 |
Claims  |
|
|
I claim:
1. A method of machine processing a first well log and a second well log,
where each log is made up of a multiplicity of log samples derived from a
device taking measurements while being passed through a borehole in an
earth formation, so as to locate components of said first log which are
likely to be incorrectly depth displaced relative to said second log and
to convert one of said logs into an improved, depth-shifted log,
comprising the following steps each of which is machine implemented:
(a) filtering the logs to locate within each log a plurality of log
elements which correspond to respective element types of a collection of
different preselected element types, where each preselected element type
corresponds to a significant well log feature and may thus correspond to a
significant earth formation feature;
(b) finding, according to the type of element, a number of representative
characteristic parameters of each of the log elements located in the
preceding step;
(c) matching the characteristic parameters found in the preceding step for
a log element with the corresponding parameters found in the preceding
step for log elements of the other log to find which log elements
correspond to each other; and
(d) depth-shifting one of said logs based on log element correspondence
found in the preceding step and producing a tangible representation of the
resulting depth-shifted log.
2. A method as in claim 1 wherein the step of matching said characteristic
parameters includes the step of establishing predetermined depth bounds on
the second log, between which may be found an element corresponding to a
specific element of the first log, and matching the parameters, of said
specific element, only with the parameters of the elements of the second
log which are located within said predetermined bounds.
3. A method as in claim 2 where the step of matching includes matching only
said parameters of elements which are of comparable types.
4. A method as in claim 3 where the step of finding representative
characteristic parameters includes detecting element boundaries which
define the extent of an element and where the step of establishing said
predetermined bounds includes arranging said element boundaries according
to pre-established criteria of corresponding positions to provide
provisional bounds for matching said parameters.
5. A method as in claim 4 where the step of establishing said predetermined
bounds includes modifying said provisional bounds according to
pre-established criteria of corresponding positions after determining that
at least two elements, one from each of said first log and second log,
correspond to each other to provide further provisional bounds for use
thereafter in matching said parameters of possible corresponding elements.
6. A method as in claim 5 where the step of matching said parameters of
selected types of said elements includes arranging said selected types of
elements to allow matching of elements of said types in a preselected
order of types.
7. A method as in claim 6 where the step of matching said parameters
includes selecting as corresponding only elements which match within
defined criteria.
8. A method as in claim 7 including the step of finding depth displacements
between at least two elements, one for each of said first log and second
log, which have been found to correspond to one another.
9. A method of machine-processing of a first well log and a second well log
which are made up of log samples derived from a device or devices taking
measurements while being passed through a borehole in an earth formation,
so as to find depth displacements between respective samples of different
logs and to use the found depth displacements to produce a tangible
representation of at least one of (i) an improved, depth-shifted log
resulting from filtering a first one of said logs on the basis of
components thereof which have been found to be depth-displaced relative to
corresponding components of the second one of said logs, and (ii) bedding
plane inclinatious of an earth formation, comprising the following steps,
each of which is machine-implemented:
(a) processing samples of said logs to locate multi-sample log elements
which conform to respective specific types of log elements;
(b) finding characteristic parameters for each of a plurality of said
located log elements;
(c) sorting said locations to provide bounds for selecting possible
corresponding elements;
(d) selecting possible corresponding elements located within said bounds;
(e) matching said characteristic parameters of one selected element of a
first one of said logs with the possible corresponding elements of a
second one of said logs to determine which elements of the first log and
second log correspond to one another; and
(f) using the results of the preceding step to produce a tangible
representation of at least one of:
(i) an improved, depth-shifted log resulting from modifying said first log
on the basis of components thereof which are due to elements of said first
log determined in the preceding step to have a selected correspondence
with elements of the second log; and
(ii) bedding plane inclinatious of said earth formation defined by elements
of said first log and second log which have been found to correspond to
each other in the preceding step.
10. A method as in claim 9 where the step of selecting elements includes
selecting only elements of comparable specific types as possible
corresponding elements.
11. A method as in claim 9 including the step of modifying said bounds
according to given criteria for corresponding elements to provide new
bounds for selecting further possible corresponding elements.
12. A method as in claim 11 where the step of selecting possible
corresponding elements includes selecting as possible corresponding
elements, only elements of types which have been selected as comparable
types of elements.
