AU609509B2 - Extended range, pulsed induction logging tool and method of use - Google Patents

Description

b i: 4 )i AU AI 2 1276/ 88 PCT WORLD INTELLECTU RO TY AN IO 9 IO LICIO LI nternalu THau CO INTERNATIONAL APPLICATION PUBLISHED UNDER THE PATENT COOPERATION TREATY (PCT)

-C~

(51) International Patent Classification 4 International Publication Number: WO 88/ 09940 G01V 31/18 Al (43) International Pulicatin Date: December 1988 (15.12.88) (21) International Application Number: PCT/US88/01775 (22) International Filing Date: (31) Priority Application Number: (32) Priority Date: (33) Priority Country: 26 May 1988 (26.05.88) 059,. )7 8 Jure 1987 (08.06.87) (74) Agent: ERICKSON, Roger, Owen, Wickersham and Erickson, 433 California Street, Ilth Floor, San Francisco, CA 94104 (US).

(81) Designated States: AT (European patent), AU, BB, BE (European patent), BG, BJ (OAPI patent), BR, CF (OAPI patent), CG (OAPI patent), CH (European patent), CM (OAPI patent), DE (European patent), DK, FI, FR (European patent), GA (OAPI patent), GB (European patent), HU, IT (European patent), JP, KP, KR, LK, LU (European patent), MC, MG, ML (OAPI patent), MR (OAPI patent), MW, NL (European patent), NO, P.O. SD, SE (European patent), SN (OAPI patent), SU, TD (OAPI pateut), TG (OAPI patent).

Published With international search report.

Before the expiration of the time limit for amending the claims and to be republished in the event of the receipt of amendments.

(71) Applicant: MPI [US/US]; 19925 Stevens Cr;eek Boulevard. Suite 176, Cupertino, CA 95014 (US).

(72) Inventors: GILL, Stephen, P. 32 Flood Circle, Atherton, CA 94025 WATSON, John, D. 645 Melville Drive, Oakland, CA 94611 BRINK, Keith, 0. 5010'Northiawn Drive, San Jose, CA 95130 (US).

AUSTRALIAN

4 JAN 989 PATENT OFFICE A e. MO ]R 98 (54) Title: EXTENDED RANGE, PULSED INDUCTION LOGGING TOOL AND METHOD OF USE D TH HOST RECORDER (57) Abstract EN R COMPUTE A pulse'd induction logging system (10) is provided includ- to Z 3 9 7 ing a logging sonde (14) with downhole microprocessor/controller circutry (28) in contact with a separate host computer/controller POW (27) at the earth'., surface, The host computer/controller (27) and SUPPLY two downhole raicroprocessor/controllers generate a digital pulsed logging coce for periodically driving a rajiation coil with an oscillating current having high peak power. A powerful primary pulse of electromagnetic energy is periodically generated /9 for irradiating the adjacent formation. The sonde also includes an antenna array (30) for detecting secondary induced fields in the formation around the borehole. The array includes a series of grouped, paired coils (AI-A 13 axially spaced along the borehole. Each group of paired coils independently detects the secondary field, Each detected signal per group of paired coils, is digitized, reformated and transmitted to the host computer/controller (27).

T!is doouinict ncontahils thi j rluenluentS rniade 1ttnck Section 49 and is corre( t for Priting.

rn 1 1 "I :I I i. I I 1-

I

'C,

V O 88/09940 PCT/US88/01775 2 EXTENDED RANGE, PULSED 3 INDUCTION LOGGING TOOL 4 AND METHOD OF USE 6

SPECIFICATION

7 8 9 This invention relates to a method and apparatus for investigating an earth formation traversed by a bore- 11 hole. More particularly the invention relates to an 12 extended range, pulsed 'induction logging system in which 13 amplitude versus time signals provided by a secondary field 14 within the adjacent formation (after irradiation by a series of powerful, primary pulses of electromagnetic energy) can 16 be detected and processed so that a formation parameter of 17 interest conductivity and/or dielectric constant, can 18 be accurately indicated.

19 By the term ".pulsed" it is meant that the primary energy comprises a single burst of high level energy at each 21 logging slstion, although the induction logging sonde may be 22 continuously moving along a borehole penetrating the forma- 23 tion.

24 BACKGROUND OF THE IVVENTION 26 Induction logging of earth formations from a 27 borehole is an established commercial procedure. In such 28 operations, a constant intensity, magnetic field was gener- 29 ated for propagation into the adjacent formation, by continuously driving a' source antenna with an alternating 31 current at a constant frequency and a low steady state power 32 level. A receiver coil assembly was usually electrically 33 balanced to respond to induced eddy currents in the adjacent 34 earth formation. The secondary magnetic field resulting from the eddy currents was then used to generate a voltage 36 signal in the receiver coil assembly. The detected voltage 37 signal varied in accordance with the conductivity of the 38 PCT/UIS88/01 77 WO 88/09940 -2- 1 adjacent formation. Usually, only the component of detected 2 signal voltage that was in-phase with drive current was 3 employed to indicate the formation conductivity or resis- 4 tivity as an amplitude vs. depth trace.

