• 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:
    Systems and methods for detecting pipe characteristics, such as detecting tubular well-bottom defects and estimating the overall thickness of downhole tubular members (e.g., pipes such as casing and / or casing). A method of detecting defects may include providing a tool for detecting faults in a wellbore, wherein the defect detection tool comprises at least one transmitter and at least one receiver; obtaining nominal parameters of properties of the pipes; the determination of the nominal responses corresponding to the nominal parameters; determining a fault profile for a plurality of pipes disposed in a wellbore; determining responses associated with a fault for the anomaly detection tool based on at least the nominal parameters and the fault profile; calculating a gradient based on at least the responses associated with a fault, the nominal responses, the nominal parameters and the fault profile; taking downhole measurements of the plurality of pipes using the defect detection tool; and calculating the parameters of the final solution of the plurality of pipes using at least the downhole measurements, the nominal responses, the gradient and the nominal parameters.
    公开号:FR3058454A1
    申请号:FR1759339
    申请日:2017-10-05
    公开日:2018-05-11
    发明作者:Baris Guner；Burkay Donderici；Ilker R. Capoglu
    申请人:Halliburton Energy Services Inc；
    IPC主号:
    专利说明:

    © Publication number: 3,058,454 (to be used only for reproduction orders) © National registration number: 17 59339 ® FRENCH REPUBLIC
    NATIONAL INSTITUTE OF INDUSTRIAL PROPERTY
    COURBEVOIE © Int Cl 8 : E 21 B 47/117 (2017.01), E 21 B 12/02
    A1 PATENT APPLICATION
    ©) Date of filing: 05.10.17. © Applicant (s): HALLIBURTON ENERGY SERVICES, © Priority: 06.11.16 IB WOUS2016060753. INC. - US. @ Inventor (s): GUNER BARIS, DONDERICI BUR- KAY and CAPOGLU ILKER R .. (43) Date of public availability of the request: 11.05.18 Bulletin 18/19. ©) List of documents cited in the report preliminary research: The latter was not established on the date of publication of the request. (© References to other national documents © Holder (s): HALLIBURTON ENERGY SERVICES, related: INC .. ©) Extension request (s): (© Agent (s): GEVERS & ORES Société anonyme.
    104 / DETERMINATION OF PIPE PROPERTIES DURING THE CORROSION INSPECTION.
    FR 3 058 454 - A1 (g /) Systems and methods for detecting pipe characteristics, such as detecting defects in downhole tubular members and estimating the overall thickness of tubular members well bottom (for example, pipes such as casing and / or production tubing). A defect detection method may include providing a defect detection tool in a wellbore, wherein the defect detection tool comprises at least one transmitter and at least one receiver; obtaining nominal parameters of pipe properties; determining the nominal responses corresponding to the nominal parameters; determining a defect profile for a plurality of pipes disposed in a wellbore; determining responses associated with a fault for the fault detection tool based on at least the nominal parameters and the fault profile; calculating a gradient based on at least the responses associated with a fault, the nominal responses, the nominal parameters and the fault profile; taking downhole measurements of the plurality of pipes using the fault detection tool; and calculating the parameters of the final solution of the plurality of pipes using at least the downhole measurements, the nominal responses, the gradient and the nominal parameters.
    DETERMINATION OF PIPE PROPERTIES DURING CORROSION INSPECTION
    BACKGROUND OF THE INVENTION For the exploration and production of oil and gas, a network of wells, facilities and other conduits can be established by connecting sections of metal pipe to each other. For example, a well installation can be completed, in part, by lowering multiple sections of metal pipe (i.e., a casing train) into a borehole, and cementing the casing train in place. In some well installations, multiple casing trains are used (for example, a multiple concentric train arrangement) to allow for different operations related to well completion, production, or enhanced oil recovery options (EOR ).
    Corrosion of metal pipes is a permanent concern. Efforts to reduce corrosion include the use of corrosion resistant alloys, coatings, treatments, and corrosion transfer, among others. Similarly, efforts to improve corrosion monitoring are underway. For downhole casing trains, various types of corrosion monitoring tools are available. One type of corrosion monitoring tool uses electromagnetic fields (EM) to estimate the thickness of pipes or other indicators of corrosion. For example, an EM logging tool can collect EM logging data, where EM logging data can be interpreted to be correlated with a level of flux leakage or EM induction with corrosion. When multiple casing trains are used together, the proper management of EM logging tools for corrosion detection and data interpretation can be complex.
    BRIEF DESCRIPTION OF THE FIGURES These drawings illustrate certain aspects of certain examples of this disclosure, and should not be used to limit or define the disclosure.
    Figure 1 is a schematic illustration of an operating environment for a fault detection tool.
    FIG. 2 illustrates an example of a flow diagram of an inversion for the electromagnetic inspection of faults / corrosion.
    Figure 3 illustrates an example flow diagram of a calculation of approximate derivatives using a perturbation technique.
    Figure 4 illustrates an example of a flow diagram of a general implementation of the perturbation technique on the scale of an oil field.
    FIG. 5 illustrates an example flow diagram of an implementation using a perturbation around an initial solution which can be used to improve the accuracy of the results.
    FIG. 6 illustrates an example of a flow diagram of a hybrid disturbance / complete inversion technique which can be applied to increase the precision when the faults are large.
    FIG. 7 illustrates an example of corrosion diagrams illustrating the results of the inversion using complete direct modeling.
    Figure 8 illustrates an example of corrosion diagrams illustrating the results of the inversion using an approximate model obtained using the disturbance.
    DETAILED DESCRIPTION This disclosure may relate, in general, to systems and methods for detecting characteristics of pipes, such as the detection of faults in tubular downhole elements and the estimation of thickness. overall well bottom tubular elements (for example, pipes such as casing and / or production tubing). More specifically, the present disclosure may relate to techniques which can improve the speed of electromagnetic detection of corrosion in metal pipes. The proposed techniques can use the nominal properties of pipes to obtain a linear approximation of changes in the response of a fault detection tool as a function of changes in pipe parameters; and then use these approximations either directly (for example, inversion of the matrix) or indirectly (for example, within the framework of a direct model of the inversion). Consequently, a complete characterization of the fault detection tool using electromagnetic modeling may only be necessary for the calculation of the disturbances around the nominal values; a complete direct model may not be called during the inversion. A variety of different implementations which can be centered on this main idea are discussed in the disclosure and may include the disturbance of certain unknown variables while applying interpolation to others, the interpolation of the disturbed responses, the resolution of the different disturbed responses and the choice of the precise response, the calculation of a new disturbance around the result if the result is different from the initial value or outside a confidence zone, hybrid disturbance techniques and complete inversion.
    Pipe corrosion can be a dangerous condition which can lead to ruptures and bursts of cased wells in oilfield applications. It may be important to detect and correct potential corrosion in cased wells in a timely manner. In addition to electromagnetic inspection techniques, there are other techniques for inspecting wells for corrosion, including acoustic tools, diameters and cameras. Of these, only electromagnetic tools can allow inspection of exterior pipes if there are multiple concentric pipes in an inspection area.
    Electromagnetic tools can be used to determine the thickness of each pipe individually in order to assess the corrosion levels of each pipe. Such a process can generally be called an inversion, since the thicknesses of the pipes can be inverted from known electromagnetic measurements. There may be other parameters that need to be inverted along with the thickness of the pipes, such as the permeability and conductivity of the pipes, since these parameters can also affect the value of electromagnetic measurements.
    The traditional inversion techniques used in the inspection of pipes can be slow and costly in computer resources, because it may be necessary to ask the direct model used to simulate the electromagnetic measurements which correspond to a given pipe configuration several times to determine which pipe configuration best matches the measurement obtained, and each pass of the direct model usually takes a long time. Alternatively, a conversion bank may be created in advance to include a potential number of pipes, pipe thicknesses, permeability and conductivity of the pipes, but such a bank may need to be very rich to account for all potential scenarios.
