METHODS AND APPARATUS FOR DOWNHOLE ACOUSTIC TELEMETRY

Abstract

An improved method and apparatus for transmitting and receiving acoustic signals is presented that enables high speed transmission without repeaters. At the transmitter, a piezoelectric transducer is used to create acoustic waves through the drill string assembly walls. Transmission occurs at more than one passband and each carrier frequency undergoes different modulation and channel coding schemes as the channel permits. At the receiver with multiple accelerometers with known distance, the acoustic signal is processed using several signal diversity techniques to increase the signal-to-noise ratio. A lock-in amplifier extracts the attenuated signals, and a sensor fusion estimation algorithm is performed to decrease the effect of noise. In addition, the wave speed determined from the multiple accelerometer configuration is used for cancellation of echoes and reflections.

Description

FIELD

Embodiments disclosed herein generally relates to a method and apparatus for transmitting, receiving and signal processing attenuated signals from a noisy downhole through mechanical acoustic transmissions without repeaters. Applicant would like to acknowledge that this present invention is based on the work supported by the Korea Agency for Infrastructure Technology Advancement grant, funded by the Ministry of Land, Infrastructure and Transport (21IFIP-B133607-05, Development of directional mud motor and the direction control technology for drilling system).

BACKGROUND

Traditionally, drilling for natural resources was achieved by going straight down to a reservoir. To measure the conditions downhole, the drilling assembly was first removed from the wellbore. Then, sensors and instruments attached to an electrical cable are lowered by a wireline truck. Measurements are performed and sent to the surface through the electrical wire. This operation, however, cannot be run while the wellbore is being drilled and cannot be conducted in real-time. As a result, mud-pulse telemetry technology was developed to allow measurement-while-drilling (MWD).

Mud-pulse telemetry uses the drilling fluid, or mud, that is pumped through the internal passage of the drill string assembly, through the drill bit, and out the annular passage between the drill string walls and the wellbore. The pressure is controlled by valves that either releases or stops the mud from flowing, which is then detected by a pressure transducer at the surface. Mud-pulse telemetry has earned commercial success in drilling operations even though it is only able to transmit at rates of about 3-6 bit per seconds (bps).

Another widely used telemetry method that can boast slightly higher transmission rates (˜6 bps) is electromagnetic telemetry wherein electromagnetic waves are transmitted from downhole sensors to the surface. However, these electromagnetic waves can be severely attenuated by the hydrocarbon formation properties around the wellbore, and also cannot be used on offshore sites. Only at niche locations, where electromagnetic propagation is sufficiently unhindered, can it perform effectively.

Acoustic telemetry tools consisting of electromechanical transducers to produce acoustic waves have been investigated to transmit downhole data to the surface. It has seen commercial application with transmission rates of about 20-30 bps. The acoustic waves would travel through the walls of a drill string assembly and are affected by the different impedances of the assembly components and boundary conditions, causing reflections and distortions. In addition to this, surface and drilling noise further corrupt the travelling acoustic signal. Different signal processing and noise cancellation techniques are employed to combat these effects.

In general, a communication system requires a transmitter and a receiver. The transmitter converts an electrical signal into a form that is appropriate for the transmission channel. For wireless transmitters used in mobile and wi-fi, the transmitter converts electrical signals into electromagnetic waves. In acoustic telemetry, the electrical signal must be converted into physical acoustic waves using electromechanical transducers such as electromagnetic, magneto-strictive or piezoelectric actuators.

In U.S. Pat. No. 5,703,836, an acoustic transmitter is described where the electromechanical transducer is a piezoelectric ceramic stack. The stack is manufactured as halves and mounted on a main mandrel with an interference fit.

In U.S. Pat. No. 6,147,932, the piezoelectric ceramic stack is sandwiched between a mandrel shoulder and an anvil for another interference fit and the piezo stack did not need to be manufactured in halves. However, these acoustic telemetry systems have a major disadvantage.

In order to transmit over greater depths, repeaters must be used to ensure the integrity of the signal. Moreover, the drilling operation needs to be stopped during the receiving signals and transmitting signals with repeaters and therefore real-time data collection is not available. These repeaters are expensive to implement and are unattractive to potential adaptors. Repeaters are smaller versions of a full acoustic telemetry tool. They detect acoustic signals from a lower downhole location and retransmit the acoustic signal. In some instances, they can also be used to measure local conditions and append data to the outgoing acoustic signal. However, even one malfunction among the repeaters is enough to compromise a MWD transmission. Efforts have been made to efficiently use the piezoelectric transducer by optimizing the power going to the load but transmission over long distances still have limited performance.

US 2015/0377017 A1 discloses an apparatus for transmitting a signal through a downhole medium. The transmitter includes a voltage source, a load, a switching circuitry to apply positive and negative polarity voltage to the load, and a charge controller to control the driving voltage. The transmitter is able to apply a complex envelope to a passband signal. A DC-DC voltage converter is used to increase the voltage from a battery source and an H-bridge is used to drive the piezoelectric load modelled as a capacitor.

In US 2019/0017371 A1, a method and system for applying a voltage across a piezoelectric transducer load is taught. The temperature, or compressive strain is monitored to keep the magnitude of the driving signal below the negative polarity limit. Alternating positive and negative electrodes are inserted between the stack elements to provide a driving voltage equally across the whole piezo transducer.

U.S. Pat. No. 6,956,791 B2 introduces an apparatus for receiving downhole acoustic signals. The acoustic signals are processed and then remotely transmitted to a monitoring station. The apparatus wraps around a drill string component accessible from the surface. The housing is secured so that the drilling operation does not cause it to fall, thereby injuring personnel or damaging equipment.

In CA 2,374,733, an acoustic telemetry system where the receiver uses two sensors to reduce the corruption caused by drilling noise is disclosed. The primary propagation mode is axial while a second propagation mode is torsional and is used to create a third signal that has reduced corruption. The system also uses normal filters that further remove the effects on the signal due to channel distortion.

U.S. Pat. No. 8,634,273 describes a multi-frequency downhole acoustic transmission on one or more passbands, and different modulation techniques on each of the transmission frequencies. In addition, the channel properties of the acoustic channel are determined using a swept frequency signal. The carrier frequency is determined using the detected passbands.

SUMMARY

With advancements in downhole sensors, high transmission rate from downhole MWD or logging-while drilling (LWD) tools is critically important for accurate steering of the drill bit, and assessment of the downhole environment. The present invention provides a viable solution to MWD operations, firstly by eliminating the need for repeaters with advanced signal processing techniques, and secondly by having the electronics and more expensive components of the acoustic transducer tool to be inline, or retrievable, and be augmented with other types of MWD or LWD tools. The acoustic transducer apparatus consists of a piezoelectric ceramic stack, a switching circuitry to drive the piezo electric load, a power supply and corresponding step-up converter, and a microcontroller, to connect to the MWD sensors and instruments, and to provide the driving signal for the switching circuit.

In a broad aspect of the invention, an acoustic telemetry system for use with a drill string having MWD sensors and instruments, while conducting MWD operations can comprise an acoustic transmitter tool and a surface receiver. In an embodiment, the acoustic transmitter tool can have a mandrel with an annulus therethrough, for securing the tool to the drill string. The acoustic transmitter tool can further comprise an annular piezoelectric actuator secured to the mandrel for generating acoustic waves for travelling longitudinally along the drill string, a pressure barrel concentrically positioned within the annulus for housing electronics, and electro-mechanical connections for communicating between the piezoelectric actuator and the electronics.

