillator circuit based on NAND gates, as shown in fig. 1 and described in more detail below. An envelope detector circuit may be used to provide the oscillation amplitude, which may then be digitized and fitted to an exponential decay curve to determine the damping coefficient. Alternatively, two voltage comparators and a timer may be used to determine the time it takes for the envelope to decay between two reference voltage levels, and this time may be used to determine the damping coefficient. Alternatively, a constant oscillation amplitude can be maintained by using an automatic gain adjustment circuit that sets the gain of the aforementioned amplifier in a closed loop. In this alternative measurement technique, the decay time is measured without stopping the oscillation; conversely, the amount of gain required to maintain a constant amplitude due to the fluid surrounding the electromechanical resonator may be used to determine the damping coefficient. The gain of an automatic gain controlled amplifier may be determined by digitizing a gain adjustment signal that controls its gain and using previous calibration measurements that correlate the gain adjustment signal to the amount of gain produced by the amplifier.

According to one embodiment, the sensor may comprise an electromechanical device in contact with the fluid and being a component of an electrical circuit that drives and detects oscillatory motion of the device as a function of time when the sensor interacts with the fluid. The motion of the electromechanical device is affected by the fluid and a quantitative relationship is directly established between the viscosity and density of the fluid and the resonant frequency and damping of the electromechanical resonator. Alternatively, if the viscosity and density of the fluid are well known functions of temperature and pressure (e.g., for methane), the temperature and pressure of the fluid can be used to establish a direct relationship between frequency and damping so that its thermodynamic state and other properties (bubble point, dew point, GOR, etc.) associated with that state can be known in real time.

In particular, a piezoelectric tuning fork may be used as the electromechanical vibration device. One example of a suitable piezoelectric tuning fork is described in U.S. patent 7,562,557 to Bennett et al, which is incorporated herein by reference in its entirety. However, the invention is not limited to a particular resonator element, as long as it can be integrated into the oscillator circuit as part of the oscillator itself. The mechanical properties of the tuning fork as a two-terminal device can be described by a Butterworth-Van Dyke model (as shown in fig. 7) as a series resistance (R), inductance (L), and capacitance (C) circuit (RLC circuit) representing the mechanical damping, mass, and compliance of the device, respectively, where the parallel capacitance (C0) represents the capacitance of the device, including the capacitance between the electrodes of the device, including any stray capacitance wires between the electrical leads, and the capacitance due to the dielectric medium surrounding the device. The model represents a resonant system and establishes a direct relationship between the mechanical and electrical domains through piezoelectric action. When C0 is much larger than C, it is difficult to oscillate a high damping resonator (e.g., a resonator in a viscous liquid). Therefore, the preferred resonator design utilizes the selection of piezoelectric materials, shapes and sizes to minimize C0 relative to C. Preferred resonators also maintain a high Q factor (quality factor) of resonance in the liquid state, for example by reducing the area in contact with the liquid. In addition, an inductor (or circuit simulating the action of the inductor) can be placed in parallel or in series with the resonator to cancel the action of C0 near the resonant frequency, thereby making the oscillator circuit easier to resonate when the tuning fork is in a high viscosity liquid. Alternatively, a reference capacitor having the same or very similar parasitic capacitance as the piezoelectric oscillator may be used in the differential measurement scheme (fig. 13). In this technique, the same input signal is fed to the piezoelectric resonator and the reference capacitor, and their outputs are subtracted from each other to cancel the influence of the parasitic capacitance. Even when the electromechanical response signal is very small compared to the parasitic signal, the subtraction and amplification can achieve oscillation even in a high damping environment. A differential amplifier may be used for the subtraction. In one embodiment, the reference capacitor may be a second tuning fork, the tines of which are clamped or held by epoxy. This ensures that the piezoelectric (electromechanical) response of the reference capacitor does not contribute significantly, but only the parasitic capacitance, around the frequency of interest. By patterning electrodes on the same type of piezoelectric substrate having the same geometry as the electromechanical resonator without the tuning fork shape, almost the same capacitor can be manufactured; it therefore does not have any resonant frequency near the resonant frequency of the tuning fork.