    The techniques described in this disclosure can be based on a disturbance to obtain the electromagnetic response of faults on the pipes for a given pipe area. This technique can be used as a direct model in an inversion; since the computation by the direct model can be appreciably improved, a resulting inversion can also be much more effective. Alternatively, and even more effectively, faults on the pipes can be resolved using regular matrix inversion.
    The proposed reversal technique can improve the speed of the reversal, of the order of several hundred, without any major loss of precision. Therefore, it can save money. In addition, the proposed inversion technique can also be combined with a regular inversion. Regular reversal may only be applied to areas that may require additional inspection, such as areas with defects, based on the results of the proposed technique.
    Figure 1 illustrates an operating environment for a fault detection tool 100 as disclosed in this disclosure. The fault detection tool 100 may include a transmitter 102 and receivers 104. The fault detection tool 100 may be operatively coupled to a transport line 106 (for example, wired line, smooth cable, coiled piping , pipe, or equivalent) which can provide mechanical suspension, as well as electrical connectivity, to the fault detection tool 100. The conveyor line 106 and the fault detection tool 100 may extend to the inside the casing train 108 to a desired depth inside the wellbore 110. The conveyor line 106, which may include one or more electrical conductors, can leave the wellhead 112, can pass around a pulley 114, can engage an odometer 116, and can be wound on a winch 118, which can be used to raise and lower the tool assembly in the borehole 110. The signals recorded by the out it 100 fault detection can be stored in a memory and then processed by a display and storage unit 120 after recovery of the fault detection tool 100 from the wellbore 110. Alternatively, the signals recorded by the the fault detection tool 100 can be sent to the display and storage unit 120 by means of the transport line 106. The display and storage unit 120 can process the signals, and the information contained in these can be displayed to an operator to be observed and stored for further processing and consultation. The display and storage unit 120 may also contain an apparatus for supplying control signals and power to the downhole tool assembly, wherein the downhole tool assembly well includes a fault detection tool 100.
    A conventional casing train 108 may extend from the wellhead 110 at or above the surface to a selected depth inside a wellbore 109. The casing train 108 may comprise a plurality of joints or casing segments, each segment being connected to the adjacent segments by a threaded collar.
    Figure 1 also illustrates a conventional pipe train 122, which can be positioned inside the casing train 108 extending over part of the distance to the bottom of the wellbore 110. The pipe train 122 may be production tubing, production tubing, tubing train, or other pipe disposed within tubing train 108. A seal 124 can generally seal the lower end of the space tubular casing annular and can fix the lower end of the pipe train 122 to the pipe train 108. The defect detection tool 100 can be dimensioned so that it can be lowered into the wellbore 110 through the pipe train 122, thereby avoiding the difficulties and costs associated with extracting the pipe train 122 from the wellbore 110.
    In logging systems, such as, for example, logging systems using the fault detection tool 100, a digital telemetry system can be used, in which an electrical circuit is used both to power the tool detection device 100 and for transferring the data between the display and storage unit 120 and the fault detection tool 100. A DC voltage can be supplied to the fault detection tool 100 by a power supply located above surface level, and the data can be coupled to the DC electrical conductor by a baseband current pulse system. Alternatively, the fault detection tool 100 may be powered by batteries located within the downhole tool assembly, and / or the data provided by the fault detection tool 100 may be stored inside the downhole tool assembly, instead of being transmitted to the surface during logging (fault detection).
    The transmission of electromagnetic fields by the transmitter 102 and the recording of signals by the receivers 104 can be controlled by an information manipulation system. The transmitter 102 and the receivers 104 may include coils.
    The systems and methods of this disclosure can be implemented, at least in part, using an information manipulation system 124. An information manipulation system 124 can include n any instrumentality or any aggregate of instrumentalities allowing to calculate, estimate, classify, process, transmit, receive, find, produce, switch, store, display, manifest , detect, record, reproduce, manipulate or use any form of information, intelligence or data for commercial, scientific, control, or other purposes. For example, an information manipulation system 124 may be a personal computer, a network storage device, or any other suitable device, and may vary in size, shape, performance, functionality, and of price. The information handling system 124 can comprise a random access memory (RAM), one or more processing resources such as a central processing unit (CPU) or a hardware or software control logic, a ROM, and / or other types of non-volatile memory. Additional components of the information manipulation system 124 may include one or more disk drives, one or more network ports for communicating with external devices, as well as various input and output (I / O) devices, such as '' keyboard, mouse and video display. The information handling system 124 may also include one or more buses for transmitting communications between the various hardware components.
    Alternatively, the systems and methods of this disclosure can be implemented, at least in part, with a non-transient computer-readable medium. Non-transient computer readable media include any instrumentality or aggregation of instrumentalities that may retain data and / or instructions for a period of time. Non-transient computer readable media may include, for example, but not limited to, storage media such as a direct access storage device (for example, a hard disk drive or a floppy drive), sequential access storage device (for example, tape drive), compact disc, CD-ROM, DVD, RAM, ROM, erasable and electrically programmable read-only memory (EEPROM) and / or memory flash; as well as communication media such as wires, optical fibers, microwaves, radio waves, and other electromagnetic and / or optical carriers; and / or any combination of the above.
    The anomaly detection tool 100 can be used for excitation of the transmitters 102. The transmitters 102 can transmit electromagnetic signals in an underground formation. The electromagnetic signals can be received and measured by the receivers 104 and processed by the information handling system 124 to determine the parameters of the pipes, such as for example the thickness of the pipes and the pipes having a defect. The defect detection tool 100 may be disposed in the wellbore 109, in which the defect detection tool 100 may include one or more transmission coils 102 and at least one reception coil 104. fault detection 100 and / or the information handling system 124 can obtain nominal parameters of pipe properties, determine the nominal response of the tool (i.e., simulated values of what the tool would measure) corresponding to nominal parameters through modeling; determining a fault profile for a plurality of pipes disposed in the wellbore 109; determining the response associated with a fault in tool 100 according to at least the nominal parameters and the fault profile by means of modeling; calculating a gradient (for example, a vector derivative) based on at least the response associated with a fault, the nominal response, the nominal parameters and the fault profile; taking measurements (for example, downhole measurements) of the plurality of pipes using the fault detection tool 100; calculating the parameters of the final solution of the plurality of pipes using at least the measurements, the nominal response, the gradient and the nominal parameters. The parameters of the final solution can be used to make an operational decision on drilling, logging, production or completion. A gradient can be defined as a vector of derivatives for a function of multiple variables.