In another broad aspect of the invention, a method for downhole telemetry and acoustic transmission of MWD data for drilling operations using a drill string assembly comprises the steps of:

(a) providing an acoustic telemetry tool having a piezoelectric transmitter and positioning the acoustic telemetry tool down a wellbore;

(b) providing a surface receiver;

(c) prior to commencing the drilling operations, determining an appropriate acoustic transmission frequency;

(d) after determining an appropriate acoustic transmission frequency, commencing the drilling operations while conducting acoustic transmission using the appropriate acoustic transmission frequency;

(e) repeating the step of determining an appropriate acoustic transmission frequency as the drill string is elongated during the drilling operations to determine a subsequent acoustic transmission frequency;

(f) continue drilling operations and acoustic transmission using the subsequent acoustic transmission frequency; and

(g) repeating steps (e) and (f) as necessary until the drilling operations are complete.

BRIEF DESCRIPTION OF THE DRAWINGS

Referring to the drawings, several aspects of the present invention are illustrated by way of example, and not by way of limitation, in detail in the figures, wherein:

FIG. 1 is a diagram of a typical directional drilling setup where an acoustic telemetry system is used. PA

FIG. 2 is a diagram of the acoustic transmitter tool that produces acoustic waves longitudinally in both directions. PA

FIG. 3 is a diagram of the operation for performing a sweep of the drill string assembly to determine the best carrier frequencies for transmission.

FIG. 4 is a diagram of the operation for simulating the drill string channel to determine the best carrier frequencies for transmission.

FIG. 5 illustrates the processes that the transmitter undergoes from the receiving the MWD data to applying a voltage across the piezoelectric stack.

FIG. 6 illustrates the processes at the receiver in order to recover the transmitted data.

FIG. 7 shows a functional block diagram of the digital signal processing steps performed by the acoustic telemetry system.

FIG. 8 illustrates the method superposing two-frequency transmission through the drill string.

FIG. 9 shows a sample waveform that contains two frequencies that coincide with the passbands of the drill string and its fast Fourier transform (FFT) plot.

FIG. 10 is a cross-sectional view of an embodiment of the invention illustrating an acoustic transmitter tool comprising a mandrel having an annulus therethrough.

FIG. 11 is a cross-sectional view of the embodiment of FIG. 10, illustrating a pressure barrel concentrically positioned within the annulus and connected to an annular piezoelectric transducer.

FIG. 12 is an isometric sectional view of the embodiment of FIGS. 10 and 11, illustrating the pressure barrel's connection to the mandrel.

FIG. 13 is an embodiment of the acoustic tool wherein the mandrel is further separated into two components.

FIG. 14 is a cross-sectional view of the embodiment shown in FIG. 13.

FIG. 15 is another embodiment of the acoustic transducer tool connected to an isolator to prevent signals from the bottom hole assembly (BHA) from interfering with the acoustic tool.

FIG. 16 illustrates the prior art of a simple convolutional encoder.

FIG. 17 is a diagram of the laboratory scale setup. The simulated drill string is made up of two steel sections and an attenuating middle section.

FIG. 18 shows the preload case of the piezoelectric actuator coupled to the load cell and the simulated drill string.

FIG. 19 plots the measured frequency response function (FRF) of the simulated drill string: a) linear plot; and b) magnitude plot in dB and the phase plot with phase wrapping.

FIG. 20 presents the attenuation of the accelerometer signal through the drill pipe. The attenuation values at 2387 Hz and 3742 Hz are highlighted.

FIG. 21 shows the measured and the estimated frequency response function used for simulation of the drill string dynamics.

FIG. 22 presents the extraction of the modal parameters from the real and imaginary frequency response functions (FRF).

FIG. 23 shows the experimental passband waveforms: a) transmitted driving signal; and b) received accelerometer signal.

FIG. 24 depicts the lock-in amplifier (LIA) demodulation in baseband waveforms: a) transmitted baseband waveform; and b) received baseband waveform. Phase rotation is experienced by the second waveform.

FIG. 25 compares the magnitudes of the baseband waveforms: a) transmitted magnitude; and b) received magnitude.

FIG. 26 shows a sample of the phase shift detection performed by the acoustic telemetry system: a) phase shift from the lock-in amplifier (LIA) output; and b) detected bits.

FIG. 27 shows the magnitude response of the simulated channel for different damping ratios; b) BER plotted with damping ratio and transmission speed for the first mode (2387 Hz); and c) BER plotted with damping ratio and transmission speed for the second mode (3742 Hz.

FIG. 28 plots the performance with and without the lock-in amplifier (LIA) demodulation.

FIG. 29 illustrates the experimental performance from convolutional coding and signal fusion.

FIG. 30 compares the performance of the system in terms of input voltage: a) comparison of each mode and the combined transmission (non-coded); b) coded vs. non-coded performance of the first mode; c) coded vs. non-coded performance of the second mode; and d) coded vs. non-coded performance of the MRC combined signal.

DETAILED DESCRIPTION

With reference to FIG. 1, a typical MWD drilling operation consists of the bottom hole assembly (BHA), the drill string assembly, and the derrick which supports the entire assembly. The BHA encompasses the drill bit components, the instruments and sensors, and the MWD transmitter of the present invention. The sensors and instruments measure the position and orientation of the drill bit, as well as other chosen formation properties, and send the data to the MWD transmitter. The MWD transmitter then processes and converts the data into a form appropriate for the medium of transmission.

For the purposes of this present invention, the downhole MWD acoustic transmitter will be referred as the acoustic transmitter tool. The purpose of the acoustic transmitter tool is to convert the sensor and instrument data into a driving signal for a piezoelectric transducer that is mounted annularly on the mandrel of the tool as shown in FIG. 2. The transducer creates acoustic waves that propagate longitudinally along both directions of the drill string axis. The acoustic waves can propagate and be received by a surface receiver system that can comprise of at least two accelerometers, or acoustic emission sensors, positioned longitudinally along the drill string, and measures the acoustic waves and sends the data to a computer for signal processing. In known embodiments, the surface receiver is connected to another downlink telemetry system (i.e. mud pulse (MP) or electromagnetic (EM) telemetry methods) that sends the carrier frequency information to the acoustic transmitter.

A. Operations

For the telemetry operations to run effectively, it is best practice to determine the best transmit frequency for use with a drill string that is being used. Determining a transmit frequency that is the most suitable can be achieved by sending a chirp, sine sweep, or impulse signal that encompasses all frequencies of interest. In an embodiment, a method for determining the transmit frequency is demonstrated in FIG. 3. As shown, an acoustic transmitter tool or transmitter is prompted from the surface by an operator of the MWD operations to transmit a control signal comprising a broad range of frequency signals that are of interest. Sending the control signal to the transmitter can be accomplished using known methodologies. Some examples include using an acoustic surface transceiver, a mud-pulse transmitter, or an electromagnetic transmitter to send the control signal.

The broadband control signal from the downhole transmitter can be received by a surface receiver and the frequency content can be determined by using a fast Fourier transformation (FFT). From the FFT plot, in embodiments, frequency ranges where the response is relatively flat over a required bandwidth can be examined. Centers of suitable frequency ranges can then be chosen to be the transmit frequencies.

Alternatively, in other embodiments, the frequency response of the drill string assembly can be simulated using finite element (FE) methods to produce a representation of the drill string channel based on boundary conditions, length, geometry, and material properties of the drill string assembly. The simulation can be performed in real-time, or it can be preprogrammed on the transmitter.