In another embodiment, the drive and sense functions of the electromechanical resonator may be decoupled. This can be done using different physical effects for driving and sensing, for example using various combinations of transduction between electrical, magnetic, mechanical and optical domains. Alternatively, if the same transduction method is used for driving and sensing, decoupling can also be achieved by spatially separating the areas of driving and sensing. The necessary separation length may vary depending on the transduction used. In the specific example of a piezoelectric tuning fork, it can be driven to a resonant state in several ways, and the resulting mechanical deformation on the piezoelectric material can be sensed as a voltage output through a patterned electrode. This approach minimizes parasitic capacitance and minimizes direct electrical signal coupling of the input signal to the output. In its simplest form, a tuning fork can be rigidly mounted on a mechanical vibrator, such as a shear piezoelectric transducer. Applying an electrical signal to the shear piezoelectric at the resonant frequency of the tuning fork, the motion of the shear piezoelectric may be mechanically coupled to the tuning fork, and the resonant mode of the tuning fork may be excited. When the tuning fork is deformed, it produces a space charge distribution along its piezoelectric crystal that can be picked up as a voltage difference between the patterned electrodes on the tuning fork. The efficiency of the coupling of the initial motion to the desired resonant frequency depends on the direction of the tuning fork relative to the direction of motion of the shear piezoelectric and the stiffness of the bond between the two objects. For example, to excite the tuning fork's scissor mode without exciting the fundamental cantilever mode of the entire tuning fork body, the direction of the shearing motion should be orthogonal to the direction of the scissor motion, as shown in FIGS. 10A-D. Alternatively, the electrodes may be patterned on the tuning fork so that driving and sensing can be performed by the distal electrodes, as shown in fig. 11A-C. In such a three-electrode scheme, the parasitic capacitance between the input and output ports may be several orders of magnitude smaller than that of a two-terminal device having a similar geometry. By including manufacturing complexity, the second set of electrodes can be patterned on the backside of the tuning fork. The front electrode can be used for driving and the back electrode can be used for sensing. In this case, as shown in fig. 12A-C, the electromechanical conversion efficiency can be further improved.

According to one aspect of the operation of such a circuit, the resonator is incorporated into the oscillator circuit and then turned on. The oscillator circuit may comprise the circuits of fig. 1 and 3, which will be described in more detail below. By means of the continuous feedback mechanism of the circuit, the tuning fork oscillates from a small fluctuation of its motion, which overcomes the damping from the environment and gradually increases until a maximum amplitude is reached. At this point, the feedback mechanism is turned off and the oscillation of the resonator is damped due to environmental damping. This process is repeated continuously and the frequency and decay time of the oscillation is obtained at each off-period.

The model describing the oscillation attenuation is a model of a damped driverless harmonic oscillator whose solution to the oscillation velocity (proportional to the current generated by the piezoelectric effect) is given by:

wherein

Is the oscillation phase, the decay time constant τ is related to the damping of the fluid, and the frequency ω is related to the effective mass of the resonator, including the additional mass of the fluid dragged by the resonator. These quantities are related to the quality factor Q of the oscillator, as follows

And

in a liquid environment, Q becomes very small, e.g., about 10. Using the above equations, and for signals at tens of kilohertz, e.g. 3x10⁴Estimated natural frequency ω of KHz₀It will be appreciated that a time constant of about one millisecond is obtained, which allows for very rapid measurements and "real-time" information of fluid properties to be calculated and reported to downstream systems, such as hardware processor-based machines that execute or otherwise implement code to configure those machines to process fluid property data received from such circuits in a production well.

