• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    Hybrid / time domain modeling can be used to calculate the synthetic transient response of an eddy current pulse for use in an efficient and high dynamic range inversion process for current pipe inspection tools pulsed Foucault. In accordance with some embodiments, the frequency domain response signals are calculated for a plurality of frequencies and transformed, by transforming the frequency-domain-time domain, into a first synthetic transient response signal. A time limit associated with the occurrence of spurious oscillations in the first synthetic transient response signal is then automatically determined, and a second synthetic transient response signal starting at the determined time limit is determined directly with a time domain digital technique. . A portion of the first synthetic transient response signal terminating at the time limit may be combined with the second synthetic transient response signal.
    公开号:FR3057901A1
    申请号:FR1758865
    申请日:2017-09-26
    公开日:2018-04-27
    发明作者:Ahmed E. FOUDA；Burkay Donderici
    申请人:Halliburton Energy Services Inc；
    IPC主号:
    专利说明:

    (57) Hybrid frequency / time domain modeling can be used to calculate the synthetic transient response of an eddy current pulse for use in an efficient, high dynamic range inversion process for inspection tools. pulsed eddy current pipe. According to certain embodiments, the frequency domain response signals are calculated for a plurality of frequencies and transformed, by the transformation of the frequency domain-into time domain, into a first synthetic transient response signal. A time limit associated with the appearance of parasitic oscillations in the first synthetic transient response signal is then automatically determined, and a second synthetic transient response signal starting at the determined time limit is determined directly with a digital time domain technique. . A portion of the first synthetic transient response signal ending at the time limit may be combined with the second synthetic transient response signal.
    HIGH DYNAMIC RANGE INVERSION FOR PIPE INSPECTION
    BACKGROUND In gas and oilfield operations, the early detection of any corrosion in the well casing and / or production columns is important to ensure the integrity and safety of the gas and oil well. State-of-the-art methods for detecting corrosion at the bottom of the well generally involve the descent of a pipe inspection tool into the production column. Different types of pipe inspection tools include mechanical feelers, ultrasonic acoustic tools, cameras, electromagnetic flux dispersing tools, and electromagnetic induction tools. Of these tools, only electromagnetic induction tools can be used to detect corrosion in the outer casings beyond where the tool is lowered.
    Existing electromagnetic induction pipe inspection tools generally include at least one transmitter coil and at least one receiver coil. The transmitter produces a time-varying primary field which induces eddy currents inside the metal pipes, and the receiver records secondary fields generated by the pipes. These secondary fields contain information regarding the electrical properties and metallic content of the pipes, and can be reversed by any corrosion or loss in the metallic content of the pipes. Electromagnetic induction tools can be frequency domain tools that operate at a set of discrete frequencies (with higher frequencies to inspect internal pipes and lower frequencies to inspect external pipes). Furthermore, electromagnetic induction tools can operate in the time domain by transmitting transient pulses and by measuring the decomposition response versus time (in which the early times correspond to the internal pipes and the posterior times correspond to the external pipes); these tools .5 are called tools for inspecting pulsed eddy current pipes.
    Regardless of the type of pipe inspection tools, a model-based inversion is generally used to estimate the physical and electrical properties of each pipe in the casing from the measured responses. Model-based inversion uses an advanced computer model that simulates the tool and well pipes and includes various adjustable parameters (such as the thickness and relative magnetic permeability of each pipe, or other pipe parameters) , and involves an iterative procedure to find values of adjustable model parameters that minimize the mismatch between the measurements and the synthetic data produced from the model. The advanced model can be invoked on the fly inside the minimization process, or, moreover, the synthetic data can be precalculated for different combinations of values of the parameters of the model and stored in a database. In both cases, an effective reversal process depends on a quick and precise model.
    BRIEF DESCRIPTION OF THE FIGURES [0004] Figure 1 is a diagram of a pipe inspection system deployed in an exemplary borehole environment, in accordance with the various embodiments.
    Figures 2A and 2B are cross-sectional and side views, respectively, of an exemplary configuration of an electromagnetic induction pipe inspection tool lowered into a pipe, in accordance with the various embodiments.
    Figure 3 is a graph of an example of a time-dependent response for a pulsed eddy current, in accordance with the various embodiments.
    Figure 4 is a flow diagram of a method for inspecting the pipe using pulsed eddy currents, in accordance with the various embodiments.
    Figure 5 is a flow diagram of a method for calculating a hybrid synthetic transient response signal, in accordance with the various embodiments.
    FIG. 6 is a graph of the percentage of error in the synthetic impulse response signal resulting from the amplification of the conductivity of the fluid in the borehole to various increased values, in accordance with the various embodiments.
    Figures 7A is a graph of a true synthetic impulse response and the impulse responses set to zero corresponding to various times, in accordance with various embodiments.
    Figure 7B is a graph of the percentage of impulse response errors set at zero in Figure 7A relative to the true impulse response, in accordance with the various embodiments.
    Figures 8 are a flow diagram of an exemplary computer system for calculating hybrid synthetic transient response signals in accordance with the various embodiments.
    FIGS. 9A-9C are graphs of the synthetic impulse response signals based on the time domain and based on the frequency domain, a hybrid impulse response signal calculated from it in accordance with the various embodiments, and corresponding frequency domain based hybrid impulse response signals, respectively, for an example pipe configuration including a pipe.
    FIGS. 10A-10C are graphs of synthetic impulse response signals based on the time domain and based on the frequency domain, a hybrid impulse response signal calculated from it in accordance with the various embodiments, and the corresponding frequency domain-based hybrid impulse response signals, respectively, for an example pipe configuration comprising two nested pipes.
    Figures 11A-11C are graphs of synthetic impulse response signals based on the time domain and based on the frequency domain, a hybrid impulse response signal calculated from it in accordance with the various embodiments , and the corresponding frequency domain based hybrid impulse response signals, respectively, for an example pipe configuration comprising three nested pipes.
    FIGS. 12A-12C are graphs of synthetic impulse response signals based on the time domain and based on the frequency domain, a hybrid impulse response signal calculated from it in accordance with the various embodiments, and the corresponding frequency domain-based hybrid impulse response signals, respectively, for an example pipe configuration comprising four nested pipes.
    DESCRIPTION The present disclosure describes an efficient, high dynamic range inversion method for pulsed eddy current (i.e., time domain) pipe inspection tools which is based on modeling. of the hybrid frequency / time domain to calculate the synthetic transient responses to an eddy current pulse.
    By convention, there are two approaches for calculating a synthetic transient response signal for a pulsed eddy current tool: a frequency-to-time domain conversion and a direct calculation of the time domain. In the first approach, a frequency domain solver (such as, for example, a semi-analytical method or a finite element method (FEM)) is used to calculate the transient response signal in the frequency domain, at from which the time domain transient response signal is then calculated using a frequency-to-time domain transformation (such as, for example, an inverse Fourrer transform). With this approach, the time domain signal is generally not stable, since it undergoes parasitic oscillations, for time delays, especially in pipe configurations with low metal content, e.g., with a small number of concentric pipes, pipes with thin walls or pipes with low real permeability. (Here, the real permeability is the equivalent permeability of an entire space which will produce a signal with the same signal quantity as a pipe which has a certain value of relative permeability. The real permeability is a function of the winding geometry, the geometry of the pipe and the relative permeability of the pipe while the relative permeability of the pipe is the property of the material and does not depend on any geometry). In the second approach, a time domain solver (such as, for example, a finite difference time domain method (FDTD)) is used to directly calculate the transient time domain response signal. This approach can be excessively slow since very small time steps are required to obtain stable responses at early times.