As shown in FIG. 4, carrier frequency information can be sent down to the acoustic transmitter tool and acoustic transmission continued. In the case that the simulation results are preprogrammed on the transmitter, sending the carrier frequency data would not be necessary.

During drilling operations, and as additional drill pipes are added to increase the length of the drill string, the channel properties of the drill string will change accordingly. Thus, as new drill pipes are added, the acoustic telemetry system can be prompted again to send a chirp signal to determine more suitable carrier frequencies and the most appropriate carrier frequencies can be chosen.

With reference to the flowchart in FIG. 5, and in an embodiment, the acoustic transmitter can be driven by an H-bridge switching circuit 21 that can change the polarity of a voltage source. As shown, a microcontroller 16 can process data received from MWD sensors and instruments 12 supported on the drill string, and the microcontroller can determine whether the acoustic transmitter tool should transmit or not. The microcontroller 16 can poll the MWD sensors and instruments using standard communication protocol (i.e., CANBUS) and can log the received data in its memory 15. The microcontroller 16 can also simultaneously log the data from a flow switch of the bottomhole assembly (BHA). Afterwards, the microcontroller 16 can determine if the flow switch is on and can encode the received data into an acoustic signal. In an embodiment, the received data is protected from errors by a channel coding scheme 28. Alternatively, flow switch can be replaced by the available surface transmitter (from the previous carrier frequency selection method) which can be used to prompt the acoustic tool to begin transmission.

Once transmission from the transmitter tool is performed, accelerometers, or acoustic emission sensors, can be configured on the surface drill string assembly to detect the longitudinal acoustic waves emanating from the acoustic transmitter tool. In embodiments, the detected acoustic waves can be converted into electrical signals which are then processed. As the acoustic waves can contain multiple frequencies, a frequency signal diversity combining method can be implemented. In embodiments, at each frequency wherein a signal was transmitted, a lock-in amplifier (LIA) can be used to perform demodulation. The signal to noise ratio (SNR) can be first measured from each frequency and a threshold set for combining of the accelerometer signals. The signal from each frequency is then compared to one another, and the highest quality signals are chosen to be the primary signal for the receiver. In addition to frequency diversity, having at least two or multiple accelerometers at known distances on the surface also creates spatial diversity. Signal diversity is advantageous as it is unlikely that reflections from joints between each drill pipe along the drill string, and the hydrocarbon formation, will destructively interfere with all the accelerometers mounted. As such, in embodiments, the accelerometer with the highest SNR after demodulation will also be used as the primary signal for detection. Additionally, the accelerometers can be combined in accordance with maximal ratio combining (MRC) if the SNRs for each branch is known.

Attenuation of the longitudinal acoustic waves travelling through the drill string is dependent on the frequency of transmission. As a result, there will be performance differences at each of the different frequency channels. To maximize the efficiency of the acoustic telemetry system, in embodiments, the different channels can be configured to utilize different encoding and modulation techniques.

As an example, a carrier frequency with a higher attenuation is more prone to the effects of noise, so a simpler modulation scheme like binary phase shift keying (BPSK) can be used and a higher performance channel coding scheme can be implemented. On the other hand, if a carrier frequency has relatively good resistance to noise, a higher order modulation scheme like quadrature phase shift keying (QPSK) or quadrature amplitude modulation (QAM) can be used and a simpler error correction method can be adopted.

Another advantage of having at least two or multiple accelerometers at known distances is the ability to detect the wave speed of the acoustic signal. Impedance changes and different geometries of the drill string channel, it is common to see reflections or echoes. Inter-symbol interference (ISI) can be introduced to the acoustic signal detected by the accelerometers. By utilizing the wave speed, a signal corrupted by echoes and reflections can be improved by cancelling the noise. In embodiments, this can be done by subtracting a time shifted weaker version of the original signal to itself which can be determined with the help of the wave speed.

With reference to FIG. 6, processes that the surface receiver undertakes to recover the original data from a single accelerometer before signal combination is shown. In an embodiment, the accelerometer signal can be conditioned and amplified for analysis and can be sent directly to a digital LIA. The resulting signal can then be split and mixed with a local oscillator, such as an orthogonal oscillator, and a phase-shifted version for dual phase detection. This process can detect a magnitude and a phase of the signal envelope relative to the reference signal. Afterwards, lowpass filters can be applied to the two outputs to eliminate the high frequency components of the demodulated signal. The processes from the transmitter can be then reversed by using a raised cosine receive filter 26, a demodulator 27, and a channel decoder 28. The receiver then checks for correctable errors and displays the data to a user. If uncorrectable errors are detected, the received data is determined to be corrupt and a replacement or new transmission is requested by the surface receiver.

An evaluation procedure is also performed by using a message sequence that is known at both the transmitter and the receiver. The transmitter is prompted using any of the methods mentioned before and the receiver assesses the received bits and compares it to the reference message. The bit error rates are extracted and is shown to the user. Additionally, different transmission speeds of the same reference message sequence can be used to determine the maximum transmission rate with a reasonable bit error rate (BER).

B. Signal Processing

Signal processing can be performed at both the transmitter tool and the surface receiver in order to ensure that the received data is the most accurate estimation of the transmitted data. In addition, because drill strings exhibit characteristics of a passband channel, the received data must be upconverted to a frequency within the passband to achieve the least amount of attenuation. To achieve this, several techniques can be applied.

In an embodiment, and as shown in FIG. 7, the three main components of the telemetry system are the transmitter, the channel, and the receiver. The aim is to recover the data transmitted at the surface receiver as accurately as possible after the signal was distorted by the channel, either through dispersion, noise, reflections, echoes, and other disturbances.

Transmitter

At the surface transmitter, multiple frequencies can be utilized based on the drill string channel properties. First, a test signal such as sine sweep, chirp, or impulse signal is transmitted. The receiver then analyzes the frequency response function (FRF) of the drill string channel at the receiver. Alternatively, finite element (FE) modelling can be performed by considering the boundary conditions, length, and material properties of the drill string. Then, based on the resulting FRF, the receiver sends a signal back to the transmitter containing the selected transmit frequencies at or near natural frequencies of overall dynamics. At each of these frequencies, depending on the fading characteristics and susceptibility to noise, different encoding and modulation techniques can be applied. This gives the system the flexibility to adapt to the changing drill string channel as the wellbore is extended. This system can also be exploited to provide redundancy by transmitting the same message data simultaneously at multiple frequencies. In general, each encoded baseband signal packets are assigned to different carrier frequencies and are superposed to create a multi-frequency time signal. For example, a modulation scheme based on binary phase shift keying (BPSK) can be used for high disturbance frequency channels, but a more complex scheme like quadrature phase shift keying (QPSK), or quadrature amplitude modulation (QAM) are used for the frequency channels that have greater signal-to-noise ratios (SNR).

Looking again at FIG. 7, the signal-processing method at the transmitter incorporates channel coding, modulation, pulse-shaping, and up-conversion. These mechanisms ensure that the signal would be transmitted through the channel as efficiently as possible, with minimal losses. In embodiments, message data from the transmitter tool can be encoded into a modulated drive signal using a forward error correction scheme (FEC) called convolutional encoding. This FEC scheme generates a series of bits that contain only parity bits, as opposed to block coding. Implementation of the coding scheme can use memory registers that calculate the modulo 2 sums (XOR logic gate) of a current sequence.

