CHARACTERIZATION SYSTEM AND METHOD FOR CEMENT EVALUATION THROUGH-TUBING

Abstract

In at least one embodiment, a well inspection method and system is disclosed. The method includes receiving signals from a casing structure using a well inspection tool, performing a Fast Fourier Transform (FFT) on the signals to generate spectrogram data, determining entropy spectra from the spectrogram data, performing component analysis based at least in part on the entropy spectra to select a subset of the entropy spectra, and determining casing loadings or cement bonding associated with the casing structure based at least in part on a subset of the entropy spectra.

Description

BACKGROUND

1. Technical Field

This disclosure relates generally to oilfield equipment and more particularly to systems and methods for evaluating cement-bonding or casing loadings through tubing or multiple casings.

2. Description of the Prior Art

To reliably evaluate cement-bonding quality and identify isolation zones of a cased hole, particularly for a plugged and abandoned (P&A) well, complexities may arise and a variety of sonic and ultrasonic logging tools may be used to address such complexities. For example, mature sonic logging tools include the variable density log tool (VDL) and cement bond tool (CBL), using which waves are excited and received in the pitch-catch mode. An ultrasonic tool, in an example, is an isolation scanning tool. In such a tool, ultrasonic waves are generated and acquired in a pulse-echo mode. Segmented radial bond logging tools also have been developed for cement evaluation, including Segmented Bond Tool (SBT) and Radial Bond Tool (RBT). However, the deployments of these tools may be restricted to a single casing well environment, due to the limitations of hardware and/or data processing. For a dual pipe string, the signals leaked from the external casing are sunk into the strong guided wave of the tubing and fluid signals, particularly in a late time of the measurements. The variances caused by the casing loading can be much lower than those caused by uncontrollable intensity changes of stimulation (source) pulses. To improve the accuracy and reliability of cement evaluation through tubing techniques, an entropy component analysis technique can be introduced to detect cement bonding condition of the casing through the tubing, acquiring logging data with a radial bond tool, of which the window length of received waveforms is extended to obtain tool responses to the external casing (tool setting modification).

SUMMARY

In at least one embodiment, a well inspection method is disclosed. The method includes receiving signals from a casing structure, using a well inspection tool. A further step includes performing, using at least one processor associated with the well inspection tool, a Fast Fourier Transform (FFT) on the signals to generate spectrogram data. The method includes determining entropy spectra from the spectrogram data. The method includes performing a component analysis based at least in part on the entropy spectra to select a subset of the entropy spectra. A determining step of the method is for casing loadings or cement bonding associated with the casing structure based at least in part on a subset of the entropy spectra.

In at least one embodiment, a system for well inspection is also disclosed. The system includes a transmitter of a well inspection tool to transmit a test signal into a casing structure and one or more receivers of the well inspection tool to receive signals from the casing structure. Further, the system includes at least one processor and memory comprising instructions that when executed by the at least one processor enable the system to receive signals from a casing structure using the well inspection tool, to perform a Fast Fourier Transform (FFT) on the signals to generate spectrogram data. The system is also enabled to determine entropy spectra from the spectrogram data, to perform a component analysis based at least in part on the entropy spectra to select a subset of the entropy spectra, and to determine casing loadings or cement bonding associated with the casing structure based at least in part on a subset of the entropy spectra.

BRIEF DESCRIPTION OF DRAWINGS

Some of the features and benefits of the present disclosure having been stated, others will become apparent as the description proceeds when taken in conjunction with the accompanying drawings, in which:

FIG. 1 is a partial cross-sectional view of a casing structure subject to a well inspection using a system or well inspect tool as detailed herein and in accordance with at least one embodiment.

FIG. 2 is a plan view of a casing structure subject to a well inspection using a system or well inspect tool as detailed herein and in accordance with at least one embodiment.

FIGS. 3A, 3B, and 3C illustrate intermediate features enabled using at least one processor executing instructions comprised in a memory of a system for well inspection, in accordance with at least one embodiment.

FIGS. 4A, 4B, and 4C illustrate truncated entropy spectra from different principal components of a component analysis, such as a principal component analysis (PCA), performed on data from entropy spectra, to support artificial intelligence/machine learning (AI/ML) features enabled by the well inspection method and system herein, in accordance with at least one embodiment.

FIG. 5 is a block diagram of computer and network aspects for a well inspection system as described in FIGS. 1-4C herein, in accordance with at least one embodiment.

FIG. 6 is a flow diagram of a well inspection method to support the descriptions with respect to FIGS. 1-4C herein and used at least in part with the block diagram of FIG. 5, in accordance with at least one embodiment.

While the disclosure will be described in connection with the preferred embodiments, it will be understood that it is not intended to limit the disclosure to that embodiment. On the contrary, it is intended to cover all alternatives, modifications, and equivalents, as may be included within the spirit and scope of the disclosure as defined by the appended claims.

DETAILED DESCRIPTION

The foregoing aspects, features and advantages of the present technology will be further appreciated when considered with reference to the following description of preferred embodiments and accompanying drawings, wherein like reference numerals represent like elements. In describing the preferred embodiments of the technology illustrated in the appended drawings, specific terminology will be used for the sake of clarity. The present technology, however, is not intended to be limited to the specific terms used, and it is to be understood that each specific term includes equivalents that operate in a similar manner to accomplish a similar purpose.

In at least one embodiment, to resolve issues in use of the entropy spectra of waveforms received by sector receivers, a further component analysis, such as principal component analysis (PCA), may be used. The entropy spectrum is a defined and computed entropy to suppress effects on pitch-catch measurements caused by unstable intensity of a stimulation pulse and reveal states of tubing and casing, as well as the interferences of vibrations of these two pipes, which are associated with the casing bounding conditions. The principal component analysis or PCA is conducted on a training data set of an entropy spectra of waveforms of sectors to extract the feature components that are related to the resonance of tubing and bonding conditions of the casing. The feature components are used to model the entropy spectra of waveforms of sectors for detecting the bonding conditions of the external casing.

FIG. 1 is a partial cross-sectional view of a casing structure or casing 100 subject to a well inspection using a system or well inspect tool as detailed herein and in accordance with at least one embodiment. Such a casing structure 100 may include barriers, such as a production tubing 102, a production casing 104, an intermediate casing 106, a surface casing 108, and a conductor casing 110. In at least one embodiment, such a casing structure 100 may include cemented casing forming some of the casing structure 100 or may include cement filling between annular spaces of the casing structure 100.

In at least one embodiment, a casing structure 100 herein may be part of a P&A well (also referred to herein as a plugged and abandoned well) or may be of any type of well, including but not limited to conventional and unconventional hydrocarbon producing wells. A well inspection tool 112 may be deployed downhole into the casing structure 100 to perform various logging functions, such as detection of various anomalies, such as casing defects, casing eccentricity, topology, integrity, and other information. In at least one embodiment, an intent for verification of casing loading or cement bonding may be to prevent flow paths in the cement that can allow hydrocarbon leakage for a P&A well, where the hydrocarbons may then appear on an environment surface. In at least one embodiment, the well inspection tool 100 may be part of or may fully include an acoustic-based imaging device for detecting integrity and defects of the casing structure 100. The casing structure 100 may therefore have one or a series of cylindrical metal casings and cement wall layers between the cylindrical metal casings or between a casing and a formation. In at least one embodiment, the cement wall layers are subject to different stresses. For example, the stress can be caused by the formation movement, which may be associated with loading from applied or natural materials, such as gas, liquids, light cement, and heavy cement.