13. A method as in claim 12 where the step of selecting possible
corresponding elements includes selecting as possible corresponding
elements, elements of a given type only after determining which elements
of a previously given type correspond to one another.
14. A method as in claim 13 where the step of selecting possible
corresponding elements includes selecting as possible corresponding
elements only elements of a given type located within the bounds provided
by modifying previous bonds according to given criteria for corresponding
elements, thereby restricting the number of possible locations of
corresponding elements of said given type.
15. A method as in claim 14 where the step of selecting possible
corresponding elements of a given type located within modified bounds
includes providing bounds modified as the result of determining which
elements of a previously given type correspond to one another.
16. A method as in claim 15 including the step of finding depth
displacements between elements determined to correspond to one another.
17. A method as in claim 15 where said given type of element is different
than said previosuly given type of element.
18. A method as in claim 15 where said given type of element corresponds to
a less significant type of element than said previously given type of
elements.
19. A method as in claim 15 where said given type of element is of the type
for which the determination of corresponding elements is less reliable
than the determination of corresponding elements of said previously given
type.
20. A method of automatically locating with a machine, characteristic
signal elements corresponding to recognizable features in sampled signals
and using the located elements to find which ones correspond to each other
and then using the found correspondence to produce an improved signal
resulting for modifying one of said signals on the basis of components
thereof which are due to samples thereof having found displacements
relative to samples of another of said signals, comprising the following
machine-implemented steps:
(a) searching samples of said signals to detect the locations of
multi-sample signal elements of a first specified type;
(b) searching said samples to detect the locations of signal elements of a
second specified type;
(c) finding characteristic parameters for said detected signal elements
according to their type;
(d) providing predetermined limits for corresponding elements and locating
possible corresponding elements within said predetermined limits;
(e) matching said characteristic parameters for said located possible
corresponding elements to determine elements which correspond to each
other; and
(f) using the determination of correspondence made in the preceding step to
modify the first signal on the basis of components thereof due to samples
thereof having displacements relative to samples of the second signals,
and producing a tangible representation of the modified first signal.
21. A method as in claim 20 including the step of finding a displacement
between elements which correspond to each other.
22. A method as in claim 21 where said predetermined limits correspond to
an assumed maximum possible displacement between said detected locations
of signal elements.
23. A method as in claim 22 including the step of modifying said
predetermined limits in accordance with the displacement between elements
which correspond to provide further limits for locating further possible
corresponding elements.
24. A method as in claim 23 including the steps of locating possible
corresponding elements within said modified limits and comparing said
characteristics to determine if any other elements correspond to each
other.
25. A method as in claim 24 where the step of locating possible
corresponding elements within predetermined limits includes locating, as
possible corresponding elements, elements which are of one of said first
and second specified type and are within said limits.
26. A method as in claim 25 where the step of locating possible
corresponding elements within said predetermined limits and within said
modified limits includes locating, as possible corresponding elements,
elements of said first specified type within said predetermined limits and
said second specified type within said modified limits.
27. A method of automatically determining with a machine characteristic
signal elements corresponding to recognizable features in sampled signals
and further processing the signals to extract more useful contents thereof
and to produce tangible representations of said more useful contents
comprising the following machine implemented steps;
(a) searching samples of said signals to detect sample patterns indicative
of locations of a plurality of specific types of signal elements;
(b) determining, according to the type of element, defined boundaries and
characteristic parameters of each element corresponding to a detected
sample pattern;
(c) establishing bounds for locating possible corresponding elements within
said bounds;
(d) locating a reference element of a preselected type and possible
corresponding elements of comparable types within said established bounds
and comparing said possible corresponding elements with said reference
element to provide a correlation coefficient for each comparison;
(e) comparing said correlation coefficient to determine which, if any, of
said possible corresponding elements correspond to said reference element;
and
(f) using the determination made in the preceding step to extract selected
more useful contents of at least one of said signals and to produce a
tangible representation of the extracted more useful contents.