    The fault detection tool 100 and / or the information manipulation system 124 can determine the nominal response corresponding to the nominal parameters by means of a well plan. The fault detection tool 100 and / or the information manipulation system 124 can determine a fault profile by determining an electrically small fault (for example, 1/1000 times the skin depth ) or large (for example, 1 time the skin depth). The fault detection tool 100 and / or the information handling system 124 can determine the calculated response (for example, associated with a fault) based on at least the nominal parameters and the fault profile via the calculation of the calculated parameters by adding the fault profile to the nominal parameters and using direct modeling on the parameters associated with a fault to calculate the calculated response (for example, associated with a fault). The fault detection tool 100 and / or the information handling system 124 can calculate a gradient based on the calculated response (for example, associated with a fault), the nominal response, the nominal parameters and the profile of default using:
    F l d (j) = P N (j) if J * i (Eq. 3, illustrated below)
    PP (j) = P N (j ') + Aj if j = i (Eq.4, shown below)) and (Eq. 5, illustrated below) where P N is the set of nominal parameters; Pfl is the set of parameters where the ith element has a defect (where i, j = l, ..., Lp), Lp is the number of parameters; and M N are the corresponding responses (or what the tool would measure). The defect detection tool 100 and / or the information manipulation system 124 can take measurements (for example, downhole measurements) using at least two spacings (for example, about 0.5 feet at about 10 feet; from 0.2 m to about 3 m) between at least one transmitter 102 and the plurality of receivers 104. The fault detection tool 100 and / or the information manipulation system 124 can calculate the parameters of the final solution using at least the measurements, the nominal response, the gradient and the nominal parameters. The fault detection tool 100 and / or the information manipulation system 124 can calculate the parameters of the final solution using the measurements and the responses of the solution calculated in intermediate steps in an iterative inversion and can calculate the parameters of the final solution using the measures and intermediate responses of the solution using the updated equation P up (i) = P (i) + —— L2 · [ = i, ..., L P , or
    P is the vector of parameters. P (i) is the ith element of the vector. LP is the total number of parameters. P up is the vector of the updated solution. I represents the “real” measurements while M represents the simulated responses corresponding to the vector P. L is the vector of gradients and d is the vector of increments for the determination of the value assumed to be updated. Furthermore, the fault detection tool 100 and / or the information manipulation system 124 can calculate the parameters of the final solution using at least the measurements, the nominal response, the gradient and the nominal parameters by forming a matrix equation, where the elements of the matrix can be composed of the nominal response, the measurements, the gradient and the nominal parameters. The defect detection tool 100 and / or the information manipulation system 124 can calculate the parameters of the final solution by solving the matrix equation and can calculate the parameters of the final solution by solving the matrix equation at l using:
    [z, ··· L / Jx (p a -p) = im n => p A = ^ ··· üj x [a ··· l4'x [Î · L /, J ^ x (7-î ' v ) + p v (Eq. 8, illustrated below) where Li is the vector of gradients assuming a perturbation on the parameter i, i = l, ..., Lp. I represents the (actual) measurements. M N represents the simulated responses corresponding to the set of nominal parameters. P N is the set of nominal parameters. P F is the parameter vector of the final solution.
    A conversion table can be used to model the variations in magnetic permeability. The conversion table can include gradient vectors and nominal responses calculated for a set of different permeabilities. The fault detection tool 100 and / or the information manipulation system 124 can also use two different fault profiles to calculate two distinct gradient vectors, calculate a solution response using a combination of the gradient for each profile default parameters of the final solution using this solution response during the inversion. The fault detection tool 100 and / or the information handling system 124 can also use two different fault profiles to calculate two separate gradient vectors, two separate responses of the solution for each fault profile, two separate parameters of the solution for each response of the solution, and a quality value for each parameter of the solution. In addition, the fault detection tool 100 and / or the information handling system 124 can select the parameters of the final solution as parameters of the solution which maximize the quality value. In certain implementations, if the difference between the parameters of the final solution and the nominal parameters are beyond a threshold (outside a confidence zone), the calculation of the gradient can be repeated around this solution solution to improve accuracy by replacing the nominal parameters with the parameters of the final solution. The fault detection tool 100 and / or the information handling system 124 can also determine the presence of a fault by calculating the difference between the parameters of the final solution and the nominal parameters, and if a fault is present. , the fault detection tool 100 and / or the information manipulation system 124 can execute a subsequent complete reversal.
    FIG. 2 illustrates a general iterative inversion algorithm for inverting the thicknesses of the pipes in EM corrosion inspection tools, such as the fault detection tool 100. The purpose of the inversion may be to minimize the so-called cost function (ε). Box 200 indicates that the starting iteration = 1, 8 m j n = where 8 m j n denotes the minimum calculated value of the cost function during the inversion. Frame 202 provides a direct model. Box 204 indicates that if the iteration = 1, P = P 19 or P = P up . Box 206 provides the calculation of a cost function. Box 208 provides the calculation of the values assumed to be updated. This cost function, in its most basic form, can include a mismatch between the measurements and the results of the simulation. For example, the cost function can be the norm of the quadratic error between the measurements and the results of the simulation. In other cases, regularization terms can be added to the inversion to vary it more regularly or to cause the model to evolve within certain physical constraints.
    At the start of the first iteration of the inversion, (8 m j n ) which represents the minimum value of the cost function obtained during the inversion can be fixed as being infinity. Direct modeling can simulate the response M of the fault detection tool (for example, illustrated in Figure 1) using an assumed initial value P 13 of the parameter vector which needs to be inverted P. Note that the top bar on these variables is intended to represent a column vector. It should also be noted that the response can be a matrix corresponding to the responses of different receiver-transmitter pairs at different frequencies, for example, for a frequency-based electromagnetic tool. However, any matrix can be converted to a vector (for example by concatenating its rows to create a column vector), and assuming that the measurements are column vectors simplifies the demonstration of the solution of the perturbation equations later; this rating was therefore adopted without any loss of generality. The parameter vector can include the thickness of each pipe {Tl, ...., TK} where K is the number of pipes in the inversion. It can also include the permeability and conductivity of each pipe. In other cases, some of these parameters may be known or may be given an approximate value. In still other inversions, certain parameters are assigned a unique value; for example, the permeability and conductivity of all pipes can be assumed to be the same.
    Once the response of the model has been calculated, it can be used in conjunction with the measurement matrix, again as an example corresponding to the responses of the different transmitter-receiver pairs at different frequencies for a tool based on the frequency, to get the cost function. The cost function can then be compared to the minimum cost function obtained at this stage, e mm . If ε is less than s m i n , which is always true for the first iteration, the supposed final value of the parameters can be updated with those used in this iteration and E m j n is fixed as being ε. Box 210 indicates that if ε <E min < P = P, E m j n = ε. The frame 212 indicates that if E min <E seu ii or iteration = iteration * 13. In the next step, a convergence check can be performed by comparing the cost function to a predetermined threshold value and the number of iterations to the maximum number of iterations. If either of these convergence conditions is satisfied, the inversion stops and returns pf as the response. The frame 214 indicates that iteration = iteration + 1. The frame 216 provides Return P up . Otherwise, the assumed values of the parameters can be updated, using techniques which are widely known, such as the Levenberg-Marquardt algorithm. The direct model can be run again using these parameters and the above steps can be repeated until the convergence criteria are satisfied.
    It can be seen that direct modeling can be the real bottleneck of the calculation in the inversion; and it can be called as many times as necessary before convergence is reached. Any improvement in the computation of the direct model can considerably increase the efficiency of the inversion.
    Disturbance technique. In the inspection of defective pipes, the nominal values of the thickness of the pipes are generally known a priori. Other parameters which may be important, such as the permeability and conductivity of the pipes, can also be known or their approximate values can be calculated by calibrating the fault detection tool 100 (for example, illustrated in Figure 1), which are not described in this document. A proposed technique can be a perturbation (i.e., the perturbation can be a change in one of the parameters and it can be used to obtain the slope of the response assuming that the response is linear, therefore, linearization).
    Inversion parameters around these nominal values in order to calculate their approximate derivatives. These approximate derivatives as well as the nominal response can be used to calculate an approximate response for any possible set of parameters as described below. This computation can be arithmetic and thus it can be computed efficiently in comparison with the execution of the complete direct model for each iteration of an inversion. In general, the inside or outside diameter of the pipes can be assumed to be fixed and the fault can only occur in a certain direction. For example, an outside diameter can be fixed and it can be assumed that any defect only changes the inside diameter of the pipe. In certain other implementations, this internal diameter or this external diameter of the pipes can be disturbed (linearized) individually. In still other cases, the inner diameter of the innermost pipe can be disturbed in addition to the thickness of each of the pipes. This approach can increase accuracy in cases where the inside diameter of the innermost pipe can be measured using a mechanical diameter.