As illustrated in FIG. 1, the well inspection tool 112 can traverse into a casing structure 100 and may be used for determining integrity of the casing structure 100. In at least one embodiment, the well inspection tool 112 can be used to determine the integrity of each of the barriers (such as casings themselves and cement layers between the casings) and/or of a number of annular spaces between the barriers. The well inspection tool 112; 520 may include an acoustic signal generator component 516 (in FIG. 5), a transmitter 518 (also, 112A or 112B), and one or more receivers 508 (also 112A or 112C). When the transmitter is able to perform as a receiver, the combination is a transceiver 112A. The well inspection tool 112; 520 may be deployed at the different depths 118 inside the casing structure 100, to secure data from different sectors 116 at such different depths. The well inspection tool 112; 520 therefore has material and structural integrity to withstand the high pressures and high temperatures at these depths.

In at least one embodiment, the well inspection tool 112 includes the transmitter 112B and the one or more receivers 112C or a transceiver 112A. The configuration of the transmitter and receiver may be in a pitch-catch mode, where a vertical separation is present between these components. In at least one embodiment, the configuration of the transmitter and receiver may be a pulse-echo configuration, where these two components are within a substantially same horizontal plane (with little to no vertical separation), thereby forming a transceiver. In at least one embodiment, therefore, signals received in the transceiver 112A may travel in a perpendicular direction 120 relative to an axis of the casing structure 100. In at least one embodiment, the signals received in the one or more receivers 112C may travel along an axial direction 122 relative to the axis of the casing structure 100.

In at least one embodiment, such a method and system can reliably evaluate casing loadings and cement bonding, to identify isolation zones of a casing structure, particularly for a P&A well. While a variety of acoustic, such as sonic and ultrasonic, logging tools have been developed for cement evaluation, matured sonic logging tools that include VDL and cement bond logging (CBL) may rely on waves that are excited and received in a pitch-catch mode. In at least one example, an ultrasonic tool may be an isolation scanning tool. For such a tool, ultrasonic waves are generated, transmitted, and are acquired in a pulse-echo mode. Further, a segmented or a radial bond logging Tool (SBT or RBT) may be used. In at least one embodiment, however, deployments of these tools may be restricted to single casing well environments, due to limitations of hardware and/or data processing methods.

In at least one embodiment, two receivers 112C may be provided on the well inspection tool 112, with one receiver at 5 feet from the transmitter 112A and with a second receiver at 3 feet from the transmitter 112A. The well inspection tool 112 is configured to operate at 20 KHz. This may represent an RBT configuration for the well inspection tool 112 providing a 3 feet vertical resolution at a low frequency of 20 Khz. Further, an SBT configuration of the well inspection tool 112 may be used to acquire compensated two direction azimuthal measurements for detecting bonding conditions. This tool may operate at 80 Khz with vertical resolution of 9 inches. Further, the SBT tool can provide evaluation of cement bond conditions for at least six azimuthal sectors. In at least one embodiment, pads are arranged circumferentially on the SBT tool, at two vertical levels, for mounting the sensors, which make the SBT tool omnidirectional.

In at least one embodiment, a single or dual casing configuration may be subject to analysis using the well inspection tool 112. The pipes may be concentric and eccentric. As the present method uses PCA, in at least one embodiment, and does not rely only on amplitude or entropy spectra, it presents accurate features of casing loading and of cement bonding conditions that may exist outside a casing. Further, the phase array measurements of the well inspection tool 112, in the pitch-catch mode, can include acquired waves having compressional mode, shear mode, flexural modes, lamb modes (A modes, S modes, etc.), that may be analyzed for determination of casing loading or cement bonding conditions. Whereas, the pulse-echo measurements of the well inspection tool 112 may have only compression waves.

In at least one embodiment, a phase array (such as, of 20 KHz) of receivers that are capable of functioning at sonic or ultrasonic low frequency may be used in the well inspection tool. Component analysis, such as PCA, may be used to identify specific feature components of an entropy spectra to be evaluated for problems or issues, such as indicative of variances in the integrity of the well. For example, the feature components are related to a resonance of tubing and bonding conditions of the casing. This may be variances associated with the integrity of a well. In at least one embodiment, a formation signal is between 8 to 17 Khz.

FIG. 2 is a plan view of a casing structure subject to a well inspection method using a system or well inspect tool as detailed herein and in accordance with at least one embodiment. To improve accuracy and reliability of cement evaluation using through-tubing techniques, a well inspection tool 112 may be fully or partly within a casing structure 100; 200 to apply the entropy and PCA analysis techniques to detect casing loadings and cement bonding of at least one cement 202 layer of the casing structure 100; 200. In at least one embodiment, such a well inspection tool may include a signal generator, transmitter, and receiver downhole, and may also include a processor to execute instructions from a memory, where these aspects are located in one or more modules 212 on a surface or up-hole environment, relative to a downhole environment that is at least partly within the casing structure 100; 200. In at least one embodiment, the well inspection tool 112 may be centered in the casing structure 100 or may be eccentric, with an offset 216, relative to the center of the casing structure 100. The center of the casing structure 100 may be represented by at least a center of the inner-most casing 102.

In at least one embodiment, a transmitter of the well inspection tool 112 can transmit acoustic signals for pulse-echo measurements or for pitch-catch measurements, of which the chirp pulse may be applied to enhance a sensitivity of measurements to external casing loadings of a casing structure excited by such acoustic signals 100; 200. In at least one embodiment, an acoustic impedance may be used to identify types of casing loadings for a single pipe. For a dual pipe string, third-interface echo (TIE) signals may be used to represent a signal reflected by the internal surface of production casing 104, and may be used in the analysis of the cement-bonding conditions and casing loadings. However, such TIE signals may be weak signals from the production casing 104 and may be sunk into the measurement noise. In at least one embodiment, such sinking of the weak TIE signals may be particularly the case for late time pulse-echo measurements.

Furthermore, variances in measured waveforms caused by the casing loading may be even lower than uncontrollable intensity changes of the source pulses generated for driving pulse-echo measurements. To address these aspects, entropy spectra generated from the signals in a receiver can suppress effects on the pulse-echo measurements or pitch-catch measurements caused the uncontrollable intensity changes of the pulses, where such an entropy spectrum may be used for determining states of a perturbed tubing and a perturbed casing, and can be also used to quantify interferences of vibrations associated with these two pipes. All this information can be used to differentiate the casing loadings, or the cement bonding conditions.

FIG. 2 also illustrates that the well inspection method herein can be used to determine cement-bonding conditions and casing loadings associated with the casing structure 100; 200, based at least in part on the entropy spectra. The cement bonding conditions pertain to the cement 202 located around the production casing 104, in at least one example. Therefore, the well inspection tool 112 provides an acoustic signal from the production tubing 102 through one or more media 206, 208 to the production casing 104. In at least one embodiment, the media may be water or hydrocarbons or mud. In at least one embodiment, as illustrated in FIG. 2, there may a free pipe or no bonding condition caused by gaps (here, illustrated as one or more slots 204A, B), a fully bonded condition (illustrated by a remainder of the cement 202), or a partial bonding condition (illustrated by intersections or physical changes from where a slot 204A; B starts and a remainder of the cement 202 occurs).