28. A well log processing method comprising:
a. deriving a first well log and a second well log on the basis of well
logging measurements produced by passing one or more well logging devices
through one or more boreholes in an earth formation;
b. machine-processing the logs to locate within each a plurality of log
elements which correspond to respective element types of a collection of
different preselected element types, where each preselected type
corresponds to a significant well log feature and is therefore likely to
correspond to a significant earth formation feature;
c. finding for each log element and according to the type of element, a
number of representative characteristic parameters thereof;
d. for each of a number of the log elements of one of the logs, matching
the characteristic parameters thereof with the corresponding parameters
found for log elements of the other log; and
e. generating and machine-storing a representation of the depth
displacement between log elements of the two logs which correspond to each
other on the basis of said matching, to machine-store thereby a
representation of the depth-shift which is likely to be required for
depth-aligning the two logs to each other.
29. A method as in claim 28 in which the matching step comprises matching
the parameters of a given element of the first one of said logs with the
parameters of only log elements of the second log which are within a
defined depth span of the second log and matching the parameters of
another given element of the first log with the parameters of only
elements of the second log which are within another depth span of the
second log.
30. A method as in claim 29 in which the matching step includes matching
only the parameters of elements which are of comparable types.
31. A method as in claim 30 in which said matching comprises arranging the
log elements in a selected order of types and matching elements of one
type before matching elements of the next type in said order. |
|
 |
|
 |
Claims |
|
|
|
Description |
|
|
BACKGROUND OF THE INVENTION
This invention relates generally to techniques used to automatically
determine correlations between corresponding recognizable signal elements.
More particularly, the invention relates to automatic identification of
sampled geophysical signal elements by type, determination of
characterizing features of these elements and the use of these features in
a search for corresponding elements within a system of related search
bounds. The determination of displacements between elements found to be
corresponding are useful in investigating subsurface formations.
The properties of subsurface formations of the earth vary considerably with
depth. This variation may occur abruptly forming boundaries separating one
earth formation from another. These boundaries vary in depth and
inclination or dip from the earth's surface. When the direction or the
degree of dip changes, structures are often formed which are potential
hydrocarbon traps. Thus the recognition and mapping of formation
boundaries is important to the oil and gas industry.
In seismic measurements acoustic waves are transmitted from the surface and
reflected by such boundaries. The reflections or events, as they are
known, are measured at the surface using horizontally spaced geophones.
U.S. Pat. No. 3,681,748 entitled, "Velocity Stack Processing of Seismic
Data" issued Aug. 1, 1972 to Emory E. Diltz illustrates a method of
employing specific event information, limited through predetermined
velocity-time patterns, to present event data in the velocity-time domain.
Since time, in such cases, may be regarded as a function of formation
depth, such presentations may reflect the inclination of formation
boundaries with depth and horizontal displacement.
A more direct method of measuring the dip and the direction or azimuth of
the dip of subsurface formations employs a dipmeter tool passed through a
borehole drilled into the subsurface formations. These tools employ
various means to obtain signals representative of variations of formation
properties and, in particular, representative of formation boundaries
intersecting the borehole. The signals are typically taken from at least
three points radially spaced apart on the surface of the borehole. One
such tool is described in the paper, THE HIGH RESOLUTION DIPMETER TOOL, by
I. A. Allaud and J. Ringot published in the May-June, 1969 issue of "The
Log Analyst".
In determining the inclination of a formation boundary from dipmeter
signals, the signals obtained from one point on the borehole surface are
correlated to determine displacements from corresponding signals obtained
from at least two additional points. Two such displacements may determine
the position of a plane representing the correlation portion of the
signals. The method assumes that the correlated portion of the signal may
represent some common feature of the formation.
The correlation of signals to determine displacements is typically
accomplished by use of relatively standard correlation techniques. A paper
describing one such technique and providing several correlation functions
for such use is COMPUTER METHODS OF DIPLOG CORRELATION by L. G. Schoonover
and O. R. Holt published in the February 1973 issue of "Society of
Petroleum Engineers Journal". To determine displacements, cross
correlation functions are applied to pairs of corresponding signals
located within identical finite-length intervals called correlation
intervals. A correlation function is used to determine the degree of
likeness or correlation coefficient for the signals in these intervals.
The finite length correlation intervals used in dipmeter correlation
usually comprise a large number of samples corresponding to about three
feet of borehole recording. A series of coefficients are determined for a
series of possible corresponding correlation intervals taken at different
displacements between the intervals. These intervals are systematically
selected within a search interval placed about some first assumed depth
displacement. Normally the search interval is also of finite length. It is
measured on one of the signals in directions both above and below the
first assumed displacement. One signal may be considered as a base or
reference signal and the other signal as a comparison or search signal.