    In various embodiments described here, a hybrid approach is used to solve the problems from which conventional approaches suffer when they are individually used. In this type of hybrid approach, first, the transient response signal of the frequency domain is calculated and transformed into a transient response signal of the time domain (e.g., conventionally) to provide a transient response of the domain synthetic temporal which is precise at early times. The time limit at which the synthetic time domain transient response signal resulting from the frequency-to-time domain transformation begins to fail (i.e., which is associated with the onset of parasitic oscillation ) is automatically detected. A time domain solver is then executed for the portion of the synthetic time domain response signal starting at the detected time limit to provide an accurate response to late times; the time step used in the time domain solver can be calculated based on the time limit. The part of the time domain signal calculated from the frequency domain signal and ending at the time limit and the part of the time domain signal directly calculated in the time domain and starting at the time limit are then combined. in a global synthetic time domain response signal which is precise for both early and late time.
    The hybrid method generally provides a greater dynamic range of the synthetic response signal (i.e., a greater ratio between the longest time and the shortest time at which the response signal synthetic is precise) rather than purely the frequency domain solution, and it is faster than the purely time domain solution since larger time steps can be used in the time domain solver, which is advantageous in various applications. For example, a high dynamic range method allows further processing of pulsed eddy current response signals, e.g., to resolve parameters such as tool or pipe eccentricity, or to simultaneously estimate the magnetic permeability and thickness of each pipe. It is well established in inversion theory that when the number of unknown (i.e., adjustable) parameters in the tool and pipe configuration model increases, more independent measurements are generally necessary to resolve these strangers. The increase in the dynamic range of the synthetic response signal provides more independent measurements (due to a longer response with more independent time samples), thus allowing the use of a large number of adjustable parameters in the model. An efficient high dynamic range process also creates a more precise planner. A planner can be used to optimize various parameters of the operation of the tool and the inversion algorithm, such as the duration of the measured transient response signal to be used in the inversion, or the weights applied in the inversion to different time samples and different transmitter / receiver networks, based on synthetic modeling of the sensitivity of the tool to a given pipe configuration. For precise planning, the model must allow the calculation of the transient response over the entire dynamic range measurable by the tool.
    The previous description will be more easily understood in the light of the following description and the attached figures.
    Figure 1 is a diagram of a pipe inspection system deployed in an exemplary borehole environment, in accordance with the various embodiments. The borehole 100 is illustrated during a wired logging operation, which is performed after completion of the drilling and removal of the drill string. As illustrated, the borehole 100 has been completed with a surface casing 102 and an intermediate casing 104, both cemented in place. In addition, a production column 106 has been installed in the borehole 100. The fluid of the ring fills the space between the production column 106 and the casing 104. The pipes 102, 104, 106 are generally manufactured in metal, usually steel. The space between the pipes can be filled with cement, mud and other fluids from the borehole.
    While three pipes 102, 104, 106 are illustrated in this example, the number of fitted pipes can generally vary, depending, for example, on the depth of the borehole 100. The methods of pipe inspection described herein are generally applicable pipe sets comprising any number of pipes, such as a single pipe or two or more nested pipes.
    Wired logging generally involves measuring the physical parameters of the borehole 100 and the surrounding formation, such as, in this case, the condition of the pipes 102, 104, 106, as a function of the depth at l inside the borehole 100. Pipe measurements can be made by lowering an electromagnetic induction pipe inspection tool 108 into the borehole 100, eg, on a wired line 110 wound around a winch 112 mounted on a logging truck. The wired line 110 is an electric cable which, in addition to lowering the tool 108 to the bottom of the well, can be used to transmit current to the tool 108 and to transmit control signals and / or data between the tool 108 and a logging unit 116 (implemented with an appropriate combination of hardware and / or software, eg, with a properly programmed general purpose computer including one or more processors and memory) located above the surface, eg ., inside the logging truck. In some embodiments, the tool 108 is lowered to the bottom of the region of interest and, subsequently, is brought up, e.g., at a substantially constant speed. During this upward journey, the tool 108 can take measurements on the pipes, either at different positions at which the tool 108 stops, or continuously when the pipes pass.
    For pulsed eddy current measurements, the tool 108 includes an electromagnetic transmitter, such as a coil carrying current 118 and an associated pilot and a measurement circuit 119, which can operate to emit electromagnetic pulses to the pipes 102, 104,
    106 to induce pulsed eddy currents therein. The electromagnetic waves diffracted from the casing trains 102, 104, 106 are received, in the form of a transient electromagnetic response signal, with a receiving coil 120 re-linked to the pilot and to a measurement circuit 119. Furthermore, changes to the The level of current flowing through the coil 118 which results from the response of the casing to the induced eddy current can be measured, allowing the transmitting coil 118 to also function as a receiving coil. In the following, no distinction is made between electromagnetic response signals acquired by a separate transmitter / receiver, since the processing methods described above are equally applicable to both.
    The tool 108 may also include a telemetry circuit 122 for transmitting the transient response signals measured to the logging unit 116 for processing and / or storage at the level thereof, or the memory (not -illustrée) for the storage of the response signals at the bottom of the well for a later recovery of the data once the tool brought to the surface. Optionally, the tool 108 may contain an analog or digital processing circuit 124 (e.g., an integrated microcontroller running appropriate software) which allows processing of the response signal measured at least partially at the bottom of the hole (e.g., before transmission to the surface). In any case, a log, i.e., a sequence of measurements correlated with the depths along the borehole 100 at which they are taken, will be produced. The computer, or other circuit used to process the measured transient electromagnetic response signals to obtain parameters of. pipes based on these, is hereinafter called the processing unit, regardless of whether it is contained inside the tool 108 in the form of processing circuit 124, whether it is supplied on a separate device such as a log unit 116, or in part in both. Collectively, the pipe inspection tool 108 and the processing unit (eg, 124 and / or 116) are referred to herein as the pipe inspection system.
    Instead of being transported to the bottom of the hole on a wired line, as described above, the pipe inspection tool 108 can be deployed using other types of transport, as will be easily understood by those skilled in the art. For example, the tool 108 can be lowered into the borehole by a smooth cable (a solid mechanical cable which generally does not allow the transmission of current and signal), and can include a battery or other independent current supply as well as a memory for storing the measurements until the tool 108 is brought back to the surface and the data recovered. Alternative means of transportation include, for example, a coiled tube, a downhole tractor, or a drill stand (eg, used as part of a tool train inside or near the downhole module during logging / measurement operations during drilling).
    Figures 2A and 2B further illustrate, in cross-sectional and side views, respectively, an example, of a configuration of an electromagnetic induction tool 108 lowered into a pipe 200 (e.g., corresponding to the column of production 106). The example of the tool 108 comprises one (as illustrated) or several networks of coaxial emitting coils 202 and receiving coils 204 with given axial lengths, and possibly a magnetic core 206 of a material with high permeability, such as the ferrite. A non-magnetic tool housing 208 may contain the coils 202, 204 and the core 206, as well as any associated circuit (such as, for example, a pilot and a measurement circuit 119, a telemetry circuit 122 and a processing circuit 124).