Various oscillator circuits may be adapted to work with electromechanical resonator sensors. The oscillator circuit must be able to oscillate in all circumstances in which a measurement is to be obtained. For a liquid environment, the Q (i.e., quality factor) of the resonator is small (e.g., about 10), and some oscillator circuits may not provide sufficient amplification to sustain oscillation. In this case, additional amplification may be employed. For example, if an unbuffered logic gate cannot provide enough amplification to sustain oscillation in a lossy environment, it may be replaced with a buffered logic gate or a plurality of logic gates in series to provide the additional amplification required for oscillation. In some cases, when the Q factor of the resonator is small due to being in a viscous fluid, it may not produce enough phase shift to cause the oscillator circuit to oscillate. In this case, additional impedances, such as reactances, may be added in parallel and/or in series with the resonators to provide additional phase shift. As shown in fig. 9, the impedance Zp is increased in parallel with the resonator X, and the impedance Zs is increased in series with the resonator X. It will be appreciated that the "sensor" symbolically illustrated by the crystal schematic in figures 1 and 3 includes an electromechanical resonator and parallel and/or series impedances that must be added to the resonator to cause the circuit to oscillate in the target fluid. Those skilled in the art will recognize that the impedance may be added as a network of one or more resistors, capacitors, and inductors, or as an active circuit that models the current-voltage relationship of these networks. Active circuits have certain advantages because they can create current-voltage relationships, whereas passive components (e.g., "negative resistances") cannot create these relationships. Small active circuits may also mimic relationships where corresponding passive components would be much larger, for example when modeling large inductors. In one embodiment, in the circuit shown in FIG. 1, the "sensor" consists of an inductor in series with the resonator to enable the circuit to operate in a fluid of higher viscosity.

In one embodiment, an oscillator circuit suitable for use in the present invention includes a means to deactivate the electromechanical resonator so that the decay of the oscillation can be observed. For example, a feedback loop containing an electromechanical device may be opened, or other circuit components in the oscillator circuit may be opened, shorted, or otherwise altered to change the resonator from a driven state to a non-driven state. The circuit of the oscillator circuit or a separate circuit may comprise means for measuring the oscillation frequency and means for measuring the decay rate of the oscillation. Two such embodiments are discussed below.

Referring to fig. 1, a circuit 100 includes a NAND logic gate 102 configured as an amplifier as described below,

resistors

104 and 106, and

capacitors

108 and 110 connected around a "sensor" 112 that includes an electromechanical resonator that both forms part of an oscillator circuit to define the oscillation of the circuit and is configured to be in direct contact with a fluid to be measured so that the effect of the fluid on the resonator performance can be measured to determine fluid characteristics. The gain in the oscillator circuit 100 is provided by a NAND logic gate (U1)102, i.e., the NAND/logic gate acts as an amplifier for the circuit. The oscillator is disabled by a logic low on the on/off input and enabled by a logic high on the on/off input. This input may be provided by a digital output of the microcontroller, for example, to control whether the oscillator circuit is in a driven mode or a non-driven mode. After oscillation, the non-driven mode is suitable for collecting damping data and determining fluid properties, for example using the above formula or other equations that benefit from damping data. The "to timer" output of the circuit has a square wave at the oscillation frequency. This output may be provided, for example, to a timer input of the microcontroller, so that the frequency of the oscillation can be accurately measured by the timer system of the microcontroller. If desired, the "to ADC" output of the circuit may be provided to an analog-to-digital converter to sample the ringing and determine damping.

Fig. 2 shows an output graph 200 that includes two

waveforms

202 and 204 for the circuit 100. When the on/off input is high 206, oscillation begins and eventually reaches a steady amplitude, as shown by waveform 202. Then, when the on/off input goes low 208 (time 0 seconds in fig. 2), the NAND gate is disabled and the oscillation decays, as shown by waveform 204. The waveforms to the output of the ADC are shown as 202, 204. The data shown in fig. 2 are measured by a sensor in vacuum. The time to start oscillation and the decay time are much faster when the sensor is in a liquid. In any case, the damping rate of the electromechanical device in the reference liquid and at a specified temperature and pressure may be obtained for benchmarking purposes (e.g., calibrating a given sensor 112).