In at least one embodiment, a well inspection tool 112 may be fully self-sufficient to determine cement-bonding conditions and casing loadings associated with the casing structure 100; 200, based at least in part on the entropy spectra. For example, pitch-catch measurements herein can be used to detect cement bonding conditions while pulse-echo measurements herein may be used for detecting casing loading. In at least one embodiment, therefore, the well inspection tool 112 may include a processor to execute instructions from an associated memory without a need to communicate signals up-hole to a computer to perform such functions. In at least one embodiment, therefore, a module 212 having such a processor and memory may be located within the well inspection tool 112, may be associated with the well inspection tool 112 downhole, or may be located up-hole and may communicate with the well inspection tool 112 via wires 210 or wirelessly.

FIGS. 3A, 3B, and 3C illustrate intermediate features 300, 350, 370 enabled using at least one processor executing instructions comprised in a memory of a system for well inspection, in accordance with at least one embodiment. To determine the entropy spectra, a Fast Fourier Transform (FFT) may be performed on the signals to generate spectrogram data. FIG. 3A illustrates a truncated amplitude spectrum 300, which is a part of a full amplitude spectrum, that would be generated from the FFT performed for signals received in a receiver from one or more azimuthal sectors, at least one depth, where such determination is performed. In at least one embodiment, in the truncated amplitude spectrum 300, the x-axis corresponds to different frequencies 302, the y-axis corresponds to different depths 304;118 at which measurements are performed for the production tubing 102 or production casing 104.

In at least one embodiment, signals are received for different depths 304, with respect to a circumference of the production tubing 102 or the production casing 104 to indicate where a test signal is provided and where from the signal associated with the test signal is received. In at least one embodiment, the varying shades in the truncated amplitude spectrum 300 (also in the shading legend 306) represent the amplitude value of each of the lines of the frequency 302 and range from 2 to 12×10⁻³.

In at least one embodiment, approaches herein include a method for well inspection using the well inspection tool 112. The method includes loading well environmental parameters and other data into a system that includes at least one processor and memory. The parameters include well name, well location, true vertical depth (TVD), measured depth (MD), borehole diameters (such as internal diameter (IDs) of one or more casings), formation thicknesses, formation densities, formation porosities, formation saturations, formation matrix compositions, mud types, mud densities, borehole fluids, completion intervals, casings outer diameters (ODs)s, casings thickness, casings lengths, casings weights, tubing OD, tubing thickness, and tubing weight.

In at least one embodiment, the other data may include load radial bond logging data, including time sampling interval times of waveforms, operation frequencies of sensors, sampling depths, and waveform array data acquired by sector sensors (such as the receiver 112C or transceiver 112A), at each sampling depth 118 of a well including the casing structure 100.

The FFT S_RX^i,j(k) of waveforms {s_RX^i,j (k), n=1, 2, . . . , NT} at the j^th sector 116 and the i^th depth 118 (representing a sampling point) is determined by Equation (1):

$\begin{array}{cc}{S}_{RX}^{i,j}\left(k\right)={\sum }_{n=0}^{N-1}{s}_{\mathrm{RX}}^{i,j}\left(n\right)W_N^{\mathrm{kn}}=A_{\mathrm{RX}}^{i,j}\left(k\right)e^{j{\psi }_{\mathrm{RX}}^{i,j}\left(k\right)}.& \mathrm{Equation}\left(1\right)\end{array}$

Further, in Equation 1,

$s_{RX}^{i,j}\left(n\right)=s_{RX}^{i,j}\left(n{T}_s\right),W_N^{kn}=e^{-j\frac{2\pi }{N}\mathrm{kn}},f_s=\frac{1}{T_s},\mathrm{and}{f}_K=\frac{k}{N}{f}_s,\mathrm{and}{A}_{RX}^{i,j}\left(k\right)$

represents the amplitude of signal at frequency f_K, s_RX^i,j(n) is the signal value at the n^th time sampling point, W_N^kn is the exponential kernel function for converting the signal from the time domain to the frequency domain. Still further, in Equation (1), i=1, 2, . . . , M_depth 118 and j=1, 2, . . . , N_sector 116.

In at least one embodiment, a sampling frequency that satisfies the Nyquist-Shannon sampling theorem is used to acquire the signals in a receiver 112C or transceiver 112A to avoid the alias of FFT spectrum of a signal that is returned. Still further, a limited length of the return signal is used in such analysis. Therefore, the data underlying the amplitude spectrum that includes the truncated amplitude spectrum 300 in FIG. 3A, is referred to herein as the spectrogram data that is output from the FFT process.

In at least one embodiment, with the FFT performed, an entropy spectrum may be determined from the spectrogram data. To do so, a sum of a power spectrum for a waveform is determined. Then, the power spectrum is normalized using the sum of the power spectrum. Finally, a contribution of the individual frequency component of a waveform to the entropy spectrum is determined based in part on a natural logarithm of the normalized power spectrum.

In at least one embodiment, therefore, the power spectrum may be represented as:

$\begin{array}{cc}{P}_{i,j}\left(k\right)={\left[A_{i,j}\left(k\right)\right]}^2.& \mathrm{Equation}\left(2\right)\end{array}$

In Equation (2), k=0, 1, . . . , K. In at least one embodiment, it may be difficult to analyze the return signals using the power spectrum at least in part because of the casing structure. For example, in some instances the acoustic signal may be strong and in other instances it may be weak. Further, it is difficult to determine whether an amplitude of a return signal is affected by the cement-bonding conditions, a pipe condition (such as pipe surface roughness or pipe eccentricity), or the source signals themselves. In at least one embodiment, instead of using the power spectrum to perform any analysis for casing loading or cement bonding, a defined entropy spectrum {E_i,j(k), k=0, 1, . . . , K} is first determined for waveforms s_i,j(n), acquired at the j^th sector 116 and the i^th depth 118 sampling point. The entropy spectra may be based on the power spectrum and is defined as:

$\begin{array}{cc}{E}_{i,j}\left(k\right)=-p_{i,j}\left(k\right)\log \left[p_{i,j}\left(k\right)\right].& \mathrm{Equation}\left(3\right)\end{array}$

In Equation (3), the p_i,j(k) value is a normalized power spectrum and may be represented as:

$\begin{array}{cc}{p}_{i,j}\left(k\right)=\frac{P_{i,j}\left(k\right)}{{\sum }_{m=0}^{K-1}{P}_{i,j}\left(m\right)}.& \mathrm{Equation}\left(4\right)\end{array}$

Further, in Equation (4), {P_i,j(k), k=0, 1, . . . , K−1} is the FFT power spectrum of waveforms s_i,j(n), acquired at the j^th sector 116 and the i^th depth 118 sampling point.

FIG. 3B illustrate intermediate features 350 enabled using at least one processor executing instructions comprised in a memory of a system for well inspection, in accordance with at least one embodiment. In at least one embodiment, the intermediate features 350 pertain to entropy spectra of a waveform at the same sector and depth as the FFT features 300 of FIG. 3A. For pulse-echo mode measurements of dual pipe (such as including production tubing and production casing), the signals received may have waveforms composed of multiple echoes reflected by the production tubing and multiple TIE signals reflected by an inner surface of production casing. The term-2 is the second time term of a waveform for determining the entropy spectra. Further, the waveform of term-2 or a later time term provides strong sensitivity to the casing loading, compared with term-1.