The search intervals are usually taken on the comparison signal.
For example, let S.sub.1 and S.sub.2 designate respectively the signals
considered as the reference signal and the comparison signal. The
correlation process considers a finite interval X of S.sub.1 and computes
the correlation coefficient for a comparison interval X' of the same
length on S.sub.2. The comparison interval is systematically moved from a
first assumed displacement to successively displaced intervals on S.sub.2
within the search intervals. A coefficient C(d) to be defined below is
computed for each such displacement.
Commonly signals are recorded digitally as discrete samples S(n) versus
constant increments of time or depth. Thus the signals S.sub.1 (n) and
S.sub.2 (n) are available as two series of discrete samples each series
varying as the value of n. One correlation coefficient C(d) computed
between given intervals X' and X' may be expressed as:
##EQU1##
where:
d is the displacement between the correlation interval X and the comparison
interval X'.
N is the number of samples in each interval, X or X'.
S.sub.1 (n) is the value of the (n)th sample of signal S.sub.1 in the
correlation interval X.
S.sub.2 (d+n) is the value of the (n)th sample of signal S.sub.2 in a
comparison interval X' displaced d samples from X.
##EQU2##
The displacement d which gives the coefficient C(d) corresponding to the
best correlation is taken as the displacement between the samples 1
through N of S.sub.1 and samples (d+1) and (d+N) of S.sub.2.
Even though such expressions may use amplitude and mean value normalization
features, they necessarily include the effects of using finite length and
arbitrarily placed intervals of the signal. In addition, the length of the
correlation interval often determines the type of signal features
represented in the value of the best correlation function.
The ends of the finite correlation intervals are usually chosen in an
automatic and arbitrary manner. Abnormal sample values occurring near the
end portions of the intervals considered in the computation may cause the
correlation coefficient to suffer from so called "end effects". These
effects may lead to ambiguous values of the correlation coefficient. An
improvement on the use of correlation techniques is described in copending
application--"Well Logging Depth Correlation Technique", U.S. Ser. No.
70,709, filed Sept. 9, 1970 by David H. Tinch et al and now abandoned.
The finite interval method of correlation requires changing the correlation
interval to include many samples of the corresponding signals in order to
compare long duration signal features and few samples in order to compare
short duration features. Further, when two features or signal elements
present on the correlation interval on one signal separated by a first
separation are compared with two corresponding features present on a
second signal but here separated by a different separation, distorted
correlation coefficients may result. Since an identical number of samples
is required in each interval, it is difficult to compare two or more
corresponding features present in the same correlation intervals but at
different separations. One attempt at handling this problem is described
in U.S. Pat. No. 3,700,815, "Automatic Speaker Verification by Non-Linear
Time Alignment of Acoustic Parameters" issued Oct. 24, 1972 to Doddington
et al. This patent describes a method of piece-wise resampling one of the
two signals within intervals between signal features. The newly formed or
warped samples are then reused in a correlation process. Unfortunately
this process also distorts displacements between corresponding features
within the warped interval.
Additional U.S. patents describing typical correlation processes and uses
of displacements between best comparing signal intervals are U.S. Pat. No.
2,927,656 entitled, "Method and Apparatus for Interpreting Geophysical
Data" issued Mar. 8, 1960 to F. J. Feagin, et al and U.S. Pat. No.
3,550,074 entitled, "Method for Determining the Static Shift Between
Geophysical Signals" issued Dec. 22, 1970 to C. W. Kerns et al. Whether
the simple amplitude difference or the more complex mean value formulas
are used to compute the correlation coefficients, each such computation is
still repeatedly performed on numerous samples within a preset correlation
interval systematically displaced on one of the corresponding signals. The
computation is performed usually without examining the type or duration of
the signal features actually present. Thus many unproductive computations
are performed on intervals which may not even contain significant signal
features. Further, the computations may be performed on features of
completely different characteristics which in addition to wasting valuable
time, may give rise to erroneous miscorrelations.