Various technical papers and publications have 6 discussed the operating principles of induction logging 7 systems as described above. If the proper precautions were 8 undertaken, the continuously detected voltage signal was 9 directly and linearly proportional to the electrical conductivity or resistivity of the logged formation normalized to 11 the range of formation values usually encountered.

12 Certain operational effects tended to adversely 13 affect the accuracy of the data provided by such prior 14 logging systems, however. One such non-linear effect resulted primarily from mutual interaction of different 16 portions of the eddy currents, a so-called "'skin effect", 17 which varied as a function of frequency of operation, the IS effective length of the source and receiver coil array, and 19 the conductivity of the adjacent foamition. -Although the occurrence of these objectionable variations could be sub- 21 stantially reduced or eliminated by proper choice of operat- 22 ing frequency and effective coil system length, such 23 restraints limited desirable objectives of the system. For 24 example, to increase the range of the logging system in a lateral direction, the effective system coil length must 26 also be increased. A larger coil spacing, viz., between the 27 source and receiver array, increased the non-linearity of 28 the resulting detected signal due to 4kin effects.

29 There have been several methods proposed to correct for the aforementioned skin effect problems. In 31 one such system,"a function circuit was used to correct the 32 detected signals in accordance with a predetermined func- 33 tion. In another system, the phase-quadrature signal (said 34 to be approxirlately equal to the skin effect component of the in-phase deection signal over a given range of conduc- 36 tivity and frequency values), was eliminated.

37 38 *j ''J SWO 88/09940 PCT/US88/01775 1 Another adverse effect also limited the accuracy 2 of results of conventional induction 1ogging systems. This 3 effect related to the fact that the adjacent formation 4 can be heterogeneous, a plurality of conducting zones may exist in the adjacent formation othet tha the true for- 6 mation conductivity) and (ii) the scrata above and below 7 the formation of interest, may effect signal response. Such 8 conditions created substantial errors in the accumulated 9 data using prior art techniques.

Yet another adverse effect arose when the borehole I was filled with a drilling mud which formed mud-cake along 12 the sidewall of the borehole and permitted a filtrate to 13 invade the formation. As a result the diameter of the bore- Shole also varied so that the logging sonde was offset different lateral distances from the mud cake as data was 16 taken, introducing yet another adverse data effect.

17 To overcome the aforementioned adverse effects in 18 part, different arrays and associated circuitry were 19 desig!ed in prioe art systems to provide a plurality of different radial logging devices. The separate devices, 21 however, had to be designed so that their operations were 22 independent in attempts to compensate for various adverse 23 effects. For example, individual signals were of.ten compen- 24 sated for either by tornado charts or by time domain computational methods, such as by .the addition of weighting 26 factors. These stored weighting factors combined in such a 27 way that the effect of regions other than the region under 28 consideration, were diminished.

29 While prior art alternating current induction logging systems were effective, their utility was greatly 31 restricted by their limited lateral range, their limited 32 vertic(l resolution, their limited accuracy in determining 33 true formation conductivity, and their inability to deter- 34 min" dip angle of beds or the range and azimuth of formation anomalies.

36 37 38, 4

YI

Accordingly it is an aim of the present invention to provide a new and improved induction logging method and apparatus whereby logs of improved accuracy, range, resolution, and reliability may be obtained.

0*

S

Sr

S..

0 SUMMARY OF THE INVENTION In a first aspect, the present invention provides a method of systematically logging an earth formation around a borehole using a logging sonde to generate a primary magnetic field in the formation at depth and detecting an induced secondary magnetic field indicative of a formation electrical parameter, comprising the steps of: generating a series of time spaced primary magnetic field pulses of high peak power at a series of 15 stations along the borehole;

S.

(ii) detecting response data relative to the secondary magnetic field induced in said formation by each said primary magnetic field pulse; and (iii) processing said response data to provide an indication of a formation electrical parameter around the borehole by modeling response of the adjacent formation using a series of assumed conductivity and bed thickness values along with actual parameters of the 25 pulsed primary field, and *6 cross-checking a forward solution of the model of step against measured detected componente of the secondary magnetic field at depth.

S**

Preferably the peak power of the primary field pulse 30 is in the range of 1 to 1000 megawatts, and typically megawatts.

Preferably the steps (ii) of detecting the secondary S magnetic field includes the sub-steps '~sing a series of antennas, each of which independently detect components of said secondary field associated with different formation regions and digitizing each detected component iIpos ndent ly.

S..

S;

S.

S

S.