Fig. 3 shows another embodiment in which a circuit 300 is provided. Circuit 300 includes analog switch 302, operational amplifier 304,

resistors

306, 308, and 310, and

diodes

312 and 314. Also, in significant part, the circuit 300 includes a "sensor" 316 that includes an electromechanical resonator that both forms part of the oscillator circuit to define the oscillation frequency of the circuit and is configured to be in direct contact with the fluid to be measured so that the effect of the fluid on the resonator performance can be measured to determine the fluid properties. The gain in oscillator circuit 300 is provided by an operational amplifier (op amp) (U2) 304. When there is a logic low level on the on/off input, the analog switch U3302 is open and does not drive oscillation (i.e., it decays). In this mode, the operational amplifier 304 acts as a current-to-voltage converter, providing a voltage on the output to the ADC proportional to the decaying current oscillations from the sensor. When a logic high level is present on the on/off input, the analog switch U3302 closes, allowing positive feedback, which results in continuous oscillation at the sensor resonant frequency. The on/off input may be provided by a digital output of the microcontroller. The "to ADC" output of the circuit may be provided to an analog-to-digital converter which samples the sustained and decaying oscillation and enables the processor to determine the frequency and decay time of the oscillation. Diodes 312 and 314(D1 and D2) may prevent the op amp output from being driven into saturation. Without these diodes, the oscillation frequency would be reduced due to the time it takes for the operational amplifier to reach saturation.

In another embodiment, damping is determined based on the amount of negative resistance that must be added in series or parallel with the resonator to maintain a constant amplitude oscillation (e.g., the circuit of FIG. 3 may be modified to accomplish this, where the resistor (308) defining the negative resistance is replaced with a variable resistor device (e.g., an N-channel enhancement mode MOSFET) and the gate voltage of the MOSFET (which determines the drain-source resistance in the linear region of operation) may be adjusted to keep the amplitude of the amplifier output constant. accordingly, FIG. 3 shows a circuit using an oscillator of an operational amplifier. the resistance provided by the MOSFET, and thus the negative resistance required to keep the amplitude of oscillation constant, may be determined by sampling the gate voltage of the MOSFET. Damping is determined based on the amount of gain that must be applied in the feedback loop of a resonator contained within the tank circuit to maintain the oscillation at a constant amplitude.

Suitable platforms that can deliver sensors to a desired location or set OF locations IN a production WELL include wireline tools as known to those skilled IN the art, such as the untethered sensors described IN co-pending U.S. patent application 15/143,128 entitled METHOD AND DEVICE FOR imaging MEASUREMENTS OF down gas procedure IN a wireline WELL filed on 29.4.2016, which application is incorporated herein by reference as if set forth IN its entirety herein, or on network nodes at different depths IN a permanently deployed network OF sensors disposed within a WELL. The structure of the device makes it particularly useful for remote downhole operations because the device can be made to fit into small packages (e.g., the volume of the resonator and circuitry can be less than 1cc) and consume little power (e.g., about 1 microjoule per measurement).

In one embodiment, the circuit of FIG. 3 is modified to implement automatic gain control as shown in FIG. 8. The resistor (308 in fig. 3) that sets the negative resistance seen at the non-inverting input of the operational amplifier may be replaced by an N-type enhancement mode MOSFET, where the gate voltage supplied to the MOSFET is the gain control voltage and is the output of the envelope detector that is applied to the oscillation at the output of the operational amplifier. The voltage obtained by the MOSFET gate therefore depends on the oscillation amplitude. If the oscillation amplitude decreases, the gate voltage will decrease. This will increase the drain-source resistance of the MOSFET, thereby increasing the negative resistance added to the resonator and increasing the oscillation amplitude. If the oscillation amplitude increases, the gate voltage will increase. This will reduce the drain-source resistance of the MOSFET, thereby reducing the negative resistance added to the resonator and reducing the oscillation amplitude. Thus, with a circuit of this arrangement, the amplitude of the oscillation will be maintained at a constant level. The negative resistance required to maintain a constant amplitude oscillation can be determined by measuring the gate voltage applied to the MOSFET in combination with a calibration curve that gives the drain-source resistance of the MOSFET as a function of its gate voltage. In this regard, the hardware processor may execute or otherwise implement code therein to identify the resistance from the calibration curve (or a function representative of the calibration curve) based on the measured gate voltage.

