RU2574329C1 - Determination of characteristics of bed components on site of works performance - Google Patents

Abstract

FIELD: measurement equipment.

SUBSTANCE: multiple instant shots of nuclear magnetic resonance (NMR) are collected from a well shaft, indicating changes in a geological bed and determining data of nuclear magnetic resonance. They also perform identification of multiple fluids in a geological bed on the basis of data of nuclear magnetic resonance, determine appropriate signatures of nuclear magnetic resonance for identified fluids on the basis of nuclear magnetic resonance data, determine apparent volumes for identified fluids on the basis of signatures of nuclear magnetic resonance and determine corrected volumes for identified fluids on the basis of apparent volumes.

EFFECT: increased accuracy and validity during determination of bed component characteristics.

26 cl, 81 dwg

Description

BACKGROUND OF THE INVENTION

Downhole logging measuring instruments (or instruments) are used in boreholes (wellbores) to perform, for example, measurements of reservoir quality indicators to obtain the petrophysical properties of the formations surrounding the wellbore and the fluids and minerals contained therein. Such downhole logging tools may include, for example, electromagnetic instruments, nuclear instruments, acoustic instruments, and nuclear magnetic resonance (NMR) instruments.

Downhole logging tools can be moved throughout the borehole suspended on an armored electric cable (logging cable) after the borehole is drilled. Such logging tools are widely used. However, it is desirable that the information when drilling a borehole comes from instruments for measuring while drilling (IPB) and logging while drilling (CVB), which are usually placed in special specialized weighted drill pipes that form part of the tool string, called the layout of the bottom of the drill string ( BHA) used for drilling, directing and logging a borehole. Depending on the circumstances, based on the results of the collection and processing of such information during the drilling process, the borehole operator can make real-time informed decisions regarding the drilling operation to optimize the trajectory or placement of the borehole and optimize the drilling characteristics, and / or regarding the performance of logging measurements before a wellbore recorded immediately behind a drill bit in a newly drilled formation, and / or formation deformation as a result of the drilling process, and / or touching I provide or guarantee logging in boreholes that are difficult to logging using the methods of logging with a probe on a cable.

Measurements during drilling can provide information on drilling parameters, such as drilling mechanics (for example, the axial force applied to the drill bit attached to the bottom of the drill string, also called the load on the drill bit, the torque applied to the drill string, also called rotational downhole moment, and downhole fluid pressure inside or outside the drill string, sometimes converted to equivalent fluid density, such as equivalent circulation density and equivalent static I density), drilling mechanics (for example, the bottom hole assembly speed, also called the drill pipe rotational speed, shock and vibration of the drill string, and fluid flow rates when using the turbine speed), the direction of the borehole (for example, geodesic or geomagnetic direction borehole and deviation of the borehole from the vertical), and other information, such as temperature in the borehole. Logging tools during drilling can provide measurements of reservoir quality parameters, such as electrical resistivity or conductivity of the formation, complex dielectric constant, natural gamma radiation, bulk density and photoelectric factor, porosity according to logging data using thermal or epithermal neutrons, transverse thermal neutron capture cross section (called SIGMA), various gamma radiation spectra induced by neutrons (inelastic spectra or spec ry capture or activation of the gamma-radiation) and acoustic travel times or velocities of acoustic waves and relaxation time distribution or diffusion constant in nuclear magnetic resonance. Instruments for measuring during drilling and logging while drilling often have components whose functions are similar to those of components located on cable-lowered devices (for example, emitting and receiving antennas), but instruments for measuring during drilling and logging while drilling usually create for working in harsh and aggressive borehole drilling conditions. The terms “measurements while drilling” and “logging while drilling” are often used on an equal footing, and it should be understood that the use of any term in this disclosure includes the collection of information about the formation and borehole, as well as data on displacement, position and geometry of the drilling assembly.

Downhole logging tools can be used to determine the volumetric parameters of the formation, that is, the amount of volumetric fraction, which can be expressed as a percentage, of each component present in a given sample of the formation being evaluated (these components, which can be considered elementary structural blocks of the formation, are also called finite elements) . Measurement of volumetric parameters of the formation may include identification of the components present and the assignment of unique characteristics or signatures of each or any component, taking into account the various logging measurements considered (signatures of these finite elements are also called endpoints). In this case, together with the corresponding model of the geological environment and the mixing rules of the measurements in question, it is possible to convert the measurements of logging tools into elemental volume fractions of one or more components. The term “model of the geological environment” extends to the geometric or spatial arrangement of the various components relative to each other, and the term “mixing rules” reflects how specific readings of logging measurements behave or change with changes in percentages of the components present (when the change obeys a linear relationship, the mixing rules are considered linear).

Summary

This summary is intended to introduce a selection of concepts, which are further discussed below in the detailed description. This summary is not intended to identify key or essential features of the claimed subject matter and is not intended to be used as a means to limit the scope of the claimed subject matter.

The method of downhole logging of a geological formation having a wellbore may include collecting a plurality of snapshots of nuclear magnetic resonance (NMR) from the wellbore showing changes in the geological formation and determining nuclear magnetic resonance data. The method may also include identifying a plurality of fluids in the geological formation based on nuclear magnetic resonance data, determining appropriate nuclear magnetic resonance signatures for the identified fluids based on nuclear magnetic resonance data, determining apparent volumes for the identified fluids based on nuclear magnetic resonance signatures, and determining adjusted volumes for identified fluids based on apparent volumes.

A related downhole logging system may include a downhole logging tool for collecting a plurality of snapshots of nuclear magnetic resonance (NMR) from a wellbore in a geological formation, showing changes in the geological formation and determining nuclear magnetic resonance data. In addition, the processor can be turned on to identify multiple fluids in the geological formation based on nuclear magnetic resonance data, determine corresponding nuclear magnetic resonance signatures for identified fluids based on nuclear magnetic resonance data, determine apparent volumes for identified fluids based on nuclear magnetic resonance signatures, and determining adjusted volumes for identified fluids based on apparent volumes.

In addition, a non-transient computer readable medium is provided. A computer-readable medium may have computer-executable instructions for causing a computer to identify a plurality of fluids in a geological formation based on a plurality of nuclear magnetic resonance (NMR) snapshots from a wellbore in the geological formation, showing changes in the geological formation and determining nuclear magnetic resonance data, determining the corresponding nuclear magnetic resonance signatures for the identified fluids based on nuclear magnetic resonance data, determining apparent volumes for identified fluids based on nuclear magnetic resonance signatures; and determining adjusted volumes for identified fluids based on apparent volumes.

Brief Description of the Drawings

In the drawings:

FIG. 1 is a view of a borehole logging system according to an example of an embodiment;

FIG. 1A is a schematic diagram illustrating possible sources and causes of changes between different snapshots (for example, at different points in time, or from different depths of the study, or at different depths of the well), according to this disclosure;

FIG. 2 is a view of an exemplary electromagnetic downhole tool or tool that can be used with the system of FIG. one;

FIG. 2A is a graph of an exemplary conventional overlay showing individual responses (vectors) of a downhole logging tool to each component in the formation, according to this disclosure;

из фиг. 2А;FIG. 2B is a graph of a modified overlay including vectors (or rectilinear segments) passing through the beginning O formed by replication (or dragging) of various vectors

from FIG. 2A;

FIG. 2C-2F are graphs illustrating the application of overlays, while the patterns and clusters of data points are approximately aligned along different linear segments, showing permutations of pairs of different x-components;

FIG. 2G-2I are graphs that follow from FIG. 2C-2F and illustrate the application of overlays, while the patterns and clusters of data points are approximately aligned along different linear segments passing through the beginning of O, while showing permutations of pairs of different x-components;

FIG. 2J-2M are graphs showing the combined information of FIG. 2C-2F and FIG. 2G-2I, showing the pattern of permutation of pairs of x-components on the original normal overlays;

FIG. 2N-2P are graphs using the same examples as in FIG. 2C-2M, which illustrate the use of the observed variation in porosity to assign an unambiguous response in logging measurements or a signature for exactly the same components that are simultaneously involved in permutations of pairs of different x-components;

и когда эти точки данных отображают при норме выше заданного порога

шума;FIG. 2Q is a graph illustrating pseudo-normalization, in which the norm is taken equal

and when these data points are displayed at a norm above a given threshold

noise

и когда эти точки данных отображают при норме выше заданного порога

шума;FIG. 2R is a graph illustrating pseudo-normalization, in which the norm is taken equal

and when these data points are displayed at a norm above a given threshold

noise

FIG. 3A-3D are graphs illustrating simulated effective water signatures compared to those obtained (assuming that the water is a wetting liquid);

FIG. 4 is a graph of a nuclear magnetic resonance (NMR) dataset used as an example in this application;

FIG. 5-13 are graphs illustrating the application of the pseudo-normalization methods described above with respect to the nuclear magnetic resonance data set shown in FIG. 4, including clusters of pseudo-normalized patterns corresponding to permutations of pairs of different x-components according to this disclosure;

FIG. 14A-14G are graphs illustrating the concept of hyperspace of porosity, defined as multidimensional space, containing in the logging measurements the end points of these components present inside the pore space of the formation, and the application of such a concept to determine porosity according to an example embodiment;

FIG. 15 is a graph illustrating an example of minimization according to an example embodiment for the case where the present fluids were water, oil and gas, and the considered measurements were bulk density, porosity according to neutron logging and thermal neutron capture cross section (SIGMA);

, используемые в случае истинного нормирования, согласно примеру варианта осуществления;FIG. 16A is a diagram in which vectors are displayed

used in the case of true rationing, according to an example of an embodiment;

для сопоставления с фиг. 2, используемое в случае истинного нормирования;FIG. 16B is a diagram showing a relation (s)

for comparison with FIG. 2 used in the case of true rationing;

, используемые в случае кажущегося нормирования;FIG. 16C is a diagram in which vectors are displayed

used in case of apparent rationing;

для сопоставления с фиг. 2А, используемое в случае кажущегося нормирования;FIG. 16D is a diagram showing a ratio (s)

for comparison with FIG. 2A used in the case of apparent rationing;

в качестве оси X и

в качестве оси Y;FIG. 17A is a diagram illustrating a new cross-slot family with

as the x axis and

as the y axis;

в качестве оси X и

в качестве оси Y, в этом случае различные наклоны будут считываться как

;FIG. 17B is a diagram illustrating a cross-slot with

as the x axis and

as the Y axis, in which case the various slopes will be read as

;

и когда эти точки данных отображают при норме выше заданного порога

шума, согласно примеру варианта осуществления;FIG. 18 is a graph illustrating the results of pseudo-normalization, in which the norm is taken equal

and when these data points are displayed at a norm above a given threshold

noise according to an example embodiment;

и когда эти точки данных отображают при норме выше заданного порога

шума, согласно примеру варианта осуществления;FIG. 19 is a graph illustrating the results of pseudo-normalization, in which the norm is taken equal

and when these data points are displayed at a norm above a given threshold

noise according to an example embodiment;

и когда эти точки данных отображают при норме выше заданного порога

шума, согласно примеру варианта осуществления;FIG. 20 is a graph illustrating the results of pseudo-normalization, in which the norm is taken equal

and when these data points are displayed at a norm above a given threshold

noise according to an example embodiment;

приравнено к объему фильтрата бурового раствора на нефтяной основе, принимающего участие в перестановках пар x-составляющих, согласно примеру варианта осуществления;FIG. 21 is a graph showing clusters of truly normalized patterns (as opposed to pseudo-normalized) corresponding to permutations of pairs of different x-components, where

is equal to the volume of oil-based drilling mud filtrate participating in permutations of x-component pairs, according to an example embodiment;

FIG. 22 is a graph illustrating how the combined effects of the hydrogen index and polarization factor are identified and then corrected according to an example embodiment;

FIG. 23 is a flowchart of an example of a method for determining volumetric fluid parameters according to an example of an embodiment;

FIG. 24 and 25 are graphs showing integration of the distribution of T ₂ in depth followed by integration over the cumulative T ₂ , respectively, according to an example embodiment;

FIG. 26A-26F are graphs showing T ₂ distributions selected over the zones of interest and based on the results of various passes (drilling passage and cleaning passage), for interband mode-related constraints according to an example embodiment;

FIG. 27A-27H are graphs illustrating selection results from different zones and passages (drilling pass and cleaning pass) displayed together, including the underlying extracted Gaussian modes shown in the graphs of FIG. 28A and 28B;

FIG. 29 and 30 are graphs representing associated apparent nuclear magnetic resonance signatures for components of various fluids normalized to unity (i.e., 100 porosity units), according to an example embodiment;

FIG. 31 is a graph showing an elemental analysis of fluid volumetric parameters for apparent fluid volumes according to an example embodiment;

FIG. 31B is a graph showing an elemental analysis of fluid volumetric parameters for adjusted apparent fluid volumes after redistributing negative values using the covariance matrix of FIG. 32;

FIG. 32 is a visual representation of a covariance matrix of calculated apparent volumes according to an example embodiment;

FIG. 33 is a graph showing an example of a final analysis of fluid volumetric parameters intended to find the true fluid volumes after correcting the apparent apparent fluid volumes for the corresponding combined effects of the hydrogen index (VI) and polarization factor;

FIG. 34 is a graph showing the combination of the results of the final analysis of the volumetric parameters of the fluids and rock mineralogy from the data of thermal neutron capture spectroscopy for the elements into a combined elemental analysis of volumetric parameters that takes into account the available fluids and the bulk of the rock, according to an example embodiment;

FIG. 35 is a structural diagram of an exemplary computer system that can be used to implement various actions described in this application; and

FIG. 36 is a flowchart illustrating aspects of an example method associated with the system of FIG. one.

Detailed description

The present description is made with reference to the accompanying drawings, in which examples of embodiments are shown. However, many other embodiments may be used, and therefore, the description should not be interpreted as being limited to the embodiments set forth in this application. More specifically, these embodiments are presented in such a way that this disclosure is complete and complete. In all the drawings, similar elements denote similar elements.