In at least one embodiment, for pitch-catch mode measurements of dual pipe (such as including production tubing and production casing), the signals received may have waveforms composed of multiple signal traveling through a material of the production tubing. Therefore, the intermediate features 350 may then pertain to an entropy spectra of the amplitude features 300 of FIG. 3A.

As illustrated, the entropy spectra having the intermediate features 350 is an image plotting frequency 352 against a depth index 354 (such as a normalized depth relative to a depth point at one or more sectors) that reflects different receivers or sensors in a pitch-catch configuration of the well inspection tool 112. There may be depth sampling points for eight receivers from depth index 0.5E+4 to depth index 3.0E+4, for instance. In at least one embodiment, the intermediate features 350 therefore represent a combined set of entropy spectra of acquired waveforms at which casing loading or cement bonding conditions are studied. This allows verification of the data from each sensor against other sensors of the same sector and same depth at which all the signals are received after a test signal is provided from a transmitter of the well inspection tool 112.

In at least one embodiment, the depth may be maintained, and different sectors may be probed, and which may vary from 0 to 360 degree (such as with respect to a 360 degree circumference of such a tubing or casing) at which a signal for pulse-echo or pitch-catch mode measurements is received and analyzed for the casing loadings or the cement bonding at that sector. Further, the shading 356 provides the amplitudes associated with the frequency 352.

The power spectra and entropy spectra of waveforms have significant differences in features related to tubing, casing, casing loadings, and a source pulse, as demonstrated in the difference between FIGS. 3A and 3B. The edges of the shading 308 of the features 300 in the 200 to 300 KHz frequency bands of FIG. 3A is less prominent than the shading 358 (entropy spectra) of the features 350 of the same frequency bands of FIG. 3B (reference numeral 308 versus reference numeral 358), and therefore, this supports that accurate determination of casing loadings may be obtained by using an entropy spectrum described herein.

In at least one embodiment, a component analysis performed from the entropy spectra allows selection of component features of the entropy spectra that most likely inform about specific cement bonding or casing loading features. At least a feature 308 of the waveforms of the return signals, in amplitude spectra, provides information that an acoustic part of the signals has source pulses of the pitch-catch measurements, at varied intensities. However, the feature 358 from the waveforms of the signal received in a receiver, from the entropy spectra, reflects those effects on pulse-echo measurements or pitch-catch measurements, caused by variances in source pulse density of a test signal may be suppressed in the entropy spectra.

In at least one embodiment, data quality analysis may be performed using the entropy spectra E_i,j(k) (Equation (3)) of waveforms s_i,j(n) and of specific depth (i) and sectors (j) 116. This is to identify floating zero point of circuit, variances in intensity of source pulses, resonances of tubing and casing, and other relevant aspects of the casing structure 100. A selection of a depth sampling point range [i₀, I_depth] of logging data for cement evaluation and a number (I_training) of depth sampling points may be used to generate a training dataset. This is to extract feature components in the signals received at the receiver of the well inspection tool 112, of which feature components may pertain to cement bonding conditions or casing loading. These feature components may be used to identify underlying data of the entropy spectra that most likely inform about specific cement bonding or casing loading features.

In at least one embodiment, the entropy spectra E_i,j(k) (Equation (3)) of multiple waveforms s_i,j(n) of different sectors in a selected depth sampling points range may be merged (see reference 350 in FIG. 3B) into one dataset giving by {E_q(k), k=1, 2, . . . , K, q=1, 2, . . . , N(I_depth−_i+1)}. Further, the dataset may be randomly sampled to provide a first part of the data. The first part of the data is a merged dataset of entropy spectra of the waveforms of the sectors and form a training dataset for use with the PCA process. The training dataset is used to extract feature components therein via the PCA process. Then, a determination can be made for an averaged entropy spectrum {e_c,0(k), k=1, 2, . . . , K} of the training dataset.

In at least one embodiment, a component analysis process, such as PCA, can be performed on the training dataset to extract feature components, represented by {e_c,r(k), r=1, 2, . . . , R}. For example, a covariance matrix X is determined from a first part of the entropy spectra that is sampled from the entropy spectra. In at least one embodiment, PCA is an aspect of singular value decomposition (SVD) performed on the covariance matrix X to identify factors causing measurement variances and to reduce dimension, as well as extract feature component that represent problems within the dataset (such as within the entropy spectra).

In at least one embodiment, the first part of the entropy spectra is the training dataset of the entropy spectra of waveforms of the sectors. The extracted feature components {e_c,r(k), r=1, 2, . . . , R} from the PCA can be validated with a second part of the entropy spectra (that is, the merged dataset of entropy spectra of waveforms or by a log analysis expert).

In at least one embodiment, when the feature components from the component analysis, such as PCA, are determined to be valid, the underlying entropy spectra may be used to determine casing loadings or cement bonding associated with the casing structure. When the features components are not valid, the selection of a depth sampling point range [i₀, I_depth] of logging data for cement evaluation and for casing loadings, as well as the number I_training of depth sampling points for the first (training) part of the dataset may be performed for a different first part of the dataset.

In at least one embodiment, the entropy spectra of waveforms of each sector in the selected depth sampling point range are assembled, as represented in Equation (5):

$\begin{array}{cc}{E}_{i,j}\left(k\right)=e_{c,0}\left(k\right)+{\sum }_{r=1}^R{z}_{i,j,r}{e}_{c,r}\left(k\right),+{\epsilon }_{i,j}\left(k\right).& \mathrm{Equation}\left(5\right)\end{array}$

Here, {z_i,j,r, i=1, 2, . . . , M_depth, j=1, 2, . . . , N_sector} are scores of the features components of entropy spectra of waveforms of sectors. For example, as illustrated in references 370 of FIG. 3C, the method herein includes determining scores 380, 382; 384 of the feature components of the entropy spectra according to the angle (in degrees) 376 of the well inspection tool 112 and the depth index 378, which represent the sector at which the signal is received in one or more receivers.

In at least one embodiment, a least-squared error method may be applied to solve an optimization problem to minimize fitting errors between raw entropy and the modeled entropy (of Equation 5). Then, the scores 380, 382; 384 are elements of a solution vector of the optimization problem. The scores 380, 382 are illustrated as shades (of a range of scores from −0.2 to 0.2) 384 but may be of different color to visualize the differences in a user interface of a well inspection system. The scores 380, 382 are therefore associated with a depth and a sector within a well where the signal is received and with areas in the entropy spectra indicative of sharp contrast changes at 20 KiloHertz (KHz) (see reference 358 in FIG. 3B, for instance).

In at least one embodiment, individual ones of the scores may be selected for which underlying ones of the feature components are to be associated within a dataset. For example, visualization and interpretation of feature components of the entropy spectra may be used to select the scores and the underlying entropy spectra may be associated to form the dataset that is the subset of the entropy spectra. Scores of feature components may be selected based in part on same or similar bonding conditions of external casing of the entropy spectra at different sectors. These feature components may be merged into one dataset. Further, clustering may be performed within the dataset for a subset of the feature components. For example, the clustering is for a subset of the feature components, to identify clusters for the subset of the feature components that is associated with the casing loadings or cement bonding of the casing structure. For example, clusters of different casing loadings or cement bonding conditions may be defined. Then, feature components that are highlighted in the PCA process implies that their underlying entropy spectra corresponds to individual ones of the defined clusters. Then, the casing loadings or cement bonding conditions is determined for the depth and sector.