It is an object of this invention to provide a new technique of determining
correlations between features or elements of sampled signals representing
variations of measured properties.
A further object is to determine at the same time reliable comparisons
between elements of sampled signals represented by varying numbers of
samples.
It is an object of this invention to provide an automatic technique of
recognizing signal elements representing a variety of features.
It is a further object of the invention to provide a new and improved
technique of comparing two or more sample intervals to determine the
degree of correspondence of these intervals.
A further object is to provide a correlation technique wherein the
intervals to be correlated are determined in a nonarbitrary method.
An additional object is to provide an efficient and accurate method of
comparing two signal elements to determine their degree of correspondence.
In particular, an object of the invention is to provide a method of
comparing signal intervals of unequal length.
A further object is to provide a technique for comparing signal elements
wherein the possibility of making an error and wasting processing capacity
in comparing elements which could not possibly correspond is reduced.
A further object of the invention is to provide a technique for properly
considering the case where an element present on one signal has no
corresponding element.
Further, it is an object to prevent the determination of a false
correlation indication in cases where there is no comparable element or
where there is only a doubtful comparison.
It is a further object to provide an improved technique of comparing
correlations for more than one possible corresponding feature or element
of a sampled signal.
It is a still further object to compare correlations corresponding to
correlation coefficients or degree of comparison to determine the
resolution of such comparisons and still further, the quality of the
comparison itself.
An additional object is to provide a technique of correlation wherein only
signal features or elements which are of comparable types are compared.
A still additional object is to provide a method of determining comparisons
of signal elements of varying significance.
A particular object is to provide a method where the more significant
elements are compared to determine reliable corresponding elements.
It is an object of the invention to produce a significant increase in the
number of reliably determined correspondences between elements of sampled
signals.
It is also an object to provide reliable correspondences between signal
elements representing large features as well as small features of sampled
signals without the necessity of recomputing with different correlation
lengths or correlation functions for this purpose.
More particularly, it is an object to provide a technique to compare only
those elements known to be within reasonable limits for displacements
between such elements, and wherein such limits are automatically narrowed
in a rational manner.
An additional object of the invention is to provide an efficient method of
automatically reducing search intervals used in the search and comparison
of possibly corresponding signal elements.
It is an object of the invention to provide a new method of determining
displacements between corresponding portions of sampled signals.
It is an additional object to determine improved displacement value between
corresponding signal elements.
It is a further object to determine corresponding signal elements and the
displacements between such elements.
It is an object to provide a new method of correlating sampled signals to
determine displacements between samples of these signals.
In accordance with the techniques of the present invention, a method for
automatically determining with a machine and without human intervention
correlations between characteristic signal elements corresponding to
recognizable features as represented by discrete samples of the signals
comprises processing the samples to recognize groups of samples
representing specific types of elements selected to correspond to
significant signal features. Characteristic parameters are determining
according to the type of element and compared to determine which elements
correspond to one another. In accordance with further features of the
invention, characteristic parameters are compared for elements located
within predetermined bounds of possible corresponding elements. These
bounds may be determined by searching and sorting boundary positions
according to pre-established laws of corresponding positions to provide
provisional bounds for use in searching for possible corresponding
elements. The parameters determined for elements located within
provisional search bounds are compared and if an acceptable comparison is
found, the corresponding bounds are modified to indicate subsequent search
bounds for use in searching for additional possible corresponding
elements.
In accordance with additional features of the invention, several specific
types of elements of varying significance are recognized. Further, the
specific types of elements are classified by using given ranges of
thresholds for identifying various sizes of elements corresponding to a
range of significance for elements of a given type. Still further, the
parameters of elements of the more significant types are compared and if
an acceptable comparison is found, the corresponding bounds are modified
to indicate bounds for use in searching for possible corresponding
elements of less significant types.
The steps of comparing parameters of elements of a given type located
within previously provided search bounds and modifying bounds
corresponding to elements found to correspond to provide further search
bounds are repeated for remaining elements until all elements have been
processed.
The displacements between elements and boundaries found to be corresponding
may be taken as representing the displacement between corresponding signal
features. If the signals are from a dipmeter tool, for example, the
displacements may be used to determine the attitude of a geological
feature relative to the position of the tool and when provided with the
tool position, they may be used to determine the strike and dip of the
geological features.