In one embodiment, the circuit may alternatively be configured to drive the resonator such that oscillation is initially established and maintained at least briefly, and then to stop driving the resonator, allowing the oscillation to decay (e.g., using the circuit of fig. 1) so that a decay measurement may be made or otherwise completed. In another embodiment, the gain of the circuit is adjusted to maintain a constant oscillation amplitude, and the amount of gain required is measured as an indication of the amount of damping. This is because greater damping requires greater gain to maintain constant oscillation. In another embodiment, the circuit is configured to emulate a "negative resistor" connected to the resonator. The amount of negative resistance will automatically adjust to maintain a constant amplitude oscillation and the amount of negative resistance required will be measured as an indication of damping. This approach can be used because a larger negative resistance is needed to counteract the larger damping energy loss, which can be considered as a (positive) resistance inside the resonator.

In one embodiment, the variable resistor circuit may be as shown in FIG. 8, where the adjustable resistor is actually the drain-source resistance R of the MOSFET_DSWhich is regulated by the gate voltage of the MOSFET. In this embodiment, the gate voltage of the MOSFET is generated by an envelope detector on the output of the amplifier, so a larger amplitude output results in a larger gate-source voltage and therefore a smaller R_DSThis results in a decrease in oscillation amplitude until R_DSDamping in the resonator can just be counteracted. The gate voltage is sampled by an A/D converter in the microcontroller, which may be correlated with R, for example, using code executed in the microcontroller that correlates with the value just described_DSIs related to the amount of damping in the resonator. The oscillation frequency may also be measured by the microcontroller. Based on the measurement andR_DSthe frequency of the undriven ringing can be calculated.

In one embodiment, a system and method are provided that can measure temperature and pressure in addition to the density and viscosity of all fluids present in multiple locations along a wellbore. Temperature and pressure measurements are achieved by including commercially available temperature (e.g., RTD) sensors and pressure sensors. Their readings are measured by an analog-to-digital converter interfaced to a microcontroller in the device so that temperature and pressure data, as well as density and viscosity, can be recorded or transmitted.

Although the viscosity is not plotted in the PVT plot, it can be used as an indication of when a particular state change, bubble point, dew point, etc. has occurred. Untethered sensor balls (e.g., the sensor ball described in co-pending U.S. patent application 15/143,128 referenced above) provide an inexpensive solution for measuring fluid properties at all pressures and temperatures between the surface and any given reservoir depth. By measuring the temperature, pressure, and density and viscosity of all fluid phases encountered traveling down the well from the surface to a selected reservoir depth, the apparatus, methods and systems of the present invention can reconstruct the viscosity and phase map information of the recovered fluid along the most important pressure-temperature trajectory (i.e., found in the wellbore).

It is recognized that knowledge of the viscosity and density of each phase can help determine flow regime and improve accuracy of the phase flow rate. The ability to measure viscosity and density at a fast sampling rate can immediately indicate which phase is present in the multiphase flow, providing a time series of phases at the sensor location. The time series can be used to determine the abundance of each phase and the size and shape of the flow structure of each phase, thereby determining the flow state. A second such time series downstream of the first may be associated with the first to determine the time it takes for each fluid packet to move between the sensors, thereby more accurately measuring the phase flow rate.

As mentioned above, the oscillator circuit incorporates an electromechanical resonator arranged within the feedback loop of the circuit such that the resonant frequency of the resonator defines the oscillation frequency of the oscillator circuit. In addition to defining the oscillation frequency of the circuit, the resonator is also in contact with the fluid to be measured. This arrangement allows the resonant frequency to be determined much faster than in other systems where the resonator is separate and distinct from the oscillator circuit driving it. Thus, the arrangement disclosed herein allows for increased measurement speed. An increase in measurement speed is beneficial in many ways. For example, in a production well, there are often multiple fluids entering the well. The increase in measurement speed results in an increase in the number of measurements that enables one to distinguish between the various fluid types in the production well and determine the correct viscosity for each fluid type. Conversely, a system with a slower measurement speed provides less accuracy because the sensor may be in more than one fluid type during the measurement.