    As a general, but non-limiting example, consider the following set of parameters which need to be inverted:
    P = {T , ..., Τ κ , μ γ , ..., μ κ , σχ, ..., σ ^} (1) where K is the number of pipes and T represents the thicknesses of the pipes as above, μ; is the permeability of the i th pipe and σ; is the conductivity of the i th pipe. Then, initially the nominal response of the tool M N is simulated using the nominal values of the parameters, P N = (Tfi, ..., Τ $, μ%, ..., μ%, σ , .. ., σ £} where
    M N = F {P n } (2) In Equation 2, F denotes the complete direct model. Then, each element of the parameter set is disturbed compared to its nominal value and the corresponding response of the tool is also recorded. The set of parameters when the i th element of the set of parameters has been disturbed can be designated by PP so that:
    P l D (j) = P N (J) ifj * i
    P l D (J) = P N (j) + to i ifj = i designated by M · 3 .
    (3) (4) where Aj is the perturbation for element i. The corresponding response can be [0019] If the number of parameters is designated by LP (equal to 3 times K in our example), this calculation can be repeated LP times. A disturbance amount can be a percentage of the nominal value of each parameter (for example, such as 10%). In other applications, the disturbance may be a fixed amount. For example, thickness measurements can be disturbed by 0.01 ”, relative permeability measurements can be disturbed by 1 (without unit) and conductivity measurements can be disturbed by 100,000 S / m. In still other applications, the disturbances may be different for each pipe. In such cases, they can be a function of the tolerances required for each pipe. Again, these examples are not intended to be limiting. In general, it can be understood that an amount of disturbance must be optimized for the configuration specific to an EM corrosion inspection tool. Once these disturbed responses have been calculated, the corresponding approximate derivatives for each element in the parameter set can be found as follows:
    (5) Figure 3 illustrates a calculation of the approximate derivatives using a perturbation technique. The frame 300 provides the calculation of a nominal response M N = F {P n }.
    Frame 302 provides the calculation of a response associated with a fault F {P } = M ; i = 1, ..., L p . The frame 304 provides the calculation of the derivatives for each parameter Lj = -4 [0021] Solution of the perturbation technique. For an arbitrary given set of parameters P G , a corresponding approximate response M G can be calculated as follows using the derivatives calculated above and the principle of superposition:
    M G = M N + x (Ë G (i) - R (0)} (6) Note that this approximate answer can be used to obtain the properties of an arbitrary pipe in two ways. In the first technique, as illustrated above, it can be directly injected into the inversion to replace the complete direct model. For example, in the inversion illustrated in Figure 2, the frame 202 (for example, the "direct model" frame ) can use Equation 6 instead of the full direct model. Since Equation 6 only includes simple arithmetic operations on already calculated responses, it can significantly speed up the inversion process. Furthermore, since the response of l the fault detection tool 100 (for example, illustrated in FIG. 1) can be linearized, the calculations of the values assumed to be updated can be carried out as follows:
    P up (i) = p (0 + d -M) ï-i (7) where d denotes a vector which can determine an amount of increment in a vector of the solution on the basis of a gradient. The specific value of d can be dependent on the implementation and the algorithm. Any number of well-known inversion algorithms (for example, Levenberg-Marquardt algorithm) can be checked to find out how the increment vector can be calculated for this particular algorithm.
    Alternatively, and even more simply, Equation 6 can be solved as being a generalized inversion of the matrix since it is a system of linear equations. It can be understood that this system may not be quadratic; that is, the number of measurements may not be equal to the number of unknown parameters. In these cases, a system of equations can be overdetermined if the number of independent measurements is greater than the number of unknowns or of undetermined elements if they are less than the number of unknowns. In any case, a generalized inversion of the matrix can become a solution by the least squares.
    To illustrate this point, it can be assumed that the measurements entered are designated by the vector /. Then, assuming that the derivative matrix is full rank and that the number of measures is greater than the number of parameters, Equation 8 can be written in the form:
    [Z ^ ··· L ip ] x (p '' - p) = 7-m
    Ρ Λ - üj x [a ··· L ^ 'x [a ··· Lz, jJx (7-Âr) + p (8) [0025] In Equation 8, T can denote the transpose of a vector —R while P can be the vector of the solution of the unknown parameters as above. For Equation 8, the set of solutions can be assumed to be real. If any of the parameters is complex, the transpose can be replaced by a conjugate transpose.
    In Figure 4, an example of the general implementation of the perturbation technique for an arbitrary well is illustrated. First, a well is divided into zones where the nominal properties remain the same. Frame 400 provides the division of a well into zones where the nominal properties remain the same. Box 402 provides the calculation of a nominal response for each zone. Box 404 provides the start of logging. For each zone, a nominal response, as well as the approximate derivatives of the unknown parameters can be determined as described from Equation 3 to Equation 5. This can be done before the logging operations begin as illustrated in Figure 4, or this can be done when a different area begins to be logged. Then, the unknown parameters can be resolved for each log point by determining the area where the log point is located and then either by injecting the approximate direct model illustrated in Equation 6 in an inversion as described in Figure 2, or using Equation 8 directly. Box 406 provides the determination of the corresponding area for each log point. Frame 408 provides the resolution of the unknown parameters using an approximate model obtained using nominal and disturbed responses. The benefits of the proposed technique may be system dependent, but the benefits may increase as the number of log points in an inversion zone is increased. As a general example, suppose there are 3 pipes in an area and only the thicknesses of the pipes are resolved. Then, the proposed perturbation technique can only call the complete direct model 4 times, once for the nominal response and once for each calculation of the approximate derivative for each thickness of the pipes. In comparison, a regular inversion can on average call the complete direct model ~ 10-15 times at each log point. If there are 300 points in an area, this can mean savings in the order of several thousand.
    Note that this basic example is intended to constitute an example of how the proposed technique can be applied; many variations may exist and additional processing steps may be undertaken in an actual logging tool. For example, data subjected to logging may require calibration as mentioned above to ensure that the model matches the measured data; that is, there is no gain drift or shift in the data.
    Alternative implementations. There are many similar alternative implementations of the proposed technique. Some of these implementations can be:
    Disturbance of some unknown variables while applying an interpolation for the others: if the responses of some of the variables change quickly, a small conversion table can be created for them instead of applying the disturbance to increase accuracy . As an illustrative example, it can be assumed that the permeability of all pipes is assumed to be the same (but unknown) and that the pipes can be interpolated and that the thicknesses of the pipes can be resolved using perturbation. Then, a number of points (for example, 10) around the nominal permeability value for a given well area can be selected for use in the interpolation. For each of these points, the nominal response (assuming the corresponding permeability) and the disturbed responses for thicknesses can be found. Then, in an inversion like the one depicted in Figure 2, the nominal and perturbed responses can be interpolated using the assumed value of the permeability at an iteration point, and these interpolated responses can be used to model a response of l The defect detection tool (for assumed thickness values) as illustrated above in Equation 6. The interpolation can be linear or be a higher order interpolation such as a cubic interpolation. The disadvantage may be that a direct inversion of the matrix is not possible using such an interpolation; moreover the number of calculations can increase in proportion to the points used to create the table for the interpolated variable.
    Interpolation of disturbed responses: as mentioned above, an approximate derivative can be calculated in the perturbation technique. This calculation can only be exact for the amount of change used in the calculation of the disturbance. Therefore, in some applications, the disturbance can be calculated for different values. Then, during the inversion, the response of the perturbation can be interpolated using these different responses and the assumed value given of the unknown parameters for this iteration. For example, suppose that only the thicknesses of the pipes are resolved. Then the disturbed responses (and the corresponding approximate derivatives) can be calculated for different amounts of disturbance; for example 1%, 10% and 50% of the nominal value. Then, during an inversion such as that explained in Figure 2 based on the assumed value of the thickness values, the corresponding disturbed responses can be interpolated from the calculated responses. Then Equation 6 can be used to model the response of a defect detection tool using these interpolated responses as above.