    FIG. 3 is a graph of the time-dependent response to a pulsed eddy current of step function, for example, measured by the receiving coil 120. During the times t <τ, an alternating current passing to through the emitting coil 118 generates an alternating primary magnetic field, which induces alternating eddy currents in the pipe or pipes surrounding the emitting coil 118. These eddy currents, in turn, create an alternating secondary magnetic field, which induces a voltage through or a current through the take-up coil 120. A constant amplitude of the current through the take-up coil 118 will produce the illustrated constant amplitude of the current in the take-up coil 120. At t = τ, the current through the send coil 118 is off (e.g., according to an approximate step function), resulting in the formation of an eddy current pulse which results in the formation of cha decaying transient secondary magnetic mps reflected in a transient decomposition response signal 300 (eg, an "impulse response" signal) measured at the receiving coil 120. The power of this transient response signal 300 at different times is sensitive to the parameters of different pipes within a set of multiple nested pipes.
    For example, the early signal is more sensitive to the innermost pipes, while the late signal is sensitive to both the inner and outer pipes. Therefore, the transient response signal 300 can be subdivided into multiple time slices (indicated by dotted lines), and sampled and inverted for different time slices to estimate the parameters of the different pipes.
    Figure 4 is a flow diagram of a method 400 for inspecting the pipe using pulsed eddy currents, in accordance with the various embodiments. The method includes placing an electromagnetic induction tool (such as tool 108) in a borehole (action 402) and acquiring a log by inducing pulsed eddy currents and measuring the signal transient response to multiple depths within the borehole (action 404), eg, at regular depth intervals when the tool is lowered into or raised from the borehole. In addition, a computer model of the tool and the pipe configuration of the borehole, which may include a single pipe or a set of multiple nested pipes, is created (action 406). The computer model is based on a priori knowledge of the configuration of the pipe (such as the number of pipes, their external diameter, nominal thickness and electrical conductivity), and includes one or more parameters of adjustable pipes whose values must be determined during the subsequent mechanical inversion process to minimize a mismatch between the measured transient response signal and a synthetic transient response signal calculated based on the model, as described in detail below. (The term "minimize" in this context describes a reduction in mismatch to a value below a defined threshold, eg, measured in terms of a cost function, and is not intended to imply that an absolute minimum is reached). The adjustable parameters can include, for example, the relative magnetic permeabilities of the pipes, which are generally not known a priori with precision and can vary significantly from one casing section to another, and the thickness of the pipe, which may deviate from the nominal thickness due, for example, to corrosion. The relative permeabilities of the pipes used in the casing of the well are generally between 1 to 120.
    In some embodiments, the log is processed in multiple parts, eg, each corresponding to a casing section, by stacking the log points of this part in an input data buffer. The log, or each buffered part of the log, can be calibrated (action 408) to compensate for any discrepancy between the actual tool and the computer model of the tool. In various embodiments, the previous knowledge of the nominal thickness of the pipes is used to determine the calibration constants of the tool. The calibration constants can be of real value or of complex value.
    Each transmitter, receiver or transmitter / receiver can have an associated constant. An average magnetic permeability for pipes- (or sections thereof) can also be determined by matching the transient response measured for a nominal pipe section (i.e., a pipe section that can be assumed as not having undergone corrosion) to the synthetic transient response calculated from the computer model of this section. The determined calibration constants are then applied to the entire portion of the log in the input data buffer, and the calculated average magnetic permeability is used as a starting point in the inversion process. In some embodiments, a resolution improvement technique which involves deconvolution of the impulse response of the tool is applied to sharpen the measured transient response signal (action 410).
    The calibrated log, with improved resolution (or part of the log) is fed point by point (in action 412) in the digital inversion process. (A "log point" refers to the transient response signal measured at a given depth). The digital inversion (action 414) is used to transform the measured transient response signals into the value of the adjustable pipe parameter (s) which minimize a mismatch between the measured and synthetic response signals. In order to calculate the synthetic response signal, an advanced model of eddy current induction and creation of the secondary magnetic field and the resulting response signal at the receiver is used. As illustrated, the digital inversion (action 414) may include an iterative process in which, during each iteration, the values for the pipe parameters adjustable to the depth at which the fed log point was acquired are defined (action 416) (starting, in the first iteration, with an initial assumption, which can be based on the nominal pipe parameters and can include the mean permeability as estimated in action 404), and the synthetic transient response signal calculated from the computer model with the values of the defined parameter is obtained (action 418). The synthetic response signal for a given set of pipe parameter values can be calculated on the fly during the respective iteration. Furthermore, synthetic response signals can be precomputed for multiple sets of pipe parameter values, and the synthetic response signal precomputed for the given set of pipe parameter values simply retrieved during the iterative process. Whether precomputed or calculated on the fly, the synthetic transient response signal (from action 418) is compared to the measured transient response signal (entered at 412) to determine the mismatch, e.g. terms of a cost function (calculated in action 420). The parameter values are adjusted iteratively until the convergence of the measured and synthetic transient response signals (as determined in action 422) as reflected, for example, in the value of the cost function which falls below a specified convergence threshold. The values of the pipe parameters which minimize the mismatch (in the sense that they have given the convergence of the measured and synthetic response signals) can then be returned. The digital inversion (action 414) is repeated for subsequent log points until the last log point has been processed (as determined in action 424).
    In some embodiments, the cost function F used to quantify the mismatch between the measured and synthetic transient response signals is formed from the linear combination of a mismatch function and a stabilization function ( also called the accrual term), e.g., as follows:
    s (x) ^ m> < --1 W ac [ m
    F 1 ^ 4 / χ X Xnom) There, X is a vector with N unknowns corresponding to the adjustable parameters of the model (eg, x = [ΐγ, ..., ίι ^ ρ, μγ, ..., μ ΝΡ ], which include the thickness and relative magnetic permeabilities of NP pipes), x name is a vector of the nominal parameters of the model, m is a vector of measurements with real values M at different time slices and receivers (with M = N Rx · N T s, where N Rx is the number of receivers N TS is the number of time slots in the transient response signal), s (x) is the corresponding M-value vector of the synthetic transient response signal , Wfl is an MXM matrix of measurement weights, W cal is an MXM diagonal matrix of real-time calibration constants, and W $ is a MXM diagonal matrix of regularization weights. The mismatch function is formed as the normalized L2 squared of the weighted difference between the calibrated measures W cal m and the data predicted by the synthetic model s (x), and it is normalized by the number of measures M to make the function of cost F independent of the number of measures. The stabilization function is formed like the normalized L1 of the weighted difference between the parameters x of the model and the nominal parameters of the model Xnom · Οθ in this way, the regularization penalizes the large variations of the thicknesses starting from the nominal thicknesses, and helps to attenuate the problem of non-uniqueness that could arise in cases involving many pipes (eg, or more).