Additionally, the increased speed of the systems and methods described herein allows one to address the separation of fluids in multiphase flows in production wells. In contrast, slower sensors can obscure the fluid properties of the various fluids they encounter during measurement. The increase in sensing speed allows the composition and structure of the multiphase flow to be determined. Also, the increase in sensor speed allows for multiple rapid measurements, which allows one sensor to sense and perceive individual fluid phases in the downhole fluid, while the other sensors are not fast enough in response to provide the necessary fine granularity in the data to perceive the individual fluid phases. In a production well, the sensors may be in different fluid types every few milliseconds as bubbles of oil, gas and brine are flooded.

According to one non-limiting example, an embodiment was tested under laboratory conditions. Referring to fig. 4, the system includes a tuning fork 400 (i.e., a wired tuning fork oscillator) as an electromechanical resonator. The tuning fork was fully characterized under simulated downhole pressure and temperature conditions. The device is electrically actuated and piezoelectrically induced using lock-in amplification techniques and direct frequency response measurements of its impedance. Formants are obtained and fitted with peak width, amplitude and resonant frequency as fitting parameters. The peak width and frequency allow to extract the damping and additional mass of the oscillator in the fluid. A fluid dynamics model was developed to calibrate the resonance response with the viscosity and density of the test fluid. For this purpose, sensors are activated in different fluids (air, water, mineral oil, hydraulic oil) and calibration parameters are obtained. The test apparatus was then run at different pressure and temperature conditions to simulate downhole conditions (see fig. 5 and 6). More specifically, fig. 5 shows the results of the measured viscosity of the ISO 15 hydraulic oil at high pressure and high temperature, and fig. 6 shows the results of the measured density of the ISO 15 hydraulic oil at high pressure and high temperature.

The device was found to be suitable for measuring viscosities in the target range (up to 50 cP).

One problem with piezoelectric electromechanical resonators is the decoupling of the electrical and mechanical response. When an electrical potential is applied between two electrodes deposited on a piezoelectric material, the material undergoes a mechanical change due to an electric field. The mechanical principle of the resonator may be shown to be similar to an RLC circuit. Fig. 14 shows a similar resonator circuit, where Lt is the effective inductance representing the mechanical inertia of the resonator, Ct is the effective capacitance representing the mechanical spring constant, and Rt is the effective resistance representing the mechanical damping. However, the two electrodes of the resonator also form a capacitor, so similar resonator circuits also need to take into account electrical coupling to accurately model the resonator. For example, in this arrangement, Cp is the parasitic capacitance, which is the effective capacitance through the piezoelectric crystal. Cd is the capacitance through the dielectric coating and Cm is the capacitance through the fluid medium in which the resonator is intended to be used. Additionally, Rd and Rm are the effective resistances of the dielectric and the dielectric, respectively.

Fig. 15A and 15B illustrate ideal resonance in terms of amplitude (fig. 15A) and phase shift (fig. 15B) of the output current when the additional capacitance is not considered. As shown, in fig. 15B, at resonance, the phase shift is zero, i.e., no phase shift hysteresis is observed. However, when considering parasitic capacitance and other capacitances, a 90 degree phase shift occurs at resonance, and the output current is the sum of currents due to the mechanical system and parasitic capacitance. Especially when the mechanical damping is large (i.e. in the liquid), the current flowing through the parasitic capacitor dominates and the resonant behavior becomes indistinguishable.

Fig. 15C and 15D show the effect of parasitic capacitance (Cp, Cm, Cd in fig. 14) under low and high damping conditions (i.e., small Rt and large Rt). Especially when using a closed loop system, it is important to obtain a clear formant with zero phase difference. As shown in fig. 15C and 15D, when the parasitic capacitance and high damping play a dominant role, the resonance behavior is weakened, and zero-phase crossing does not occur.

A differential circuit as shown in fig. 13 can help eliminate the effect of parasitic capacitance. Differential circuit 1300 includes a tuning fork resonator 1302 and a reference capacitor 1304. The electrical output of the resonator 1302 is fed to a first input of a differential amplifier 1306 and the electrical output of the reference capacitor is fed to a second input of the differential amplifier 1306. The output of the differential amplifier is therefore proportional to the difference between the outputs of the resonator 1302 and the reference capacitor 1304. A large output from the differential circuit 1300 means a large difference in output between the resonator and the reference capacitor. This signal can then be used to determine and compensate for the difference in capacitance between the resonator 1302 and the reference capacitor 1306.