S

00 I i .I x ~05; 0

S

S.

S.

0i

SS

Si

S)

In a further aspect, the present invention provides a method for systematically logging an earth formation around a borehole using a logging sonde to generate a primary magnetic field in the formation at depth and detecting an induced secondary magnetic field indicative of a formation electrical parameter, comprising the steps of: generating a series of time spaced primary magnetic field pulses of high peak power at a series of stations along the borehole by selectively driving a radiation coil with a pulsed current, the selective driving of said radiation coil being controlled by a digital logging code, which is generated by a host computer/controller at the earth's surface in association with at least one microprocessor/controller on-board the logging sonde; 15 (ii) detecting response data relative to a secondary magnetic field induced in said formation by each said primary magnetic field pulse; and (iii) processing said response data to provide an S indication of a formation electrical parameter around the .O0 borehole.

In a third aspect, the present invention provides an S apparatus for systematically logging an earth formation around a borehole using a logging sonde by generating a primary magnetic field in the formation at depth and A t detecting an induced secondary magnetic field indicative of a formation electxical parameter, comprising: means for generating a series of time spaced primary magnetic field pulses of high 'eak power at a series of stations along the borehole; (ii) means for detecting response data relative to :e a secondary magnetic field induced in said formation by each S said primary magnetic field pulse; and (iii) means for processing said response data to provide an indication of a formation electrical parameter around the borehole, wherein said processing means includes fi Go 0@ 0 .5

S

1

*S

*S

S

if7- It p; I I f 6- 6 means for modeling the response of the adjacent formation using a series of assumed conductivity and bed thickness values along with actual parameters of the pulsed primary field, and means for cross-checking a forward solution of against detected components of the secondary magnetic field.

In a fourth aspect, the present invention provides an apparatus for systematically logging an earth formation around a borehole using a logging sonde by generating a primary magnetic field in the formation at depth and detecting an induced secondary magnetic field indicative of a formation electrical parameter, Imprising: rmeans for generating a series of time spaced primary magnetic field pulses of high peak power at a series of stations along the borehole, said generating means including a radiation coil and a power source network in selective operative contact therewith, said power source S0: network including a magnetohydrodynamic (MHD) source and a switch network; (ii) means for detecting response data relative to Sthe secondary magnetic field induced in said formation by each said primary magnetic field pulse; and (iii) means for processing said response data to provide an indication of at least one electrical parameter see. of the formation around the borehole.

In accordance with method aspects oi a preferred embodiment of the present invention, the final displays result from modeling the logged formation at each logging station using a series of assumed conductivity values and bed thickness (along With the actual parameters of pulsed primary field), and when cross-checking the forward solution of the model against the digital field data actually detected by the logging system. Operations stop when convergence occurs. In the steps and amplitude v. time representations of the detected signals are transformed to depth/amplitude-frequency domain values.

1 s/EM 0 7 The resulting matrix can undergo additional processing involving "frequency slicing", i.e. obtaining "frequency slices" of the matrix centered at a particular frequency, to yield accurate indications of resistivity and dielectric values of the formation. Thus bed, dip and azimuthal resolution can be divoced from the ranging requirements of conventional tools.

It should be noted thfat "frequency slices" concentrated on low frequencies, say 10 to 20 kilohertz, yield accurate conductivity values; at higher frequencies, between 1 to 10 megahertz, true dielectric values are indicated.

s' Other advantages and features of the invention will S become apparent from the following detailed description thereof presented in conjunction with the accompanying drawing.

O0

S

S

I

S

I~

S,

S

0 BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a partial sectional view of a borehole penetrating an earth formation illustrating operations of a pulsed indication logging system in accordance with the present invention;

S

4 p0 88/09940 PCT/US88/01775 3 1 FIG. 2 is a side elevation of the logging sonde of 2 FIG. 1; 3 FIG. 2A is a detail in side elevation of a radia- 4 tion coil for generating a pulsed primary magnetic field of extremely high power level.

6 FIG. 2B is a detail in elevation of an antenna 7 array within the sonde of FIG. 2.

8 FIG. 3A, 3C-3F are a series of waveforms illus- 9 trating various operating characteristics of the present invention, 11 FIG. 4 is a time v. depth display illustrating 12 pulsed logging operations in detail.

13 FIG. 5 is a circuit diagram in block form of 14 control and primary field generating circuitry elements for generating the primary pulse of electromagnetic energy ii 16 accordance with the invention.

17 FIG. 6 is a further block diagaam of the field 18 generating current elements of FIG. 5 including a L-R-C 19 power circuit.

FIG. 7 is a flow diagram of operation o£ the 21 circuit block diagrams of EIGS. 5 and 6.

22 FIG. 8 is a circuit diagram of the L-R-C power 23 circvit of FIG. 6.

24 FIG. 9 is a side elevation of a plasma generator for generating the pulse primary field of the invention as 26 an alternate embodiment therefor.