An example is shown in FIG. 16 where memory registers act as a sliding window that registers a current sequence. The output of the sliding window are parity bits that contain information regarding the sequence of binary data. Encoding is necessary because it is inevitable for the channel and the electronics noise inherent to the system to distort and disperse the transmitted data. These disturbances can cause bits to be flipped and become errors at the surface receiver. In an alternate embodiment, an improved coding scheme called Turbo codes can be employed. Turbo codes use two convolutional encoders where one input is interleaved to protect from burst errors. The output of the Turbo code comprises of the output of the two convolutional encoders and the message data itself. Note that the channel coding scheme of the acoustic telemetry system are not limited to convolutional encoders because different forward error correcting codes can be utilized on different frequency channels.

Multiple methods exist for wireless transmission of binary data through a physical channel. As wireless communication utilizes sinusoidal waves transmitted in different media, one commonly used method is to manipulate the parameters of these waves. The sinusoidal function can be written in the form

s(t)=A cos(2πf_ct+θ)  (1)

where t is time, A is amplitude, f_c is frequency, and θ is phase. The sinusoidal component of the function is sometimes referred to as the carrier frequency in passband modulation. In vibrations, equation (1) is used as the forcing function that determines the response of the mechanical system. Modulation onto the sinusoidal carrier frequency is required to adapt the signal to the passband frequencies of the channel. As an example, binary phase shift keying (BPSK) uses the phase θ to convey digital information onto the carrier. In embodiments, the present invention also allows higher rate modulation schemes such as quadrature phase shift keying (QPSK) and quadrature amplitude modulation (QAM). The waveform for a BPSK transmission that alternates between a 0° and 180° phase shift can be described by

s₁(t)=A cos(2πf_ct)  (2)

and

s₂(t)=A cos(2πf_ct+π) or −A cos(2πf_ct)  (3)

where s₁(t) and s₂(t) represents a bit “1” or a bit “0” respectively. Applicant notes that the phase shift can be represented either as 180° (π) phase shift or a negative amplitude.

For real physical systems, any input can cause transient responses that decay over time. Transmission of binary data through a wireless physical channel resemble multiple sequential inputs. Accordingly, if the transient responses overlap with the next input, then inter-symbol interference (ISI) will occur. To minimize this, a raise-cosine filter that satisfies the Nyquist criterion for ISI which can be described as

$\begin{array}{cc}x\left(t\right)=\frac{\sin \left(\pi t/T_s\right)}{\left(\pi t/T_s\right)}\left[\frac{\cos \left(\mathrm{\pi \alpha }t/T_s\right)}{1-4{\alpha }^2{t}^2/T_s^2}\right]\mathrm{or}x\left(t\right)=\sin c\left(t/T_s\right)\left[\frac{\cos \left(\mathrm{\pi \alpha }t/T_s\right)}{1-4{\alpha }^2{t}^2/T_s^2}\right]& \left(4\right)\end{array}$

where T_s is the symbol period and a is the roll-off factor. The roll-off factor determines the bandwidth of the filter through the expression

$\begin{array}{cc}B=\frac{1}{2T_s}\left(1+\alpha \right)& \left(5\right)\end{array}$

In addition to minimizing ISI, the raise-cosine filter can shape the BPSK signal to fit into the selected bandwidth, thereby allowing for more effective transmission through the channel. The choice for a roll-off factor is an optimization problem between excess bandwidth and ease of implementing timing recovery. Depending on the filter span, the filter can introduce a known delay in the time domain which can be compensated for at the receiver. The resulting waveform after filtering is referred to as a baseband BPSK signal because it has not yet been mixed with the carrier.

In embodiments where a coherent detection is not feasible, another modulation scheme called differential phase-shift keying (DPSK) can be implemented. DPSK encodes data on the phase shift itself rather than the actual phases so that the surface receiver does not need to perfectly track the carrier frequency of the signal. It simply detects when a phase-shift occurs to detect the coded message. The data bits are first differentially encoded where the previous symbol is subtracted from the current symbol. Without differential encoding, an error located in the middle of the packet may corrupt subsequent bits. An example of needing non-coherent demodulation is when the phase rotation due to the frequency offset precludes traditional BPSK demodulation.

It is known that a frequency response contains distinct passbands and the passbands are believed to have been caused by the impedance mismatches at the tool joints inducing echoes and reflections. Within these, passbands are also structures resembling a comb filter that are related to the number of drill pipes in assembly.

In embodiments, a baseband signal can be upconverted in order to take advantage of these passbands. Up-conversion can be performed by “mixing” or multiplying the baseband signal to a carrier frequency. The baseband signal acts as the envelope for the carrier frequency. Furthermore, a drill string will have multiple passbands from which the carrier frequency for the acoustic wave can be selected. By transmitting multiple frequencies from these passbands at the same time, the transmission rate can be increased, or redundant data can be used to reduce uncertainty due to error and noise.

With reference to FIG. 8, a multi-frequency transmission can be achieved where two carrier frequencies contain the data streams x₁(t) and x₂(t). The resulting passband signals are superposed and transmitted. This results in a signal with two main frequency components corresponding to the peaks of the drill string channel which can is shown in FIGS. 9a and 9b.

In greater detail, FIG. 9a illustrates a time domain waveform contains two different oscillations that correspond to the carriers. This can be more clearly seen in FIG. 9b where the frequency content of the signal is determined by applying an FFT. Applicant notes that the tool is not limited to two frequencies, but can transmit data with multiple frequencies as the drill string channel allows. The resulting electrical signal is then used to drive the piezoelectric transducer to create acoustic waves through the drill string.

Receiver

The surface receiver of the system can comprise at least two or multiple accelerometers positioned longitudinally along the drill string that act as antennas for the transmitted acoustic signal. At least two accelerometers are arrayed along an axis parallel to the rotation axis of the drill string such that it can measure the longitudinally propagating acoustic waves. The accelerometers are each connected to a high dynamic reserve lock-in amplifier (LIA) that acts as a homodyne detector. The demodulated accelerometer signals are fused using maximal ratio combining (MRC), equal gain combining (EGC) or selection combining (SC) to improve the signal-to-noise ratio (SNR) to improve the accuracy. Finally, the signal is decoded for error correction to recover back the original message data.

As the acoustic waved travel longitudinally along the drill string, it is known that the acoustic waves are affected by noise and disturbances, either from the channel itself or from the various the electronic components positioned along the drill string. Accordingly, and in embodiments, multiple accelerometers or acoustic emission sensors can be mounted in a configuration that allows for the detection of longitudinal acoustic waves at the surface receiver.

In forced vibrations, it is known that an output displacement should have the same frequency as the input force. However, because of impairments caused by the transducers and the dynamics of the drill string, a received waveform of the longitudinal acoustic wave measured at the surface receiver will be considerably different than a transmitted waveform. Furthermore, dissipation of acoustic energy as it travels through the drill string can also lead to attenuation. By itself, attenuation is not a major problem as commercial accelerometers can measure very small signals. However, when noise is added, attenuation increases the probability of corrupting any transmitted data.

Thus, in embodiments, a lock-in amplifier (LIA) at the surface receiver for both carrier frequencies can be employed. Lock-in amplifiers can be used in situations where the desired signal is completely buried in noise, such as during atomic force microscopy (AFM) measurements. In combination with a reference oscillator that can be tuned to the frequency of the desired signal can be used to track its amplitude and phase. The LIA can reject all other frequencies that are not the same as the frequency of the reference oscillator.

In embodiments, the LIA can be tuned to the exact frequency of the received signal such that the reference oscillator of the LIA is precisely the same as the received signal, and thereby allow an ideal operation to be performed on the received acoustic signal.