Also, the displacements may be used to align displaced signals by applying
alignment corrections. The signals then aligned on common geological
features may be properly combined and used for further evaluation of
subsurface formations.
For a better understanding of the present invention, together with other
and further objects thereof, reference is had to the following description
taken in connection with the accompanying drawings, the scope of the
invention being pointed out in the appended claims.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1A illustrates one application of the present invention.
FIG. 1B illustrates simplified steps in the correlation and displacement
determination processes.
FIGS. 2A and 2B illustrate identification of types of signal elements known
as bumps and depressions.
FIG. 3 shows how a slope function of a sampled signal might be calculated
for a given sample.
FIGS. 3A through 3F illustrate various sampled signal models and
corresponding slope function values.
FIG. 4 illustrates identification of types of elements known as peaks-bumps
and peak-depressions.
FIG. 5 illustrates identification of a type of element known as a surge.
FIGS. 6A and 6B illustrate certain features of the concept of bounds and
boundaries described in the disclosure.
FIG. 7 illustrates symbols and notations used in the explanation of the
invention.
FIGS. 8A through 8D are processing diagrams illustrative of steps used to
recognize specified elements and record their representative boundary
positions and characteristic parameters.
FIGS. 8E through 8G are processing diagrams illustrative of the steps used
to sort elements and boundaries to provide references for subsequent
processes.
FIGS. 9A through 9C are processing diagrams illustrative of steps used to
search for corresponding elements within provisional search bounds and the
modification of these bounds.
FIG. 9D illustrates steps which may be used to derive displacements.
FIGS. 10 and 11 represent illustrative intervals of two signals which are
useful to demonstrate certain features of the invention.
FIGS. 12, 13, 14 and 15 represent reference tables useful in the
description of the search for possible corresponding elements located on
the two signals of FIGS. 10 and 11.
FIG. 16 illustrates upper and lower bounds for boundaries used to guide the
search for possible corresponding elements.
FIGS. 17A through 17D' illustrate steps in the process of modifying upper
and lower bounds corresponding to two of the elements illustrated in FIG.
16 which were found to correspond.
FIG. 18 illustrates a possible result of modifications illustrated in FIGS.
17A through 17D'.
The method illustrated in the following description may be applied to
correlate signal elements obtained from any number of sources. The
illustrated signals serve to demonstrate the usefulness of the method as
applied to signals obtained from geophysical instruments. The signals
might well have been obtained from biomedical instruments, for example.
Referring now to FIG. 1A, there is illustrated a method of obtaining and
processing signals obtained from a borehole investigating device commonly
known as a dipmeter. A more complete description of this device may be
obtained from U.S. Pat. No. 3,521,154 issued July 21, 1970 to J. J.
Maricelli.
A borehole apparatus 18 is lowered into a borehole 10 for investigating
earth formations 11. Typical earth formations are represented by shale
formations 13 and 14 with an intervening sand formation 15. Typical
boundaries 16 and 17 are shown between the different formations. The
downhole investigating device 18 is adapted for movement through the
borehole 10 and includes four pads designated 19, 20, 21 and 22 (the front
pad member 19 obscures the view of the backpad member 22, which is not
shown). The pad members 19 through 22 are adapted to derive measurements
at the wall of the borehole.
The pads 19 through 22 each include a survey electrode Ao. One of the pads
herein designated pad 19 may contain an additional survey electrode Ao'.
Each survey electrode is surrounded by an insulation material 48. The
insulation material and thus also the survey electrodes are surrounded by
a main metal portion 45 of the pad. The main metal portion 45 of each pad,
along with certain other parts of the apparatus, comprise a composite
focussing element for confining the survey current admitted from the
various survey electrodes to a desired critical pattern. Survey signals
representative of changes in the formations opposite each pad are obtained
from circuits comprising the Ao electrodes, focussing elements and current
return electrode B.
The upper end of the device 18 is connected by means of armored
multiconductor cable 30 to a suitable apparatus at the surface for raising
and lowering downhole investigating device through the borehole 10.
Mechanical and electrical control of the downhole device may be
accomplished with the multiconductor cable which passes over a shieve 31
and then to a suitable drum and winch mechanism 32.