    Figure 5 is a flow diagram of a method 500 for calculating a synthetic transient response signal, in accordance with the various embodiments. As mentioned with reference to FIG. 4, this method 500 can be used to calculate the synthetic transient response signals on the fly during the iterative digital inversion (action 418), or to precompute the synthetic response signals for a subsequent recovery of one of the precomputed signals during each iteration. Method 500 takes a set of pipe parameter values (e.g., pipe diameters, thicknesses and actual magnetic permeabilities) as input, and first includes performing advanced modeling of the frequency domain for calculating frequency domain response signals for a plurality of frequencies (action 502) and using the frequency domain-to-time domain transformation to transform the plurality of calculated frequency domain response signals into a first synthetic transient response signal of the time domain (action 504). From the first synthetic transient response signal of the time domain, a time limit associated with the occurrence of parasitic oscillations inside the signal is automatically determined (action 506). Advanced time domain modeling (eg, using an FDTD simulation or other digital time domain technique) is ultimately used to calculate a second synthetic transient response signal starting directly at the time limit determined (action 508). The time step and the spatial discretization used in the advanced modeling of the time domain can be determined, in action 510 (before action 508), based on the time limit; in some embodiments, this involves amplifying the electrical permittivity of the material filling the ring between the pipes to allow for a larger time step (as explained in more detail below). According to the calculation of the time domain of the second synthetic transient response signal, part of the synthetic transient response signal ending at the level of the first limit is associated with the second synthetic transient response signal to give a global synthetic transient response signal (action 512). In some embodiments, the first and second transient response signals, and therefore also the combined transient response signal, are impulse responses, which correspond to the signal acquired at the receiver in response to a brief pulse from the transmitter. In this case, the combined transient response can be integrated (action 514) to obtain a synthetic impulse response signal, which captures the way in which the reception signal breaks down over time following an abrupt power-off (in stages) of the 'transmitter. In another embodiment, the first and second transient response signals are from the transitional response signals.
    With more details, a frequency domain solver, such as a semi-analytical method or a FEM, is used to calculate the response in the frequency domain (actions 502), and then a technique for transforming the frequency domain-in time domain, such as the inverse Fourier transform, is used to calculate the response signal of the transient time domain (action 504). The time domain response signal produced suffers from instability, in the form of oscillations, and late times due to the finite precision of the frequency-to-time domain transformation technique. These instabilities tend to be more pronounced in scenarios involving few pipes (eg, only one or two pipes) and / or pipes with low real permeability. The time when the time domain response signal obtained by transforming the frequency-into time domain begins to oscillate (i.e., the time boundary between the stable and unstable parts of the signal) is automatically detected, in action 506, and hereinafter designated by t max (indicating the maximum time until which the response of the calculated time domain can be considered to be precise, and after which the response becomes unstable). The time limit t max can be determined algorithmically, e.g., by examining a synthetically generated impulse response IR, which should, theoretically, be a function of monotonic decomposition of time, and the choice of the earliest time for which a increase in the impulse response can be observed, so that IR (t max + 5t)> IR (t max ), in which 5t is a small fixed time step.
    Following the determination of the time limit t max (action 506), a time domain solver, such as an FDTD, a finite element time domain (FETD), or another method based on the time domain which is precise and provides a high dynamic range at late times, is used · to directly calculate the late time parts of the response (ie, the part for t> t max ). To guarantee numerical stability, the time step used by the solver is chosen so as not to exceed a maximum value determined by the well-known stability condition of Current:
    Vêr fi, i, i C ° vjàz 2 Ay 2 Δζ 2 in which c 0 is the speed of light in free space, e r is the relative permittivity of the material which fills the ring between the pipes, and Δχ, Δγ and Δζ are the spatial discretizations in the directions x, y, and z, suitably chosen to solve the geometrical details of the modeled problem (eg, detail the configuration of the tool and the pipe) and / or the principal components of the spectrum of wavelengths propagating in the digital grid. In accordance with the various embodiments, the fact that the time domain solver is executed only for times greater than t max is exploited to increase the spatial discretization and / or to amplify the relative permittivity in order to allow larger time steps always satisfying the Current condition.
    In a dissipation medium, the spatial discretization can be chosen based on the frequency spectrum of the signal, for example, so that the depth of the skin (inside the pipes) corresponding to the frequency the highest in the spectrum of interest is sampled by at least ten grid points. Using the FDTD to calculate the late time portion of the transient response signal, c. that is, the part of the signal for t> tn, ax , the maximum frequency of interest is inversely proportional to t max . Furthermore, for a given take-up coil and pipe, a measurement of the frequency domain ^ frequency (f) at frequency f is, with a good approximation, proportional to a measurement of the time domain (1 / f) at time t = l / f, due to the highly dispersive nature of the problem. Therefore, the maximum frequency of interest can be taken as l / t max (or a substantially similar value, such as any value between 1 / (2t max ) and 2 / t max ). From f max , in combination with the material properties of the pipes (such as magnetic permeability, electrical resistivity and electrical permittivity), the depth of the skin can be calculated. The calculated skin depth, in turn, determines the spatial discretization. A smaller maximum frequency of interest corresponds to a greater minimum wavelength of interest and, consequently, a greater skin depth and a greater spatial discretization. Thus, the use of a frequency domain solver to calculate the early time portion of the response (i.e., the response for t <t max ) can significantly speed up the FDTD simulation by allowing discretization larger spatial and, therefore, a larger time step. In addition, when the simulation progresses towards late times, the maximum frequency of interest decreases. In certain embodiments, the spatial discretization and the time step are adaptively increased, one or more times during the calculation of the second synthetic response signal, based on this decrease in the maximum frequency of interest for late times, which accelerates all the more the FDTD simulation.
    As is apparent from the current stability conditions described above, the time step for the FDTD simulation, and, thus, the total simulation time for calculating the response for a given duration, is proportional to the square root of the relative permittivity G r . It is therefore desirable to amplify the permittivity to reduce the simulation time, as long as this can be done without impacting the physics of the problem. From Maxwell's equation (in particular, Ampère's law) the complex electrical permittivity of the fluid in the borehole filling the annular spaces between the pipes is given by jù) in which e 0 is the electrical permittivity in l 'free space (8. 85 x 10' 12 F / m), e r is the permittivity Te IêfffLv e ~ T3tf fluid of the thunderstorm (e.g. mud), ω is the angular frequency (ω = 2π / ), and Ob_fi is the electrical conductivity of the borehole fluid. For the frequency range generally contained in the transient response to pulsed eddy currents in the application scenarios considered here, the second imaginary term in the above-mentioned equation is generally much larger than the first term. Therefore, for the purposes of the FDTD simulation, the current value e r of the relative permeability can be amplified safely (ie, without affecting physics) to a "mathematical" value e a mpif. " E r (eg, e amp u f _> 10 e r ) as long as (b max e o e ampUf " a b fl , in which is the greatest angular frequency 2π used in the calculation of the response FDTD: For example, £ a mpiif can be chosen as:
    _ ^ bfltmax € amplif. ~ 2neo . 1Q 'which ensures that eampiif. does not exceed 10% of the imaginary second term within the frequency range of interest. With such an amplified value, the FDTD simulation will provide a digitally stable impulse response for times t> t max . Advantageously, the amplified permittivity accelerates the simulation of the time domain by allowing larger time steps. In certain embodiments, the efficiency of the simulation is further increased by implementing an adaptive amplification of the permittivity, c. i.e., Eampiif. is gradually increased (one or more times) as the modeling of the time domain progresses. For example, at certain times t> t max , be recalculated in accordance with _ Gbjit eamplif. - 2ji £ q . 1θ to use successively larger time steps at late times.