In embodiments, the reference signals of the LIA can be expressed as

v_X(t)=cos(2πf_rt)  (6)

and

v_Y(t)=sin(2πf_rt)  (7)

where f_r is the frequency of the reference signal. In embodiments, assuming a single frequency, the products of mixing the BPSK signal from equation (1) with both the LIA references can be expressed as

s(t)v_X(t)=A cos(2πf_ct+θ(t)cos(2πf_rt)  (8)

or

s(t)v_X(t)=½A[cos(2π(f_c−f_r)t+θ(t))+cos(2π(f_c+f_r)t+θ(t))]

and

s(t)v_Y(t)=A cos(2πf_ct+θ(t)sin(2πf_rt)  (9)

or

s(t)v_Y(t)=½A[sin(2π(f_c+f_r)t+θ(t))−sin(2π(f_c−f_r)t+θ(t))]

which were derived from trigonometric identities. If f_c and f_r are the same, then

s(t)v_X(t)=½A(cosθ(t))+cos(4γf_ct))  (10)

and

s(t)v_Y(t)=½A(sin(θ(t))+sin(4πf_ct))  (11)

The two waveforms can be passed through the lowpass filter to remove the double-frequency components so that only the phase-dependent components of the signal remain. With reference to the equations above, these are referred to as the X (in-phase) and Y (quadrature) outputs of the LIA.

For BPSK, the phase θ(t) alternates between 0° and 180° which implies that the only the X output alternates between a positive and negative voltage while the Y output remains at zero.

Detection of the transmitted signal can be performed through various methods. In an embodiment, a phase-locked loop (PLL) can be used to control the orthogonal oscillator of the LIA. The PLL can first be used to track the preamble of the transmitted signal and once the preamble has been tracked, the oscillator can be fixed to track the rest of the packet.

However, in BPSK transmission, phase shifts that are inherent in the scheme can disrupt the tracking ability of the PLL. To resolve this, in embodiments, the system can use one of two ways and still maintain coherent detection. In an embodiment, a Costas loop can be used, which ignores the symbols modulated onto the carrier frequency and just tracks the carrier itself. In another embodiment, an accelerometer signal can be squared to remove the phase transitions. Squaring the accelerometer signal can create a frequency component at double the carrier frequency, which the PLL can use to track the envelope.

In embodiments where if coherent detection cannot be achieved, non-coherent methods such as differential binary phase shift keying (DBPSK) can be applied. DBPSK differs from BPSK only from the fact that the binary data needs to be differentially encoded at the transmitter and then decoded at the receiver. Instead of the actual state, or phase, the data is encoded onto the phase shifts of the signal. Even with phase rotation due to the frequency offset, these phase shifts can be easily detected using a phase comparator.

In embodiments, after the received acoustic transmitter signals have been demodulated into the baseband waveform, a timing synchronizer can be used.

In embodiments, the synchronizer can employ a decision-directed approach where a cluster variance can be used as an error signal. A sampling time is then corrected by using interpolation. In another embodiment, an early-late gate synchronizer that produces an early and delayed sample δ at the sampling instant can be employed. When there is a timing offset, the amplitudes of the early and late samples can be different and the difference can be used as the error signal that controls a voltage-controlled clock. In embodiments, the synchronization steps produce the estimated amplitudes of the pulses that contain the digital data of the transmitted acoustic waves.

In embodiments, to combine the multiple signals, a MRC scheme was implemented on the calculated phase for each of the frequencies. MRC uses the SNR values as the weighting factors for each diversity branch, which are the optimal factors for an Additive White Gaussian Channel (AWGN).

For example, the combined received signal for the two frequency diversity branches of the setup can be calculated by:

$\begin{array}{cc}r\left(z\right)=\frac{SN{R}_1*r_1\left(t\right)+SN{R}_2*r_2\left(t\right)}{SN{R}_1+SN{R}_2}& \left(12\right)\end{array}$

where r₁(t) and r₂(t) are received signal values for the first and second modes. SNR₁ and SNR₂ are the calculated SNR values for the first and second modes, respectively. The weights are normalized by dividing the signal with the sum of the SNR values.

After demodulation, the bits can be detected from the baseband signal through sampling and synchronization, or through a matched raised cosine receive filter.

Further, and in embodiments, channel decoding can be performed to recover the estimated received data. In further embodiments, the Viterbi algorithm is a time invariant trellis-based decoder can be used to calculate minimum Hamming distances for all the possible sequences that a sliding window of the convolutional encoding scheme could have gone through. Any errors that may have occurred would increase the Hamming distances and would therefore be less likely to be chosen as the estimated message bits. After the recovery of the message bits, the data is displayed to the user in their chosen format.

C. Mechanical Design

Once again, FIG. 1 depicts a typical directional drilling setup. The drill string assembly is supported by the derrick and is rotated by a top-drive motor. The drill string itself is a combination of drill pipes and collars connected at specially threaded tool joints. At the end of the drill string assembly is the bottom hole assembly (BHA) where the instruments, sensors, transmitters, motors, and the drill bit can be located. As the drill bit bores deeper underground, additional drill pipes are added to the drill string assembly, thereby elongating the drill string channel. In an embodiment, an acoustic telemetry or transmitter tool can be positioned at a BHA location closest to the surface, where it can transmit acoustic signals through the drill string walls without passing through the BHA components. At a downhole end of the acoustic transmitter tool an acoustic isolator is inserted to prevent acoustic interference from the other instruments of the BHA.

The main body of the acoustic transmitter tool can be a hollow steel mandrel radially symmetrical about a central axis. Likewise, the internal passage or annulus through the mandrel is symmetrical on the same axis. An outer surface of the mandrel can be manufactured so that the piezoelectric ceramic stack and thermal compensators are appropriately accommodated. In an embodiment, a second hollow steel component or an inner mandrel can be threaded where it has contact with acoustic transducer housing, providing a preload required by the piezoelectric stack. The stack is electrically coupled to a pressure barrel centralized inside the mandrel that houses the electronics of the acoustic telemetry tool. A rubberized isolator which has low stiffness and high damping (absorbs the wave energy) can be used to minimize the transmission of acoustic waves through it. In another embodiment, the mandrel can be further separated into two components.

With reference to FIG. 10, the acoustic transmitter tool 1 is shown. The main supporting structure of the transmitter tool is a mandrel 4 having an annulus therethrough, on which an acoustic transducer 2 is mounted onto. The acoustic transducer can be a piezoelectric or a magneto-strictive element that converts an electrical signal into a mechanical acoustic wave that will then propagate through the drill string assembly and to the surface.

In an embodiment, the acoustic transducer is a PZT ceramic stack that expands or contracts depending on the voltage applied between each element of the stack. In addition, and in embodiments, the stack 2 can be preloaded between two brass cylinders to compensate for thermal expansion during operation.

During assembly, the mandrel 4 can be stretched and the transducer assembly 2 is mounted through an interference fit. Using an anvil 3, the stack is preloaded and restrained. As shown, the acoustic tool can be threadably connected to the drill string assembly using typical box and pin threaded connections. During operation, the tool allows the passage of drilling fluid through its annulus by entering through the inlet 6.

Known acoustic telemetry tools mount their electronics on the mandrel. However, when drilling operations fail, it is difficult to retrieve the often expensive electronics as they are directly mounted on the drill strings walls.