27 FIG. 10 is a circuit equivalent of the plasma 28 generator of 'IG. 9 illustrating additional charging/dis- 29 charging circuitry.

FIGS. 11A and 11B are waveforms of the primary 31 wave generated by the circuit of FIG. 32 FIG. 12 is a circuit diagram in block form of 33 control and secondary circuit detection elements including 34 paired coil antenna groups for detecting the induced secondary.field in an adjacent earth formation.

36 3

'I

9 FIGS. 13A and 13B illustrate a modification of the paired coiled antenna group of FIG. 12 to form a 3-component antenna group; FIG. 14 is a preamplification circuit used by and forming part of the circuit of FIG. 12; FIGS. 15 and 16 are elements of a mi-ing circuit used by the circuit of and forming a part of FIG. 12.

FIG. 17 is a flow chart of steps related to detection of the secondary induced field in the logged earth formation.

FIG. 18 is a flow chart illustrating digital control steps for initializing and identifying command codes downhole.

FIG. 19 is a flow diagram of A/D conversion.

FIGS. 20 and 21 are flow charts of previous steps for enchaning signals so as to provide improved results.

FIG. 22 is a matrix display of results in accorda ce with the present invention.

D. DTAILED DESCRIPTION OF THE INVENTION 20 FIG. 1 shows diagrammatically a pulsed induction S logging system 10 for investigatin an earth formation 11 using a series of periodic pulses 12 of electromagnetic (EM) energy as the primary magnetic field for induction logging purposes.

By the "pulsed" as previously indicated, it is meant that the primary field pulses 12 being irradiated into formation 11 are discontinuous with timw but their energy level per pulse interval is substantia7ly constant, Moreover, although the movement of sor,de 14 along borehole 15 may or may not be continuous, there must be movement from at least one logging station indicated in solid line at 16 to at least another logging station indicated in phantom line at 17, at which tiNte a second pulse 12 of primary energy is generated. Thus, the pulse generation occurring S I SIPEN, 05 JUL 1989 1 at the stations 16 and 17 in conjunction with movement of 2 the sonde, defines a pulsing sequence (ON-OFF) in which OFF 3 time (between two adjacent ON pulses 12 coincident with 4 stations 16 and 17), must be longer than the width (time) of the pulses 12 themselves.

6 Sonde 14 is initially lowered into the borehole 7 15 and then is raised by reeling cable 19 over pulley 22 8 via winch 20 at the earth's surface 21. Slip rings 23 9 electrically connect power supply 24 to the logging sonde 14. Logging depth is indicated by encoder 25 to generate 11 digital pulses for recorder 26.

12 As the sonde 14 moves relative to the borehole 13 15, the high peak power EM pulses 12 are generated via a 14 digital logging code generated at host computer/controller 27 at the earth's surface and received at a downhole 16 microprocessor/ .ntroller 28 in the sonde 14. The downhole 17 microprocessor/controller controls a power circuit (not 18 shown) which is selectively enabled so as to trigger the 19 periodic pulses 12 via source radiation, coil system 29.

Each pulse 12 induces a secondary field in the formatfIn 21 11, such induced signals then being detected by an antenna 2 2array 30 within the sonde 14. After being digitized and 33 reformatted via a second downhole microprocessor/controller 24 (not shown), the detected signals are transmitted uphole to the host computer/controller 27. Since the sonde 14 has 26 important system responsibilities, a detailed description 2 of its construction and operation is presented below in 28 2 conjunction with FIG. 2.

29 CONSTRUCTION OF SONDE 14 31 As shown, the sonde 14 is divided into a command 32 section 35 supported at a upper end 36 by cable 19.

Connected at an opposite lower end 38 of the command 34 section 35, is a radiation section 40. The radiation section is of similar diameter as command section 35 and 36 they are mechanically coupled by a collar 41A. As shown, 37 3 radiation section 38 PCT/US88/01775 WO 88/09940 1 40 in turn is coupled to a spacer section 44 which in turn Sis coupled to a receiver antenna section 45 by similar 3 coupling collars 41B and 41C, respectively. In turn, 4 receiver antenna section 45 .A *,apled in similar fashion by collar 41d to a conventional directional instrument section 6 46 containing sets of magnetometers and accelerometers that Sresult in the generation of a directional output for trans- 8 fer uphole. Such section 46 is a product of Tensor Incorpo- 9 rated, Austin Texas.

In forming the radiation and antenna sections 11 45, care is taken to form them of non-magnetic material, 12 while spacer section 44 is formed from material of suffi- 13 cient strength to withstand operational stress without undue 14 increases in weight, such as fiberglass. Within the radiation section 40 is a single inductive radiation source coil 16 50, also shown in detail in FIG. 2A.