    Resolution of the different disturbed responses and choice of the best: as another variant, since the proposed solution shown in Figure 4 can be fast, it can be applied several times using different values of the disturbance. Then, on the basis of the results of these different solutions, a disturbance value which is considered to be the closest to the actual fault can be determined and only the solution corresponding to this disturbance value can be returned as being the output. For example, let us again assume that the thicknesses of the pipes are resolved and that the disturbed responses have been calculated for a quantity of disturbances corresponding to 1%, 10% and 50% of the nominal value. Then, if the true deviation on a pipe is 15%, the results of all solutions should generally be around 15% but the one using a disturbance of 10% may be the most accurate since it is the closest to the real difference. Note that in this technique, solutions can be obtained using simple matrix inversion as described in Equation 8.
    Calculation of a new perturbation around the result if the result is very different from the initial value or outside of a confidence zone: it may be possible to determine a "confidence zone" on the basis of the properties of l tool for detecting faults and nominal properties of the pipes that are examined; and predicting whether the result of the perturbation solution can be confidently considered to be accurate. For example, in some cases, a confidence zone may include a pre-specified interval around the assumed initial values (i.e. the nominal values). In other cases, a confidence zone may be adjusted based on the value of the amount of disturbance applied. The length of the interval can be dependent on the operating frequency, the properties of the pipes, the noise level, etc. If any of the inverted parameters are outside the confidence zone for this parameter, the results can be considered to be inaccurate. In other implementations, an accuracy test may be based on a combination of certain parameters rather than on each parameter individually. For example, the deviation of the total thickness from the nominal value can be checked rather than the deviation of the individual pipes separately to determine the precision or a weight can be applied to the precision of each parameter. If the result is in a region that is predicted to have low accuracy using the initial solution; a new approximate derivative can be calculated by applying the perturbation around this initial result and a more precise solution can be obtained. This process can be repeated until the solution is predicted to be in a confidence zone; for example as being within a certain threshold of the initial supposed value. Note that the approximate derivatives around the nominal value should not be deleted during this operation since these may be necessary to resolve other log points within this same area. Figure 5 shows an example of implementation consisting in using a perturbation around an initial solution which can be used to improve the accuracy of the results. The frame 500 provides obtaining a solution P F around the nominal values P 19 . Frame 502 provides l s , | <Th. Acc z —f
    H pie U - '. Frame 504 provides the return of P 1 . Frame 506 provides the step of setting P F = P 19 ; calculation of the response and the approximate derivatives around the new P 19 . The frame 508 provides obtaining a new solution P F using an updated perturbation. In this document, initially a solution can be performed using the pre-calculated perturbation around the nominal value for this area and a solution (P F ) is obtained. Then, this result can be compared to the initial assumed value (i.e., nominal values), (normalized to the initial assumed value to account for differences in magnitude) and if the difference is less than a threshold Th acc , P F can be returned as the answer. Otherwise, a new disturbance can be applied to the result and the process can be repeated until a satisfactory answer is found. Although the norm of the difference in results from the assumed initial values is compared to a threshold, a more general approach using a "confidence zone" can be used as described above to determine the accuracy of the result.
    Hybrid perturbation and complete inversion technique:
    as mentioned above, in some cases the proposed technique can be used to determine regions of interest (i.e., regions with defects) and a complete inversion (i.e., an inversion using the full direct model instead of the model approximate obtained using the perturbation) can be applied to these points. This process can be automated as in Figure 6. Figure 6 shows that a hybrid disturbance / full inversion technique can be applied to increase accuracy when faults are large. The frame 600 provides obtaining a solution P F by using a perturbation around the nominal values P ig . Frame 602 provides II P te II - '. The framework
    604 provides the return of P F. Box 606 provides the application of an inversion using the full direct model. Box 608 provides the return of the solution P F P m obtained by using this inversion as the result.
    In this approach, initially a perturbation solution can be obtained. The normalized difference between the result and the initial assumed value can be compared to a threshold to determine the imprecision. As described in the previous section, this criterion can be provided as an example and just one of the many criteria that can be used to determine the vagueness of the result. A more general approach can use a trust zone as described above. In some implementations, the accuracy of each parameter can be checked individually. For example, if there is a large defect in any of the pipes; a complete reversal can be applied. The results of the complete inversion (P F > f m ) can be returned as the solution to the problem.
    Consequently, the systems and methods are provided for detecting the characteristics of the pipes, such as detecting faults in tubular downhole elements and estimating the overall thickness of the tubular downhole elements. (for example, pipes such as casing and / or production tubing). The systems and methods may include any of the various features of the systems and methods disclosed in this disclosure, including one or more of the following statements.
    Item 1: defect detection method comprising: the installation of a defect detection tool in a wellbore, in which the defect detection tool comprises at least one transmitter and at least one receiver ; obtaining nominal parameters of pipe properties; determining the nominal responses corresponding to the nominal parameters; determining a defect profile for a plurality of pipes disposed in a wellbore; determining responses associated with a fault for the fault detection tool based on at least the nominal parameters and the fault profile; calculating a gradient based on at least the responses associated with a fault, the nominal responses, the nominal parameters and the fault profile; taking downhole measurements of the plurality of pipes using the fault detection tool; and calculating the parameters of the final solution of the plurality of pipes using at least the downhole measurements, the nominal responses, the gradient and the nominal parameters.
    Statement 2: Method according to statement 1, further comprising the use of the parameters of the final solution to make an operational decision on drilling, logging, production or completion.
    Item 3: method of detecting faults according to item 1 or item 2, in which the determination of the nominal responses corresponding to the nominal parameters comprises the use of a well plan.
    Statement 4: method of detecting faults according to any one of the preceding statements, in which the determination of a defect profile comprises the determination of a defect, in which the defect is 1 times a skin depth.
    Statement 5: method of detecting faults according to any one of the preceding statements, in which the determination of a fault profile comprises the determination of a fault, in which the fault is 1/1000 times a depth of skin 6: method of detecting faults according to any one of the preceding statements, in which the determination of responses associated with a fault according to at least the nominal parameters and the fault profile comprises the calculation of the associated parameters to a fault by adding the fault profile to the nominal parameters and using direct modeling on the parameters associated with a fault to calculate the responses associated with a fault.
    Statement 7: method of detecting faults according to any one of the preceding statements, in which the calculation of a gradient based on at least the responses associated with a fault, the nominal responses, the nominal parameters and the profile of default includes the use of Equations (3) and (4).
    Statement 8: method of detecting faults according to any one of the preceding statements, in which taking measurements at the bottom of the well comprises at least two spacings between at least one transmitter and at least one receiver, in which the spacings are between about 0.5 feet and about 10 feet.
    Statement 9: method of detecting faults according to any one of the preceding statements, in which the calculation of the parameters of the final solution using at least the measurements at the bottom of the well, the nominal responses, the gradient and the nominal parameters includes determining the parameters of an initial solution and responses of the solution based on at least the nominal measurements, the solution parameters, the nominal parameters and the gradient, and calculating the parameters of the final solution using the measurements at the bottom of the well and the responses of the solution.
    Statement 10: method of detecting faults according to statement 8, in which the calculation of the parameters of the final solution using the measurements at the bottom of the well and the responses includes the use of Equation 7.
    Statement 11: method of detecting faults according to any one of the preceding statements, in which the calculation of the final parameters of the solution using at least the downhole measurements, nominal responses, gradient and nominal parameters includes training of a matrix equation where elements of the matrix include the nominal responses, the gradient and the nominal parameters, and the calculation of the parameters of the final solution by solving the matrix equation.
    Statement 12: method of detecting faults according to any one of the preceding statements, in which the calculation of the parameters of the final solution by solving the matrix equation 8.
    Statement 13: method of detecting faults according to statement 12, in which the nominal parameters are updated only when the difference between the parameters of the final solution and nominal parameters is greater than a threshold.
    Statement 14: method of detecting faults according to any one of the preceding statements, in which the nominal parameters are chosen as being the parameters of the solution of a previous iteration.