In an embodiment and as shown in FIG. 11, the electronics can be housed inside a pressure barrel 7. In a preferred embodiment, the pressure barrel 7 can be made from beryllium copper that can withstand high temperature and high pressure commonly encountered kilometers underground. The pressure barrel houses the electronics mechanically and electrically coupled to the walls of acoustic tool. A conduit or passage 9 operatively connects driving circuitry of the acoustic transducer 2 to the electronics inside the pressure barrel 7. Furthermore, as drilling fluid should be able to freely pass through the annulus, the pressure barrel 7 can be concentrically positioned within the annulus of the mandrel 4 by way of centralizers 10. As shown, and in embodiments, the electronic coupling can be mounted on the mandrel by bolts 8.

In an embodiment, the pressure barrel can house a power source such as a battery, a voltage converter/amplifier, a driving circuit, a microcontroller unit, and a port for receiving instrument and sensor data from the LWD modules. The battery can be responsible for providing power for all the components of the acoustic transmitter tool. The voltage of the battery for the preferred embodiment is generally in the 28-32 V range. A DC-DC converter or a bipolar amplifier is used to provide power to a driving circuit, such as an H-bridge circuit, to actuate the piezoelectric transducer. To provide power to the other components, a separate battery pack may be used, or alternatively, a step-down converter. The driving circuit is controlled by a microcontroller unit (MCU) where the process of converting the message data into an electrical driving waveform is performed. The pressure barrel can be connected to the annularly mounted piezoelectric transducer through a channel on the mandrel. The connection is held secure using specialized bolts that can withstand the downhole environment.

In embodiments, the piezoelectric ceramic stack 2 can be made of lead zirconate titanate (PZT) that converts electrical energy to mechanical energy thereby creating an acoustic wave along the walls of the drill string that can be detected by an accelerometer on the surface. In an embodiment, the piezoelectric stack can be driven by an H-bridge switching circuit connected to a DC-DC converter that boots the preprocessed electrical signal into a higher voltage. The digital signal processing (DSP) component of the microcontroller performs channel coding, modulation, pulse-shaping, and up-conversion on binary data logged by the sensors and instruments attached to the module. In an embodiment, a power supply is also seated on the pressure barrel that houses the transducer electronics. Depending on the electrical coupling chosen, an embodiment could also have a power supply annularly seated on a mandrel. In another embodiment, and induction type coupling can be created between the primary and secondary coils of a transformers to drive piezoelectric stack that would not require a hard-wired electrical connection.

At the end of the pressure barrel 7, an intermodular connector is employed to transfer information from the sensors to the transmitter. FIG. 12 shows an isometric section view of one embodiment of the acoustic telemetry tool where the components can be more clearly seen.

With reference to FIG. 13, in an embodiment, the mandrel 4 can further comprise a bottom sub 11, a top sub 12 and an inner mandrel 13. The bottom sub 11, can be adapted to fit the inner mandrel 13 at about a downhole end thereof. Once the inner mandrel 13 is fit to the bottom sub 11, o-rings can be used to secure the components to each other and prevent the mud or fluid from escaping an annulus therethrough. Once secured, the acoustic transducer or piezoelectric stack 2 can be slid onto the inner mandrel 13 to mount against shoulders of the bottom sub 11. To secure the piezoelectric stack 2 and prevent it from sliding off the inner mandrel 13, an anvil 15 can be slid onto the inner mandrel 13 to sandwich the piezoelectric stack against the shoulder of the bottom sub 11. Anti-torsional alignment blocks 14 can be aligned with grove on an inside surface of the anvil 15 to ensure that torsional forces are not applied to the fragile PZT layers of the stack. In embodiments, and as shown, the alignment blocks 14 can be screwed onto the inner mandrel 13. However, the anti-torsional alignment blocks 14 can be attached using any method. This particular embodiment assists in preventing torsional forces form being applied to the piezoelectric stack.

As shown, the top sub 12 can have two threaded tapered surfaces: one for the bottom sub 11 and one for the anvil 15. Accordingly, by using applying a known amount of torque, a preload can be applied on the piezoelectric stack 2.

In an embodiment, an isolator can be positioned at a downhole end of the acoustic transmitter tool to limit, minimize or otherwise prevent the effects of acoustic interference, both from the BHA noise and other transmitter reflections.

Referring back to FIG. 11, the acoustic isolator would be connected between the tool itself and the rest of the BHA. In this embodiment, the piezoelectric transmitter can achieve maximum transmission efficiency as the acoustic signal will on only propagate on one direction. In addition, any noise coming from the BHA operations will be reflected away from the transmitter, thereby further improving the performance of the acoustic telemetry tool.

Alternatively, a high strength, high damping material (i.e. rubber) can be introduced between the acoustic tool's tool joints and the BHA drill string assembly to attenuate acoustic waves passing through it from both directions as shown in FIG. 15.

D. Laboratory Experiments

Developing and designing an appropriate experimental setup is necessary for any successful research. A drill string assembly, with the drill collars and the BHA, is often a few kilometers long. In a laboratory setting, it is difficult to implement such a length. Instead, the drill pipes and tool joints were scaled down to a more manageable size. Threaded steel pipes were connected by a coupling nut to simulate a drill string assembly. To account for the different properties of the drill string (i.e. length), the setup allows replacement of the middle section with different materials. For example, attenuation can be achieved by inserting a high damping material such as a plastic or rubber pipe between the two steel pipes. The simulated drill string itself is suspended on elastic cords that resemble a free-free boundary condition. The laboratory experiments provide a scaled down prototype of the present invention.

i. Overall Design

As previously discussed, in acoustic telemetry, data from the sensors and instruments on the BHA is converted to an electrical signal. A high frequency electromechanical transducer can be used to convert the electrical signal to longitudinal acoustic waves. In embodiments and as discussed above, transducers can be made of piezoelectric or magneto-strictive materials.

In the experimental setup, a piezoelectric actuator made from PZT was used. A voltage difference applied across a PZT element can cause it to expand accordingly. A diagram of the laboratory setup can be seen in FIG. 17. The total length of the drill string is approximately 152.4 cm.

A preload was applied to the PZT actuator, before operation, by using a preload bolt. The preload case designed for the actuator is shown in FIG. 18. The maximum allowable preload for dynamic applications of approximately 15 MPa was applied using the preload bolt. Using epoxy, two load plates were attached to both ends of the actuator. On one side, a bolt with a flat face was used to apply a preload, and the other was pressed to a load cell. The load cell was then used to measure the preload applied. The other side of the load cell was coupled with the beginning of the simulated drill string. The voltage difference (drive signal) was applied to the piezoelectric actuator (stimulating the piezoelectric transducer) using the two lead wires coming out of the preload case.

Most of the digital signal processing was performed on a computer. A data acquisition (DAQ) device was used to convert digital signals to analog and vice versa. An input/output (I/O) DAQ was connected to the computer to allow it to send/receive signals. For the transmitter, the output of the DAQ was connected to a bipolar amplifier that then drove the piezo actuator. The driving signal was given a DC offset as the piezo actuator can only expand in one direction over the range of 0-1000 V.

At the receiver, two accelerometers measure the acoustic signal. The accelerometers were fed into a signal conditioner for power and signal amplification. Afterwards, a LIA extracted and demodulated the acceleration signal and its output was connected back to the Input/output (I/O) DAQ system. FIG. 17 describes the connections between the components of the setup.

A program was developed to connect to the I/O DAQ using the provided application programming interface and drivers. Dual symmetrical power supplies were used to provide power needed by the bipolar amplifier. The materials used for the simulated drill string in the laboratory setup is tabulated in.