17 18 INDUCTIVE COIL 19 In FIG. 2A, note that the purpose of inductive coil 50 is to receive a powerful pulse of amplitude-varying 21 current to generate a corresponding powerful pulse of pri- 22 mary electromagnetic energy.

23 The magnetic energy associated with current in 24 the radiator coil 50 can be characterized as a magnetic dipole (strength equal to the product nf current, cross i 26 sectional area, and number of turns in the ci, 50). As a 27 result, enormous transient pulses cav, Lh generated 28 because the limitations of coil heating and a large continu- 29 ous power supply are not present. Power is simply generated by the discharge of stored energy, which may be in 31 electrical or in chemical form.

32 Thus, the coil 50 must be 1 c+trz=agnetically able -3 to Withstand large current pulses needed to produce the 34 intense electromagnetic field of the present invention.

Consequently, the turns 51 of the 'coil 15 wound about a 36central mandrel 52 must be large in cross section to 37 WO 88/09940 PCT/US88/01775 WO 88/09940i2, 1 provide sufficient surface area for this requirement.

2 Headers 53 complete the mechanical construction. FIG. 2A 3 shows the radiation antenna coil 50 polarized in the 4 vertical direction: that is, generating a magnetic field directed along the axis of the tool and borehole. The 6 antenna coil 50 may also be oriented in a horizontal direc- 7 tion, if desired.

8 Antenna array 54 within the antenna array section 9- 45 is a series of axially spaced receiving coils generally indicated at 54 as an antenna array and shown in detail in 11 FIG. 2B. As shown, pairs of receiving coils 54 over a 12 central region 55 each forms a single antenna group A 1 13 A 2

A

2 These groups A 1

AI

2 are comprised of two 14 inductive pickup coils 56, 57 wound in opposition about 13 mandrel 58 so as to cancel out the common mode of the 16 induced signal. The signals from adjacent antenna groups 17 are mixed electronically so that the digitized signal 18 represents the second derivative or curvature of the mag- 19 netic field. Coils 59 at the end of the array together form a similar antenna group A 13 In each group (wound in 21 opposite direction) the number of turns is inversely 22 proportional to the cube of the distance of the group AI-A 1 3 23 relative to the mid-point of the source coil 50 of FIG 2A to 24 facilitate the first differencing. The rumber of antennas groups is limited only by the available power to run the 26 tool electronics, pulse power supply, and motors.

27 28 PRINCIPLES OF SIGNAL. DETECTION 29 The information on the distribution of the resistivity in the layered formation is contained in the 31 curvature of the magnetic field. For high spatial resolution 32 applications, this information is typically 80 to 120 dB 33 down from the peak magnetic field measured at each group 34 location. Because available electronic amplifiers have a 36 37 I 3%4( vl.- 7 S O V88/09940 PCT/US88/01775 rVO 88/09940 '3' 1 dynamic range of about 100 to 110 dB, the signal must be 2 differenced by the opposing coil method to avoid exceeding 3 the dynamic range limitation of the front end amplifiers.

4 As further indicated, each group Al...A 12 is axially zpaced from adjacent neighboring group by a 6 constant distance D. In detecting induced secondary fields 7 within the adjacent formation, the present invention 8 requires very high precision measurements of the magnetic 9 fields by the antenna array 54. As previously indicated, information characterizing the formation is ,at levels from 11 80 dB to 120 dB down in the measured signal. Once in digi- 12 tal form, however the data can be accurately processed and 13 transmitted up the wireline with no further degradation in 14 quality. If the data were not digitized in the tool, the signals would be degraded by transmission losses and the 16 information would be irretrievably lost.

17 In this regard, signal acquisition is defined in 18 this application as the process of acquiring a signal by 19 the array 54, amplifying it, mixing the signal with other signals, digitizing and storing the combination in an 21 on-board memory before transference uphole to the host 22 computer/controller.

23 24 DATA ACQUISITION As previously indicated, pulse generation and data 26 acquisition is carried out in a systematic manner as the 27 sonde transverses the borehole using a host com- 28 puter/controller at the earth's surface in conjunction with 29 two on-board microprocessor/controllers. Together, they form a computer network that creates a logging (command) ON/OFF 31 pulse code, a part of which is shown at 76 (FG. 4) that 32 systematically controls logging operations.