    As reflected in the preceding formula, the highest value that e boost can assume is mainly limited by the conductivity of the fluid due to the borehole. In many practical applications, this conductivity is rather low; eg, for petroleum-based mud, it can be as low as about 10 “ 3 S / m, which results in long simulation times (possibly prohibitively long). It has been observed, however, that the physics of the pipe inspection problem changes only slightly if an unrealistically high "mathematical" muddy conductivity (or more generally, a conductivity of the borehole fluid) is used instead. real conductivity. This is illustrated in the graphs of Figure 6, which demonstrate, for the mathematical conductivities of 10 2 S / m, 10 3 S / m, and 10 4 S / m, respectively, the percentage of error in the synthetic pulse signal with respect to a synthetic impulse response signal calculated for a typical, realistic conductivity value of 10 -2 S / m, reported over a period of time of 200 ms. The legend also indicates the total simulation time for the three reported cases, which decreases from 135 minutes for the smallest mathematical conductivity to 14 minutes for a mathematical conductivity which is two orders of magnitude higher. The simulation underlying Figure 6 is based on a configuration of four nested pipes with properties which are summarized in Table 1 below. As can be seen in Figure 6, a conductivity of 10 3 S / m ensures an error below 10% over the entire time interval of interest while obtaining a reasonable simulation time. Therefore, in various embodiments, the conductivity & b_fi of the borehole fluid is amplified, e.g., by a factor of at least 10 (and in some cases by 2 or more orders of magnitude), while maintaining a resulting error in the synthetic transient response signal below a tolerance threshold, eg, below 10%.
    By calculating the signal of the response of the transient time domain for the early times (t <t max ) coming from the response signals of the frequency domain p and for the late times (t> t max ) directly by a simulation of the time domain, the balance between precision and efficiency can be greatly improved, compared to using only modeling based on the frequency domain or only modeling based on the time domain. From the first transient response signal based on the frequency domain for t <t max and the second transient signal based on the time domain for t> tmax, a hybrid transient response signal can be calculated (action 512). If the hybrid response signal is an impulse response, it can also be integrated (action 514) to give the hybrid synthetic impulse response signal. The hybrid synthetic impulse response signal can then be compared to a measured impulse response signal to minimize a mismatch between the two in a digital optimization procedure, whereby the measured response is inverted to give values for the adjustable pipe parameters of the model.
    When calculating the hybrid impulse response signal by integrating the hybrid impulse response signal, any slight mismatch between the parts of early time (based on the frequency domain) and late time (based on the time domain) of the hybrid impulse response can cause a significant error in the impulse response at late times. To alleviate this problem, the fact that the impulse response in a dissipating medium decomposes monotonically over time (due to physics) can be used to anchor the impulse response signal to zero, at a sufficiently large time. as the maximum interest time in the synthetic signal, without compromising the accuracy of the solution in the interest interval time. (The maximum time of interest corresponds to the end of the time interval used to calculate the mismatch between the synthetic and measured transient response signals). FIGS. 7A and 7B illustrate such an anchoring at zero. The simulation underlying Figures 7A and 7B is based on a configuration of four nested pipes with properties which are summarized in Table 1 below. In FIG. 7A, an example of a real impulse response is illustrated with three curves for the impulse response anchored at 200 ms, 300 ms and 400 ms, respectively. The legend also indicates the respective simulation times, which increase for the late zero anchoring times. In Figure 7B, the percentage of error between the impulse responses true and anchored to zero and reported for the three cases for the time interval of interest, which is taken as the interval up to 200 ms. As can be seen, anchoring to zero at time t anchoring > 300 ms allows an error below 10% across the entire interval of interest. In some embodiments, t anchoring is chosen to be at least twice the maximum time of interest.
    Figure 8 is a flow diagram of an example processing unit 800 for calculating the hybrid synthetic transient response signals in accordance with the various embodiments. The processing unit 800 can be implemented, for example, in a surface logging unit 116 or a computer communicating with the surface logging unit, or in a processing circuit 124 integrated in the inspection tool of pipe 108. The processing unit 800 includes one or more processors 802 (eg, a conventional central processing unit (CPU), a graphics processing unit, or other components) designed to execute the stored software programs in memory yes 804 (which can be, for example, random access memory (RAM), read-only memory (ROM), flash memory, etc.). The processor (s) 802 can be single-core or multi-core processors. Multiple processor cores can be advantageously used, for example, to speed up the simulation of the time domain by executing it by parallelization in the cores. In some embodiments, the processing unit 800 also includes user input / output devices 806 (e.g., screen, keyboard, mouse, etc.), permanent data storage devices 808 ( including, e.g., computer readable semiconductor, optical and / or magnetic media such as hard disks, CD-ROMs, DVD-ROMs, etc., device interfaces 810 for communicating directly or indirectly with the pipe inspection tool 108, a network interface 814 which facilitates communication with other computer systems and / or data repositories, and a bus system (not shown) through which the other components of the processing unit 800 communicate. The processing unit 800 can, for example, be a general-purpose computer with appropriate software installed to implement the calculation methods described here. While only one unit is illustrated, the processing unit t Raitement 800 can also be distributed to multiple computers connected to each other through a wireless or wired network, such as a local area network or the Internet.
    The software programs stored in memory 804 include instructions executable by the processor to carry out the methods described here, and can be implemented in any of the various programming languages, e.g., without limitation, C, C ++, Object C, Pascal, Basic, Fortran, Matlab and Python. The memory 804 can also store, in addition to or as a part of the software programs, data on which the instructions operate, such as the calculation model 820 of the tool and the set of pipes, which includes one or more parameters of adjustable pipes, and transient response signals measured 822. The instructions themselves can be grouped into various functional modules, eg, for the purpose of reusing and sharing the functionality of certain modules with other modules. According to the illustrated embodiment, the modules include, for example, a calibration module 824 and a resolution improvement module 826 for preparing the logs for a later inversion (in accordance with actions 408, 410 of the method illustrated in the Figure 4); an inversion module 828 for numerically optimizing the adjustable pipe parameters of the calculation model 820 based on a cost function quantifying the mismatch between the measured and synthetic transient response signals; a frequency domain modeling module 830, a time domain modeling module 832 and a frequency domain-to-time domain transformation module 834 for calculating synthetic transient response signals based on the frequency domain and based on time domain (ie, the first and the second); and a hybrid response module 840 for combining the first and second transient response signals in the hybrid transient response signal. The hybrid response module 840 can include multiple sub-modules, such as a time limit module 842 for determining the time limit at which time domain modeling begins; a discretization module 8.44 for determining the time step and the spatial discretization used by the time domain modeling module 832, which can include the amplification of the electrical permittivity and (possibly) the conductivity of the fluid of the borehole entering in the calculation of the time step; a zero anchor module 846 for matching the parts based on the frequency domain and based on the time domain of the synthetic impulse response signals; and an integration module 848 for obtaining the synthetic impulse response signals from the impulse response signals.
    Of course, the calculation functionality described here can be grouped and organized in several different ways, the illustrated grouping being only one example. In addition, the various calculation modules illustrated in FIG. 8 must not all be part of the same software program or even be stored on the same computer. Instead, some groups of modules can operate independently of the others and provide data output that can be stored and subsequently supplied as input to the other modules. For example, modules 830, 832, 834, 836 for calculating synthetic response signals can be run on a computer to pre-calculate synthetic response signals for various combinations of pipe parameter values and to store response signals synthetics in a database. The database can later be accessed by the 828 inversion module, which can be stored on another computer, to find the synthetic response signal for a set of pipe parameter values used in a given iteration of the process. inversion. Furthermore, as will be readily understood by those skilled in the art, software programs implementing the methods described here (e.g., organized into functional modules as illustrated in Figure 8) can be stored, separately from any unit processing, on one or more computer-readable non-volatile media (such as, without limitation, a semiconductor, optical or magnetic storage medium), from which they can be downloaded into this (volatile) memory system d 'a processing unit for execution.