Generally speaking, this application provides a method for borehole logging of a geological formation having a borehole, which may include collecting snapshots of nuclear magnetic resonance (NMR), such as periodically obtained snapshots or snapshots of nuclear magnetic resonance from multiple depths of investigation (MGI) ) showing changes in the geological formation. The method may also include identifying the difference between different data recorded at different points in time or from different depths of exploration (GI), and identifying a variety of fluids in the geological formation, including fluids combined by the signature of the hydrogen index (VI) and polarization factor, based on the collected nuclear magnetic resonance data, and the method may include pseudo- and true normalization methods and methods of statistical analysis using multi-dimensional histograms (e.g., histograms grams of nuclear magnetic resonance distributions) or classification schemes based on neural networks (NS), or factor analysis, or analysis of principal components (AGCs) to identify mutual component replacement signatures, and this may include determining on-site the boundary values of the nuclear magnetic distribution resonance. The method may also include determining appropriate apparent, true, and effective coupled nuclear magnetic resonance signatures for the identified fluids based on nuclear magnetic resonance data periodically collected or obtained from multiple depths of the study, for example, T ₂ distributions, and using empirical curve fitting methods (coupled or not bound by constraints), or factor analysis or analysis of the main components (bound or not bound by constraints), or calculating the intersection of two or more lines connected to the same point in two- or multidimensional space, determining the apparent volumes for the identified fluids based on the T ₂ signatures and using the associated nuclear magnetic resonance data or corresponding source echo signals, determining the corrected apparent volumes for identified fluids based on apparent volumes and using covariance of apparent volumes, and determining true volumes for identified x fluids based on the apparent volume adjusted and the hydrogen index (VI) and polarization factors identified fluids. The method may also include identifying the porosity space from nuclear magnetic resonance data, which can be represented by a set of factors, which can also be represented as a distribution with the same number of components as the considered distributions of T ₂ nuclear magnetic resonance, based on the collected data nuclear magnetic resonance, and at least one of the signatures identified T ₂ non-wetting fluids or effective signatures T ₂ of wetting fluid (e.g., GRU new signatures), determination of apparent porosity independent of snapshots based on nuclear magnetic resonance data (independent of the type of fluid) based on the identified pore space of nuclear magnetic resonance and collected nuclear magnetic resonance data, collection of well log data showing the mineralogy of the geological formation (such such as thermal neutron capture spectroscopy data or mud data), and the determination of the true independent of snapshots porosity according to nuclear magnetic resonance data, corrected for the influence of mineralogy of the bulk of the rock, based on calculated apparent porosity independent of snapshots of nuclear magnetic resonance.

In addition, associated well logging systems and non-transient computer-readable media are provided. Used in this application, the expression "nuclear magnetic resonance data", used throughout this summary, is presented in a general form for agreement with the scope of possible sequential actions in nuclear magnetic resonance. In addition, current nuclear magnetic resonance data obtained using KPMG sequences (Carr-Parsell-Maybum-Gill sequences) usually include the original echo signals, which can then be converted into window sums, which in turn then can be transformed into one-dimensional distributions in time of longitudinal relaxation (the so-called distribution of T ₁ ), or one-dimensional distributions in time of transverse relaxation (so-called distribution of division T ₂ ), or one-dimensional distributions of constant diffusion, or multidimensional combinations of them (in this case, the expression “distribution” can be replaced by the expression “map”), which in turn can be used to calculate porosity bins according to nuclear magnetic resonance data. They are linearly connected with each other and therefore it should be understood that the expression “nuclear magnetic resonance data” covers, for example, the original echo signals, window sums, distributions or maps and porosity bins.

In FIG. 1 shows a system 40 at a well location in which various embodiments may be implemented. In the example shown, the location of the well is a location on land, but the methods described in this application can also be used to locate the location of the well in the sea, on a swamp or shelf. For example, in this exemplary system, a wellbore 41 is formed in a subterranean or geological formation 42 by rotary drilling. As will be described below, in some embodiments, directional drilling may also be used.

The drill string 43 is suspended in the well bore 41 and has a bottom hole assembly (BHA) 44 that, for example, includes a drill bit 45 at a lower end. In addition, for example, the system 40 includes a platform and a rig 46 assembly located above the wellbore 41. For example, the assembly 46 includes a rotary table 47, a lead pipe 48, a drill hook 50, and a swivel 51. The drill string 43 can be rotated using the rotary table 47, which is engaged with the lead pipe 48 at the upper end of the drill string. For example, the drill string 43 is suspended from a drill hook 50 that is attached to a tackle block (not shown) through a lead pipe 48 and a swivel 51, which allows the drill string to rotate relative to the drill hook. For example, a top drive system (not shown) can also be used to rotate and axially move the drill string 43.

In the illustrated example, system 40 may also include flushing fluid or drilling fluid 52 stored in a well 53 formed therefor at the location of the well (or in the reservoir). A pump 54 delivers drilling fluid 52 into the interior of the drill string 43 through an opening in the swivel 51, causing the drilling fluid to flow down the drill string, as shown by directional arrow 55. The drilling fluid exits the drill string 43 through holes or nozzles (not shown) in the drill bit 45, and then goes up through the annular space (annulus) between the outer surface of the drill string and the borehole wall, as shown by directional arrows 56. The drilling fluid lubricates a flat bit 45 and carries fragments of cuttings to the surface, where it is cleaned and returned to the well 53 for reuse.

According to the shown embodiment, the bottom of the drill string assembly 44 may include a while-drilling logging module (CVM) 57, a drilling while measuring module (IPB) 58, a variable-speed directional drilling system, or an engine 60 and a drill bit 45.

Logging module 57 during drilling may be placed in a drill collar of a special type known in the art and may include one or more types of downhole logging tools. In addition, it should be understood that optional logging modules 61 while drilling and / or measuring while drilling can also be used in some embodiments. (Everywhere when referring to the module at position 57, the module at position 61 may also be meant.) The logging module 57 during drilling may have the functionality of measuring, processing and storing information, as well as exchanging information with ground equipment, for example, with a logging and control unit 62, which may include a computer and / or other processors for decoding information transmitted from the modules 57, 58 logging during drilling and measurements during drilling, as well as registration and calculation of parameters based on information. Information provided from the logging modules 57, 58 while drilling and measuring while drilling may be transmitted to a processor 34 (which may be located outside the well site or, in some embodiments, may be located on the well site as part of the well logging and control unit 62 and etc.) to determine volumetric parameter information related to the components in the geological formation 42, which will be further described below.

In the illustrated embodiment, the drilling module 57 may include electromagnetic or nuclear or acoustic devices, and a nuclear magnetic resonance (NMR) measuring device similar in configuration to a device sold under the trademark proVISION ™ or MR Scanner ™, which are trademarks of Schlumberger Technology Corporation, Sugar Land, Texas, USA. The signals may be transmitted from the drill string 43 to the logging and control unit 62 using a pressure transducer 63 via a wired or wireless communication line to the logging and control unit, or there may be any other signal transmission system known in the art, applicable for transmitting signals from a device located in a borehole to a surface unit, for example, a transmission system via a wire integrated into the drill pipe (BTP). See, for example, US patent No. 6866306 (Boyle et al.), Which is fully incorporated into this application by reference.

The drilling measurement module 58 may also be placed in a special type of drill collar known in the art and may include one or more devices for performing measurements of drilling conditions and parameters, such as measurements of drilling mechanics, measurements of drilling dynamics and other downhole conditions. An instrument for measuring while drilling may also include an (not shown) installation of generating electrical energy for the downhole system. It may include a turbine generator driven by the flow of the drilling fluid, it being understood that other energy and / or battery systems may be used. The measurement module 58 during drilling may include one or more measuring devices of the following types: for example, a bit load measuring device, a torque measuring device, a shock and vibration measuring device, a pressure and / or temperature measuring device, a speed measuring device ( revolutions per minute), a device for measuring the flow rate of the drilling fluid and a device for measuring direction and / or inclination.

An example of a resistivity logging tool 57 during propagation-based drilling is shown in FIG. 2. In the example shown, between the upper and lower radiating antennas TX ₁ and TX ₂ there are upper and lower receiving antennas RX ₁ and RX ₂ . Antennas are formed in the recesses of a modified drill collar and are located on magnetic cores or insulating material. The phase shift of the electromagnetic wave between the receivers gives an indicator of the formation resistivity at a relatively small depth of study (and called the resistivity by phase shift, USF), and the attenuation of the electromagnetic wave between the receivers gives an indicator of the formation resistivity at a relatively large depth of study (and called the resistivity of attenuation, HSS). US Pat. No. 4,899112 (Clark et al.), Incorporated herein by reference in its entirety, may be consulted for further details regarding the example of the measuring instrument described above. During operation, signals representing the attenuation and signals representing the phase are supplied to a processor, the output of which can be connected to a telemetry circuit.

Some other electromagnetic (EM) logging tools use one or more oblique or transverse antennas with or without axial antennas. These antennas may be emitters or receivers. An oblique antenna is an antenna whose dipole moment is neither parallel nor perpendicular to the longitudinal axis of the device. A transverse antenna is an antenna whose dipole moment is perpendicular to the longitudinal axis of the device, and an axial antenna is an antenna whose dipole moment is parallel to the longitudinal axis of the device. A three-axis antenna is an antenna in which three antennas (i.e., antenna frames) are mutually orthogonal. Often one antenna (frame) is axial, and the other two are transverse. It is said that two antennas have equal angles if their dipole moment vectors cross the longitudinal axis of the device at the same angles. For example, two oblique antennas have the same inclination angles if their dipole moment vectors, having their origin, conceptually attached to a point on the longitudinal axis of the device, lie on the surface of a straight circular cone, centered on the longitudinal axis of the device and having a vertex at this reference point. Transverse antennas have equal angles of 90 °, and this is true regardless of their azimuthal orientations relative to the device.

This application describes a system and method for identifying and optionally calibrating at least some of the responses (signatures) of a downhole logging tool to formation components based on measurements only. Before attempting to identify the signatures (also called endpoints) of the individual components (also called finite elements), for example, fluids and minerals entering the underground reservoir, you can try to identify the patterns resulting from the replacement of mutual components (x-components). When substitution occurs in pairs (that is, when one component of I replaces another component of J), then, ceteris paribus, the procedure is equivalent to comparing one component with respect to another. If one of the responses of the device to a component is fully characterized, it is possible to obtain responses of the device with respect to the other component.

Consonant dimensions can be used (i.e., truly consonant or essentially consonant). True consonant measurements are defined as measurements with the same depth of investigation (GI) and, therefore, they are equally influenced, for example, by penetration of the drilling fluid filtrate (penetration of the drilling fluid filtrate refers to the movement and invasion of drilling fluids into the pore spaces of the formation, while the pore space acts like a filter that passes solid particles present in drilling fluid compositions and captures the filtrate, naturally displacing reservoir fluids). Measurements of the thermal neutron capture cross section (also called SIGMA), density according to neutron gamma-ray logging (also called PNGH) and porosity according to thermal neutron logging (such as PCTN), obtained from Schlumberger Technology Corporation's EcoScope ^TM , can be read truly consonant dimensions. Essentially, consonant measurements are defined as measurements at different depths of study and, nevertheless, are equally influenced by penetration of the mud filtrate due to, for example, adaptive processing methods, such as correction methods for the influence of penetration, or consonant measurements may result from well logging measurements in the reservoir of the same type, despite the fact that the actual depth of the study may be different. For example, this is the case when the measurements are very weakly affected by penetration, and the readings are entirely carried out in the untouched zone (the zone not disturbed by the penetration of the drilling fluid filtrate), or are very strongly affected by penetrations, and the readings are entirely carried out in the zone penetration (zone damaged by the penetration of the mud filtrate). The characteristics of the reservoir present at a particular depth can be measured, and the logging data sets are collected twice or multiple times at different points in time or from different depths of the study (each such logging data set to be collected will hereinafter be referred to as a snapshot in a collective value) when formation changes are expected to occur, for example, due to penetration of the mud filtrate. To describe the characteristics of such a change (s) that may occur, measurements from the same snapshot may be consistent, however measurements from different snapshots may not be consistent with each other. Measurements can simultaneously be truly consonant and essentially consonant when they have the same depth of study and also reflect the washed zone or untouched zone. Since nuclear magnetic resonance data in various forms (i.e., related to the original echo signals or window sums, or distributions, or porosity bins) can be considered to include numerous components, each of which can be considered as a separate and different measurement, nuclear magnetic data resonances constitute an ideal dataset of truly consonant measurements. In addition, when using gradient nuclear magnetic resonance devices (such as the proVISION ^™ or MR Scanner ^™ devices discussed earlier), the resonant volume of the nuclear magnetic resonance is a thin cylindrical shell surrounding the device, which can actually be a mask for penetrating the mud filtrate ( that is, it may not be influenced by penetration), while shallow penetration remains below the depth of investigation when measuring nuclear magnetic resonance, or the volume is completely influenced by the penetration of the drilling fluid filtrate when penetrating to a depth exceeding the depth of investigation when measuring nuclear magnetic resonance, and the data of nuclear magnetic resonance in this case can be considered true or essentially consonant.

You may notice that at first glance, the complexity of the problem should increase with the transition to the identification of patterns resulting from the replacement of x-components, and should be commensurate with the number of available components, but this is not the case. For example, with Z components, Z × (Z-1) permutations of pairs of x-components will be possible, the number of which is much larger than Z (Z × (Z-1) ordered pairs, or half the number of disordered pairs), but in its own way the nature and practice of a very small number of such permutations of pairs of x-components will correspond to the case under consideration. Due to the migration and replacement of fluids throughout the geological time scale and due to the fact that the relative permeability of various fluids in the reservoir increases depending on the saturation of the corresponding fluid, the distribution of the currently existing natural fluids inside the subterranean formations is such that at any one depth in such strata, one of the natural stratal fluids in the stratum is predominantly mobile, while others are already displaced during geological time after being delayed. ia precipitation and burial, and the maturation and migration of hydrocarbons. In addition, any penetrating fluid that upsets or changes the initial balance of formation fluids (i.e., the equilibrium distribution of fluids corresponding to the geological time scale) can be known because it can be injected from the surface or mined to the surface.