    Statement 15: method of detecting faults according to any one of the preceding statements, in which a conversion table is used to model the variations in magnetic permeability.
    Statement 16: method for detecting faults according to any one of the preceding statements, further comprising the use of two different fault profiles to calculate two distinct gradients, different responses of the solution for each fault profile and the parameters of the final solution using all measures of the solution.
    Statement 17: method of detecting faults according to any one of the preceding statements, further comprising the use of two different fault profiles to calculate two distinct gradients, the responses of the solution for each fault profile, the parameters of the solution for each response of the solution, and a quality value for each parameter of the solution, and to select the parameters of the final solution as parameters of the solution which maximize the quality value.
    Statement 18: method for detecting faults according to any one of the preceding statements, further comprising determining the presence of a fault by calculating a difference between the parameters of the final solution and the nominal parameters and if a fault is present, by performing a subsequent complete reversal.
    Statement 19: fault detection system comprising: a fault detection tool, in which the fault detection tool comprises at least one transmitter and at least one receiver; and an information handling system configured to: obtain nominal parameters of pipe properties; determining the nominal responses corresponding to the nominal parameters; determining a fault profile for a plurality of pipes disposed in a wellbore; determining the responses associated with a fault based on at least the nominal parameters and the fault profile; calculating a gradient based on at least the responses associated with a fault, the nominal responses, the nominal parameters and the fault profile; take downhole measurements; and calculating the parameters of the final solution using at least the measurements, the nominal responses, the gradient and the nominal parameters.
    Statement 20: fault detection system according to statement 19, in which the information handling system is configured to determine the responses associated with a fault according to at least the nominal parameters and the fault profile by calculating the parameters associated with a fault by adding the fault profile to the nominal parameters and using direct modeling on the parameters associated with a fault to calculate the responses associated with a fault.
    Statement 21: fault detection system according to statement 18 or statement 19, in which the information handling system is configured to carry out any of the process steps from statement 2 to 1 Statement 18.
    In order to facilitate a better understanding of the present embodiments, the following examples of certain preferred embodiments are given. The following examples of this type should not be construed in any way to limit or define the scope of the disclosure.
    EXAMPLES A case given as an example was simulated to demonstrate the efficiency and the precision of the proposed process. An EM frequency fault detection tool (eg fault detection tool 100 illustrated in Figure 1) was used in this example. It has been assumed that the fault detection tool 100 comprises a single transmitter (for example, transmitter 102 illustrated in Figure 1) and 6 receivers (for example, receivers 104 illustrated in Figure 1) and that it operates at 4 separate frequencies. The fault detection tool 100 is used inside 5 concentric pipes. The pipe parameters are summarized in Table 1. The 4th pipe has 3 faults of 2 feet (0.6 m) while the 5th pipe has a major fault with a length of 6 feet (1.8 m) and a smaller defect, 1 foot (0.3 m), adjacent to it as shown in Table 1. The inversion is only applied to the thickness of each individual pipe and to the permeability of the first pipe. The permeability of the rest of the pipes and the conductivity of all the pipes are fixed at their nominal values.
    Pipe 1 2 3 4 5 Diameter 2.8 inch 7.0 inch 9.6 inch 13.4 inch 18.6 inch outside (7.3 cm) (17.78 cm) (24 cm) (34 cm) (47.3 cm) Thickness 0.2 inch(0.5 cm) 0.3 inch(0.8 cm) 0.5 inch(1.3 cm) 0.5 inch(1.3 cm) 0.4 inch(1.1 cm) Mu valuerelative 74 74 74 74 74 Conductivity(MS / m) 4 4 4 4 4 Length 20 feet(6.1 m) 20 feet(6.1 m) 20 feet(6.1 m) 20 feet(6.1 m) 20 feet(6.1 m)
    Defaults)
    No
    No
    No
    0.1 inch x 0.1 inch x 2 feet (0.2 cm 6 feet (0.3 cm x 0.6 m), line x 1.8 m), line central to central to 5 feet (1.5 m) 10 feet (3 m) (17.5%); (31%); 0.05 inch x 0.03 inch x 2 feet 1 foot (0.1 cm (0.13 cm x x 0.3 m), line 0.6 m), line central to central to 13.5 feet 9 feet (2.7 m) (4 m) (7%) (10%); 0.03 inch x 2 feet (0.1 cm x 0.6 m), line central to 13 feet (4 m) (6%)
    Table 1: Parameters of the pipes Figure 7 shows the results of an inversion using a complete direct model. A logging area is limited to an area of ~ 22 feet (6.7 m). Line 1 shows the inverted thickness while line 2 is the true value. A sampling interval is 1/3 foot (0.1 m): There are a total of 66 data points on the diagram. The first five sub-representations (from the left) are the thicknesses of the pipes starting from the innermost pipe. The corrosion diagrams show the inverted thickness and the true value (i.e., nominal value). Finally, the rightmost diagram is the mismatch, the norm of the cost function, which is a measure of the quality of the inversion. It can be seen that the faults on the pipes are precisely reversed.
    Figure 8 shows the results when the disturbance is applied to the thicknesses of all the pipes and to the permeability of the first pipe. Line 1 shows the inverted thickness while line 2 is the true value. The amount of disturbance was set at 0.1 inch (0.3 cm) for thicknesses and 0.1 inch (0.3 cm) for relative permeability. The approximate direct model obtained using the perturbation was introduced into the direct inversion model and the same inversion as in the case illustrated in Figure 7 was applied except for the replacement of the direct model. It can be seen that the results obtained were almost identical to complete inversion; no loss of precision was observed. However, the computation time has been reduced by ~ 20 times in this case.
    The foregoing description provides various examples of the systems and methods of use disclosed in this disclosure which may contain different process steps and alternative combinations of components. It is understood that, although the individual examples may be addressed in this disclosure, this disclosure covers all combinations of the disclosed examples, including, without limitation, the various combinations of components, combinations of process steps and properties of the system . It is understood that the compositions and methods are described in terms of "comprising", "containing", or "including" various components or steps, the compositions and methods may also "consist essentially of" or "consist of" various components and steps. Furthermore, the indefinite articles "a" or "an", as used in the claims, are defined here to mean one or more of the element which is introduced.
    For the sake of brevity, only certain ranges are explicitly disclosed in the present disclosure. However, ranges from any lower limit can be combined with any upper limit to state a range not explicitly stated, similarly, ranges from any lower limit can be combined with any other lower limit to state a range not explicitly stated, similarly, ranges from any upper limit can be combined with any other upper limit to state a range not explicitly stated. In addition, whenever a numeric range with a lower limit and an upper limit is disclosed, any number and any included range falling within the range are specifically disclosed. In particular, each range of values (of the form, "from about a to about b" or, equivalently, "from approximately a to b" or, equivalently, "from approximately ab") disclosed in this document is to be regarded as indicating all numbers and ranges included within the widest range of values, even if they are not explicitly stated. Therefore, each individual point or value can play the role of its own lower or upper limit combined with any other individual point or value or any other lower or upper limit, to state a range not explicitly stated.
    Therefore, the present examples are well suited to achieve the ends and advantages mentioned as well as those which are inherent here. The particular examples disclosed above are only illustrative, and can be modified and practiced in different but equivalent ways obvious to a specialist in the field and who benefits from these lessons. Although individual examples are discussed, the disclosure covers all combinations of all examples. In addition, there is no limitation to the construction or design details described herein, other than those described in the claims below. In addition, the terms in the claims have their clear and ordinary meaning, unless explicitly stated otherwise and clearly defined by the patent owner. It is therefore obvious that the particular illustrative examples disclosed above can be altered or modified and all of these variations are considered within the scope and spirit of these examples. In the event of a conflict relating to the use of a word or term in this memorandum and one or more patents or other documents which may be incorporated in this document for reference, the definitions which are compatible with this memorandum must be adopted.