    In general, the processing unit which performs the calculation functionality described here (possibly organized into various functional modules) can be implemented with any suitable combination of hardware, firmware and / or software. For example, the processing unit can be permanently configured (eg, with a wired circuit) or temporarily configured (eg, programmed), or both in part, to implement the functionality described. A tangible entity configured, permanently and / or temporarily, to function in a certain way or to perform certain operations described here, is here called a "module implemented on hardware" or "a hardware module", and a hardware module using one or more processors is called a “processor implemented module”. Hardware modules can include, for example, a specialized circuit or logic that is permanently configured to perform certain operations, such as a programmable pre-broadcast network (FPGA), application-specific integrated circuit (ASIC), or another specialized processor · A hardware module can also include programmable logic or circuitry, such as a general-purpose processor, which is temporarily configured by software to perform certain operations. Looking at the exemplary embodiments in which the hardware modules are temporarily configured, the hardware modules collectively implementing the functionality described must not all coexist at the same time, but can be configured or instantiated at different times. For example, when a hardware module includes a versatile processor configured by software to implement a specialized module, the versatile processor can be configured for different versatile modules respectively at different times.
    Figures 9A-12C are graphs of the synthetic and hybrid impulse response signals based on the frequency domain and based on the time domain calculated in accordance with an embodiment for an example of pipe configuration. Figures 9A-9C illustrate the response signals for a configuration of four nested pipes, Figures 10A-10C for three nested pipes, Figures 11A-11C for two nested pipes, and Figures 12A-12C for a single pipe. The calculation model underlying the calculation of the response signals is summarized in Tables 1 and 2, Table 1 showing the pipe parameters of the four pipes and Table 2 showing the dimensions of the transmitter and receiver of the pipe inspection tool.
    Pipe Diameteroutside[inches] Thickness[inches] Conductivity[S / m] Permeabilityrelative 1 1 4 2 0.2 0.410 ζ 80 2 5 7 8 0.2 0.410 7 80 3 5 9 8 0.2 0.410 7 80 4 3 13 8 0.2 0.4 ΊΟ 7 80
    TABLE 1
    Diameter[inches] Length[Diameter] Transmitter 1 9 Receiver 0, 5 9
    TABLE 2
    The response signals based on the time domain were calculated by FDTD modeling for a spatial domain defined at 15 m in the radial direction 3057901 (measured from the center of the borehole) and 30 m in the axial direction. The spatial discretization grid was automatically generated by sampling the skin depth corresponding to the highest frequency of interest in each support by at least 10 grid points. The coils with a finite length have been modeled by placing fifteen independent transmitting and receiving coils also spaced within the extent of the extended coils. The tool was modeled on an iron-free tool (i.e., without a magnetic core), which presents a more difficult test case than a tool with a magnetic core since it usually results in a response with a higher dynamic range.
    Figures 9A, 10A, 11A and 12A demonstrate synthetic impulse response signals calculated with the time domain and frequency domain solvers. The frequency domain based impulse response signals for the four pipe configurations demonstrate late time instability, manifested by parasitic oscillations. The amplitude and spread of these oscillations increase with the decrease in the number of pipes. The FDTD solution makes it possible to have stable impulse response signals without oscillation for t> t max . In early times, the spatial discretization chosen is not adequate to resolve the shallow skin depth associated with the high frequency spectrum of the decomposition response, and, therefore, the signal based on FDTD demonstrates parasitic oscillations. Figures 9B, 10B, 11B, and 12B illustrate the hybrid impulse response signals, which are stable for both early and late times. The hybrid impulse response signals calculated from the integration impulse response signals are illustrated in Figures 9C, 10C, 11C and 12C, with, for comparison, the frequency domain based impulse response signals.
    The following numbered examples are illustrated embodiments.
    1. A method comprising: the use of an electromagnetic induction tool placed inside a set of pipes, inducing eddy currents pulsed in the set of pipes and the measurement of a signal time dependent transient response; creating a model for calculating the tool and the set of pipes, the model comprising one or more pipe parameters; and the use of digital inversion to determine the values of one or more pipe parameters which minimize a mismatch between the measured transient response signal and a synthetic transient response signal calculated based on the model, the calculating the synthetic transient response signal comprising calculating the frequency domain response signals for a plurality of frequencies, and using the frequency domain-time domain transformation to transform the. plurality of frequency domain response signals calculated as a first synthetic transient response signal, automatic determination of a time limit associated with the appearance of spurious oscillations in the first synthetic transient response signal, calculation of a second synthetic transient response signal starting at the time limit determined directly by a digital time domain technique, and combining part of the first synthetic transient response signal ending at the time limit with the second synthetic transient response signal .
    2. The method of Example 1, in which the first and second synthetic transient response signals are impulse response signals.
    3. The method of Example 1, in which the first and second synthetic transient response signals are impulse response signals, the method also comprising the integration of the first and second synthetic transient response signals for obtain a synthetic impulse response signal.
    4. The method of Example 3, also comprising the anchoring of the synthetic impulse response signal to zero at a point in time greater than a maximum interest time.
    5. The method of any of the preceding examples, in which the second synthetic transient response signal is calculated using spatial and temporal discretizations determined at least in part based on a maximum frequency of interest, the maximum frequency of interest being defined at least initially at substantially an inverse value of the time limit.
    6. The method of Example 5, also comprising increasing the spatial and temporal discretizations at least once during the calculation of the second synthetic transient response signal based on a reduction in the maximum frequency of interest for late times.
    7. The method of any of the preceding examples, in which the second synthetic transient response signal is calculated using a temporal discretization determined based at least in part on a relative permittivity of a material filling a space between the pipes, the relative permittivity being amplified by a factor of at least 10, the factor being determined based at least in part on the time limit and a conductivity of a borehole fluid inside the pipes.
    8. The method of Example 7, in which the conductivity of the fluid in the borehole is amplified by a factor of at least 10, an error in the second synthetic response signal resulting from the maintenance of the amplification below 10%.
    9. The method of Example 7 or Example 8, in which the factor is adjusted to a higher value at least once during the calculation of the second synthetic transient response signal.
    10. The method of any one of the preceding examples, in which the digital inversion comprises the adjustment by iteration of the values of one or more pipe parameters, the calculation of the synthetic transient response signal at from the adjusted parameters and the determination of the mismatch between the measured transient response signal and the synthetic transient response signal.
    11. The method of any one of the preceding examples, in which the digital inversion comprises the precomputation of a plurality of synthetic decomposition response signals for a respective plurality of the sets of values of one or more multiple pipe parameters, and selecting from the plurality of precomputed synthetic transient response signals the synthetic transient response signal that minimizes mismatch.
    12. A system comprising: an electromagnetic induction tool for placement inside a set of pipes, the tool comprising at least one transmitter, at least one receiver and the associated circuit collectively configured to induce Pulsed eddy currents in the set of nested pipes and the measurement of a time-dependent transient response signal; a processing unit designed to store a calculation model of the tool and the set of pipes, the model comprising one or more pipe parameters, and the use of the numerical inversion to determine the values of one or more several pipe parameters which minimize a mismatch between the measured transient response signal and a synthetic transient response signal calculated based on the model, the calculation of a synthetic transient response signal including the calculation of domain response signals frequencies for a plurality of frequencies, and using the frequency domain-to-time domain transformation to transform the plurality of response signals from the calculated frequency domain into a first synthetic transient response signal, automatically determining a time limit associated with the appearance of parasitic oscillations in the first transient response signal e synthetic, the calculation of a second synthetic transient response signal starting at the time limit determined directly by a digital time domain technique, and by combining part of the first synthetic transient response signal ending at the limit time with the second synthetic transient response signal.