On the other hand, it is usually difficult to directly identify the signatures of the components of the individual fluids, because regardless of the reservoir balance discussed above, they may not be present at all or they may not be in sufficient quantity in the reservoir volume measured by logging tools. This usually refers to the case of repression drilling operations (that is, the case when the hydrostatic pressure of the drilling fluid exceeds the pressure of natural reservoir fluids), and this problem is complicated by the usual logging with a probe on the cable. Instead, use depression drilling (at which the hydrostatic pressure of the drilling fluid is less than the pressure of natural formation fluids) or use well logging measurements that are suitable for existing correction methods for penetration effects (such as the method described in US Pat. No. 8005618 B2 called ″ Logging while drilling system ″ (Kais Gzara), which is fully incorporated into this application by reference), the situation will be different, and one predominant component of the fluid will make the largest contribution to the measurement logging tools. However, even in such a situation, the lack of information regarding the exact amount of the fluid component present will be an obstacle to obtaining response characteristics from this fluid during well logging. In addition, in the pictures obtained as a result of the replacement of the x-components, it can be seen that other unreplaced components that remain in place may not contribute to the measurements with a logging tool, and this reduces the complexity that would otherwise arise from an attempt to find a solution for many components at the same time. As in the case of substitution of x-components by pairs, the pattern resulting from the replacement of x-components is simply related to the signature of the replacement component during downhole measurements minus the signature of the replaced component during downhole measurements.

A special situation arises when the signature of the logging measurements for a particular component is not completely defined (or is not unique), but changes depending on the time or depth of the study, even when the x-components are replaced in pairs. For example, this refers to the case of nuclear magnetic resonance measurements, when the signature of the wetting fluid usually depends on the saturation (the partial volume of the pore space occupied by a particular fluid) of this particular fluid component. For example, when one such fluid fills the pores of a well-sorted sandstone formation, the measurements of nuclear magnetic resonance, for example the T ₂ distribution, will contain one response (for example, the only maximum response located in the fluid-free part of the T ₂ distribution). However, if there is another non-wetting and immiscible fluid in the pore spaces or when penetrating into the pore spaces, displacing the thin-film residue (and remaining behind it) of the initial wetting fluid that covers and laminates the surface of the sand grains, the response of the nuclear magnetic resonance measurements to the residual fluids will be different ( that is, there may be a single maximum response located on the edge of a portion of the associated fluids of the distribution of T ₂ ). The converse is also true.

This feature of nuclear magnetic resonance can be considered and processed in at least two ways. It can be considered as a substitution of the x-components, in which case the substitution takes place in triplets, whereby the wetting fluid that initially filled the pores of the formation is replaced by a penetrating non-wetting, immiscible fluid in addition to the residual wetting fluid treated as a separate component. In this case, the three components have a well-defined single-peak response in log measurements of nuclear magnetic resonance, which essentially does not change. A feature of nuclear magnetic resonance can also be considered the substitution of x-components, in which case the substitution occurs in pairs, due to which the wetting fluid that initially filled the pores of the formation is hypothetically divided into free and bound fluids, so that the free fluid is replaced by a penetrating non-wetting, non-mixing fluid, and the bound fluid remains in place. However, in this case, the component of the bound fluid cannot be brought into correspondence with the well-defined single-peak response in log measurements of nuclear magnetic resonance. However, when one of the snapshots of nuclear magnetic resonance includes only the component of the bound fluid without any free wetting fluid, the observed pattern of substitution of the x-components can become quite definite despite the change in the signature of the wetting fluid in the interval between individual snapshots, which will further shown in the present description below.

Another possible reason why the response in the log measurements of nuclear magnetic resonance to the component of a particular fluid may vary between different downhole logs is the occurrence of limited diffusion when the saturation of such a particular fluid falls below a certain threshold. The converse is also true. Whether specific processing procedures will entail diffusion effects, this may depend on the technical characteristics of the nuclear magnetic resonance logging tool and the type of measurement used. Diffusion effects can be a factor when the gas is one of the components present in the formation, however, the response in the log measurements of nuclear magnetic resonance to the gas also turns out to be relatively small in amplitude compared to the response from the liquid due to the combination of a small hydrogen index (VI) and low polarization factor (PF), which can mask possible diffusion effects in gas-bearing formations.

Another reason why the response in logging measurements of nuclear magnetic resonance to a specific component can vary in the interval between different snapshots is associated with data recording and data processing. These operations can be extended to the differences in the responses in the log measurements of nuclear magnetic resonance to numerous components, starting with one snapshot of nuclear magnetic resonance and ending with the following, despite the fact that in reality changes in the components may not occur. These differences are usually manifested as residual low-amplitude coherent pulsations or waves (peaks and troughs) when the distribution of T ₂ nuclear magnetic resonance, for example, from various snapshots of nuclear magnetic resonance is taken and subtracted from each other. Changes from one snapshot to the next usually relate to an increase or decrease in the corresponding mode width (mode widths) on nuclear magnetic resonance distributions, while relatively small or due to properties shifts of the positions of the mode peak (mode peaks) are created for each case separately. A correlation between such related modes from different snapshots should be expected, and therefore, the effects of changes due to logging and processing can be avoided. The method depends on observing the data in the distribution space in the relaxation time of nuclear magnetic resonance or in the space of the amplitudes of the spin echoes of nuclear magnetic resonance, with each observation having its own advantages. Observation of nuclear magnetic resonance data in the space of spin echo amplitudes allows us to study effects that are generally associated with data processing parameters or, in particular, with various regularization parameters. Observation of nuclear magnetic resonance data in the distribution space over the relaxation time helps with mode-controlled (or even completely predictable) simulations.

Changes in registration parameters, types of downhole logging tools, and environmental conditions should be excluded, if possible, or, in opposite cases, taken into account. If changes are unavoidable, a new baseline well log should be recorded immediately after it is practically feasible, the idea being to measure the formation twice with essentially no changes or with the minimum expected changes between the components of the formation, under two different sets of conditions, to record changes in the responses of the downhole logging tool due to changes in conditions only. And again, observing nuclear magnetic resonance data in a space of relaxation time or constant diffusion in the general case and methods of statistical analysis or factor analysis or analysis of principal components, in particular, can also be used to facilitate identification of a related mode and / or a controlled mode (or even completely predictable) modeling.

Depending on the many possible reasons for the observed changes in the composition of the formation between different snapshots, the method and system disclosed in this application can be applied to a wide range of situations. Indeed, the observed changes may be due to displaced fluids displaced from the formation by fine-grained solid particles, penetration of finely dispersed material from the drilling fluid, changes in the phase of the pore fluid (such as changes caused by changes in pressure or temperature), chemical reactions, dissolution or sedimentation (e.g., precipitation asphaltene, scale deposition, salt dissolution, acid treatment) of solid components or changes in compaction or pressure conditions or voltage, and the timeline for observed changes may vary depending on the situation in question. Possible sources and causes of changes observed between different snapshots are shown in FIG. 1A. In the system and method, it is possible to eliminate changes depending on time (that is, when the same formation volume is measured at different points in time, where the first moment of time can be extended to the base log diagram), shown along the X axis, changes depending on the radial depth (that is, when the volumes of the same formation located deeper and deeper are measured at one point in time at which different sets of consonant measurements can be obtained for each of the deeper and deeper measured volumes), while are given along the Y axis and changes depending on the depth of the borehole (when the same component is present and takes part in the expected replacements of the pair of x-components), shown along the Z axis. The semi-cylindrical regions 71 in FIG. 1A shows the components of an underground formation surrounding a borehole, in which the composition of the formation may be different on a single radially axisymmetric boundary relative to the borehole, with the measurement volume or the study area in the formation shown by a rectangle 72. This can happen, for example, when the borehole fluid ( that is, the mud filtrate) penetrates the reservoir and displaces the natural reservoir fluids with different severity, depending on the different characteristics of the wellbore fluid and a and parameters such as composition, viscosity, pressure, porosity, permeability and saturation. The composition of the formation may vary in steps at such a radial axisymmetric boundary, or the change may be gradual. In addition, depending on changes in the wellbore fluid, or depending on various phenomena related to fluid permeability, or depending on the reason for the observed changes (e.g., chemical reactions), the composition of the formation can vary at numerous radial axisymmetric boundaries (not only one) relative to borehole. Other combinations of scenarios, such as changes with time or changes with radial depth, may also be present. In addition to this, the organization of logging in production wells and the optimization of drilling in comparison with the assessment of the formation are associated in themselves, respectively, with the contents of the borehole during production or injection and with the contents of the borehole during drilling, in contrast to the components of the formation. The system and method described in this application can be applied in the field of logging in production wells and / or in the optimization of drilling (for example, when cleaning a well and detecting an outburst).

Changes over time can be induced, for example, by injection, production, or thermomechanical deposition. Induced by injection changes can occur throughout the small timeline or large timeline. Examples of injection-induced changes within a small timeline are related to the dynamics of penetration (for example, passage during drilling as opposed to passage during cleaning or logging during repeated passage in the case of conventional repression drilling), methods of influencing the formation or injecting the solvent (e.g. penetration associated with the dynamics of the chemical reaction) and logging-injection-logging methods (i.e., numerous penetration cycles corresponding to the intended use penetrate into the formation fluids). An example of a change within a large timeline is associated with reservoir monitoring in injection wells. Similarly, mining-induced changes can occur throughout the small timeline or large timeline. Examples of production-induced changes within a small timeline include changes in the downhole drilling process and pressure-induced changes (such as gas expansion, formation of a condensate zone, gas exit from the solution, formation of a watering cone and formation of absorption zones). An example of a change within a large timeline is associated with reservoir monitoring in production wells. Changes induced by thermomechanical deposition include temperature-induced changes within a small time scale (such as melting and melting of ice or hydrates) and temperature-induced changes within a large time scale (such as changes in the properties of heavy oil using thermal stimulation methods). In addition, changes induced by thermomechanical deposition may include voltage induced changes within a large time scale.

Similar examples can be given for changes depending on the radial depth. The case of changes in the depth of the borehole is somewhat illogical and is usually applicable when a specific exact same component is actually present at many different depths of the borehole, and other components are present separately at other depths of the borehole (i.e., situations where two or more components are available simultaneously). In this case, measurements performed at any one depth may simply be the source data, as opposed to a hypothetical situation in which the same component occupies the entire volume of the formation. Even when the specific nature of this same component is known approximately, the fact of taking measurements in the presence of the same component is sufficient to use the present method. In practice, in the case when exactly the same component present at different depths along the borehole is a rock mineral, the composition of the same rock mineral can be determined or differentiated based on borehole logging data, in which there is mainly a response on rock minerals. Examples of such logging data include, but are not limited to, thermal neutron capture elemental spectroscopy and natural gamma ray spectroscopy. In addition, the composition of rock minerals can be differentiated based on observation of the surface, for example, but without limitation, according to core analysis and analysis of drilling fluid samples (in particular, analysis of cuttings). In addition, in the case when the same component, available at different depths along the borehole, is a fluid, the type of the same fluid can be differentiated in practice based on well logging data, in which there are responses mainly to fluids such as reservoir test data. In addition, the type of fluid can be differentiated, for example, based on observations of the surface, based on an analysis of produced fluids or analysis of samples of the drilling fluid and analysis of the returning drilling fluid. In addition, the type of such a fluid, such as mud filtrate, can be easily determined since it is pumped from the surface in the case of conventional repression drilling.

If the composition of the same mineral (i.e. rock mineralogy) can be determined with certainty, changes in the logging measurements can be used to identify and distinguish different fluids from each other. In addition, changes in the type of fluid, including noticeable variations in porosity (partial volume of pore space in the reservoir), can also be used to assign an unambiguous signature to the composition of the rock mineral mentioned above, taking into account the various logging measurements considered. On the other hand, if the same type of fluid can be determined with certainty, changes in the logging measurements can be used to identify and distinguish the compositions of various rock minerals. In addition, changes in mineralogy, including noticeable variations in porosity, can also be used to give an unambiguous signature to the type of fluid mentioned above, taking into account the various logging measurements considered.

, которое согласовывает соответствие различным рассматриваемым измерениям m₁, m₂,…m_α, m_β,…m_n (от 1 до n), а моментальные снимки можно обозначить с использованием векторных обозначений в виде

(от 1 до N), тогда как однозначные сигнатуры различных составляющих пласта с учетом рассмотренных различных каротажных измерений можно обозначить с использованием аналогичных векторных обозначений в виде

(от А до Z). Кроме того,

может представлять само

или любое линейное преобразование его. Когда сигнатуры объема и некоторых составляющих в каротажных измерениях известны априори, обозначение

может также охватывать такие преобразования, которыми из

удаляются вклады этих известных составляющих с получением менее сложного вектора

, который зависит от остающихся неизвестных составляющих.To identify and classify changes that occur, in this example, you can enter a vector symbol

, which coordinates the correspondence to the various measurements under consideration m ₁ , m ₂ , ... m _α , m _β , ... m _n (from 1 to n), and snapshots can be indicated using vector notation in the form

(from 1 to N), while the unique signatures of the various components of the reservoir, taking into account the various logging measurements considered, can be denoted using similar vector notation in the form

(A to Z). Besides,

may represent itself

or any linear transformation of it. When the signatures of the volume and some components in the logging measurements are known a priori, the designation

may also cover transformations from which

the contributions of these known constituents are removed to produce a less complex vector

, which depends on the remaining unknown components.