    权利要求:
    Claims (20)
    [1" id="c-fr-0001]
    The claims relate to the following:
    1. Defect detection method comprising:
    installing a fault detection tool in a wellbore, in which the fault detection tool comprises at least one transmitter and at least one receiver;
    obtaining nominal parameters of pipe properties; determining the nominal responses corresponding to the nominal parameters;
    determining a defect profile for a plurality of pipes disposed in a wellbore;
    determining responses associated with a fault for the fault detection tool based on at least the nominal parameters and the fault profile;
    calculating a gradient based on at least the responses associated with a fault, the nominal responses, the nominal parameters and the fault profile;
    taking downhole measurements of the plurality of pipes using the fault detection tool; and calculating the parameters of the final solution of the plurality of pipes using at least the downhole measurements, the nominal responses, the gradient and the nominal parameters.
    [2" id="c-fr-0002]
    2. The method of claim 1, further comprising using the parameters of the final solution to make an operational decision on drilling, logging, production or completion.
    [3" id="c-fr-0003]
    3. A method of detecting faults according to claim 1, in which the determination of the nominal responses corresponding to the nominal parameters comprises the use of a well plan.
    [4" id="c-fr-0004]
    4. A method of detecting defects according to claim 1, in which the determination of a defect profile comprises the determination of a defect, in which the defect is 1 times a skin depth.
    [5" id="c-fr-0005]
    5. A method of detecting defects according to claim 1, wherein the determination of a defect profile comprises the determination of a defect, in which the defect is 1/1000 times a skin depth
    5
    [6" id="c-fr-0006]
    The method of detecting faults according to claim 1, wherein determining responses associated with a fault based on at least the nominal parameters and the fault profile comprises calculating the parameters associated with a fault by adding the fault profile. nominal parameters and using direct modeling on the parameters associated with a fault to calculate the responses associated with a fault.
    [7" id="c-fr-0007]
    The method of detecting faults according to claim 1, wherein calculating a gradient from at least the responses associated with a fault, the nominal responses, the nominal parameters and the fault profile comprises the use of ^ 0) = ^ 0) if 1 * 1
    P ° (y) = P N (fi) + Δ (if j = i and Li = - l -; where P N is the vector of nominal parameters, P
    15 is the vector of parameters associated with a fault where the i th element has a fault, Δ; is the amount of disturbance on the ith parameter, M is the simulated response of the tool corresponding to P (response associated with a defect), MP N is the nomianl response, and i, j = 1, ..., Lp where Lp is the number of parameters (i. e length of P N ).
    20
    [8" id="c-fr-0008]
    8. A method of detecting faults according to claim 1, in which taking measurements at the bottom of the well comprises at least two spacings between at least one transmitter and at least one receiver, in which the spacings are between approximately 0.5 feet and about 10 feet.
    [9" id="c-fr-0009]
    9. A method of detecting defects according to claim 1, in which the calculation of the parameters of the final solution using at least the measurements at the bottom of the well, the nominal responses, the gradient and the nominal parameters comprises determining the parameters of a initial solution and solution responses based on at least the measurements
    5 nominal, solution parameters, nominal parameters and gradient, and calculating the parameters of the final solution using downhole measurements and solution responses.
    [10" id="c-fr-0010]
    10. A method of detecting faults according to claim 8, in which the calculation of the parameters of the final solution using the downhole measurements and the responses comprises the use of P up (i) = P (t) + - ---- 4 -——; i = 1, ..., L p , where P is the vector of parameters of the solution at an intermediate stage during the inversion, P up is the vector of parameters of the updated solution, Lp is the number of variables, / is the vector of measures, M is the vector of responses corresponding to P, Lj is the vector of gradients for
    15 the i th parameter, d is the vector of increments.
    [11" id="c-fr-0011]
    11. The method of detecting defects according to claim 1, in which the calculation of the final parameters of the solution using at least the bottom-of-well measurements, the nominal responses, the gradient and the nominal parameters comprises the formation of a
    20 matrix equation where the elements of the matrix include the nominal responses, the gradient and the nominal parameters, and the calculation of the parameters of the final solution by solving the matrix equation.
    [12" id="c-fr-0012]
    12. A method of detecting defects according to claim 1, in which the calculation of the parameters of the final solution by solving the matrix equation comprises [Z · · · Û] x (ρ λ - P *) = 7 - âT => p z = Γ ([ζ7- · - L /. © x [z © · ü /> | 'x [z; ··· L / .J'lx ^ -Z / ^ + px 7 where P is the vector of nominal parameters, P is the vector of parameters of the final solution,, is the vector of gradients assuming a defect of the parameter I where i = lLp, Lp is the number of parameters, I is the measurement vector, M is the response vector corresponding to ”.
    [13" id="c-fr-0013]
    13. A method of detecting faults according to claim 12, in which the nominal parameters are updated only when the difference between the parameters of the final solution and the nominal parameters is greater than a threshold.
    [14" id="c-fr-0014]
    14. A method of detecting faults according to claim 1, in which the nominal parameters are chosen as being the parameters of the solution of a previous iteration.
    [15" id="c-fr-0015]
    15. A method of detecting defects according to claim 1, in which a conversion table is used to model the variations of the magnetic permeability.
    [16" id="c-fr-0016]
    16. The method of detecting faults according to claim 1, further comprising using two different fault profiles to calculate two distinct gradients, different responses of the solution for each fault profile and the parameters of the final solution using all the measurements. of the solution.
    [17" id="c-fr-0017]
    17. The method of detecting faults according to claim 1, further comprising the use of two different fault profiles to calculate two distinct gradients, the responses of the solution for each fault profile, the parameters of the solution for each response of the solution, and a quality value for each parameter of the solution, and to select the parameters of the final solution as parameters of the solution that maximize the quality value.
    [18" id="c-fr-0018]
    18. The method for detecting faults according to claim 1, further comprising determining the presence of a defect by calculating a difference between the parameters of the final solution and the nominal parameters and if a defect is present, by performing a subsequent complete reversal.
    [19" id="c-fr-0019]
    19. Defect detection system comprising:
    a fault detection tool, wherein the fault detection tool comprises at least one transmitter and at least one receiver; and an information handling system configured to: obtain nominal parameters of pipe properties; determining the nominal responses corresponding to the nominal parameters;
    determining a fault profile for a plurality of pipes disposed in a wellbore;
    determining responses associated with a fault based on at least the nominal parameters and the fault profile;
    calculating a gradient based on at least the responses associated with a fault, the nominal responses, the nominal parameters and the fault profile;
    take downhole measurements; and calculating the parameters of the final solution using at least the measurements, the nominal responses, the gradient and the nominal parameters.
    [20" id="c-fr-0020]
    20. A fault detection system according to claim 19, in which the information handling system is configured to determine the responses associated with a fault based on at least the nominal parameters and the fault profile by calculating the associated parameters. to a fault by adding the fault profile to the nominal parameters and using direct modeling on the parameters associated with a fault to calculate the responses associated with a fault.