33 34 36 37 ©Vr I-f O 88 0 PCT/US88/01775 WO 88/09940 1 LOGGING CODE 2 As shown i!,n FIG. 4, adjacent logging station 1 3 and 2 are preferably ov erlapping. Each of the antenna 4 groups A 1 A 1 3 togther with radiation coil 50 normalized to axial lengths Ll, L 2 provides for a seri4es of overlapping 6 magnetic field measurements at each logging station. In that 7 way, lateral range requirement (known to be a function of a the spacing Le between radiation coil 50 and antenna array 9 54) can be decou)led from the system logging requirement since incremZtal vertical response of the formation (adja- 11 cent each antenna group-A 2

A

13 is directly related to the 12 distance between each antenna group A,...AI 3 relative to 13 radiation coil 14 For conductivity values usually encountered in logging operations, assuming an operating frequency of about 16 7500 Hertz, an antenna spacing of 1 meter and an average 17 coil spacing Le (See Fig. 2) of 8 meters, experience indi- 18 cates the maximum lateral range of the tool or sonde is 19 about 100 meters with a vertical layer resolution of about 1 meter. 'or an operating frequency of 15 kHz, an antenna 21 spacing of 8 inches and an."average coil spacing Le (See Fig.

22 2) of 3 meters, experience indicates the maximum lateral 23 range of the tool is about 15 meters with a vertical layer 24 resolution of about 8 inches.

As shown in FIG. 4, operations associated with 26 pulsincg, detection, amplification and digitization steps 27 define a constant logging interval 77 that is. long 28 compared to pulse interval 78 but much less than the idle 2, interval 76. The pulse width of pulses 77 is so small compared to the other operations requirements, to 31 transmitting the data uphole that the duty factor for 32 operation is very small, .0001 or less. Additionally, 33 sufficient time is available for digitizing the detected A signals between logging stations using one of the more 36 7 C) 1 cl, 05 JUL 1989 1 downhole controller/microprocessors. Also the pulse 2 interval must be short enough that the upward motion of the 3 sonde is negligible.

4 The velocity of the sonde during a logging run is Sbetween 10-30 feet per second. The pulse logging code 76 6 defined by such a run is not normalized to sonde speed but 7 is made to depend on logging station depth. Depth (or 8 logging station location) is constantly and accurately 9 monitored by the host computer coi.troller 27 (FIG. 1) at the earth's surface 21 through the depth encoder 25. Since 11 the depth of each logging station along the entire run can 12 be predetermined and such data placed in memory of the host 13 computer/controller 27, firing commands for the sonde via 14 the logging code 76 is automatically generated, 16 WAVEFORM CHARACTERISTICS 17 FIGS. 3A ant 3C illustrate the nature of primary 18 electromagnetic field in more detail. In these waveform 19 diagrams the horizontal axis is time normalized to a pulse interval used in the induction logging procedure while the 21 vertical axes represent various electrical quantities.

22 FIG. 3A illustrates the general character of the 23 exciting drive current and drive voltage for the source 24 coil.

At the time to, the drive voltage instantaneously 26 changes from 0 to V-cap and the current waveform 61 begins 27 to undergo change in accordance with the depicted damped 28 Soscillation, lagging behind the voltage Waveform 62 as 29 Sshown. .The result is that a high intensity, time-varying magnetic field 65 of FIG. 3C is generated. Its relatively Shigh peaks 66 result in a pulse waveform of relatively high Spower content substantially above the ambient noise level, 33 Abrupt steepness in the shape of its waveform is also noteworthy. Such characteristics aids in the generation Sof a full-range of frequency components. Induced magnetic 36 37 fields finally detected by the antenna array have 38 ,r I *ll' 01775 SUL 1989 1 similar frequency characteristics and such signals can be 2 related to a formation parameter of interest 3 conductivity, resistivity or dielectric constant).

4 As well understood by those skilled in the art, the field 65 of FIG. 3C has the same amplitude variation as 6 the current waveform 61 of FIG. 3A. As is well known, the 7 magnetic field 65 will create a primary field of magnetic 8 induction B having a time variation in field strength which 9 corresponds to that of the H field. In the absence of eddy currents, the only inductive field will be the source 11 inductive field created by the primary magnetic field 12 Therefore the voltage induced in the receiver antenna array 13 would be in accordance with the rate of change with time of 14 the strength of that field.

16 BED RESPONSE 17 FIGS. 3D, 3E, and 3F illustrate how field 18 responds in different environments.

19 In this regard, note that the secondary inductive field induces a voltage signal in the receive' coil 21 assembly that varies in accordance with the conductivity of 22 22 the adjacent formation. Since such compornnt is in-phase 23 with drive current waveform 61 FIG. 3A, its change vis-a- 24 vis the primary induced field is directly indicative of 2 formation conductively or resistivity.

26 As shown in the FIGS. 3D, 3E and 3F, eddy 27 currents substantially affect the peak values 69 of the 28 detected signal waveforms 71, 72 and 73 associated with 29 high, intermediate and low resistivity formations. Since the flow of such eddy currents create a secondary inductive 31 field that opposes the primary currents, the high 32 33 conductivity environments create the largest secondary 33 inductive field to oppose the priiiary inductive field.

34 Such a secondary field will act upon each antenna group and induce therein a component of voltage having a waveform 36 correspondence to the voltage waveform 62 of FIG. 3A.