, и

, и

. Приведенным выше представлен пример возможного отображения многокомпонентных данных ядерного магнитного резонанса, а термин «распределение» («распределения») используется в этой заявке относительно связанных (соответствующих) кривых. В настоящем описании рассмотренные консонантные каротажные измерения m₁, m₂,…m_α, m_β,…m_n также могут быть сделаны безотносительными к единицам измерения или безразмерными путем нормирования измерений на значение шума, обязательно присутствующего в измерении каждого вида. Это способствует сохранению различных измерений выше уровня присущего шума и предотвращению искажения шумом достоверной информации. Это также полезно при отображении рассмотренных выше векторов или функций в нейтральном или не зависящем от пользователя масштабе. Заметим, что упомянутое выше нормирование измерений отличается от других нормирований, применяемых позднее в настоящем описании, таких как псевдонормирование сигнатур или истинное нормирование сигнатур (включая два отдельных вида истинного нормирования с использованием кажущегося или истинного объема фильтрата бурового раствора в соответствии с примером варианта осуществления).In addition, the above vector can be displayed as curves for all n data points, receiving a number of values m _1, m _2, ... m _α, m _β, ... m _n, and in this case the vector notation can be discarded and replaced by the corresponding symbols of the curves

, and

, and

. The above is an example of the possible display of multicomponent nuclear magnetic resonance data, and the term “distribution” (“distribution”) is used in this application with respect to related (corresponding) curves. In the present description, the considered consonant log measurements m ₁ , m ₂ , ... m _α , m _β , ... m _n can also be made irrespective of the units of measurement or dimensionless by normalizing the measurements to the noise value necessarily present in the measurement of each type. This helps to keep various measurements above the level of inherent noise and prevent noise distortion of reliable information. It is also useful when displaying the vectors or functions discussed above in a neutral or user-independent scale. Note that the above rationing of measurements is different from other rationing used later in the present description, such as pseudonorming of signatures or true rationing of signatures (including two separate types of true rationing using the apparent or true volume of mud filtrate in accordance with an example embodiment).

векторов), и хотя они отображены на фиг. 2А в двумерной плоскости (то есть, на поверхности страницы, на которой они напечатаны), предполагается, что наложение находится в n-мерном пространстве для согласования с n рассмотренными каротажными измерениями m₁, m₂,…m_α, m_β,…m_n. Эти наложения используют для нанесения на кроссплот точек

данных, соответствующих каждому из моментальных снимков, которые фактически являются линейной смесью различных конечных точек

составляющих, взвешенных в соответствии с выраженным в процентах объемом соответствующей составляющей, присутствующей в исследуемом объеме на соответствующем моментальном снимке. (Серые) линии, соединяющие различные векторы, соответствуют направлению, параллельному перестановке пары соответствующих x-составляющих (замещению x-составляющих), например, линии, заканчивающиеся в А, В, I, J и Z. Это дополнительно детализировано на фиг. 2В.In FIG. 2A shows conventional overlays displaying individual logging responses for each component (signatures

vectors), and although they are displayed in FIG. 2A in a two-dimensional plane (i.e., on the surface of the page on which they are printed), it is assumed that the overlay is in n-dimensional space to match the n considered log measurements m ₁ , m ₂ , ... m _α , m _β , ... m _n These overlays are used for drawing on the crossplot points

data corresponding to each of the snapshots, which are actually a linear mixture of different endpoints

components weighted in accordance with the percentage expressed in the volume of the corresponding component present in the test volume in the corresponding snapshot. The (gray) lines connecting the various vectors correspond to the direction parallel to the permutation of the pair of corresponding x-components (replacement of x-components), for example, lines ending in A, B, I, J and Z. This is further detailed in FIG. 2B.

можно выразить в виде линейного сочетания векторов

как:If we assume that measurements are taken using linear mixing rules, then in this case the changes

can be expressed as a linear combination of vectors

as:

,

,

,

,

являются взаимозависимыми. Например, введенное обозначение

относится к различию каротажных измерений

между соответствующими моментальными снимками под номером j и под номером i (то есть, различию между

и

), например, введенное обозначение

относится к выраженному в процентах объему составляющей J, представленной в объеме исследуемого пласта соответствующим моментальным снимком под номером j, и, например, введенное обозначение Δ_ij(V_I) относится к отличию выраженного в процентах объема составляющей I, представленной соответственно в объеме исследуемого пласта соответствующими моментальными снимками под номером j и под номером i.it is predicted that the above expression is not unique, since the vectors

are interdependent. For example, the designation entered

refers to the difference in logging measurements

between the corresponding snapshots at j and i (i.e., the difference between

and

), for example, the notation

refers to the percentage expressed in the volume of component J represented in the volume of the studied formation by the corresponding snapshot at number j, and, for example, the designation Δ _ij (V _I ) entered refers to the difference in the percentage expressed in the volume of component I, represented respectively in the volume of the studied formation by the corresponding snapshots at j and i.

Below is the expression for the case of a permutation of one pair of components I and J:

, можно нанести на кроссплот в зависимости от модифицированного наложения из фиг. 2В, которое также предполагается наложением в n-мерном пространстве. Модифицированное наложение, включающее в себя векторы (или прямолинейный сегмент), обозначенные AB, BA,…,JZ, SJ,…, начиная с исходной точки О, можно получить репликацией (или перемещением) различных векторов

из фиг. 2А на кроссплот, представленный на фиг. 2В, с началом в точке О. То есть, отображаются те же самые линейные сегменты из фиг. 2А, но линейные сегменты перенесены параллельно относительно начала, находящегося в той же самой точке, то есть, в центре кроссплота. Обозначением и кодированием различных линейных сегментов отражены соответствующие начальные и конечные точки из соответствующей фиг. 2А. Картины точек данных, получившиеся в результате перестановок пар x-составляющих, точно лежат на каждом соответствующем линейном сегменте.To improve the initial understanding of the ongoing x-component permutation, the resulting measurement difference, such as

can be applied to the crossplot depending on the modified overlay from FIG. 2B, which is also assumed to be superimposed in n-dimensional space. A modified overlay including vectors (or a straight segment) denoted by AB, BA, ..., JZ, SJ, ..., starting from the starting point O, can be obtained by replicating (or moving) different vectors

from FIG. 2A to the crossplot shown in FIG. 2B, starting at point O. That is, the same line segments from FIG. 2A, but the linear segments are moved in parallel with respect to the origin located at the same point, that is, in the center of the cross-slot. The designation and coding of the various linear segments reflect the corresponding start and end points from the corresponding FIG. 2A. Pictures of data points resulting from permutations of pairs of x-components exactly lie on each corresponding linear segment.

In FIG. 2C-2F shows how overlays are used, with the patterns and corresponding clusters of data points 80c-80f aligned substantially along different linear segments, showing permutations of different pairs of x-components. In FIG. 2C-2F shows how the method described in this application is applied to changes (depending on depth) between zones, with the same component being present and participating in predictive replacements of pairs of x-components. They showed that in the case where the rock mineralogy can be distinguished with certainty, ordering without an apparent structure arises in the cloud of data points from individual downhole logging measurements. In FIG. 2C shows an example of data points representing the results of measurements with logging tools displayed in a conventional overlay, where in this case the X axis and Y axis correspond to the measurements of porosity and bulk density, respectively, according to neutron logging data. The lines shown are combined reference components that may occur in underground formations (corresponding to such x-component replacement lines), but such reference components may not actually be present in the underground formation. The reference components considered were sandstone (labeled P), limestone (labeled I), dolomite (labeled D) and water, the graphic image of which is outside the boundaries of the graph shown (the water point should be located to the right of the graph shown), and therefore the three lines shown should be connected at the point of water. The fact that the data points displayed for example do not fall on the reference lines means that, in the presence of real formation components, the reference components are displayed. In FIG. 2D-2F shows the same data points from FIG. 2C, but with the separation of these data points separately, respectively, of the mineralogy A, B and C of the rock, while the mineralogy of the rock can be distinguished with certainty. Data points are displayed along with the hypothetical response of the rock formation itself.

FIG. 2G-2I follow from FIG. 2C-2F and they show how modified overlays are used together in patterns and clusters of data points 80g-80i aligned relatively accurately along different linear segments showing different permutations of the pairs of x-components. The three segments separately show linear segments corresponding to patterns resulting from permutations of pairs of x-components, including three rock mineralogies A, B and C, respectively. In FIG. G it can be seen that the rock mineralogy A is one of the constituents and there are identified fluids of four different species. Similarly in FIG. 2H shows linear segments corresponding to patterns resulting from permutations of pairs of x-components, while rock mineralogy B is one of the participating components and there are identified fluids of three different types. Similarly in FIG. 2I shows linear segments corresponding to patterns resulting from permutations of pairs of x-components, while rock mineralogy C is one of the participating components, and again three different types of fluid are identified.

In FIG. 2J-2M presents the combined information of FIG. 2C-2F and FIG. 2G-2I, which displays the pattern of permutations of pairs of x-components on the original normal overlays FIG. 2J-2L are similar to FIG. 2G-2I, using conventional and modified overlays, respectively. They suggest how linear segments can converge to the common points shown in FIG. 2M along with a cloud of 80m merging points revealing more ordered pictures. For example, FIG. 2N and 2P include clouds 80n and 80p, which follow from similar examples in FIG. 2C-2M and illustrate how noticeable variations in porosity can be used to calibrate at the place of work the signature of the same component in logging measurements, while participating in predictive permutations of pairs of different x-components.

As shown in FIG. 2N, when the mineralogy of the rock can be distinguished with confidence, but the corresponding signature in the logging measurements is not necessarily available at the place of work with a high degree of reliability, and when changes in the type of fluid are also accompanied by noticeable variations in porosity, then different linear segments are mapped to different segregated clusters of data points, corresponding permutations of various pairs of x-components can be extrapolated to the connection at one point, showing the signature of the corresponding mineralogy logging measurements. In contrast, as shown in FIG. 2P, when the fluid composition can be confidently distinguished, but the corresponding signature in the logging measurements is not necessarily available at the workplace with a high degree of reliability, and when changes in the mineralogy of the rock are also accompanied by noticeable variations in porosity, then different linear segments are mapped to different segregated clusters of data points, corresponding to permutations of pairs of different x-components, can be extrapolated to the connection at one point, showing the signature of the corresponding type of flui yes in logging measurements.

In addition, situations may arise in which the same component participating in prognostic permutations of pairs of x-components can be determined with a high degree of reliability without the need to have detailed information on its unique signature taking into account the various logging measurements considered. The method described in this application can also be used for these cases. From a mathematical point of view, it includes the use of new technologies for simultaneously matching multiple segregated clusters of data points in the presence of convergence of two or more lines connected at a common point in two- or multidimensional space. When working with nuclear magnetic resonance data, you can expand the terminology and talk about convergent distributions (or other forms of multicomponent nuclear magnetic resonance data), combined into a common distribution based on the understanding that nuclear magnetic resonance data can also be considered as vectors or data points in multidimensional space.

(различие каротажных измерений

между моментальными снимками по номером j и под номеров i, то есть между

и

), предназначенные для сборки в кластеры вместе с векторами из фиг. 2В (и идентификации, в какой паре составляющие I и J пласта замещают друг друга, между получаемыми периодически моментальными снимками или рассматриваемыми моментальными снимками при многочисленных глубинах исследования под номером j и под номером i), заранее известны, то, чтобы на практике эффективно различать эти кластеры друг от друга, может оказаться желательным вычисление гистограмм точек данных для каждого телесного угла в n-мерном пространстве или нормирование векторов

точек данных на единичную длину (то есть, проецирование их на n-мерную единичную сферу) в соответствии с выражением:Because data points are from the resulting measurement differences, such as

(difference in logging measurements

between snapshots at number j and at numbers i, i.e. between

and

) intended for assembly in clusters together with the vectors from FIG. 2B (and identifications in which pair the components of the I and J layers replace each other, between periodically obtained snapshots or considered snapshots at numerous exploration depths under the number j and under the number i), it is known in advance to effectively distinguish between these clusters from each other, it may be desirable to calculate histograms of data points for each solid angle in n-dimensional space or normalization of vectors

data points per unit length (that is, projecting them onto an n-dimensional unit sphere) in accordance with the expression:

точек данных, превышающими заданный порог

шума, а норму

можно задавать рядом способов. То есть, также возможны альтернативное псевдонормирование или другие способы проецирования. Псевдонормированием можно определенно выявлять наличие некоторых картин замещения x-составляющих, когда замещение вытекает из заметных различий

между моментальными снимками под номером j и под номером i. Кроме того, можно использовать другие статистические методы, такие как, например, факторный анализ или анализ главных компонентов (как описано в заявке №13/658502 на патент США под названием «Identifying formation, matrix and fluid related characteristics from subsurface data using factor analysis» (Jain et al.), которая была подана 23 октября 2012 года и переуступлена правопреемнику настоящей заявки и тем самым полностью включена в эту заявку путем ссылки, факторный анализ или анализ главных компонентов особенно подходит к данным ядерного магнитного резонанса). Факторный анализ или анализ главных компонентов можно применять для идентификации размерности соответствующего векторного пространства, вытекающей из различий измерений, таких как

(то есть, числа главных факторов и главных компонентов, которые могут учитываться для большой части набора данных). Эта размерность может соответствовать числу составляющих, которые замещают друг друга между рассматриваемыми моментальными снимками под номером i и под номером j за вычетом единицы. Следует понимать, что размерность также может быть установлена с помощью посторонних эмпирических данных, таких как, например, данные каротажа по буровому раствору или обоснованные предположения. Кроме того, способы наклонного вращения можно применять для образования картин замещения различных имеющихся x-составляющих. Кроме того, технологии искусственных нейронных сетей (НС) можно использовать для автоматической сегрегации пласта на зоны в соответствии с картинами, получаемыми при обучении искусственной нейронной сети. Упомянутые выше статистические методы можно применять ко всему рассматриваемому интервалу глубин, или от зоны к зоне, или к скользящему интервалу глубин.Such rationing, which may be called pseudonorming below, can be limited by vectors

data points exceeding a given threshold

noise, and the norm

can be set in a number of ways. That is, alternative pseudo-rationing or other projection methods are also possible. Pseudo-feeding can definitely detect the presence of certain patterns of substitution of x-components, when the substitution results from noticeable differences

between snapshots j and i. In addition, other statistical methods can be used, such as, for example, factor analysis or principal component analysis (as described in US Patent Application No. 13/658502, entitled “Identifying formation, matrix and fluid related characteristics from subsurface data using factor analysis” (Jain et al.), Which was filed October 23, 2012 and assigned to the assignee of this application and thereby is fully incorporated into this application by reference, factor analysis or analysis of the main components is especially suitable for nuclear magnetic resonance data). Factor analysis or analysis of the main components can be used to identify the dimension of the corresponding vector space resulting from differences in dimensions, such as

(i.e., the number of major factors and major components that can be considered for most of the data set). This dimension can correspond to the number of components that replace each other between the considered snapshots under number i and under number j minus one. It should be understood that the dimension can also be established using extraneous empirical data, such as, for example, mud log data or reasonable assumptions. In addition, oblique rotation methods can be used to form substitution patterns of various available x-components. In addition, the technology of artificial neural networks (NS) can be used for automatic segregation of the formation into zones in accordance with the pictures obtained during the training of an artificial neural network. The statistical methods mentioned above can be applied to the entire considered interval of depths, or from zone to zone, or to a moving interval of depths.