FIGS. 4A, 4B, and 4C illustrate truncated entropy spectra from different principal components of a principal component analysis (PCA) performed on data from entropy spectra, to support artificial intelligence/machine learning (AI/ML) features enabled by the well inspection method and system herein, in accordance with at least one embodiment. In at least one embodiment, multiple features may be further extracted from an entropy spectrum for classifying the casing loadings with the presence of the tubing.

In FIG. 4A, truncated entropy spectra 400 is provided from a first principal component of a PCA process on data that may be a subset of the entropy spectra 350 from FIG. 3C. This information may be used to support artificial intelligence/machine learning (AI/ML) features enabled by the well inspection method and system herein, in accordance with at least one embodiment. In at least one embodiment, a first raw segment in the truncated entropy spectra 400 is of one angle record waveforms from eight sensors. Then, the well inspection tool 112 may be rotated by an angle, such as 90 degrees. Therefore, a second raw segment in the truncated entropy spectra 400 represents one angle record of waveforms from eight sensors at the new angle. Therefore, each row in the truncated entropy spectra 400 is for one of the eight sensors. This may be similarly the case for the truncated entropy spectra 430, 460.

In at least one embodiment, the truncated entropy spectra 400, 430, 460 is for different PCA components for the same sector and depth of the signal received in the well inspection tool 112. Therefore, the y-axis is a depth index 404 and the x-axis is the frequency 402 of responses of the well inspection tool 112 (20 KHz). The label 406 provides a range of values from about −0.15 to 0.1 for the truncated entropy spectra 400, 430, 460, where the negative value indicates the influence of the scores. Further, each PCA component may highlight certain problem areas or clusters 408, 410, 412 in the entropy spectra. In at least one embodiment, the problem areas 408, 410, 412 may be identified according to FIG. 3C, where the score (shade) contrasts or changes may be used to select individual ones of the scores 380, 382 for which underlying ones of the feature components are within the problem areas 408, 410, 412 and are to be associated within a dataset that is the subset of the entropy spectra. Then, clustering may be performed within the dataset for a subset of the feature components, wherein the clustering identifies clusters 408, 410 for the subset of the feature components that is associated with the casing loadings or cement bonding of the casing structure.

In at least one embodiment, such a clustering 408, 410, 412 demonstrates that various data points from the entropy spectra can be used in an AI/ML algorithm to train a classifier to infer the type of a casing loading or cement bonding condition from the entropy spectra. For example, a classification algorithm is enabled to classify entropy data associated with the plurality of clusters 408, 410, 412 into distinct classes or categories. The classification algorithm may be based at least in part on a relationship between from clustering after the PCA is performed, in at least one embodiment. Further, the AI/ML algorithm then enables the determination of the casing loadings or cement bonding based at least in part on the entropy data classified within the distinct classes. In at least one embodiment, a first class corresponds to flow paths in the cement and a second class corresponds to no flow paths or channels in the cement.

In at least one embodiment, a K-means classifier, SVM classifier, kNN classifier, or a nearest neighbor classifier may be trained using the distinct classes 408, 410. Further, such a trained classifier may be used with newly received data of statistical features from the entropy spectra. Such a trained classifier can then classify the newly received into the distinct classes 408, 410 to enable the determination of the casing loadings or cement bonding based at least in part on PCA applied to the entropy data that is newly obtained for a P&A well or other well that includes concentric casing and/or tubing and that is cemented at one or more annular spaces. In at least one embodiment, further, such distinct classes 408, 410, 412 can provide information about fluid and cement loading or can provide information to be used in quality control during calibration of a system for determining casing loading using entropy spectra. In at least one embodiment, part of a training dataset from entropy spectra of simulated or modeled pulse-echo or pitch-catch measurements with different casing loadings or cement bonding may be used to train a classifier, whereas other parts of the training dataset can be used for calibration of the classifier.

In at least one embodiment, the statistical features are extracted features that are directly related to the dumpling effect on the vibration of the casing caused by its loading. As such, the method and system herein does not need the assumption for the mud weight. Further, the entropy spectra can be used to conduct the data quality control (QC) of pulse-echo measurements. In at least one embodiment, such QC can reveal a floating zero-point issue, monitoring failures of establishments of resonances of tubing and casing, quantifying the interferences of vibrations of casing and tubing, and removing effects on pulse-echo measurements caused by unstable intensity of stimulate pulses. In at least one embodiment, therefore, the method and system herein are reliable and not sensitive to intensity of the stimulation pulse. Further, the banded entropy contrasts between fluid and cement loading of casing for the high frequency band can be more than 10%. The defined entropy features can be used to classify casing loadings, using machine learning methods.

In at least one embodiment, the disclosure herein uses a new classification method in a supported system to evaluate quality of cement external to at least one casing or production tubing using, based at least in part on clusters from PCA applied to the entropy spectra, where such clusters represent distinct classes applied by a classification algorithm. The clusters can classify new PCA entropy spectra based at least in part on the PCA entropy spectra falling within the clusters. In at least one embodiment, the method herein is also applicable for both fresh and mature wells and can be performed fully downhole and in a combination of a downhole and a surface-based system.

In at least one embodiment, computer, and network aspects 500 for a downhole system as illustrated in FIG. 5, may be used as described herein. In at least one embodiment, these computer and network aspects 500 may include a distributed system. In at least one embodiment, a distributed system 500 may include one or more computing devices 512, 514. In at least one embodiment, one or more computing devices 512, 514 may be adapted to execute and function with a client application, such as with browsers or a stand-alone application, and are adapted to execute and function over one or more network(s) 506.

In at least one embodiment, a server 504, having components 504A-N may be communicatively coupled with computing devices 512, 514 via network 506 and via a receiver device 508, if provided. In at least one embodiment, components 512, 514 include processors, memory, and random-access memory (RAM). In at least one embodiment, server 504 may be adapted to operate services or applications to manage functions and sessions associated with database access 502 and associated with computing devices 512, 514. In at least one embodiment, a server 504 may be associated with one or more receivers 508 of a downhole tool 520.

In at least one embodiment, server 504 may be at a wellsite location, but may also be at a distinct location from a wellsite location. In at least one embodiment, such a server 504 may support a downhole tool or wireline system 520 for analysis of a casing structure in a downhole environment and to perform casing structure inspection of a well 522. Such a well inspection tool 520 may be partly downhole and partly up-holes, such as at a surface level. Such a downhole tool 520 may include a well inspection tool that may be a combination of one or more of a signal-generator 516, a transmitter 518, and one or more receivers 508 to perform at least part of the functions described throughout herein.

There may be a modelling system or signal generator 516 within a downhole tool 520 or separate from the downhole tool 520, at a server 504. In at least one embodiment, such a signal generator may be provided to perform part of the first set of steps of the algorithm to analyze or study casing loadings or cement bonding described throughout herein. In at least one embodiment, the signal generator 516 may be pre-calibrated on a surface level using known casing loadings and cement bonding conditions or simulations (or other representations, including images) thereof.