    / 7
    129
    类似技术:
    公开号 | 公开日 | 专利标题
    FR3058454A1|2018-05-11|DETERMINATION OF PIPE PROPERTIES DURING CORROSION INSPECTION
    US10502044B2|2019-12-10|Multistage processing and inversion of corrosion detection tools
    US20170176243A1|2017-06-22|Mud pulse detection using distributed acoustic sensing
    FR2719386A1|1995-11-03|Method and apparatus for measuring resistivity of formations in tube holes
    US11112524B2|2021-09-07|Method for estimating the material properties and the individual thicknesses of nested pipes
    FR3037149A1|2016-12-09|
    FR3061237A1|2018-06-29|ESTIMATE OF THE SLAVE OF QUASI-STATIC STONELEY
    US20190302058A1|2019-10-03|Remote-field eddy current characterization of pipes
    FR2795521A1|2000-12-29|Study of resistivity of geological formation around borehole with metal casing, comprises applying current to casing, shunting current, and deducing leakage current from difference of voltage drops
    US20200378240A1|2020-12-03|A Method For Unbiased Estimation Of Individual Metal Thickness Of A Plurality Of Casing Strings
    FR3041199A1|2017-03-17|
    FR2793031A1|2000-11-03|METHOD AND APPARATUS FOR DETERMINING THE RESISTIVITY OF A FORMATION CROSSED BY A TUBE WELL
    US10996199B2|2021-05-04|Artifact identification and removal method for electromagnetic pipe inspection
    Zhai et al.2009|A new method for continual forecasting of interwell connectivity in waterfloods using an extended kalman filter
    FR3057903A1|2018-04-27|EDGE CURRENT TOOLS WITH FAR FIELD
    US20200333500A1|2020-10-22|Multi-Zone Processing Of Pipe Inspection Tools
    US11181659B2|2021-11-23|Pipe thickness inversion using a fast forward model
    FR3065253A1|2018-10-19|METHOD FOR LOCATING A ROD MASS POSITION
    EP2806101B1|2016-08-24|Method for exploiting a fracture medium from a wedge reservoir model for wells selected using an equivalent transmissivity pattern
    US10989045B2|2021-04-27|Multi-tubular inversion with automatic cost functional optimization
    Larbi Zeghlache et al.2020|Enhanced Time Domain EM Technology for Multiple Casing Corrosion Monitoring
    FR3041371A1|2017-03-24|ZONAL REPRESENTATION FOR FLOW VISUALIZATION
    FR3062674A1|2018-08-10|DISTANCE UP TO MULTI-LAYER BASE LIMIT | WITH MULTIPLE INITIAL SUPPOSED VALUES
    FR3068067A1|2018-12-28|TOTAL THICKNESS DIAGRAM BASED ON FOUCAULT RUNNING FIELD
    FR3067055A1|2018-12-07|REMOTE FIELD EDGE CURRENT TECHNIQUE FOR INSPECTION OF CORROSION OF MULTIPLE PIPES INCLUDING RETRACTION SECTIONS
    同族专利:
    公开号 | 公开日
    GB2570409A|2019-07-24|
    WO2018084865A1|2018-05-11|
    GB201903841D0|2019-05-01|
    BR112019006603A2|2019-07-02|
    GB2570409B|2022-01-26|
    US10317331B2|2019-06-11|
    US20190086320A1|2019-03-21|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    WO2015157268A1|2014-04-10|2015-10-15|Halliburton Energy Services, Inc.|Multi-string monitoring using electro-magnetic corrosion detection tool|
    EP2950038A1|2014-05-26|2015-12-02|Services Pétroliers Schlumberger|Electromagnetic assessment of multiple conductive tubulars|
    US6060884A|1994-03-14|2000-05-09|Meyer, Jr.; Wallace Harold|Method and apparatus for measuring electromagnetic properties of materials in borehole environs and simultaneously determining the quality of the measurements|
    US5533572A|1994-06-22|1996-07-09|Atlantic Richfield Company|System and method for measuring corrosion in well tubing|
    US6114857A|1999-03-08|2000-09-05|Baker Hughes Incorporated|System and method for monitoring corrosion in oilfield wells and pipelines utilizing time-domain-reflectometry|
    AU2011366229B2|2011-04-18|2015-05-28|Halliburton Energy Services, Inc.|Multicomponent borehole radar systems and methods|
    CA2848956A1|2011-09-26|2013-04-04|Halliburton Energy Services, Inc.|Acoustic sensor apparatus, systems, and methods|
    RU2507393C1|2012-08-31|2014-02-20|ТиДжиТи Ойл энд Гэс Сервисиз ФЗЕ|Method of electromagnetic flaw detection in multistring wells and electromagnetic well flaw detector|
    US9678237B2|2012-12-19|2017-06-13|Halliburton Energy Services, Inc.|Method and apparatus for optimizing deep resistivity measurements with multi-component antennas|
    EP2792843A1|2013-04-17|2014-10-22|Services Pétroliers Schlumberger|Measurement Compensation Using Multiple Electromagnetic Transmitters|
    MX2016002362A|2013-10-03|2016-08-12|Halliburton Energy Services Inc|Downhole inspection, detection, and imaging using conformable sensors.|
    MX2016002360A|2013-10-03|2016-08-19|Halliburton Energy Services Inc|High resolution downhole flaw detection using pattern matching.|
    EP3129589A4|2014-04-10|2017-11-15|Halliburton Energy Services, Inc.|Casing string monitoring using electro-magnetic corrosion detection tool and junction effects correction|
    WO2016007642A1|2014-07-11|2016-01-14|Halliburton Energy Services, Inc.|Holographic techniques for corrosion evaluation of wellbore pipes|
    WO2016007883A1|2014-07-11|2016-01-14|Halliburton Energy Services, Inc.|Evaluation tool for concentric wellbore casings|
    MX2016016486A|2014-07-11|2017-04-10|Halliburton Energy Services Inc|Casing defect determination using eddy current techniques.|
    EP3134614A4|2014-07-11|2018-01-10|Halliburton Energy Services, Inc.|Slickline deployed casing inspection tools|
    EP3149519A4|2014-07-12|2017-12-20|Halliburton Energy Services, Inc.|Using an array of sensors between two transmitters in an eddy current logging environment|
    US10450856B2|2014-07-12|2019-10-22|Halliburton Energy Services, Inc.|Detecting defects in non-nested tubings and casings using calibrated data and time thresholds|
    US10458230B2|2015-02-27|2019-10-29|Halliburton Energy Services, Inc.|Formation resistivity measurement apparatus, systems, and methods|
    WO2017011078A1|2015-07-10|2017-01-19|Halliburton Energy Services, Inc.|High quality visualization in a corrosion inspection tool for multiple pipes|
    AU2016406342A1|2016-05-12|2018-08-30|Halliburton Energy Services, Inc.|Electromagnetic defect detection methods and systems with enhanced inversion options|US11029283B2|2013-10-03|2021-06-08|Schlumberger Technology Corporation|Pipe damage assessment system and method|
    WO2015187923A1|2014-06-04|2015-12-10|Schlumberger Canada Limited|Pipe defect assessment system and method|
    WO2017100387A1|2015-12-09|2017-06-15|Schlumberger Technology Corporation|Fatigue life assessment|
    US11237132B2|2016-03-18|2022-02-01|Schlumberger Technology Corporation|Tracking and estimating tubing fatigue in cycles to failure considering non-destructive evaluation of tubing defects|
    WO2019046101A1|2017-08-30|2019-03-07|Halliburton Energy Services, Inc.|Artifact identification and removal method for electromagnetic pipe inspection|
    US10989045B2|2018-11-13|2021-04-27|Halliburton Energy Services, Inc.|Multi-tubular inversion with automatic cost functional optimization|
    WO2020112091A1|2018-11-27|2020-06-04|Halliburton Energy Services, Inc.|Pipe thickness inversion using a fast forward model|
    KR102145267B1|2019-01-28|2020-08-18|한국원자력연구원|Corrosion test apparatus suitable to deep borehole|
    US11111773B1|2020-06-18|2021-09-07|Saudi Arabian Oil Company|Systems and methods for testing wellbore completion systems|
    法律状态:
    2018-09-28| PLFP| Fee payment|Year of fee payment: 2 |
    2019-10-30| PLFP| Fee payment|Year of fee payment: 3 |
    2020-04-17| PLSC| Publication of the preliminary search report|Effective date: 20200417 |
    2021-05-07| RX| Complete rejection|Effective date: 20210329 |
    优先权:
    申请号 | 申请日 | 专利标题
    IBWOUS2016060753|2016-11-06|
    PCT/US2016/060753|WO2018084865A1|2016-11-06|2016-11-06|Determining pipe properties in corrosion inspection|
    [返回顶部]