37 38 S:A.iTUTE SHE T vu>U IPEA/US T 0yl 1 I PCvS

IPEA.

.J775 r-i ,-r 05,ju" is SOURCE FIELD GENERATION FIGS. 5-11 illustrate circuitry for generation of t e high-intensity primary pulse magnetic field for logging purposes.

As shown in FIG. 5, host computer/controller 27 is connected to a downhole ccitroller/microprocessor 80 via power/data bus 79 and thence to power source/switching network 81. Prior to discharge, L-R-C circuit network 82 must have been energized by low power trickle charging current using downhole power supplies 83 under control of controller network 84. Power bus 85, as shown, is used to provide the electrical power path through downhole controller/microprocessor 80 and thence to the source/switching network 82. Control signals pass between the same circuits but use the data bus generally indicated at 86. The data bus 86 is under control of controller 87, via microprocessor 88 and memory 89. A feedback loop that includes flux measuring coil 91 about radiation coil provides an identification of the amplitude vs. time waveform of the magnetic field at the coil 50. A potential difference between a surface electrode and a conventional aonde electrode, is indicated at spontaneous potential measuring circuit 92.

CONTROLLER 84 FIG. 6 illustrates controller 84 of power source/switching network 81 in more detail.

As shown, the purpose of controller network 84 is to provide for monitoring of high and low voltage supplies 93, 94. A regulator 95, a voltage monitor 96 and timer 97 in conjunL.ion with A/D convertors (not shown) within voltage monitor 96 and flux monitor 98, are used.

Flux monitor 98 comprises a passive RC integrator connected to a A/D convertor (not shown) within monitor 96.

Thus a voltage proportional to flux is digitized for control purposes. Regulation 95 includes a voltage JNiTUTE SHEET

IPEA/US

A

|E r k0 5 0 1 7 7 1ts

IA/O

1 attenuation circuit in series with an operational 2 amplifier. As unbalanced condition occurs, the operational 3 amplifier output is converted to a digital signal at the 4 A/D converter. Similarly, voltage monitor 96 includes a attenuation circuit that monitors the voltage output of 6 supply 93 via the same A/D convert and thence provides a 7 signal to the microprocessor/controller 80 of FIG. 8 Timing circuit 97 includes a counter controlling a switch 9 as explained below to better shape the primary field waveform.

11 12 MICROPROCESSOR INTERACTION 13 Interaction of the operations of the 14 controller/microprocessor 80 of FIG 5 with controller 84 of FIG. 6 is instructive and can be best illustrated with 16 reference to flowchart 100 of FIG. 7.

17 Assume that a logging station is at hand. A 18 reset command is first generated at 103, within 19 controller/microprocessor 80 after which a load switch enable time is loaded into a register naot shown) as 21 indicated at step 102 of FIG. 7. Thereafter, there is a 22 loading of a delay interval time within another register 23 as indicated at 103. After loading the desired breakdown 24 voltage level into yet another register as indicated at 104, the d.c. supply means are energized as indicated at 2 105 and then the voltage is monitored and tested at 106, 27 107, and 108 after which an acquire command and a rest 28 command is issued at 109 and 110, respectively.

29 In order to cross-check the nature of testing commands at 106, 107 and 108, the voltage levels are 31 Sdetermined at the regulator 95, voltage monitor 96 and 32 timer 97 of controller 84 of FIG. 7 as previously indicated and then the results communicated uphole to the command 34 controller/Ticroprocessor 80. Their purpose is to \ontrol the 36 37 38 1PWU3

T

r A i 3 d 1 Ua C 7 7 S0 5 JUL 189 1 shape, amplitude and duration of the radiated pulse at the 2 source coil 50 of FIG. 5 after enabling L-R-C power circuit 3 112.

4 L-R-C POWER CIRCUIT 6 FIG. 8 illustrates L-R-C power circuit 112 in 7 more detail.

8 Its purpose is to permit accumulation of charge 9 at a low power level, 50 watts for a selected time interval, then discharge the charge through the radiation 11 coil 50 over a very short time, e.g. 60 microseconds. Peak 12 power can be at a level of about 100 megawatts with 13 megawatts being typical. it is to say, the present 14 invention delivers extremely high peak power levels at moderate average power by operating at a very low duty 16 cycle; pulse discharge time is very much faster than pulse 17 charge time.