The subsystems 508, 516, 518 of the downhole tool may be encased in one or more computing devices having at least one processor and memory so that the at least one processor can perform functions based in part on instructions from the memory executing in the at least one processor. In at least one embodiment, even though illustrated together, the system boundaries of each module 508, 516, 518 may be around a distributed system having the subsystems in different geographic locations, including downhole and surface areas.

In at least one embodiment, a receiver 508 of a downhole tool 520 is provided to receive return signals from a downhole environment of a well 522. In at least one embodiment, a system for determining casing loadings or cement bonding includes a wireline system for the analysis, where such a system may be adapted to transmit, either through wires or wireless, information received therein, from the receiver 508 back to the surface. In at least one embodiment, acoustic signaling performed using a signal generator 516 may be associated with calibration or pre-calibration of signals intended to test one or more casing loadings (and representations thereof) that may be recorded within a receiver 508 or the well inspection tool or downhole tool 520.

The signal generator 516 can communicate to a transmitter 518 and to one or more receivers 508 to enable determination of casing loading or cement-bonding conditions of a well 522 using specific acoustic signals described throughout herein and using entropy with PCA analysis of signals received to classify statistical features therein against known casing loading or cement-bonding conditions associated with distinct classes. For example, specific signals received associated with specific ones of the casing loading or cement-bonding conditions can be verified as classifying in specific ones of the distinct classes using the above-described approaches.

Such return signals may include aspects of the transmitted acoustic signals to be applied from an acoustic transmitter 518. Detected return signals to the receiver 508 may be used to determine the casing loadings. In at least one embodiment, therefore, the system 500 enables the at least one processor (such as from components 504A-N, 512, 514 or in a server 504 or fully within a well inspection tool 520) to access the return signals, the entropy features extraction, the statistical features extraction, and to access classes of known casing loading or cement-bonding conditions to perform the classification of the statistical features into the known classes.

In at least one embodiment, such returned signals may be received in the receiver 508 and transmitted from there if such a receiver 508 is a transceiver. In at least one embodiment, a server 504 may function as a setting device (with the acoustic transmitter of the well inspection tool providing the actual acoustic signal and the receiver receiving a signal). In at least one embodiment, however, the server 504 may also perform other functions described throughout herein and at least as to the algorithm described herein.

In at least one embodiment, one or more component 504A-N may be adapted to function as a signal provisioning device within a server 504 to enable the acoustic transmitter to transmit its signals at the frequency ranges described throughout herein. In at least one embodiment, one or more components 504A-N may include one or more processors and one or more memory devices adapted to function as a detector, sensor, or receiver, while other processors and memory devices in server 504 may perform other functions.

In at least one embodiment, a server 504 may also provide services or applications that are software-based in a virtual or a physical environment (such as to support the simulations referenced herein). In at least one embodiment, when server 504 is a virtual environment, then components 504A-N are software components that may be implemented on a cloud. In at least one embodiment, this feature allows remote operation of a system for determining casing loading or cement-bonding conditions for formation using a wireline system that includes a well inspection tool, as discussed at least in reference to all the figures herein. In at least one embodiment, this feature also allows for remote access to information received and communicated between any of aforementioned devices. In at least one embodiment, one or more components 504A-N of a server 504 may be implemented in hardware or firmware, other than a software implementation described throughout herein. In at least one embodiment, combinations thereof may also be used.

In at least one embodiment, one computing device 510-514 may be a smart monitor or a display having at least a microcontroller and memory having instructions to enable display of information monitored by a receiver 508. In at least one embodiment, one computing device 510-512 may be a transmitter device to transmit directly to one or more receivers or to transmit via a network 506 to one or more receivers 508 and to a server 504, as well as to other computing devices 512, 514.

In at least one embodiment, other computing devices 512, 514 may include portable handheld devices that are not limited to smartphones, cellular telephones, tablet computers, personal digital assistants (PDAs), and wearable devices (head mounted displays, watches, etc.). In at least one embodiment, other computing devices 512, 514 may operate one or more operating systems including Microsoft Windows Mobile®, Windows® (of any generation), and/or a variety of mobile operating systems such as iOS®, Windows Phone®, Android®, BlackBerry®, Palm OS®, and/or variations thereof.

In at least one embodiment, other computing devices 512, 514 may support applications designed as internet-related applications, electronic mail (email), short or multimedia message service (SMS or MMS) applications and may use other communication protocols. In at least one embodiment, other computing devices 512, 514 may also include general purpose personal computers and/or laptop computers running such operating systems as Microsoft Windows®, Apple Macintosh®, and/or Linux®. In at least one embodiment, other computing devices 512, 514 may be workstations running UNIX® or UNIX-like operating systems or other GNU/Linux operating systems, such as Google Chrome OS®. In at least one embodiment, thin-client devices, including gaming systems (Microsoft Xbox®) may be used as other computing device 512, 514.

In at least one embodiment, network(s) 506 may be any type of network that can support data communications using various protocols, including TCP/IP (transmission control protocol/Internet protocol), SNA (systems network architecture), IPX (Internet packet exchange), AppleTalk®, and/or variations thereof. In at least one embodiment, network(s) 506 may be a networks that is based on Ethernet, Token-Ring, a wide-area network, Internet, a virtual network, a virtual private network (VPN), a local area network (LAN), an intranet, an extranet, a public switched telephone network (PSTN), an infra-red network, a wireless network (such as that operating with guidelines from an institution like the Institute of Electrical and Electronics (IEEE) 802.11 suite of protocols, Bluetooth®, and/or any other wireless protocol), and/or any combination of these and/or other networks.

In at least one embodiment, a server 504 runs a suitable operating system, including any of operating systems described throughout herein. In at least one embodiment, server 504 may also run some server applications, including HTTP (hypertext transport protocol) servers, FTP (file transfer protocol) servers, CGI (common gateway interface) servers, JAVA® servers, database servers, and/or variations thereof. In at least one embodiment, a database 502 is supported by database server feature of a server 504 provided with front-end capabilities. In at least one embodiment, such database server features include those available from Oracle®, Microsoft®, Sybase®, IBM® (International Business Machines), and/or variations thereof.

In at least one embodiment, a server 504 is able to provide feeds and/or real-time updates for media feeds. In at least one embodiment, a server 504 is part of multiple server boxes spread over an area but functioning for a presently described process for analysis of a porous formation. In at least one embodiment, server 504 includes applications to measure network performance by network monitoring and traffic management. In at least one embodiment, a provided database 502 enables information storage from a wellsite, including user interactions, usage patterns information, adaptation rules information, and other information.

FIG. 6 is a flow diagram of a well inspection method 600 to support the descriptions with respect to FIGS. 1-4C herein and used at least in part with the block diagram of FIG. 5, in accordance with at least one embodiment. In at least one embodiment, a well inspection method 600 includes transmitting (602) a test signal from a well inspection tool into a casing structure. The method 600 includes receiving (604) signals from a casing structure using the well inspection tool. A step in the method 600 is for performing (606), using at least one processor associated with the well inspection tool, a Fast Fourier Transform (FFT) on the signals to generate spectrogram data. A step in the method 600 includes verifying (608) that a complete spectrogram data is generated.