, и когда эти точки данных отображаются при норме выше заданного порога

шума. Такое псевдонормирование позволяет проецировать точки данных, соответствующие замещению одной и той же пары x-составляющих, на одну и ту же точку единичного круга (или единичной сферы). Для согласования с рассматриваемыми n каротажными измерениями показанное на фиг. 2Q предполагается находящимся в n-мерном пространстве. На фиг. 2R показано псевдонормирование, при котором норму принимают равной

, и когда эти точки данных отображаются при норме выше заданного порога

шума. Показанное на фиг. 2R также предполагается находящимся в n-мерном пространстве.In FIG. 2Q shows how pseudonorming works, in which the norm is taken equal

, and when these data points are displayed at a rate above a given threshold

noise. Such pseudonorming allows you to project data points corresponding to the replacement of the same pair of x-components, on the same point of the unit circle (or unit sphere). To align with the n logs in question, shown in FIG. 2Q is assumed to be in n-dimensional space. In FIG. 2R shows pseudo-normalization in which the norm is taken equal

, and when these data points are displayed at a rate above a given threshold

noise. Shown in FIG. 2R is also assumed to be in n-dimensional space.

The feature of nuclear magnetic resonance described earlier can lead to the following modified equations:

,

,

(in which "non-wetting fluids" means "non-wetting fluids" and "wetting fluids" means "wetting fluids"), which in the case of non-wetting fluid and under the assumption that the second, made after a period of time or from another of the many depths of the study, is instantaneous the picture includes only the “bound fluid” component, with the exception of any “free wettings fluid”, they are reduced to an equation of the form:

,

,

(in which ″ bwf ”stands for“ boundary wetting fluid ”) and it is again written as:

.

.

, показанный выше, обычно связан с

в соответствии с выражением:Also a member

shown above is usually associated with

in accordance with the expression:

,

,

можно использовать другие, точно связанные выражения, которые можно получать на основании физики измерения ядерного магнитного резонанса.(in which the subscript “lm” means “logarithmic mean”), where “a” can be considered a relatively constant parameter (for example, about 4.5), and the formulas are valid within the corresponding volumetric relaxation times of the wetting fluid in question. When using the logarithmic mean T _{1, the} formula provides the advantage of freedom from diffusion effects. The above formulas were obtained by equating K _SDR ≈ K _TC (respectively known in the industry as the Schlumberger-Doll Research Center Permeability Transformation (SDR) and Timur-Coates Permeability Transformation (TS)). Also for member

You can use other precisely related expressions that can be obtained on the basis of the physics of measuring nuclear magnetic resonance.

The expressions given above serve to demonstrate that the wetting fluid actually behaves like a component with an effective, well-defined signature in logging measurements of nuclear magnetic resonance, as represented by the expression below:

,

,

which includes a counterintuitive two-peak response in nuclear magnetic resonance measurements, instead of just one peak, and where one of the peaks is positive and the other peak is negative instead of only two positive peaks. In this case, the methods of clustering data points, pseudo-normalization methods, or other statistical methods and classification methods discussed earlier can also be applied to nuclear magnetic resonance measurements.

In FIG. 3A-3D show sets of simulated and obtained effective signatures 90a-90d of the T ₂ distribution _of water from nuclear magnetic resonance measurements, from which it can be seen that they are comparable. Depending on the actual thresholds or T ₂ boundary values for the associated fluid, the simulated signatures retain the same shape, but shift to the right or left, depending on whether the T ₂ boundary values _{of the} associated fluid are more or less, respectively. These unique two-peak effective signatures (including one positive peak and one negative peak with correlated amplitudes) help distinguish between water (e.g., wetting fluid) and other non-wetting fluids present (e.g., oil or gas). In FIG. 3A shows various simulated effective water signatures minus processing artifacts, shown similar within the horizontal shift. In FIG. 3B shows simulated regularization processing artifacts illustrating which additional peaks and / or troughs can be obtained in practice. The smaller the pore spaces in the formations, the more pronounced are the effects that appear on a graph such as in FIG. 3B.

FIG. 3C contains graphs showing an effective water signature obtained using the example well log data shown in FIG. 4 and the preparation methods described below. Curve 91 is identical to curve 213, which is shown in FIG. 21, and curve 92 represents the signature of nuclear magnetic resonance measurements for oil-based mud filtrate (BRNO) obtained at the same depth using the methods described below. Curves 91, 92 were normalized using the hydrogen index and polarization factor obtained from FIG. 22 and combined to reconstruct curve 93 using the reconstruction methods described below. In FIG. 3D shows the simulated effective water signature of FIG. 3C, shown for comparison with the actual signature of FIG. 3C, and shows a good fit, expressed by curve 94.

In FIG. 4 shows nuclear magnetic resonance data sets used as examples in this application. The tracks, respectively, from left to right, display the gamma-ray curve (track 100), the distribution of T ₂ nuclear magnetic resonance according to the data recorded during drilling of the borehole (that is, the first measurement is track 101), the distribution of T ₂ nuclear magnetic resonance according to the results the passage during cleaning (that is, the second measurement is track 102), and the differential distribution of T ₂ nuclear magnetic resonance between the two mentioned above (that is, obtained by subtracting measurements while drilling from measurements after the passage during cleaning, - track 103). Note that in the case of a passage during cleaning, certain types of non-wetting and wetting fluids (non-wetting filtrate of oil-based drilling mud and type (s) of residual hydrocarbon), in contrast to the type (s) of wetting bound water, are separated in accordance with the boundary value T _{2 of} nuclear magnetic resonance of approximately 90 ms.

, и на которых отображены точки данных при норме выше заданного порога

шума.In FIG. 5-10 show the application of the pseudo-rationing methods described above. In FIG. 5 presents a family of graphs 105 of deep logs, which illustrate pseudonorming, in which the norm is taken equal

, and on which data points are displayed at a norm above a given threshold

noise.

, и на которых отображены точки данных при норме выше заданного порога

шума.In FIG. 6 shows a family of depth logging charts 106 that illustrate pseudo-normalization, in which the norm is taken equal

, and on which data points are displayed at a norm above a given threshold

noise.

, и на которых отображены точки данных при норме выше заданного порога

шума.In FIG. 7 presents a family of graphs 107 of deep logs, which illustrate pseudo-normalization, in which the norm is taken equal

, and on which data points are displayed at a norm above a given threshold

noise.

, и на которых отображены точки данных при норме выше заданного порога

шума.In FIG. 8 is a family of depth logs 108 that illustrate pseudo-normalization in which the norm is assumed to be

, and on which data points are displayed at a norm above a given threshold

noise.

, и на которых отображены точки данных при норме выше заданного порога

шума.In FIG. 9 presents a family of graphs 109 of deep logs, which illustrate pseudo-normalization, in which the norm is taken equal

, and on which data points are displayed at a norm above a given threshold

noise.

, и на которых отображены точки данных при норме выше заданного порога

шума.In FIG. 10 shows a family of plots 110 for logging, which illustrates pseudo-normalization, in which the norm is taken equal

, and on which data points are displayed at a norm above a given threshold

noise.

. На фиг. 12 показаны выбранные картины 112, являющиеся результатом перестановок пар x-составляющих, полученные при использовании

. На фиг. 13 показаны выбранные картины 113, являющиеся результатом перестановок пар x-составляющих, полученные при использовании

.In FIG. 11 to 13 show pseudo-normalized patterns corresponding to permutations of pairs of different x-components. In FIG. 11 shows the selected patterns 111 resulting from permutations of pairs of x-components obtained using

. In FIG. 12 shows selected patterns 112 resulting from permutations of pairs of x-components obtained using

. In FIG. 13 shows the selected patterns 113 resulting from permutations of pairs of x-components obtained using

.

Changes in logging or processing of logging data lead to various problems. Changes in data recording can be adjusted by making the specific assumption that nuclear magnetic resonance measurements based on the first logging measurement can be converted to equivalent measurements based on the second measurement or vice versa. This method is essentially a comparison of nuclear magnetic resonance data sets recorded together with hypothetical similar parameters. In general, the changes caused by the processing can be dealt with by considering the resulting low-amplitude peaks and troughs as noise only, and a corresponding increase in the threshold level for noise is required.

One consequence of pseudo-normalization is that clusters of x-component substitution data points corresponding to parallel vectors may not differ from each other. In addition, clusters of x-component displacement data points collected around the beginning of O and corresponding to a pair of components with similar properties, such as natural reservoir oil displaced by oil-based mud filtrate (BRNO), or natural reservoir water displaced by drilling mud filtrate on water-based (BRVO), may not differ significantly from other clusters of data points corresponding to other permutations of pairs of x-components. These circumstances are discussed below.

In another example, it is assumed that it is possible to evaluate the quantitative indicator of one of the components, for example, component J, which takes part in the replacement of the x-component. Depending on the situation, this can be done in a number of ways, for example, using measurements of nuclear magnetic resonance, resistivity, thermal neutron capture cross section, dielectric measurements, gamma radiation measurements, lithium-density / neutron measurements, or acoustic measurements. In this case, other properties of the components can be accurately reconstructed and therefore the expression “true rationing” is introduced, as opposed to pseudo-rationing, in accordance with mathematical expressions:

and

или ограниченный контраст между составляющими I и J приводит к векторам

.where component I is another component that takes part in the substitution, and can also be extended to other components. During this reconstruction, the circumstances considered above are taken into account, namely, substitutions of various x-components lead to parallel vectors

or limited contrast between components I and J leads to vectors

.

при истинном нормировании. Точки данных, получающиеся в результате истинного нормирования, которые до истинного нормирования могли быть объединены в кластеры вдоль линий замещения различных x-составляющих, теперь к тому же перегруппировываются и срастаются вокруг крупных отметок, показанных на фигуре. На фиг. 16В показано соотношение (соотношения)

для сопоставления с фиг. 2А, иллюстрирующее, каким образом реконструируются сигнатуры надлежащих составляющих в каротажных измерениях. (Фиг. 16А была повернута на 180° для получения члена

).In FIG. 16A shows vectors

with true rationing. Data points resulting from true normalization, which before true normalization could be clustered along the replacement lines of various x-components, now also regroup and grow together around the large marks shown in the figure. In FIG. 16B shows the relationship (s)

for comparison with FIG. 2A illustrating how the signatures of the appropriate components in the logging measurements are reconstructed. (Fig. 16A was rotated 180 ° to obtain a member

)

In addition, a situation may arise as when working with measurements of nuclear magnetic resonance, when the quantitative parameter of one of the components that take part in the replacement of the x-components is known accurate to a constant factor, but not accurate to an absolute value. For example, the volume of drilling fluid filtrate shown on nuclear magnetic resonance distributions, in the case of conventional repression drilling, is known to within the hydrogen index and polarization factor, which may also not be known in the borehole conditions at the place of work. In this case, the volume of the mud filtrate on the distribution of nuclear magnetic resonance can be called the apparent (apparent) volume of the mud filtrate, as opposed to the true volume of the mud filtrate after correction for the influence of the hydrogen index and polarization factor for the above reason.

In this case, the equations given above can be converted to the following form:

,

,

,

,

AND

Or

,

,

и

обозначают для составляющих I и J нормированные сигнатуры каротажных измерений ядерного магнитного резонанса (что в данном случае означает нормированные относительно 100 единиц пористости или в конкретном случае означает распределение ядерного магнитного резонанса с общей площадью под распределением ядерного магнитного резонанса, равной 100 единицам пористости), то есть, с исключением влияний водородного индекса и поляризационного фактора.where HI is the hydrogen index and Pol is the polarization factor, and

and

denote for components I and J the normalized signatures of log measurements of nuclear magnetic resonance (which in this case means normalized to 100 units of porosity or, in a particular case, means the distribution of nuclear magnetic resonance with a total area under the distribution of nuclear magnetic resonance equal to 100 units of porosity), i.e. , with the exception of the effects of the hydrogen index and polarization factor.

для случая кажущегося (apparent) нормирования (в противоположность истинному нормированию) с использованием кажущегося объема фильтрата бурового раствора. Точки данных, полученные в результате моделирования, теперь объединены вокруг увеличенных (или уменьшенных, в зависимости от конкретного случая) крупных отметок, показанных на фигуре, по сравнению с истинным нормированием, которое включает в себя использование подходящего фильтрата бурового раствора. На фиг. 16D отображено соотношение (соотношения)

в противоположность фиг. 2А, иллюстрирующей, каким образом реконструируются кажущиеся сигнатуры каротажных измерений для составляющих. (Фиг. 16С повернута на 180° для получения члена

). Как можно видеть, реконструированные каротажные отклики все же располагаются вдоль векторов, показанных на заднем плане, это означает и подтверждает то, что они отличаются от истинных сигнатур каротажных измерений для составляющих постоянным или масштабирующим множителем, а именно, HI_J·Pol_J. В таком случае достаточно, чтобы каротажный отклик ядерного магнитного резонанса от одной из составляющих, участвующих в замещениях x-составляющих, был известен, чтобы можно было определять масштабирующий множитель, используемый для реконструкции других истинных сигнатур каротажных измерений для связанных составляющих.In FIG. 16C mapped vectors

for the case of apparent (apparent) rationing (as opposed to true rationing) using the apparent volume of the mud filtrate. The data points obtained from the simulation are now grouped around the enlarged (or reduced, as the case may be) large marks shown in the figure, compared to true normalization, which includes the use of a suitable mud filtrate. In FIG. 16D displays the ratio (s)

in contrast to FIG. 2A illustrating how apparent logging measurements for components are reconstructed. (Fig. 16C is rotated 180 ° to obtain a member

) As can be seen, the reconstructed logging responses still located along the vectors shown in the background, and it is confirmed that they differ from the true signature log measurements for constituents constant or scaling factor, namely, HI _J · Pol _J. In this case, it is sufficient that the logging response of nuclear magnetic resonance from one of the components involved in the substitutions of the x-components be known, so that it is possible to determine the scaling factor used to reconstruct other true logging measurements for the associated components.