TABLE 1
Simulated drill string components
Simulated
Drill String Material
End Sections 60.96 cm AISI 1026 Carbon Steel
             Tube 1.27 cm × 0.3048 cm (×2)
Attenuating  30.48 cm Neoprene rubber tubing
Material
Coupling     M14 × 2 Steel coupling nuts

ii. FRF Analysis

The simulated drill string was composed of three threaded pipes. One pipe was coupled to the piezo actuator and another one was connected to an accelerometer. In the setup, the middle pipe was selected to be made from a neoprene rubber tube. To determine the properties of the channel, a frequency sweep was performed using the LIA. This provided a FRF plot that showed the response of the system through a range of frequencies. A sweeping signal ranging from 100 Hz to 10 kHz was chosen. The LIA was also found to be phase sensitive, which allowed for the extraction of the phase response. To measure the displacement response, an accelerometer was placed at the end of the simulated drill string. The load cell was placed between the piezo actuator and the pipe measured the applied force. The FRF obtained is shown in FIG. 19.

The attenuation of the accelerometer signal was calculated by placing two accelerometers at the beginning and at the end of the drill pipe. A first accelerometer was placed near the piezoelectric actuator outside of the preload case nearest to the source of the acoustic wave. The second accelerometer was kept at the end of the simulated drill string. The frequency response of the output was divided by the frequency response of the input to obtain the attenuation over the desired frequency range. The measurements and attenuation plots are shown FIG. 20.

In the FIG. 20, the attenuation for first and second passbands were found to be −41 dB and −26 dB, respectively. Based on prior art, the attenuation through a typical drill string assembly is known to be about 21 dB/km. Accordingly, the estimated equivalent length of the simulated drill string was calculated to be approximately 1.95 km for the first frequency, and 1.24 km for the second frequency with a maximum 40 V input.

To obtain a model of the channel for simulation, a system identification application was performed on the FRF plot. The application produced a set of coefficients for a discrete transfer function that imitates the properties of the channel. The compared values of the estimated channel to the actual channel measured, using the frequency sweep of the LIA is shown in FIG. 21. The system identification application worked by first determining a continuous transfer function and then converting it into a discrete transfer function in the form

$\begin{array}{cc}H\left(z\right)=\frac{b_0+b_{1}z+\cdots b_N{z}^N}{a_0+a_{1}z+\cdots a_{N-1}{z}^{N-1}+a_N{z}^N}& \left(13\right)\end{array}$

where b_n and a_n, n=0, 1, 2, . . . N are the coefficients of the transfer function. The coefficients found by the application are listed in Table 2. These coefficients allow us to model the drill string channel and investigate the effects of different encoding and modulation techniques through simulation.

TABLE 2
Estimated discrete transfer function coefficients
	n = 6	n = 5	n = 4	n = 3	n = 2	n = 1	n = 0
Numerator	—	8.38e−16	4.23e−14	2.07e−13	2.06 −	4.17e−14	8.20e−16
(bn)					13
Denominator	1.00	−3.59	6.67	−7.97	6.61	−3.52	0.97
(an)

Using the estimated plots, the modal parameters of the simulated drill string can be extracted. This is advantageous in investigating how the different parameters of the channel can affect the transmission of acoustic waves. The magnitude and phase plots from FIG. 21 can be converted into a real and complex frequency response function and shown in FIG. 22.

The damping ratio of the system can then be extracted using the equation, assuming the modes are far apart:

$\begin{array}{cc}\zeta \cong \frac{{\omega }_{m2}-{\omega }_{m1}}{2{\omega }_n}& \left(14\right)\end{array}$

where ω_n is the natural frequency in radians per second. The local minima and maxima are represented by ω_m1 and ω_m2 and m=1, 2 corresponds to the first and second modes, respectively. The stiffness can then be obtained using the equation

$\begin{array}{cc}{R}_{imag}=\frac{1}{2k\zeta }& \left(15\right)\end{array}$

where R_imag is the magnitude of the imaginary component of the frequency response. The extracted modal parameters are tabulated in Table 3.

TABLE 3
Modal parameters from the estimated frequency response function
	Modal Parameter	First Mode	Second Mode
	Natural Frequency (ωn)	14,567 rad/s	23,721 rad/s
	Damping Ratio (ζ)	0.014	0.013
	Stiffness (k)	1.08e11 N/m	4.14e11 N/m

The high stiffness values correspond to the attenuation experienced by the mechanical system. In the next sections, we use the above information to simulate the mechanical properties of the acoustic channel and investigate its effects on the transmission performance.

iii. Performance Results

A sample experimental transmission was performed using the acoustic telemetry setup for a single frequency and is shown in FIG. 23. The maximum allowable voltage to drive the piezoelectric actuator was set to 40 V peak-to-peak to avoid thermal issues during operation. The transmitted and received waveform for lock-in amplifier demodulation is depicted in FIGS. 24a and 24b. Before up-conversion to the carrier frequency in FIG. 24(a), the Y component is zero. However, in FIG. 24(b), the LIA outputs at the receiver clearly displayed both X and Y components due to the frequency offset. However, comparing the magnitudes in FIG. 25 show that the waveforms of the transmitted and received signals are similar. This is expected because the phase of the signal should not affect the overall magnitude.

The architecture of the LIA did not allow for the implementation of the Costas Loop for BPSK demodulation. As a result, a non-coherent DBPSK modulation scheme was performed. For this scheme, the phase shifts caused by the BPSK transitions is detected instead. Even with phase rotation caused by the frequency offset, there was a significant phase difference whenever the shift occurred, which can be detected by a phase comparator.

An example is shown in FIG. 26 where the phase transitions are indicated by the dotted vertical line. The locations where transitions occurred were recorded and used to exact the transmitted bits. A raised cosine receive filter was found not to be necessary because the phase shifts was more easily detected directly from the LIA outputs. Finally, to extract the original message data, the differential encoding was removed by subtracting the current bit from the previous bit or by using an XOR logical operation.

To investigate the effects of the drill string properties on the performance of acoustic telemetry, a simulated drill string channel with natural frequencies at 2387 Hz and 3742 Hz was constructed using the modal parameters from Table 3. The damping ratio was modified to increase or decrease the bandwidth of the simulated channel, while the transmission speed increased the bandwidth of the transmitted signal. The stiffness values remained unchanged for both modes. Noise was not added to ensure that the effects of only the damping ratio and transmission speed were observed. The BER was calculated for different combinations of the damping ratio and transmission speed, and the simulation results are presented in FIG. 27.

From FIG. 27, decreasing the damping ratio and increasing the transmission speed increased the BER. This implied that even without the effects of attenuation and noise, the system will still experience errors.

The bandwidth of the channel can be imagined as a tunnel through which the data needs to pass through. Data with a higher transmission rate contain higher frequency content and thus have higher bandwidth. If the bandwidth of the transmitted data is greater than the bandwidth allowed by the channel, the signal will become distorted and error-free recovery at the receiver becomes impossible, as was previously demonstrated. A second mode showed better performance because it had a greater bandwidth when compared to the first mode. For this reason, a transmission speed of 64 bits per second was selected for the rest of the experiments, which corresponds to a channel bandwidth of 66 Hz.

A third passband was used for the transmission of acoustic chirp signals. The width of the passband was approximately 130 Hz, which is almost two times greater than the passband from the experimental setup and twice the possible transmission rate. However, drilling will change the boundary conditions of the channel. The contact points of the drill string with the well as the borehole goes deeper will change the boundary conditions and this may alter the frequency response. The bandwidth of the channel may become wider or narrower and will need further investigation in the field.