However, the differential circuit approach is less suitable for applications where the total contribution to parasitic capacitance depends on the dielectric constant of the fluid medium. Fig. 16A and 16B show example frequency responses of resonators with varying parasitic capacitance and damping in the case of differential output. Cm2 represents the parasitic dielectric capacitance of the reference. The parasitic dielectric capacitance of the resonator is defined as 5 pF. As shown, when the resonator capacitance is different from the median capacitance (5pF), the resulting current output does not peak at the resonant frequency, or zero phase shift at the resonant frequency. When the resonator capacitance is equal to 5pF, a resonance peak and zero phase shift occur.

Therefore, in order to obtain the best results from the differential approach, the reference capacitor needs to be adjusted for each medium change, and this makes it difficult to maintain a zero phase difference between the feedback signal and the electromechanical response at all times in a closed loop system.

In another embodiment of the closed loop differential circuit shown in fig. 18, an automatic gain control is implemented, which allows the damping to be measured without turning off the drive signal. In this circuit 1800, digital

variable resistors

1805, 1810 are used as feedback resistors for respective operational amplifiers 1815, 1820 controlled by a microprocessor (uC)1830 to adjust the gain. The sinusoidal voltage amplitude on the analog node (i.e., the input or output of the electromechanical oscillator) may be obtained by feeding the AC-coupled oscillation signal to a comparator 1840 whose known reference value (Vref) is less than the desired constant voltage level on the analog node. When the circuit is activated, the microprocessor 1830 increases the gain until the comparator outputs a pulse train indicating that the oscillating voltage amplitude is greater than the reference voltage. The microprocessor then determines the pulse width and frequency of the pulse train by precisely timing the captured rise and fall events so that the duty cycle can be accurately measured. For sine wave oscillation, the oscillation amplitude is related to the duty cycle and the selected reference voltage. A control algorithm implemented by the microprocessor (e.g., proportional-integral-derivative control) can maintain a constant duty cycle by adjusting the feedback gain according to a known relationship. The gain value may be used as a measure of damping.

The use of identical resonator pairs, one of which is fixed and used as a reference, may also help to obtain very close parasitic capacitances. However, variations in packaging and small differences in the dielectric properties across the capacitor can easily interfere with such a system. Thus, minimizing the parasitic capacitance of the resonator achieves optimal device operation results for sensors used in fluids with varying dielectric constants. The reduction in the variation in the dielectric constant is obtained by a) first coating the electrode with a dielectric material and then b) coating the dielectric material with a metal film. This prevents the electric field generated in the resonator from spreading into the medium. Referring again to fig. 14, the use of a conductive metal coating eliminates Cm, reduces Rm, and eliminates the dependence of parasitic capacitance due to the fluid medium. By fixing the capacitance of the resonator device, it is again possible to use a fixed, metal-coated identical resonator as a reference, or a well-tuned fixed capacitor as a reference.

FIG. 17 shows an exemplary embodiment of a piezoelectric tuning fork resonator 1700 according to the present invention. As shown in the cross-sectional view of fig. 17, the resonator includes several layers of material. At the bottom is an electrode 1702 made of a conductive material that is coupled to control circuitry for operating the resonator (not shown). A non-conductive dielectric layer 1704 is positioned over the electrodes and a conductive coating 1706 is positioned over the dielectric layer 1704. The dielectric layer 1704 and the conductive coating 1706 can be coated on the electrodes 1702 by various deposition techniques including, for example, sputtering, physical vapor deposition, chemical vapor deposition, or electroplating. The materials (e.g., thin metal films) of the dielectric layer 1704 and the conductive layer 1706 are selected with respect to each other to ensure adhesion between the layers. A thin adhesion promoting film, such as titanium or chromium, may be deposited between the layers. Dielectric layer 1704 is preferably uniform, which can be achieved using chemical vapor or atomic layer deposition methods. The dielectric material may be, but is not limited to, silicon dioxide, aluminum oxide, or silicon nitride. Indium Tin Oxide (ITO), gold, aluminum are potential candidates for the conductive layer.