In at least one embodiment, upon positive verification in step 608, a step of the method 600 is for determining (610) an entropy spectrum from the spectrogram data. Otherwise, step 606 in the method 600 may be repeated till complete spectrogram data is generated. In at least one embodiment, the method 600 includes performing (612) a component analysis (such as, PCA) based at least in part on the entropy spectra to select a subset of the entropy spectra. In at least one embodiment, step 612 may be performed to provide classification of at least a training subset or a first part of the entropy data, such as described in reference to FIGS. 3C and 4A-4C. In at least one embodiment, the method 600 includes determining (614) casing loadings or cement bonding associated with the casing structure based at least in part on a subset of the entropy spectra.

In at least one embodiment, the method 600 includes a further step or substep for determining a covariance matrix from the entropy spectra. The principal component analysis (PCA) is performed to extract feature components from the entropy spectra. Scores of the feature components are determined, where the scores are associated with a depth and a sector within a well where the signal is received. The method 600 includes a further step or substep for selecting individual ones of the scores for which underlying ones of the feature components are to be associated within a dataset that is the subset of the entropy spectra. Further, the method 600 includes a further step or substep for performing clustering within the dataset for a subset of the feature components. The clustering identifies cluster for the subset of the feature components that is associated with the casing loadings or cement bonding of the casing structure

The method 600 includes a further step or substep for the feature components to have one or more of bonding conditions of different actual or test casing or different entropy spectra of different sectors at one or more depths in an actual or test well. In at least one embodiment, the method 600 includes a further step or substep for determining the covariance matrix from a first part of the entropy spectra that is sampled from the entropy spectra. The method 600 includes a further step or substep for validating the first part of the entropy spectra with a second part of the entropy spectra based in part on the feature components.

In at least one embodiment, the method 600 includes a further step or substep for providing the well inspection tool as part of a production tubing. The production tubing is to be internally within a casing and the casing loadings or cement bonding are associated with cement located externally relative to the casing. The method 600 includes a further step or substep for providing the well inspection tool internally within a casing, where the casing loadings or cement bonding are associated with cement located externally relative to the casing.

In at least one embodiment, the method 600 includes a further step or substep for enabling the well inspection tool to comprise a transmitter and at least one receiver in a pitch-catch or pulse-echo configuration. The method 600 includes a further step or substep for transmitting, using the transmitter, a test signal from different circumferential positions of the well inspection tool towards walls of the casing structure. The method 600 includes a further step or substep for receiving, using the at least one receiver, the signals that are associated with the test signal. The signals travel in a perpendicular direction relative to an axis of the casing structure or travel along an axial direction relative to the axis of the casing structure.

In at least one embodiment, the method 600 includes a further step or substep for determining a power spectrum for at least one of the return signals. The method 600 includes a further step or substep for determining a sum of the power spectrum and for normalizing the power spectrum using the sum of the power spectrum. The method 600 includes a further step or substep for determining the entropy spectra based in part on a natural logarithm of the normalized power spectrum.

In at least one embodiment, the method 600 includes a further step or substep wherein the casing loadings comprise one of a flow path condition or different material loadings. The method 600 includes a further step or substep for providing the well inspection tool internally within a production tubing, where the production tubing is internally within a casing. A media is provided within the production tubing and within an annular space of the casing and the production tubing.

While techniques herein may be subject to modifications and alternative constructions, these variations are within spirit of present disclosure. As such, certain illustrated embodiments are shown in drawings and have been described above in detail, but these are not limiting disclosure to specific form or forms disclosed; and instead, cover all modifications, alternative constructions, and equivalents falling within spirit and scope of disclosure, as defined in appended claims.

When introducing elements of various embodiments of the present invention, the articles “a,” “an,” “the,” and “said” are intended to mean that there are one or more of the elements. The terms “comprising,” “including,” and “having” are intended to be inclusive and mean that there may be additional elements other than the listed elements. Any examples of operating parameters and/or environmental conditions are not exclusive of other parameters/conditions of the disclosed embodiments. Additionally, it should be understood that references to “one embodiment”, “an embodiment”, “certain embodiments,” or “other embodiments” of the present invention are not intended to be interpreted as excluding the existence of additional embodiments that also incorporate the recited features. Furthermore, reference to terms such as “above,” “below,” “upper”, “lower”, “side”, “front,” “back,” or other terms regarding orientation are made with reference to the illustrated embodiments and are not intended to be limiting or exclude other orientations.

Recitation of ranges of values herein are merely intended to serve as a shorthand method of referring individually to each separate value falling within range, unless otherwise indicated herein and each separate value is incorporated into specification as if it were individually recited herein. In at least one embodiment, use of a term, such as a set (for a set of items) or subset unless otherwise noted or contradicted by context, is understood to be nonempty collection including one or more members. Further, unless otherwise noted or contradicted by context, term subset of a corresponding set does not necessarily denote a proper subset of corresponding set, but subset and corresponding set may be equal.

Conjunctive language, such as phrases of form, at least one of A, B, and C, or at least one of A, B and C, unless specifically stated otherwise or otherwise clearly contradicted by context, is otherwise understood with context as used in general to present that an item, term, etc., may be either A or B or C, or any nonempty subset of set of A and B and C. In at least one embodiment of a set having three members, conjunctive phrases, such as at least one of A, B, and C and at least one of A, B and C refer to any of following sets: {A}, {B}, {C}, {A, B}, {A, C}, {B, C}, {A, B, C}. Thus, such conjunctive language is not generally intended to imply that certain embodiments require at least one of A, at least one of B and at least one of C each to be present. In addition, unless otherwise noted or contradicted by context, terms such as plurality, indicates a state of being plural (such as, a plurality of items indicates multiple items). In at least one embodiment, a number of items in a plurality is at least two but can be more when so indicated either explicitly or by context. Further, unless stated otherwise or otherwise clear from context, phrases such as based on means based at least in part on and not based solely on.

Operations of methods in FIG. 6 and the algorithm herein can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. In at least one embodiment, a method includes processes such as those processes described herein (or variations and/or combinations thereof) that may be performed under control of one or more computer systems configured with executable instructions and that may be implemented as code (e.g., executable instructions, one or more computer programs or one or more applications) executing collectively or exclusively on one or more processors, by hardware or combinations thereof.

In at least one embodiment, such code may be stored on a computer-readable storage medium. In at least one embodiment, such code may be a computer program having instructions executable by one or more processors. In at least one embodiment, a computer-readable storage medium is a non-transitory computer-readable storage medium that excludes transitory signals (such as a propagating transient electric or electromagnetic transmission) but includes non-transitory data storage circuitry (such as buffers, cache, and queues) within transceivers of transitory signals. In at least one embodiment, code (such as executable code or source code) is stored on a set of one or more non-transitory computer-readable storage media having stored thereon executable instructions (or other memory to store executable instructions) that, when executed (such as a result of being executed) by one or more processors of a computer system, cause computer system to perform operations described herein.

In at least one embodiment, a set of non-transitory computer-readable storage media includes multiple non-transitory computer-readable storage media and one or more of individual non-transitory storage media of multiple non-transitory computer-readable storage media lack all of code while multiple non-transitory computer-readable storage media collectively store all of code. In at least one embodiment, executable instructions are executed such that different instructions are executed by different processors—in at least one embodiment, a non-transitory computer-readable storage medium store instructions and a main central processing unit (CPU) executes some of instructions while other processing units execute other instructions. In at least one embodiment, different components of a computer system have separate processors and different processors execute different subsets of instructions.