или

, чтобы также идентифицировать эти картины замещения x-составляющих, которые можно упускать при использовании только псевдонормирования (то есть когда перестановки различных x-составляющих приводят к образованию параллельных векторов

или когда ограниченный контраст между составляющими I и J приводит к образованию векторов

).At this point, certain conclusions can be drawn. For example, in the methods of statistical analysis and classification described earlier (using histograms, classification schemes based on artificial neural networks or the methodology of factor analysis or analysis of principal components), one can also apply the following true or apparent normalization (as opposed to pseudo-normalization), which is expressed accordingly as

or

to also identify these patterns of substitution of x-components, which can be missed using only pseudo-normalization (that is, when permutations of different x-components lead to the formation of parallel vectors

or when limited contrast between components I and J leads to the formation of vectors

)

In fact, it is enough that the signature of the measurement of nuclear magnetic resonance of one of the components (usually water) involved in the substitutions of the x-components is known, so that it is possible to reconstruct other signatures of measurements of the nuclear magnetic resonance of the related components, in accordance with which refinements and corrections for the hydrogen index and polarization factor is determined first, and actual responses are determined secondarily. This can be expressed as:

,

,

,

,

,

,

обозначает пористость, определяемую по данным ядерного магнитного резонанса (то есть общую площадь под распределением

в конкретном случае распределений ядерного магнитного резонанса), а член

обозначает различие пористости, определяемое по данным ядерного магнитного резонанса, между моментальными снимками под номером j и под номером i. Вновь введенные члены Slope_{IJ pair exchange} и Slope_{KJ pair exchange} имеют отношение к новому семейству кроссплотов с

по оси X и

по оси Y. Наклоны (slopes) соответствуют прямым линиям, проходящим через начало.where J is a component that replaces components I or K, and component I is a component with a priori known HI _I · Pol _I , in which case HI _J · Pol _J is determined first, and the other HI _K · Pol _{K are} determined secondarily . Newly entered member

denotes the porosity determined by nuclear magnetic resonance data (i.e., the total area under the distribution

in the specific case of nuclear magnetic resonance distributions), and the term

denotes the difference in porosity, determined by nuclear magnetic resonance data, between snapshots at j and i. The newly introduced members of the Slope _{IJ pair exchange} and Slope _{KJ pair exchange} are related to the new cross-slot family with

along the x axis and

along the Y axis. Slopes correspond to straight lines passing through the beginning.

по оси x и

по оси y. Точки данных, полученные в результате замещения (замещений) различных x-составляющих, образуют кластеры вдоль линий с различным наклоном в соответствии с HI·Pol каждой составляющей. Двумерную (2М) диаграмму из фиг. 17А можно распространить на многие измерения. В случае многомерных отображений можно рассматривать кластеры

,

векторов и гистограммы, построенные для каждого телесного угла. В другом полезном кроссплоте

отложены по оси y, а

отложены по оси x, и в этом случае, как показано на фиг. 17В, различные наклоны считываются как

. На фиг. 17В показаны точки данных, полученные в результате замещения (замещений) различных x-составляющих, образования кластеров вдоль линий с различными наклонами в соответствии с HI·Pol каждой составляющей.In FIG. 17A shows a new cross-slot family with

along the x axis and

along the y axis. Data points obtained as a result of substitution (s) of various x-components form clusters along lines with different slopes in accordance with the HI · Pol of each component. The two-dimensional (2M) diagram of FIG. 17A can be extended to many dimensions. In the case of multidimensional mappings, we can consider clusters

,

vectors and histograms plotted for each solid angle. In another useful crossplot

plotted along the y axis, and

plotted along the x axis, in which case, as shown in FIG. 17B, various slopes read as

. In FIG. 17B shows data points resulting from the substitution (s) of various x-components, the formation of clusters along lines with different slopes in accordance with the HI · Pol of each component.

в каротажных измерениях ядерного магнитного резонанса вследствие неблагоприятных влияний обработки, которые упоминались ранее, в общем случае, и регуляризации, в частности. Особенностью ядерного магнитного резонанса определяется, что альтернативно отклики на различные флюиды при каротаже ядерного магнитного резонанса можно получать с использованием связанных ограничениями нелинейных способов подбора кривой на основе наименьших квадратов, примененных в примерах из фиг. с 26А по 26F и с 27А по 27Н, чтобы исключать эти неблагоприятные влияния. Эквивалент из фиг. 21 также можно выполнять в пространстве амплитуд спинового эха. То есть вместо вычитания распределений друг из друга сначала можно вычесть амплитуды последовательности эхо-сигналов друг из друга и затем выполнить обращение разностей последовательности эхо-сигналов. Некоторые ограничения можно ввести для исключения нестабильности инверсии от положительных и отрицательных амплитуд, одновременно присутствующих в такой разностной последовательности эхо-сигналов. Кроме того, в данном случае можно применить факторный анализ или анализ главных компонентов, в том числе связанные ограничениями.In FIG. 21, 22, and 3B, using a practical example, successively shows how this works. However, although the corrections obtained for the effects of the hydrogen index and polarization factor shown in FIG. 21 may be considered reliable, the same may not necessarily be true with respect to the responses received.

in logging measurements of nuclear magnetic resonance due to the adverse effects of processing, which were mentioned earlier, in the General case, and regularization, in particular. A feature of nuclear magnetic resonance determines that, alternatively, responses to various fluids during nuclear magnetic resonance logging can be obtained using constrained non-linear least squares curve fitting methods applied in the examples of FIG. 26A to 26F and 27A to 27H to eliminate these adverse effects. The equivalent of FIG. 21 can also be performed in the space of spin echo amplitudes. That is, instead of subtracting the distributions from each other, you can first subtract the amplitudes of the echo sequence from each other and then reverse the differences in the sequence of echo signals. Some restrictions can be introduced to eliminate the instability of the inversion from the positive and negative amplitudes that are simultaneously present in such a difference sequence of echo signals. In addition, in this case, factor analysis or analysis of the main components, including those associated with limitations, can be applied.

(то есть кажущемуся объему свободного флюида при использовании граничного значения Т₂, составляющего 90 мс), и на этих графиках отображены точки данных при норме выше заданного порога

шума. На фиг. 19 представлено семейство графиков 190 глубинного каротажа, которыми иллюстрируются результаты псевдонормирования, где норма выбрана равной

, и на этих графиках отображены точки данных при норме выше заданного порога

шума. На фиг. 20 представлено семейство графиков 199 глубинного каротажа, которыми иллюстрируются результаты псевдонормирования, где норма выбрана равной

, и на этих графиках отображены точки данных при норме выше заданного порога

шума. На фиг. 21 показаны истинные нормированные картины 210, соответствующие перестановкам пар различных x-составляющих, где объем фильтрата бурового раствора на нефтяной основе, принимавшего участие в перестановках пар x-составляющих был приравнен к

. Распределения, которые являются активными в коротком спектре распределения времени Т₂, были исключены для защиты от шума и артефактов вследствие несоответствия по глубине. Картины 211-213 соответствуют арифметическому среднему исходных (то есть, менее затененных) кривых. На картинах 210 виден оттенок на двух различных несмачивающих углеводородных флюидах и одном смачивающем флюиде (воде).In FIG. 18 is a family of 180 depth logs illustrating pseudo-normalization results in which the norm is assumed to be

(that is, the apparent volume of free fluid when using a boundary value of T _{2 of} 90 ms), and these graphs show data points at a rate above a given threshold

noise. In FIG. 19 presents a family of plots of 190 depth logs that illustrate the results of pseudo-normalization, where the norm is chosen equal to

, and these graphs show data points with a norm above a given threshold

noise. In FIG. 20 presents a family of plots 199 of deep logs, which illustrate the results of pseudo-normalization, where the norm is chosen equal to

, and these graphs show data points with a norm above a given threshold

noise. In FIG. Figure 21 shows the true normalized patterns 210, corresponding to permutations of pairs of different x-components, where the volume of oil-based drilling mud filtrate that participated in the permutations of pairs of x-components was equated to

. Distributions that are active in the short spectrum of the T ₂ time distribution were excluded for protection against noise and artifacts due to depth mismatch. Pictures 211-213 correspond to the arithmetic mean of the original (i.e., less shaded) curves. In pictures 210, a shade is visible on two different non-wetting hydrocarbon fluids and one wetting fluid (water).

последний может быть загрязнен остаточными свободными флюидами в результате выполнения бурения.In FIG. Figure 22 shows how the combined effects of the hydrogen index and polarization index are identified and then adjusted. Data points corresponding to fluids with the same HI · Pol value form clusters along lines with identical slopes or they can lead to the formation of characteristic peaks (or vibrations) in the histogram of the corresponding slopes. In this case, the X axis corresponds to the apparent volume of the oil-based drilling mud filtrate according to the results of the passage during cleaning, and the Y axis corresponds to the change in porosity according to nuclear magnetic resonance between the passage during drilling and the passage during cleaning. The water point is taken to reflect the combined hydrogen index and polarization factor making up the unit (i.e., 100 units of porosity), in which case it determines the HI · Pol value for the oil-based drilling mud filtrate. In turn, this value can be extended to the other two unknown identified hydrocarbon fluids. One relevant consideration in implementing this procedure is that when estimating Δ _ij (V _J ) ^apparent (i.e., the volume of oil-based drilling mud filtrate in this example) based on

the latter may be contaminated with residual free fluids resulting from drilling.

Although it is of interest to identify the patterns resulting from the replacement of the x-components when the substitution occurs in pairs, there is at least one exceptional case, which suggests that when individual logging measurements for the components are assigned, the actual borehole logging measurements can be converted to elemental volumetric share. Now a counterintuitive example will be explained in which the logs of the logging measurements for the various components are not individually calibrated, and the group of components can be fully characterized and assigned one new defined group signature at the same time, and the replacement will not occur in pairs. A specific example is used to describe how this will work with respect to true porosity.

In FIG. 14A to 15 illustrate this situation leading to the determination of the characteristics of the porosity space (or the porosity hyperplane in the case of a space with n-1 dimension), which is the concept defined below. In FIG. 14A shows the various endpoints Fld _A , Fld _B , Fld _I , Fld _J , Fld _{Z of the} fluid and a single endpoint Mtx of the equivalent bulk of the rock (i.e., the only point representing the mineralogy of the rock in the subterranean formation) instead of various individual endpoints (e.g. Min _A , Min _Z ) mineral. In addition, the point M of the measurements is shown. When the mixing rules for the various considered borehole logging measurements m ₁ , ... m _α , ... m _n are linear (in terms of volumetric parameters), the measurement point M corresponds to the center of gravity of the above-mentioned fluid and the end points of the bulk of the rock, weighted by the corresponding volume fractions in percent present in the volume of the measured formation. In FIG. 14B shows what is called a porosity space that is covered by the end points of Fld _A , Fld _B , Fld _I , Fld _J , Fld _Z fluid.

In FIG. 14C shows a geometric visualization of a measurement of the true porosity of a subterranean formation, calculated as 1 minus the ratio of the distance of the measurement point M to the porosity space to the distance of the end point Mtx of the bulk of the rock to the porosity space. In FIG. 14D shows that displacements or displacements of the end points Fld _A , Fld _B , Fld _I , Fld _J , Fld _{Z of the} fluid relative to the space of porosity in no way affect the calculated true porosity. Therefore, in general, true porosity depends on the position and orientation of the porosity space, but not on the specific position of the individual fluid end points in this porosity space. In the case of working with measurements of nuclear magnetic resonance, you can apply the very concepts and methods. However, it should be noted that in practice the porosity space should be covered by the corresponding nuclear magnetic resonance signatures for non-wetting fluids and effective nuclear magnetic resonance signatures for wetting fluids that were introduced and whose existence was shown earlier in this description.

In FIG. 14E shows that a hyperplane can be well defined, for example, by a vector orthogonal to it and a point belonging to it (the Fld _Z point in this figure). Although this definition is very convenient, it is clear that it is not unique. Strictly speaking, n parameters can be used to determine a hyperplane with dimension n-1 inside n dimensions. In the general case, when n fluid components are present inside the porosity, the bulk of the rock is represented as the only equivalent point in the bulk of the rock and n measurements are considered, the calculation of porosity may include n known parameters as opposed to n ² parameters that are used to characterize individually n final points of fluid components.

In FIG. 14F shows how, in practice, the porosity hyperplane can be characterized. When the pore space in the sample of the measured formation is occupied by various fluids that can replace each other, and the measurements are repeated, then the measurement point M will move relative to the hyperplane parallel to the porosity hyperplane. By measuring or combining the displacement vectors of such measurement points, the orientation of the porosity hyperplane is uniquely determined (i.e., the direction of the vector orthogonal to such a porosity hyperplane is uniquely determined). In addition, fluid components are not rearranged in pairs, but possibly in triplets, etc.

More specifically, the mathematical expression for the distance of any point of M measurements to the porosity hyperplane is simply a linear sum of the form:

d (M) = p ₀ + p ₁ · m ₁ + ... + p _α · m _α + ... + p _n · m _n ,

where p ₀ , p ₁ , ... p _α , ... p _n are constant parameters that describe the orientation and position of the porosity hyperplane in space (for (p ₁ ) ² + ... (p _α ) ² + ... (p _n ) ² = 1 )

The corresponding parameters p ₀ , p ₁ , ... p _α , ... p _n can be determined using various methods. In one such method, porosity from core analysis data is used as a reference and true porosity is expressed as a linear combination of available measurements with the addition of a constant. After that, two data sets can be aligned.

In relation to this example, it is possible to perform logging while drilling the entire interval of the borehole (snapshot during the passage during drilling). Logging of the same interval can be repeated after drilling (a snapshot during the passage during cleaning or during the passage when expanding the wellbore), while assuming normal repression drilling, different fluids can occupy the same pore space at different points in time. You can study the specific parameters p ₀ , p ₁ , ... p _α , ... p _n , which minimizes the difference between d (M ^drilling ) and d (M ^cleaning ) throughout the entire logging interval under the limiting condition d (M _(FldZ) ) = 0 . This operation is used to uniquely determine the orientation and position in space of the porosity hyperplane.