    13. The system of Example 12, in which the processing unit is configured to determine the spatial and temporal discretizations for calculating the second synthetic transient response signal based at least in part on a maximum frequency of interest , the maximum frequency of interest being defined at least initially at substantially an inverse value of the time limit.
    14. The system of Example 12 or of Example 13, in which the processing unit is configured to determine a time step for the calculation of the second synthetic transient response signal based at least on part on a relative permittivity of a material filling a space between the pipes, the relative permittivity being amplified by a factor of at least 10, the factor being determined based at least in part on the time limit and a conductivity of a borehole fluid inside the pipes.
    15. A computer-readable medium storing instructions executable by a processor which, when executed by one or more processors, cause one or more processors to simulate the induction of pulsed eddy currents in a set of one or more pipes characterized by one or more adjustable pipe parameters, and to calculate a synthetic transient response signal resulting for a plurality of sets of values of one or more adjustable pipe parameters, calculating the synthetic transient response signal for each of the sets of values comprising: calculating the frequency domain response signals for a plurality of frequencies, and using the frequency domain-to-time domain transformation to transform the plurality of frequency domain response signals calculated as a first synthetic transient response signal; the automatic determination of a time limit associated with the appearance of parasitic oscillations in the first synthetic transient response signal; the calculation of a second synthetic transient response signal starting at the time limit determined directly by a digital time domain technique and the combination of part of the first synthetic transient response signal ending at the time limit with the second synthetic transient response signal.
    16. The computer-readable medium of Example 15, also storing a model for calculating the clearance of one or more pipes and an electromagnetic induction tool placed inside the clearance of one or more of several pipes.
    17. The computer-readable medium of Example 15 or of Example 16, also storing a measured transient response signal and instructions executable by a processor which, when executed by one or more processors, cause the one or more processors to determine values of the one or more pipe parameters which minimize a mismatch between the measured transient response signal and the synthetic transient response signal.
    18. The computer-readable medium of any one of Examples 15 to 17, in which the instructions executable by the processor cause the processor (s) to calculate the second synthetic transient response signal using spatial and temporal discretizations determined based at least in part on a maximum frequency of interest, the maximum frequency of interest being defined at least initially at substantially a value inverse of the time limit.
    19. The computer-readable medium of any of examples 15 to 18, in which the instructions executable by the processor cause the processor or processors to increase spatial and temporal discretizations at least once during the calculation of the second synthetic transient response signal based on a decrease in the maximum frequency of interest for late times.
    20. The computer-readable medium of any one of Examples 15 to 19, in which the instructions executable by the processor cause the processor or processors to calculate the second synthetic transient response signal using a determined time step based at least in part on a relative permittivity of a material filling a space between the pipes, the relative permittivity being amplified by a factor of at least 10, the factor being determined based at least in part on the limit and a conductivity of a borehole fluid inside the pipes.
    Several variations can be made in the systems, tools and methods described and illustrated here without departing from the scope of the inventive object. Consequently, the embodiments and the specific examples described are intended to be illustrative and not limiting.
    权利要求:
    Claims (12)
    [1" id="c-fr-0001]
    Claims
    1. Method (4 00) for pipe inspection comprising:
    the use (402, 404) of an electromagnetic induction tool (108) placed inside a set of pipes (102, 104, 106; 200), inducing pulsed eddy currents in the set of pipes and measuring a time dependent transient response signal;
    creating (406) a calculation model (820) of the tool and the set of pipes, the model comprising one or more pipe parameters; and using digital inversion (414) to determine values of one or more pipe parameters which minimize a mismatch between the measured transient response signal and a synthetic transient response signal calculated based on the model, the calculation of the synthetic transient response signal comprising:
    calculating (502) the frequency domain response signals for a plurality of frequencies, and using the transformation of the frequency domain into the time domain (504) to transform the plurality of the frequency domain response signals calculated into a first synthetic transient response signal;
    automatically determining (506) a time limit associated with the appearance of parasitic oscillations in the first synthetic transient response signal;
    computing (508) a second synthetic transient response signal starting at the time limit determined directly by a digital time domain technique; and combining (512) a portion of the first synthetic transient response signal ending at the time limit with the second synthetic transient response signal.
    [2" id="c-fr-0002]
    The method (400) for pipe inspection according to claim 1, wherein the first and second synthetic transient response signals are impulse response signals.
    [3" id="c-fr-0003]
    3. Method (400) for pipe inspection according to claim 1, in which the first and second synthetic transient response signals are impulse response signals, the method also comprising:
    integrating (514) the first and second synthetic transient response signals combined to obtain a synthetic impulse response signal; and anchoring the synthetic impulse response signal to zero at a point in time greater than a maximum time of interest.
    [4" id="c-fr-0004]
    The method (400) for pipe inspection according to claim 1, wherein the second synthetic transient response signal is calculated using spatial and temporal discretizations determined based at least in part on a maximum frequency of interest. , the maximum frequency of interest being defined at least initially at substantially an inverse value of the time limit.
    [5" id="c-fr-0005]
    5. Method (400) for the pipe inspection according to claim 4, also comprising increasing the spatial and temporal discretizations at least once during the calculation of the second synthetic transient response signal based on a decrease in the frequency d maximum interest for late times.
    [6" id="c-fr-0006]
    The method (400) for pipe inspection according to claim 1, wherein the second synthetic transient response signal is calculated using a temporal discretization determined based at least in part on a relative permittivity of a filling material. a space between the pipes (102, 104, 106; 200), the relative permittivity being amplified by a factor of at least 10, the factor being determined based at least in part on the time limit and a conductivity of a borehole fluid inside the pipes.
    [7" id="c-fr-0007]
    7. A method (400) for pipe inspection according to claim 6, wherein the conductivity of the fluid of the borehole is amplified by a factor of at least 10, an error in the second synthetic response signal resulting from the maintenance amplification below 10%.
    [8" id="c-fr-0008]
    The method (400) for pipe inspection according to claim 6, wherein the factor is adjusted to a higher value at least once during the calculation of the second synthetic transient response signal.
    [9" id="c-fr-0009]
    The method (400) for pipe inspection according to claim 1, wherein the digital inversion (414) comprises at least one of, the iterative adjustment (416) of the values of one or more pipe parameters, calculating (418) the synthetic transient response signal from the adjusted parameters, and determining the mismatch (420) between the measured transient response signal and the synthetic transient response signal; and precalculating a plurality of synthetic decomposition response signals for a respective plurality of sets of values of one or more pipe parameters, and selecting from the plurality of precomputed synthetic transient response signals the signal synthetic transient response that minimizes mismatch.
    [10" id="c-fr-0010]
    10. Pipe inspection system comprising:
    an electromagnetic induction tool (108) to be placed inside a set of pipes (102, 104,
    106; 200), the tool comprising at least one transmitter (118), at least one receiver (120), and an associated circuit collectively configured to induce pulsed eddy currents in the set of nested pipes to measure a dependent transient response signal time ;
    a processing unit (800) adapted to store a calculation model (820) of the tool and the set of pipes, the model comprising one or more pipe parameters; and use the digital inversion to determine values of the pipe parameter (s) which minimize a mismatch between the measured transient response signal and a synthetic transient response signal calculated based on the model, calculating the synthetic transient response signal including:
    computing frequency domain response signals for a plurality of frequencies, and using the frequency domain to time domain transformation to transform the plurality of frequency domain response signals calculated into a first transient response signal synthetic;
    automatically determining a time limit associated with the appearance of parasitic oscillations within 1 1 of the first synthetic transient response signal;
    calculating a second synthetic transient response signal starting at the time limit determined directly by a digital time domain technique; and combining a portion of the first synthetic transient response signal ending at the time limit with the second synthetic transient response signal.