In at least one embodiment, computer systems are configured to implement one or more services that singly or collectively perform operations of processes described herein and such computer systems are configured with applicable hardware and/or software that enable performance of operations. In at least one embodiment, a computer system that implements at least one embodiment of present disclosure is a single device or is a distributed computer system having multiple devices that operate differently such that distributed computer system performs operations described herein and such that a single device does not perform all operations.

In at least one embodiment, even though the above discussion provides at least one embodiment having implementations of described techniques, other architectures may be used to implement described functionality, and are intended to be within scope of this disclosure. In addition, although specific responsibilities may be distributed to components and processes, they are defined above for purposes of discussion, and various functions and responsibilities might be distributed and divided in different ways, depending on circumstances.

In at least one embodiment, although subject matter has been described in language specific to structures and/or methods or processes, it is to be understood that subject matter claimed in appended claims is not limited to specific structures or methods described. Instead, specific structures or methods are disclosed as example forms of how a claim may be implemented.

From all the above, a person of ordinary skill would readily understand that the tool of the present disclosure provides numerous technical and commercial advantages and can be used in a variety of applications. Various embodiments may be combined or modified based in part on the present disclosure, which is readily understood to support such combination and modifications to achieve the benefits described above.

It should be appreciated that embodiments herein may utilize one or more values that may be experimentally determined or correlated to certain performance characteristics based on operating conditions under similar or different conditions. The present disclosure described herein, therefore, is well adapted to carry out the objects and attain the ends and advantages mentioned, as well as others inherent therein. While a presently preferred embodiment of the disclosure has been given for purposes of disclosure, numerous changes exist in the details of procedures for accomplishing the desired results. These and other similar modifications will readily suggest themselves to those skilled in the art and are intended to be encompassed within the spirit of the present disclosure disclosed herein and the scope of the appended claims.

Claims

1. A well inspection method, comprising: receiving signals from a casing structure using a well inspection tool;performing, using at least one processor associated with the well inspection tool, a Fast Fourier Transform (FFT) on the signals to generate spectrogram data;determining entropy spectra from the spectrogram data;performing a component analysis based at least in part on the entropy spectra to select a subset of the entropy spectra; anddetermining casing loadings or cement bonding associated with the casing structure based at least in part on the subset of the entropy spectra.

2. The well inspection method of claim 1, further comprising: determining a covariance matrix from the entropy spectra;performing the component analysis to extract feature components from the entropy spectra;determining scores of the feature components of the entropy spectra, wherein scores are associated with a depth and a sector within a well where the signals are received;selecting individual ones of the scores for which underlying ones of the feature components are to be associated together to be within a dataset; andperforming clustering within the dataset to identify at least a cluster for the subset of the feature components, the cluster being associated with the casing loadings or cement bonding of the casing structure.

3. The well inspection method of claim 2, wherein the feature components comprise one or more of bonding conditions of different actual or test casing or of different entropy spectra of different sectors that are at one or more depths in an actual or test well.

4. The well inspection method of claim 2, further comprising: determining the covariance matrix from a first part of the entropy spectra that is sampled from the entropy spectra; andvalidating the first part of the entropy spectra with a second part of the entropy spectra based in part on the feature components.

5. The well inspection method of claim 1, further comprising: providing the well inspection tool as part of a production tubing, the production tubing being internally within the casing structure, wherein the casing loadings or cement bonding are associated with cement located externally relative to the casing structure.

6. The well inspection method of claim 1, further comprising: providing the well inspection tool internally within the casing structure, wherein the casing loadings or cement bonding are associated with cement located externally relative to the casing structure.

7. The well inspection method of claim 1, further comprising: enabling the well inspection tool to comprise a transmitter and at least one receiver in a pitch-catch or pulse-echo configuration;transmitting, using the transmitter, a test signal from different circumferential positions of the well inspection tool towards walls of the casing structure; andreceiving, using the at least one receiver, the signals that are associated with the test signal, wherein the signals travel in a perpendicular direction relative to an axis of the casing structure or travel along an axial direction relative to the axis of the casing structure.

8. The well inspection method of claim 1, further comprising: determining a power spectrum for at least one of the signals;determining a sum of the power spectrum;normalizing the power spectrum using the sum of the power spectrum; anddetermining the entropy spectra based in part on a natural logarithm of the normalized power spectrum.

9. The well inspection method of claim 1, wherein the casing loadings comprise one of a flow path condition or different material loadings.

10. The well inspection method of claim 1, further comprising: providing the well inspection tool internally within a production tubing, the production tubing being internally within the casing structure, wherein a media is provided within the production tubing and within an annular space of the casing structure and the production tubing.

11. A system for well inspection, comprising: a transmitter of a well inspection tool to transmit a test signal into a casing structure;one or more receivers of the well inspection tool to receive signals from the casing structure to the well inspection tool; andat least one processor and memory comprising instructions that when executed by the at least one processor enable the system to: perform a Fast Fourier Transform (FFT) on the signals to generate spectrogram data;determine entropy spectra from the spectrogram data;perform a component analysis based at least in part on the entropy spectra to select a subset of the entropy spectra; anddetermine casing loadings or cement bonding associated with the casing structure based at least in part on the subset of the entropy spectra.

12. The system of claim 11, wherein the memory comprising the instructions that when executed by the at least one processor further cause the system to: determine a covariance matrix from the entropy spectra;perform the component analysis to extract feature components from the entropy spectra;determine scores of the feature components of the entropy spectra, wherein scores are associated with a depth and a sector within a well where the signals are received;select individual ones of the scores for which underlying ones of the feature components are to be associated together to be within a dataset; andperform clustering within the dataset to identify at least a cluster for the subset of the feature components, the cluster being associated with the casing loadings or cement bonding of the casing structure.

13. The system of claim 12, wherein the feature components comprise one or more of bonding conditions of different actual or test casing or of different entropy spectra of different sectors that are at one or more depths in an actual or test well.

14. The system of claim 12, wherein the memory comprising the instructions that when executed by the at least one processor further cause the system to: determine the covariance matrix from a first part of the entropy spectra that is sampled from the entropy spectra; andvalidate the first part of the entropy spectra with a second part of the entropy spectra based in part on the feature components.

15. The system of claim 11, wherein the well inspection tool is part of a production tubing, the production tubing being internally within the casing structure, wherein the casing loadings or cement bonding are associated with cement located externally relative to the casing structure.

16. The system of claim 11, wherein the well inspection tool is to be internally within the casing structure, wherein the casing loadings or cement bonding are associated with cement located externally relative to the casing structure.

17. The system of claim 11, wherein the well inspection tool comprises the transmitter and the one or more receivers in a pitch-catch or pulse-echo configuration.

18. The system of claim 11, wherein the test signal is transmitted from different circumferential positions of the well inspection tool towards walls of the casing structure, and wherein the signals received are associated with the test signal, wherein the signals travel in a perpendicular direction relative to an axis of the casing structure or travel along an axial direction relative to the axis of the casing structure.

19. The system of claim 11, wherein the memory comprising the instructions that when executed by the at least one processor further cause the system to: determine a power spectrum for at least one of the return signals;determine a sum of the power spectrum;normalize the power spectrum using the sum of the power spectrum; anddetermine the entropy spectra based in part on a natural logarithm of the normalized power spectrum.

20. The system of claim 11, wherein the well inspection tool is to be internally within a production tubing, the production tubing to be internally within the casing structure, wherein a media is to be within the production tubing and within an annular space of the casing structure and the production tubing.