An expression can be obtained when the considered measurements m ₁ , ... m _α , ... m _{n are} first of all converted into the corresponding measurements Phi ₁ ^(a) , ... Phi _α ^(a) , ... Phi _n ^{(a) of the} apparent porosity. The process of converting measurements to apparent porosities is based on the assumed a priori composition of the bulk of the rock and the estimated a priori composition of the fluid occupying the pore space of the rock. For the sake of proof, we can assume that the bulk of the rock is limestone, and assume that the fluid is water with a specific salinity. In practice, this means that when the bulk of the rock is actually limestone, and the fluid is actually water, then the measurements Phi ₁ ^(a) , ... Phi _α ^(a) , ... Phi _n ^{(a) of the} apparent porosity will be equal to each other and equal true porosity of the reservoir.

In practice, the following considerations can be taken into account when choosing the a priori assumed types of the bulk of the rock and fluid to convert measurements into apparent porosities. The estimated fluid used in apparent porosities should be one of the fluids present. If free water is present in the form of formation water or a water-based drilling mud filtrate, this water can be used to produce apparent porosities. For the apparent bulk of the rock, you can use the prevailing lithology that will occur.

In FIG. 14G shows how the preceding FIG. (14A-14F) can be modified when working with apparent porosity measurements, and not with actual measurements. Essentially, the endpoint of the limestone is now located at the origin, and the endpoint of the water has coordinates (1, 1, ..., 1). This is very convenient since a new apparent porosity PHI ^(a) _P1, ... _Pα, ... _Pn , defined as a linear mixture of traditional apparent porosities, can be introduced:

PHI ^(a) _P1, ... _Pα, ... _Pn = P ₀ · Phi ₁ ^(a) + ... + P _α · Phi _α ^(a) + ... P _n · Phi _n ^(a) .

In addition, the above procedure for minimizing the difference between d (M ^drilling ) and d (M ^cleaning ) can be replaced by searching for specific parameters P ₁ , ... P _α , ... P _n , which minimizes the difference between the apparent porosities (PHI ^(a) _P1, ... _Pα, ... _Pn ) ^(drilling) and (PHI ^(a) _P1, ... _Pα, ... _Pn ) ^cleaning throughout the logging interval under the limiting condition PHI ^(a) _P1, ... _Pα, ... _Pn (water) = P ₀ + ... + P _α + ... + P _n = 1.

In FIG. 15, a graph 150 is shown as a real example of such minimization, while the fluids present were water, oil, and gas, and the measurements considered were bulk density, neutron logging porosity, and thermal neutron capture cross section. The figure shows the apparent porosities (PHI ^(a) _P1, ... _Pα, ... _Pn ) ^(drilling) and (PHI ^(a) _P1, ... _Pα, ... _Pn ) of ^cleaning along the vertical axis depending on the depth along the horizontal axis after the optimization process, described above. As you can see, there is a real imposition of apparent porosities calculated from measurements during the passage during drilling (curve 151) and the passage during cleaning (curve 152).

To obtain the resulting apparent porosity, PHI ^(a) _P1, ... _Pα, ... _Pn, it is not required that the oil and gas parameters be known (which are also very difficult to model in practice), and it can be easily extended to neighboring reference wells without the need requirement that the new pass when cleaning whenever the registration (i.e., after the parameters P _1, ... P _α, ... P _n are defined for the fluids set throughout the length of the well, the re-definition of the adjacent bearing hole can be eliminated when also awaiting the tsya that will meet the fluids of the same species).

Although now the curves for the results of the passage during drilling and the passage during cleaning are superimposed, some residual statistical noise may remain. Therefore, we can take the average porosity during the passage during drilling and passage during cleaning as the final apparent porosity, and the statistical error between porosities during the passage during drilling and passage during cleaning can be used to assess the accuracy of such a calculated final true porosity.

Using the above procedure to minimize differences between apparent porosities from different snapshots, despite the occurrence of x-component substitutions, results in a final apparent porosity (i.e., resulting identical apparent porosities during drilling passage and cleaning passage), which gives a new the advantage is that there is no dependence on the actual composition of the fluid present in the pore space, and dependence on the mineralogy of the rock. In this case, the apparent porosity PHI ^(a) _P1, ... _Pα, ... _Pn can be converted to true porosity by performing the final computational operation below (according to Fig. 14C, the porosity is calculated as 1 minus the ratio of the distance of the measurement point M to the porosity hyperplane to the distance of the end point Mtx of the bulk of the rock to the porosity hyperplane):

.

.

Note that although the consideration in this application relates to fluids occupying the pore space, the method described in this application is also suitable for the case in which the mineral salt partially clogs the pore space and then passes into the solution when the filtrate of the water-based drilling fluid penetrates into the reservoir. In this case, although there is no fluid, the salt is considered along with the components included in the porosity hyperplane. The requirements for determining true porosity are particularly satisfied by an appliance sold under the trademark EcoScope ™ (which is a trademark of Schlumberger Technology Corporation), since the various measurements obtained from such an instrument are consistent and the instrument provides up to four different apparent porosity measurements that can be mixed to find true porosity relative to four different fluids occupying the pore space. In practice, these four porosities are not used simultaneously, and the number of measurements used corresponds to the maximum number of different fluids occupying the pore space of the formation.

The automatic method should begin, for example, in the presence of apparent porosity according to neutron logging data to select the number of measurements used and to determine whether the measurements are consistent during the passage during drilling and passage during cleaning. If they are not consistent, then in addition, the porosity is considered according to the density logging data and it is determined whether previous measurements can be mixed to make consistent measurements during the passage during drilling and the passage during cleaning. If this still does not work, then the apparent porosity is additionally considered according to the measurement data of the thermal neutron capture cross section and it is determined whether three types of measurements can be mixed to make consistent measurements during the passage during drilling and passage during cleaning. The above process can be continued using the parameters of other additional measurements. An additional method may be neutron spectroscopy (for example, thermal neutron capture spectroscopy) or mud logging, which provides information about the bulk of the rock, allowing you to calculate PHI ^(a) _P1, ... _Pα, ... _Pn (Mtx).

And in this case, when mineral salt, etc. partially clogs the pore space and then passes into the solution, while the water-based mud filtrate penetrates the formation, the resulting penetration profile becomes unusual, two-interval (that is, including two interfaces or boundaries) with an axisymmetric penetration profile relative to the borehole, the volume of the formation located between the wellbore and the first interface closest to the borehole includes a salt-free zone and an intact ph water-based drilling mud liquor, the next deeper reservoir volume located between the two interfaces includes intact salt clogging the pore space and water-based drilling fluid filtrate saturated with salt, and the remaining external reservoir volume located after the second, the deepest interface, includes an undisturbed layer, which can be examined and determined using a specific set of consistent measurements from a device such as p EcoScope ^TM , including when performing registration of data from periodic observations (that is, data recorded by the results of the passage during drilling and passage during cleaning).

The following describes the application of methods for analyzing logging measurements for autonomous nuclear magnetic resonance data using the recording of periodic observation data or the recording of data from numerous depths of research (or any combination of them). The successful implementation of the methods and procedures described above and additional related operations in the particular case of downhole logging measurements of nuclear magnetic resonance will be described. In addition, a sequence of actions is presented with a completely linear linear interpretation (inversion) of the volumetric fluid parameters using a covariance-based method of volume redistribution.

The final answer resulting from the successful implementation of the methods and procedures described above can be expressed as:

Elemental spectroscopic logging data (for mineralogy) plus nuclear magnetic resonance logging data (for fluid volumetric parameters) provide complete final formation volumetric parameters.

According to the above: (a) a sourceless logging option has been developed (that is, a well logging option that does not include chemical radioactive sources deployed inside the well); (b) two sets of measurements are used (neutron spectroscopy and nuclear magnetic resonance); (c) it is possible to determine the volumetric parameters of three-phase formation fluids (gas, oil and water); (d) and provides substantially immunity over a significant range to the effects of water salinity (e.g., varying salinity of water or demineralized water).

The direct application of the methods described in relation to logging measurements of nuclear magnetic resonance follows from the following. In general, nuclear magnetic resonance measurements contain numerous components that can be interpreted to constitute a multitude of individual and true consonant individual measurements. In addition, nuclear magnetic resonance measurements from gradient instruments (in which the sensitive measurement volume has a magnetic field gradient), in particular, from a probe having a relatively small concentrated, well-defined geometric volume that can be located a few inches from the borehole wall (e.g. 2.75 inches (7 cm) with a borehole diameter of 8.5 inches (21.6 cm) with the proVISION ™). Therefore, the measurement is insensitive to penetration by the first few inches. In this regard, logging measurements during drilling, when they are combined with measurements of the mechanical penetration speed during rapid drilling, and / or when the logging tool during drilling is the lowest device in the layout of the bottom of the drill string, can be considered as a measurement of an unfilled untouched zone (undisturbed zones) with the provision of a given contrast, while penetration can significantly affect the measurement during the passage during cleaning.

Nuclear magnetic resonance signatures for fluids can be relatively simple to identify up to a normalization factor, since different fluid signatures are in different and well-defined areas of the considered nuclear magnetic resonance distribution spectra. The rules for mixing measurements of nuclear magnetic resonance can be accepted linear, which may not apply to other logging measurements, such as resistivity, porosity according to neutron logging, or acoustic logging. Nuclear magnetic resonance data can be displayed in various ways, for example, in the form of distributions, sequences of echo signals or any linear transformation of them, to simulate one-dimensional or multidimensional bins or window sums of echo signals.

The described example corresponds to a detrital medium (i.e., shale-sandstone formations) with formation gas and demineralized water (four parts of salt per thousand parts) present in the pore spaces, and drilled for repression using an oil-based drilling mud. Present formation components may include clay mineral (s) wetting water bound by clay, siltstone (quartz mineral), wetting water capillary bound in silt pores, sandstone (quartz mineral), non-wetting fluid, formation gas in sandstone pores in undisturbed untouched zone, wetting natural water in sandstone pores in an undisturbed untouched zone, non-wetting residual gas in the washed zone (also called the filtrate penetration zone, damaged by the penetration of the drill filtrate new solution), wetting the residual film of bound water, covering the grains of sandstone in the washed zone, and oil-based drilling mud filtrate.

In the future, we will proceed from the assumption of the presence of a natural reservoir gaseous hydrocarbon and water-based fluids, although traces of an additional type of fluid, such as liquid hydrocarbon, were shown in the previous figures and in the classification methods. Although natural reservoir water in sandstone pores was interpreted as the only one and one dominant signature was assigned, for greater detail, numerous water signatures corresponding to different pore sizes can be introduced.

In FIG. 23 shows one embodiment of a sequence of steps in interpreting nuclear magnetic resonance. The actions in the sequence are presented on the left side of the sequence diagram, and the corresponding figures for these actions are shown on the right.

In FIG. 24 and 25 show the integration of T ₂ depth distributions (FIG. 24), followed by integration over the T ₂ distribution function (FIG. 25). Data on the results of the passage during drilling are represented by graphs 240, 250, and data on the results of the passage during drilling are represented by graphs 241, 251 in FIG. 24 and 25. The position of the point at which the curves for passage during drilling and passage during cleaning deviate from each other in FIG. 24 corresponds to the T ₂ boundary value of the associated bound fluid intended for use (in this case 33 ms). The intervals with respect to the distributions of T ₂ below this boundary value can be interpreted, for example, as including clay mineral (s), siltstone (quartz mineral), clay bound water and capillary bound water and silt pores.

In FIG. 26A through 26F show the distributions of T ₂ for drilling passage and cleaning passage, selected from the zones of interest, with mode extraction limited to interband extraction. In FIG. 27A through 27H and 28A, 28B show the results of the selection of curves from the zones and the passages are displayed together, including the underlying extracted Gaussian modes. For nonlinear selection using the nonlinear square method, the same Gaussian modes between different zones are used (for the same measurement pass), the same Gaussian center, but possibly with a different width between different passes during the measurement (in the same zone), and the same (or correlated) Gaussian area between different passages (in the same zone) when the Gaussian modes are related to fluids.

In FIG. Figures 29 and 30 show the coupled apparent nuclear magnetic resonance signatures for various fluid components, normalized to unity (i.e., relative to 100 porosity units). These responses were presented in a T ₂ distribution format or a spin echo amplitude format, but they can be presented in other possible formats depending on the above two by linear conversion, such as one- or multidimensional porosity bins or window sums of echo signals. In FIG. 29 are shown in T ₂ format for nuclear magnetic resonance signatures for various fluid components (obtained from FIGS. 27A through 27H and 28A, 28B). It is noteworthy how changes in fluid signatures between different snapshots (i.e., between a passage during drilling and a passage during cleaning) are taken into account by assigning a signature during a connected passage during drilling and a passage during cleaning (which includes in this case 80 combined components from 40 components during drilling with the addition of 40 components of the distribution of T ₂ nuclear magnetic resonance during cleaning). The mapped signatures are normalized to unity (i.e., 100 porosity units). In FIG. 29 and 30 lines or points 290-297 of the graphs successively correspond to clay bound water; capillary bound water in the pores of silt; free water present in the sandstone pores displaced between the passage during drilling and the passage during cleaning; residual water present in the pores of sandstones during the passage during cleaning; formation gas displaced by the oil-based drilling mud filtrate between the passage during drilling and the passage during cleaning; the residual formation gas present in the sandstone pores during the passage during cleaning; oil-based drilling mud filtrate present in sandstone pores during the passage during drilling; and an additional oil-based drilling fluid filtrate that has penetrated the sandstone pores between the passage during drilling and the passage during cleaning.

In FIG. 30 shows various apparent nuclear magnetic resonance signatures for various fluid components, converted from the T ₂ format specified above to the spin echo amplitude format and reflecting the above Carr-Purcell-Maybum-Gill nuclear magnetic resonance sequence used during the data recording operation. Since the thermal noise present in the initial amplitudes of the spin echoes can be considered uncorrelated, the corresponding covariance matrix is diagonal, which makes the sequence of echo signals more suitable for the subsequent operation of elemental interpretation (inversion) of the volumetric fluid parameters. On the other hand, the T ₂ distributions are the result of an incorrect linear inversion requiring regularization methods that affect the resulting covariance matrix of the corresponding T ₂ components. Working with T ₂ distributions in identifying and classifying the various fluid components present is fairly common, but the echo sequences are better suited for the operation of elemental analysis of fluid volumetric parameters.