The role of the conductive layer 1706 is to prevent electrical interaction with the fluid medium, thereby improving sensor reliability and robustness in varying environments, particularly in conductive media. In some embodiments, the material of the conductive layer is selected to also have corrosion resistance. Alternatively, in other embodiments, additional protective coatings may be used for conformal encapsulation of the conductive and dielectric layer 1706 to prevent corrosion, abrasion, and erosion of the dielectric and conductive materials in harsh environments. For example, silicon carbide is one of a group of materials that can be used to encapsulate conductive layers.

The above-described subject matter is provided by way of illustration only and should not be construed in a limiting sense. Various modifications and changes may be made to the subject matter described herein without following the example embodiments and applications illustrated and described, including for example the specific circuit values illustrated in the drawings, and without departing from the true spirit and scope of the present invention.

Claims

1. An apparatus for determining an uncharacterized property of a downhole fluid, comprising:

an oscillation driver circuit comprising an amplifier having an output and an input, a feedback loop between the output and the input of the amplifier or logic gate;

an electromechanical resonator arranged within the feedback loop such that the resonator is driven by the oscillation driver circuit, wherein a resonant frequency of the resonator defines an oscillation frequency of the oscillator circuit; and

switching means for causing the oscillator circuit to cease driving the resonator, thereby enabling observation of a decay rate of oscillations of the electromechanical resonator in the uncharacterized fluid,

wherein the electromechanical resonator is encapsulated in a conductive layer to protect the resonator from capacitive effects of the downhole fluid.

2. The apparatus of claim 1, wherein at least one property of the uncharacterized fluid may be determined from an observed decay rate of oscillations of the electromechanical resonator.

3. The apparatus of claim 1, wherein the electromechanical resonator is a piezoelectric device having at least two electrodes and a dielectric layer covering the at least two electrodes, the conductive layer surrounding the dielectric layer.

4. The apparatus of claim 3, wherein the electromechanical resonator further comprises an anti-corrosion layer over the conductive layer.

5. A piezoelectric device for determining an uncharacterized property of a downhole fluid when driven by an alternating voltage, comprising:

at least two piezoelectric electrodes formed in the shape of tuning forks, the piezoelectric electrodes of the fork-shaped portion being operable to mechanically vibrate at a characteristic resonance frequency in response to electrical excitation by an alternating voltage;

a dielectric layer covering the at least two piezoelectric electrodes; and

an electrically conductive layer covering the dielectric layer and protecting the piezoelectric device from capacitance effects of the downhole fluid.

6. The piezoelectric device of claim 3, further comprising an anti-corrosion layer over the conductive layer.

7. The piezoelectric device of claim 4, wherein the conductive layer is comprised of a metal thin film.

8. A method of determining a property of a fluid, comprising the steps of:

a. exposing an oscillator circuit to an uncharacterized fluid, the oscillator circuit comprising:

i. an amplifier having an output and an input;

a feedback loop between the output and the input of the amplifier or logic gate; and

an electromechanical resonator arranged within the feedback loop such that a resonant frequency of the electromechanical resonator defines an oscillation of the oscillator circuit, the electromechanical resonator being encased in a conductive layer;

b. activating the oscillator circuit such that the electromechanical resonator reaches a resonant frequency in the uncharacterized fluid;

c. determining a damping ratio of the electromechanical resonator in the uncharacterized fluid when the oscillator circuit is continuously activated; and

d. calculating at least one property of the uncharacterized fluid with reference to a gain or negative resistance required to maintain a constant oscillation amplitude by an automatic gain or negative resistance control system.

9. The method of claim 8, wherein the conductive layer is comprised of a thin metal film.

10. The method of claim 9, wherein the electromechanical resonator is a piezoelectric device having at least two electrodes and a dielectric layer covering the at least two electrodes, the conductive layer surrounding the dielectric layer.

11. The method of claim 9, wherein the electromechanical resonator further comprises an anti-corrosion layer over the conductive layer.