The performance of the acoustic telemetry system was evaluated by comparing the SNR versus the BER at the receiver. The SNR was calculated by dividing the root mean squared (RMS) power of the signal by the RMS power of the accelerometer noise. The RMS of the signal was measured by subtracting the RMS power of the noise from the accelerometer signal containing the acoustic wave. It was found that by decreasing the voltage applied to the piezoelectric actuator, the SNR for acoustic transmission can be controlled. The BER was calculated by counting the number of errors at the receiver and then dividing it by the total transmitted bits. The system was left to run and transmit a few thousand bits until an estimate of the BER performance was achieved.

Lock-in Amplifier

A comparison between LIA and non-LIA demodulation is shown in FIG. 28. LIA allows for higher fidelity demodulation of the X and Y components from the accelerometer signal. Without it, noise from the errors greatly corrupted the transmitted signal, causing more errors at the surface receiver. In the plot, a first mode shows a performance that is worse than the performance of a second mode. This may be due to the bandwidth increase caused by the pulse-shaping transmit filter and a roll-off factor. A non-LIA case was demodulated using a matched receive filter that determines the optimal sampling instance for each symbol. The LIA demonstrated better performance when demodulating DBPSK modulated signals.

Channel Coding and Fusion

In experiments, different frequencies travelled through the different passbands of the simulated drill string. Applicant believes that this would imply that the waveform at 2387 Hz and 3742 Hz would experience different attenuation and noise, as shown in FIG. 20. Applicant found that by using convolutional coding, the errors caused by the drill string dynamics and disturbances can be further reduced. Comparison of the performance improvement on the LIA demodulated signal when using channel encoding and fusion is shown in FIG. 29. MRC, as outlined in equation (12), was used to fuse both frequencies to increase the SNR, therefore increasing the information available for detection while lessening the impact of noise. The calculated SNRs for the corresponding peak-to-peak input voltage was tabulated and summarized in Table 4. As shown, in MRC, the combined SNR is the sum of the individual SNR as seen on the table. By transmitting the same signal through different carrier frequencies, and then combining the phases of the demodulated signal, significant performance increase was achieved.

TABLE 4
Calculated SNR values.
	Peak-to-
	peak	1st	2nd
	Input	mode	mode	MRC
	Voltage	SNR	SNR	SNR
	(V)	(dB)	(dB)	(dB)
	40	−1.36	−3.53	0.70
	20	−7.56	−9.78	−5.52
	10	−13.65	−15.94	−11.64
	5	−19.60	−21.89	−17.59
	2.5	−25.17	−27.43	−23.14

For all three cases (1^st mode, 2^nd mode, and combined) error-correction showed significant improvement but came at a cost of bandwidth or lower transmission speed. Since error-correcting codes can only correct a few errors at a time, better performance can be seen at higher SNRs on FIG. 29. Applicant has found that combining the signals from each frequency further improves the performance of the system. As discussed, as a result of signal fusion, the phase shifts are more distinguished, thereby making detection easier while increasing its resistance to noise. Lower SNR needed for transmission was beneficial for an MWD telemetry system because space is limited at the BHA. Smaller battery packs may be required, freeing up space for other useful sensors. The improvement from channel coding and diversity combining can be further seen by plotting the BER with the peak-to-peak input voltage as shown in FIG. 30. Individually, channel coding exhibited better performance than the non-coded transmission.

Lock-in amplifiers are widely used in demodulating signals with a known carrier frequency. In the present invention, it was used to detect and demodulate acoustic waves that carry sensor information from the BHA to the surface. In addition, multifrequency transmission enabled signal diversity combining techniques that further improved the BER performance of the acoustic telemetry system. This, combined with convolution, increased the robustness of the system against errors. The presented experimental results show the benefits arising from the novelties of the invention.

Claims

1. Acoustic telemetry system for use with a drill string having MWD sensors and instruments, while conducting MWD operations comprising:

an acoustic transmitter tool, comprising: a mandrel for securing the tool to the drill string, the mandrel having an annulus therethrough; an annular acoustic transducer secured to the mandrel for generating acoustic waves for travelling longitudinally along the drill string; a pressure barrel concentrically positioned within the annulus for housing electronics; and electro-mechanical connections for communicating between the piezoelectric actuator and the electronics; and

a surface receiver.

2. The system of claim 1, wherein the acoustic transducer can further comprise a piezoelectric actuator comprises a stack of a plurality of ceramic rings composed of lead zirconate titante (PZT).

3. The system of 1, wherein the acoustic transducer can further comprises a magneto-strictive transducer.

4. The system of claim 1, wherein the electronics receives sensor data from the MWD sensors, and further comprises at least a voltage transformer, switching circuitry, and a microcontroller for converting binary signals to a modulated drive signal.

5. The system of claim 1, wherein the surface receiver further comprises at least two accelerometers or acoustic emission sensors positioned and spaced longitudinally along the drill string for receiving and detecting acoustic signals transmitted from the tool.

6. The system of 5, wherein the surface receiver further comprises a lock-in amplifier.

7. The system of 6, wherein the lock-in amplifier further comprises:

orthogonal oscillators;

phase-lock loop; and

a plurality of low-pass filters.

8. The system of claim 1, further comprising an isolator at a downhole end of the tool, for minimizing acoustic interference.

9. The system of claim 1, wherein the mandrel further comprises a bottom sub, a top sub, and an inner mandrel.

10. The system of claim 1, wherein the pressure barrel is a beryllium copper pressure barrel.

11. A method for downhole telemetry and acoustic transmission of MWD data for drilling operations using a drill string assembly, the method comprising the steps of:

(a) providing an acoustic transmitter tool having a piezoelectric transmitter and positioning the acoustic telemetry tool down a wellbore;

(b) providing a surface receiver;

(c) prior to commencing the drilling operations, determining an appropriate acoustic transmission frequency;

(d) after determining an appropriate acoustic transmission frequency, commencing the drilling operations while conducting acoustic transmission using the appropriate acoustic transmission frequency;

(e) repeating the step of determining an appropriate acoustic transmission frequency as the drill string is elongated during the drilling operations to determine a subsequent acoustic transmission frequency;

(f) continue drilling operations and acoustic transmission using the subsequent acoustic transmission frequency; and

(g) repeating steps (e) and (f) as necessary until the drilling operations are complete.

12. The method of claim 11, wherein providing a surface receiver further comprises configuring acoustic emission sensors on the drill string assembly to receive and detect acoustic signals transmitted from the acoustic telemetry tool.

13. The method of claim 12 wherein configuring acoustic emission sensors further comprises positioning at least two accelerometers longitudinally along the drill string assembly at known distances.

14. The method of claim 11, wherein determining an appropriate acoustic signal transmission frequency prior to commencing drilling operations, further comprises:

prompting the acoustic telemetry tool to transmit a broadband acoustic signal over all frequencies of interest;

receiving the broadband signal with the surface receiver; and

converting the broadband signal into electrical signals and processing the electrical signals to determine the appropriate acoustic transmission frequency.

15. The method of claim 11, further comprising processing electrical signals at both the surface receiver and the piezoelectric transmitter to increase accuracy of MWD data received at the surface receiver.

16. The method of claim 15, wherein processing electrical signals at the transmitter further comprises converting MWD data into a modulated drive signal for the transmitter involving channel coding, modulation, pulse-shaping, and up-conversion.

17. The method of claim 16, wherein modulation further comprises convolutional encoding.

18. The method of claim 16, wherein processing electrical signals at the surface receiver further comprises demodulating the modulated drive signal at the surface receiver.