In FIG. 31A shows an elemental analysis of fluid volumetric parameters designed to find apparent volumes (i.e., not adjusted for the effect of the hydrogen index and polarization factor). Clay-bound water (curve 310), capillary-bound water in sandstone pores (curve 311), water in sandstone pores during drilling passage (curve 312), oil-based drilling mud filtrate already present during passage at drilling (curve 313), the moving gas present during the passage during drilling (curve 314), and the residual gas observed during the passage during cleaning (curve 315). A simple linear inversion was performed, which by its nature is quickly interactive and easy to implement. In addition, various characteristics of the measurement of input noise and cross-correlation (or lack thereof) are taken into account explicitly by including the corresponding covariance matrix in the inversion. However, since in traditional measurements of the volumetric parameters of the formation, the volumes of the components of the formation are limited to 0≤V≤1, instead of a direct linear solution, such as presented in this application, negative values can be obtained. This may be the result of noise in the input data, the existence of inaccuracies in the obtained nuclear magnetic resonance signatures for fluids, or small depth mismatches between different snapshots. These negative values are transferred to the ordering of the inverse curves, which can be observed in the form of small changes in the presented figure (for example, curve 314 sometimes passes below curve 315 or curve 312 sometimes passes below curve 311). The method proposed to eliminate these negative values consists in redistributing them among the other components of the formation in accordance with the cross-correlation existing between the reversed volumes of the components of the formation. This is quickly achieved by spreading the covariance matrix from the input measurements of nuclear magnetic resonance into the output apparent volumes shown in FIG. 32. The redistribution of volumes is also a linear process and does not affect the speed of the proposed method, interactivity or ease of implementation.

In FIG. 31B shows an elemental analysis of fluid volumetric parameters designed to find corrected apparent volumes after redistributing negative values using the covariance matrix of FIG. 32. Note the lack of ordering of the inverse curves in this case. In FIG. 32 is a visual display of the covariance matrix of the initial apparent volumes (before correction), on which the following components are represented by lines 320-326, respectively: clay bound water; capillary bound water in the pores of silt; water present in sandstone pores; oil-based drilling mud filtrate present in sandstone pores during the passage during drilling; formation gas displaced by the oil-based drilling mud filtrate between the passage during drilling and the passage during cleaning; petroleum-based drilling mud filtrate present in sandstone pores during passage during refining; and residual formation gas present in the pores of the sandstone during the passage during cleaning.

In FIG. Figure 33 shows the final elemental analysis of fluid volumetric parameters designed to find the true fluid volumes after correcting the corrected apparent fluid volumes for the corresponding combined effects of the hydrogen index and polarization factor. From bottom to top, clay-bound water (hatching 330), capillary-bound water in silt pores (hatching 331), water in sandstone pores during drilling passage (hatching 332), oil-based drilling mud filtrate present during drilling passage ( hatching 333), the moving gas present during the passage during drilling (hatching 334), and the residual gas observed during the passage during cleaning (hatching 335). The porosity scale has a value of 0-30% (or units of porosity). In FIG. Figure 34 shows the successful implementation of the petrophysical equation (or position): logs of elemental spectroscopy (for mineralogy) (lane 340) plus logs of nuclear magnetic resonance (for elemental measurement of fluid volumetric parameters) (lane 341) provide everything necessary for a complete assessment of reservoir volumetric parameters (lane 342).

In FIG. 35, an example of a computing system 200 is shown. Computing system 200 may be an individual computer system 201A or an arrangement of distributed computer systems. Computer system 201A may include one or more analysis modules 202 that are configured to perform various tasks in accordance with some embodiments, such as the tasks shown in FIG. 23. These various tasks, analysis modules 202 can perform independently or in conjunction with one or more processors 204, which can be connected to one or more storage media 206. In addition, the processor (s) 104 may be connected to a network interface 208 to provide communication of the computer system 201A via the data network 209 with one or more additional computing systems and / or computer systems such as 201B, 201C and / or 201D (note that in computer systems 201B, 201C and / or 201D, the same architecture may be shared or not used as in computer system 201A, and they can be located at different physical places, for example, computer systems 201A and 201B can stay on board a vessel sailing in the ocean, in an office building on land or at the location of the well and communicate with one or more computer systems, such as 101C and / or 101D, which are located in one or more data centers onshore, on other vessels and / or located in different countries on different continents.

A processor may include a microprocessor, a microcontroller, a processor module or processor subsystem, a programmable integrated circuit, a programmable gate array, or other control or computing device.

Storage media 206 may be implemented as one or more computer-readable or computer-readable storage media. Note that although in the example embodiment of FIG. 35, storage media 206 is shown in computer system 201A, in some embodiments, storage media 206 may be distributed within and / or across multiple internal and / or external enclosures of computing system 201A and / or additional computing systems. Storage media 206 may include storage devices of one or more different kinds, including semiconductor storage devices such as dynamic or static random access memory (DRAM or SRAM), erasable programmable read-only memory (EPROM), electrically erasable and programmable read-only memory (EEPROM) and flash memory; magnetic disks, such as non-removable disks, floppy disks, and removable disks; other magnetic media, including tape; optical media such as compact discs (CDs) or digital versatile disks (DVDs); or other types of storage devices. Note that the instructions discussed above may be contained on a single computer-readable or machine-readable storage medium or may be contained on multiple computer-readable or machine-readable storage media distributed in a large system, possibly having multiple nodes. Such a computer-readable or machine-readable medium (or media) is considered to be part of the product (or product). A product or a production product can be any manufactured single component or multiple components. The storage medium (or media) of the data may be located in a machine that executes machine-readable instructions, or located at a remote location from which machine-readable instructions can be downloaded over the network for execution.

It should be understood that the computing system 200 is an example of a computing system and that the computing system 200 may have more or fewer components than those shown, may combine additional components not shown in the example embodiment of FIG. 35 and / or computing system 200 may have a different configuration or arrangement of the components shown in FIG. 35. The various components shown in FIG. 35 may be implemented in hardware, software, or a combination of hardware and software, including one or more signal processing integrated circuits and / or applied orientation integrated circuits.

In addition, actions in the processing methods described above can be implemented by executing one or more functional modules in an information processing device, such as general-purpose processors or specialized integrated circuits, such as application-oriented integrated circuits, user-programmable gate arrays, programmable logical devices or other suitable devices. These modules, combinations of these modules and / or a combination thereof with general-purpose hardware are included in the scope of protection of the invention.

An example of a well logging method that can be performed by the processor 34 of FIG. 1 will now be described with reference to the flowchart 360 of FIG. 36. Starting from block 360, the method includes, for example, collecting (block 362) snapshots of nuclear magnetic resonance (for example, nuclear magnetic resonance data from periodic observations or from numerous depths of research) showing changes in the geological formation as described above. The method also includes, for example, identifying (block 363) a plurality of fluids in the geological formation based on nuclear magnetic resonance data and determining (block 364) the corresponding nuclear magnetic resonance signatures for the identified fluids based on nuclear magnetic resonance data, which is also described above. In addition, apparent corrected volumes can be determined (block 365) for the identified fluids based on nuclear magnetic resonance signatures and true volumes can be determined (block 366) for the identified fluids based on apparent volumes, which is also described above. The method ends according to block 367.

Numerous modifications and other embodiments may come to mind to a person skilled in the relevant art, benefiting from the ideas presented in the above description and the related drawings. Therefore, it is understood that various modifications and embodiments are intended to be included within the scope of the appended claims.

Claims (26)

1. The method of borehole logging of a geological formation having a wellbore, the method comprising the steps of:
collecting many snapshots of nuclear magnetic resonance (NMR) from the wellbore, showing changes in the geological formation and determining the data of nuclear magnetic resonance;
identify many fluids in the geological formation based on nuclear magnetic resonance data;
determining appropriate nuclear magnetic resonance signatures for the identified fluids based on nuclear magnetic resonance data;
determining apparent volumes for the identified fluids based on nuclear magnetic resonance signatures and
adjusted volumes are determined for the identified fluids based on apparent volumes. 2. The method of claim 1, wherein the snapshots of nuclear magnetic resonance comprise periodically obtained snapshots of nuclear magnetic resonance. 3. The method of claim 1, wherein the snapshots of nuclear magnetic resonance comprise snapshots of nuclear magnetic resonance (NMR) from multiple depths of investigation (MGI). 4. The method according to p. 1, in which the definition of adjusted volumes also contains:
determining adjusted apparent volumes for the identified fluids based on apparent volumes and
determination of true volumes for identified fluids based on adjusted apparent volumes. 5. The method of claim 1, further comprising determining a fluid catalog based on at least one of a hydrogen index and a polarization factor; and wherein the determination of adjusted volumes also includes the determination of adjusted volumes also based on a fluid catalog. 6. The method according to claim 1, wherein determining the adjusted volumes comprises determining the adjusted volumes by redistributing the apparent volumes using the covariance matrix of volumes. 7. The method according to p. 1, in which the geological formation along the length of the wellbore has at least one composition of the bulk of the rock associated with it, and in which the definition of adjusted volumes includes determining the adjusted volumes also based on at least one composition of the bulk breed. 8. The method according to p. 7, in which the wellbore crosses the length of many zones in the geological formation, each of which has a corresponding composition of the bulk of the rock associated with it, and in which the definition of adjusted volumes also includes the determination of adjusted volumes also based on the corresponding compositions the bulk of the rock for each zone. 9. The method of claim 1, wherein determining the corresponding nuclear magnetic resonance signatures for the identified fluids comprises determining appropriate nuclear magnetic resonance signatures for the identified fluids also based on a Gaussian function. 10. The method of claim 1, further comprising converting the nuclear magnetic resonance signatures to the echo region before determining apparent volumes for the fluids. 11. The method of claim 1, wherein identifying the plurality of fluids in the geological formation comprises identifying the plurality of fluids based on boundary values of nuclear magnetic resonance. 12. Downhole logging system containing:
downhole logging tool for collecting multiple snapshots of nuclear magnetic resonance (NMR) from the wellbore in the geological formation, showing changes in the geological formation and determining the data of nuclear magnetic resonance; and
processor for
identifying multiple fluids in the geological formation based on nuclear magnetic resonance data,
determining appropriate nuclear magnetic resonance signatures for the identified fluids based on nuclear magnetic resonance data,
determining apparent volumes for the identified fluids based on nuclear magnetic resonance signatures and
determining adjusted volumes for identified fluids based on apparent volumes. 13. The downhole logging system of claim 12, wherein the nuclear magnetic resonance snapshots comprise periodic nuclear magnetic resonance snapshots. 14. The borehole logging system of claim 12, wherein the snapshots of nuclear magnetic resonance comprise snapshots of nuclear magnetic resonance (NMR) from multiple depths of investigation (MGI). 15. The borehole logging system of claim 12, wherein said processor determines the corrected volumes also by
determining adjusted apparent volumes for identified fluids based on apparent volumes and
determining true volumes for identified fluids based on adjusted apparent volumes. 16. The well logging system of Claim 12, wherein said processor determines adjusted volumes based on apparent volumes and a fluid catalog, wherein the fluid catalog is based on at least one of a hydrogen index and a polarization factor. 17. The well logging system of claim 12, wherein said processor determines adjusted volumes by redistributing apparent volumes using a covariance matrix of volumes. 18. The well logging system of claim 12, wherein the geological formation along the length of the wellbore has at least one composition of the bulk of the rock associated with it, and in which the process determines the adjusted volumes also based on at least one composition of the bulk breed. 19. Computer-readable media having computer-executable instructions for causing a computer to
identifying a plurality of fluids in the geological formation based on a plurality of snapshots of nuclear magnetic resonance (NMR) from the wellbore in the geological formation, showing changes in the geological formation and determining nuclear magnetic resonance data;
determining appropriate nuclear magnetic resonance signatures for the identified fluids based on nuclear magnetic resonance data;
determining apparent volumes for the identified fluids based on nuclear magnetic resonance signatures and
determining adjusted volumes for identified fluids based on apparent volumes. 20. The computer-readable medium of claim 19, wherein the nuclear magnetic resonance snapshots comprise periodically obtained nuclear magnetic resonance snapshots. 21. A computer-readable medium according to claim 19, wherein the snapshots of nuclear magnetic resonance comprise snapshots of nuclear magnetic resonance (NMR) from multiple depths of investigation (MGI). 22. The computer-readable medium of claim 19, wherein the adjusted volumes are determined by
determining adjusted apparent volumes for identified fluids based on apparent volumes and
determining true volumes for identified fluids based on adjusted apparent volumes. 23. The computer-readable medium of claim 19, wherein the adjusted volumes are also determined based on the fluid catalog, wherein the fluid catalog is based on at least one of a hydrogen index and a polarization factor. 24. The computer-readable medium of claim 19, wherein the adjusted volumes are determined by redistributing the apparent volumes using a covariance matrix of volumes. 25. A computer-readable medium according to claim 19, wherein the geological formation along the length of the wellbore has at least one composition of the bulk of the rock associated with it, and on which the adjusted volumes are also determined based on at least one composition of the bulk of the rock. 26. A method for borehole logging of a geological formation having a wellbore, the method comprising the steps of:
collecting many snapshots of nuclear magnetic resonance (NMR) from the wellbore, showing changes in the geological formation and determining the data of nuclear magnetic resonance,
identifying a group nuclear magnetic resonance signature for fluids present in the geological formation based on nuclear magnetic resonance data;
determining the apparent porosity independent of snapshots from nuclear magnetic resonance data based on the group signature of nuclear magnetic resonance and nuclear magnetic resonance data;
collecting downhole data showing the mineralogy of the geological formation; and
determine the instant porosity independent of the snapshots from the data of nuclear magnetic resonance based on the snapshot-independent apparent porosity of the data from nuclear magnetic resonance and formation mineralogy.

Images