    [11" id="c-fr-0011]
    The pipe inspection system according to claim 10, wherein the processing unit (800) is configured to determine the spatial and temporal discretizations for calculating the second synthetic transient response signal based at least in part on a frequency of maximum interest, the maximum frequency of interest being defined at least initially at substantially a value inverse to the time limit.
    [12" id="c-fr-0012]
    The pipe inspection system according to claim 10, wherein the processing unit (800) is configured to determine a time step for the calculation of the second synthetic transient response signal based at least in part on a relative permittivity of a material filling a space between the pipes (102, 104, 106; 200), the relative permittivity being amplified by a factor of at least 10, the factor being determined based at least in part on the time limit and conductivity of a borehole fluid inside the pipes.
    1/13
    MIS .1
    2/13
    类似技术:
    公开号 | 公开日 | 专利标题
    FR3057901B1|2019-11-08|HIGH DYNAMIC RANGE INVERSION FOR PIPE INSPECTION
    EP1081510A1|2001-03-07|Method of determining a model of a geological formation constrained by dynamic and static data
    FR3058454A1|2018-05-11|DETERMINATION OF PIPE PROPERTIES DURING CORROSION INSPECTION
    WO2015168538A1|2015-11-05|Distributed acoustic sensing gauge length effect mitigation
    EP2253970A1|2010-11-24|Method for imaging a target area of the subsoil using walkaway data
    CA2943397A1|2015-10-01|Method for assessing whether or not a measured value of a physical parameter of an aircraft engine is normal
    CA2810697A1|2012-03-29|Method for estimating elastic parameters by inverting 4d seismic measurements
    FR3053382A1|2018-01-05|METHOD AND DEVICE FOR DEEP POSITIONING OF A DIAGRAPHY TOOL IN PRODUCTION AND ASSOCIATED MEASUREMENT DIAGRAPH OF A HYDROCARBON WELL.
    US20190302058A1|2019-10-03|Remote-field eddy current characterization of pipes
    FR3060174A1|2018-06-15|WAVELET ESTIMATION FOR A QUADRIDIMENSIONAL CHARACTERIZATION OF SUBSURFACE PROPERTIES BASED ON DYNAMIC SIMULATION
    FR3067747A1|2018-12-21|APPLICATION OF THE MULTIPHYSICAL PRINCIPLE TO THE EVALUATION OF WELL INTEGRITY IN A MULTIPLE TRAIN CONFIGURATION
    FR3072412A1|2019-04-19|DETECTION OF BLOCKED PIPING
    CA2805688A1|2012-01-26|Method for estimating elastic parameters through inversion of 4d seismic measures
    FR3060049A1|2018-06-15|PROCESSING ACOUSTIC LOGGING DATA USING AMPLITUDE AND WAVEFORM PHASE
    FR3077142A1|2019-07-26|OPTIMIZING A REFERENCE LENGTH FOR SIGNAL PRESERVATION AND PROCESSING A REFERENCE LENGTH FOR DISTRIBUTED VIBRATION DETECTION PURPOSES
    FR3035145A1|2016-10-21|
    FR3057604B1|2019-11-22|IMPROVED MCI DIAGRAPHY FOR THE TREATMENT OF WELL BOTTOM MEASUREMENTS
    FR3057903A1|2018-04-27|EDGE CURRENT TOOLS WITH FAR FIELD
    EP3224653B1|2020-06-24|Method and device for processing well data
    FR3035723A1|2016-11-04|ANISOTROPIC PARAMETER ESTIMATION FROM AUTONOMOUS VSP DATA USING DIFFERENTIAL EVOLUTION
    FR3065253A1|2018-10-19|METHOD FOR LOCATING A ROD MASS POSITION
    FR3080190A1|2019-10-18|Systems and methods for acquiring orthogonal waveform pairs for well acoustic logging
    EP1360524A1|2003-11-12|Method for determining ocean current and associated device
    EP2875386B1|2019-04-03|Method of measuring the activity of a photon emission source
    FR3061976A1|2018-07-20|SEISMIC WELL LINKS USING LOCKOUT SCHEMES
    同族专利:
    公开号 | 公开日
    GB2567395B|2021-06-30|
    FR3057901B1|2019-11-08|
    BR112019005225A2|2019-06-11|
    US10823873B2|2020-11-03|
    WO2018080462A1|2018-05-03|
    GB201902225D0|2019-04-03|
    US20190369285A1|2019-12-05|
    GB2567395A|2019-04-10|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US4292589A|1979-05-09|1981-09-29|Schlumberger Technology Corporation|Eddy current method and apparatus for inspecting ferromagnetic tubular members|
    US5670878A|1993-06-21|1997-09-23|Atlantic Richfield Company|Inspecting a conductive object with a steady state magnetic field and induced eddy current|
    GB9620679D0|1996-10-04|1996-11-20|Halliburton Co|Method and apparatus for sensing and displaying torsional vibration|
    US7595636B2|2005-03-11|2009-09-29|Baker Hughes Incorporated|Apparatus and method of using accelerometer measurements for casing evaluation|
    DK2064413T3|2006-09-21|2018-02-12|Tüv Rheinland Sonovation Holding B V|DEVICE AND PROCEDURE FOR DETECTING AN ANOMALY IN A COLLECTION OF A FIRST AND SECOND ITEM|
    WO2014077721A1|2012-11-15|2014-05-22|Baker Hughes Incorporated|Apparatus and method for downhole transient resistivity measurement and inversion|
    US20150355358A1|2014-06-06|2015-12-10|Schlumberger Technology Corporation|Generalized spectral decomposition|
    MX2016016486A|2014-07-11|2017-04-10|Halliburton Energy Services Inc|Casing defect determination using eddy current techniques.|
    MX2016016849A|2014-07-11|2017-03-27|Halliburton Energy Services Inc|Casing defect determination using stored defect response information.|US11143780B2|2016-03-01|2021-10-12|Halliburton Energy Services, Inc.|Detecting and evaluating eccentricity effect in multiple pipes|
    US11150374B2|2018-09-10|2021-10-19|Halliburton Energy Services, Inc.|Mapping pipe bends in a well casing|
    US10989045B2|2018-11-13|2021-04-27|Halliburton Energy Services, Inc.|Multi-tubular inversion with automatic cost functional optimization|
    WO2021162726A1|2020-02-12|2021-08-19|Halliburton Energy Services, Inc.|Identifying anomalies in well-environment flexible pipes|
    US20210396125A1|2020-06-18|2021-12-23|Halliburton Energy Services, Inc.|Inversion-based array processing for cement-bond evaluation with an lwd tool|
    CN112538900A|2020-12-01|2021-03-23|刘斌|Building block type universal steel structure square tube|
    法律状态:
    2018-07-18| PLFP| Fee payment|Year of fee payment: 2 |
    2019-02-15| PLSC| Search report ready|Effective date: 20190215 |
    2019-09-26| PLFP| Fee payment|Year of fee payment: 3 |
    2021-06-11| ST| Notification of lapse|Effective date: 20210506 |
    优先权:
    申请号 | 申请日 | 专利标题
    PCT/US2016/058694|WO2018080462A1|2016-10-25|2016-10-25|High-dynamic-range inversion for pipe inspection|
    IBPCT/US2016/058694|2016-10-25|
    [返回顶部]