Electrical connections between various conductors of the multiconductor
cable which are also connected to the previously described electrodes, and
various electrical circuits at the surface of the earth are accomplished
by means of a suitable multielement slipring and brush contact assembly
34. In this manner, the signals which originate from the downhole
investigating apparatus are supplied to the signal processing circuits 39
which in turn supply the signals to a signal conditioner 40 and recorder
41. Additionally, a suitable signal generator 42 supplies current downhole
via transformer 50 and to signal processing circuits located at the
surface. The details of these circuits are described in the aforementioned
Maricelli patent.
The signals obtained from the downhole device may be recorded graphically
by a film recorder 41. One such recorder is described in U.S. Pat. No.
3,453,530 issued to G. E. Attali on July 1, 1969. In addition, the signals
may be processed to obtain discrete samples and recorded on tape. On such
tape recorder is described in U.S. Pat. No. 3,648,278 issued to G. K.
Miller et al on Mar. 7, 1972. The signals or samples thereof may also be
transmitted directly to a computer. One such transmission system is
described in U.S. Pat. No. 3,599,156 issued to G. K. Miller et al on Aug.
10, 1971.
The recorded or transmitted signals may also be processed as sampled data
by general purpose digital computing apparatus properly programmed in a
manner to perform the process described herein or by special purpose
computers composed of standard modules arranged to accomplish the same
process.
Alternatively as shown in FIG. 1A, the signals may be processed at the well
site again using conventional digital computing apparatus interfaced to
the signal conditioner 40. One such computing apparatus is the Model
PDP-11/20 obtainable from Digital Equipment Corporation. Suppliers of such
equipment may also supply signal conditioning circuits 40 and signal
conversion means 60 for conditioning and converting analog signals to
digital samples suitable for subsequent digital storage and processing.
Further, such computing apparatus ordinarily includes memory 54 for
storing data and information such as parameters, coefficients and controls
used and generated by the processing steps.
A brief description of the process is illustrated by blocks 62 through 92
of FIG. 1A. The process will be described later in greater detail. The
recorder 94 may be of the same type as recorder 41 or of the type of tape
recorder previously referenced. Also the common analog or incremental X-Y
Plotter may be used. Therefore the details and circuits of such apparatus
which are available elsewhere will not be described herein.
A summary of the processing steps indicated by FIG. 1A will be given at
this point.
The processing may be described as four smaller processes which may be
performed in sequence or optionally in parallel, at least in part. The
first process is indicated by Blocks 62 and 64, the second by Blocks 68
and 70, the third by Blocks 72 through 82 and the fourth by Blocks 90 and
92. Block 62 represents an element detection process wherein samples of
the signals obtained from signal conversion means 60 are searched to
recognize groups of samples representing specific types of elements. This
search is conducted using search patterns and threshold values stored in
memory 54. The detection process includes boundary determination and
element typing. Once an element is detected, characteristic parameters are
computed as represented by Block 64 according to control procedures
referenced from the memory. The detected elements, along with their
boundaries and characteristic parameters are subsequently stored in memory
54.
The first process may continue, as indicated by Branch 66, until a number
of elements have been detected and processed as indicated by Blocks 62 and
64. Then, either in sequence with or in coincidence with the previously
described processes, the second process may be performed. The detected
elements may be sorted, as represented by Block 68, to provide
cross-references to additional elements and boundaries according to their
type, which are then stored in memory 54. Additionally, as indicated by
Blocks 70, search bounds may be generated for each boundary creating
additional cross-references. Further details for the above described
processes may be found in reference to the description of FIGS. 8A through
8G.
The start of the next process is represented by Block 72 and begins with a
selection of a type of element known to provide a desired type of
correlation. For example, the elements corresponding to outstanding
features known to provide reliable correlations might be selected first.
Once the desired type is selected one such element previously having been
detected on one of the signals, here designated as a reference signal, is
located in storage means 54. Possible corresponding elements of comparable
types previously detected on a comparison signal within the previously
established search bounds are also located within said storage means.
Then, as indicated by Block 74, the element located on the reference signal
is compared with the elements located on the comparison signal. Each such
comparison generates a correlation coefficient which is subsequently
stored in the memory. The comparison process continues, as indicated by
Branch 76, until all such elements have been compared.
| | |