18 In more detail, L-R-C circuit 112 comprises a 19 charging circuit 113 whereby energy storage capacitors 114 are charged to a predetermined voltage. The circuit 112 21 also includes discharge circuit ill through which the 22 energy in the capacitors is rapidly discharged, creating a 23 high intensity transient current in the radiation coil 24 Initially switches 115, li and 117 are open. On digital 2command, switch 115 in the charging circuit is closed and 26 capacitor 114 are charged by a standard high voltage 27 charging supply 118. When a predeterminea charging voltage 28 on the capacitors 114 is attained, say 5000 volts, switch 29 115 is opened and the pulsed power supply is ready for firing. Upon the firing command from the host 31 computer/controller 27 at the earth's surface, switch 116 32 is closed. Tb,e energy from the capacitors 114 discharges 33 through the main discharge circuit 111 which consists of 34 Sthe capacitors 114 the inductive radiator coil 50 and Svarious parasitic resistances and inductances associated with these elements.

37 38

SW\

f5Y. S i? 1 1 775 IPA/US os5 1 The discharge is characteristic of gh L-R-C 2 circuit as shown in FIG. 3A. If the parasitic resistances 3 are small, the current waveform oscillates and decays 4 slowly. This has the advantage of concentrating the radiated energy around the fundamental frequency of the 6 oscillations. The spectrum of such a discharge is shown in 7 FIG. 3G, peaking at 15 kHz.

8 The discharge waveform can be terminated at any 9 desired time to change the frequency spectrum of the discharge current. This may be desirable to broaden the 11 frequencies available for analysis. Terminating the 12 discharge pulse is accomplished by closing switch 117 at 13 the appropriate time. This diverts the current in the 14 discharge circuit 114 through a large power resistor 120 which rapidly absorbs energy and damps the current 16 waveform.

17 18 PPMHD SOURCE 122 19 FIGS. 9-11 illustrate an alternative pulse power source for L-R-C power circuit 112 of FIG. 8 and comprises 21 a pulsed plasma magnetohydrodynamics (PPMHD) source 122.

22 The advantage of a pulsed plasma 23 magnetohydrodynamics (PPMHD) source 122 is that the major 24 fraction of the discharge circuit energy is obtained from the chemical energy of an explosive cartridge. In 26 practical well logging devices, electrical power is limited 27 to a few hundred watts, most of which is required to run 28 the various electronics and motors associated with the 29 Stool. Qn the other hand the MHD source 122 requires only a small portion of electrical energy but still can emit an 31 31 extremely powerful pulse )f EM energy and is best used for 1 32 2 ultra-extended range applications with limited resolution 33 and minimal requirements on the number of pulses per 34 logging run. Ten to thirty kilojoules can be discharged, for example, in a few tens of microseconds into the 36 radiator coil, leading to peak power 37 38 J-ITUTE SHEU!

IPE

US

I

KA'r ot I S8 PCT/US88/01775 WO 88/09940 1 levels from 500 to 5000 million watts. See U.S Patent 2 Number 3,878,409 for "Explosively Driven Electric Power 3 Generation System", incorporated herein by reference.

4 In operation, source 122 is activated by detonation. As shown in FIG. 9, chemical explosive 126 and an 6 argon gas in space 127 explode. The gas becomes ionized and 7 flows at high velocity into channel 128. The plasma flowing 8 through a magnetic field generated by coils 124 produces 9 power at electrodes 129 connected to a source coil (not shown).

11 12 IGNITION 13 FIG. 10 illustrates the ignition process.

14 As shown, a small auxiliary capacitor chargedischarge circuit 130 creates the magnetic field for the 16 source 122. A small energy storage capacitor 123, is charged 17 and upon the firing command, is discharged through a field 18 coi. 132 surrounding the MHD (magnetohydrodynamic) source 19 122. When the current in thi. auxiliary field doil reaches a maximum, the cartridge 123 is detonated and drives the 21 highly conductive plasma ,at high velocity as previously 22 indicated. The plasma moving at high velocity across the 23 applied magnetic field, develops a powerful electrical 24 output via a Faraday MHD process. This output current pulse is either transformed to a lower current at higher voltage 26 or fed directly to the radiator coil to produce the magnetic 27 dipole.

28 With switch 133 open, switch 134 is closed to 29 charge the small capacitor 131. When the capacitor 131 is charged, switch 134 is opened Uipon receiving the firing 31 command, switch 133 is closed and the capacitor 131 dis- 32 charges through time field coil 132. When the field coil 33 current reaches a maximum, the explosive cartridge 123 is 34 fired and the MHD source begins to generate current. The MHD source switches on when the pulse of plasma arrives at the 36 electrodes which are bathed in the applied magnetic field.

37 8 0 i g M 6 g ,ii PCT/US88/01775 WO 88/09940 /US88/1775 1 A typical current pulse 135 is shown in FIG. 11A.

2 The corresponding frequency spectrum 136 is shown in FIG.

3 11B. This type of pulse produces a broad band frequency 4 spectrum. Because of the large amplitude of the current pulse, the spectral energy densities are high and useable 6 over a range from a faw hundred Hertz to 50 kilohertz.

7 8 ANTENNA ARRAY OPERATION 9. FIGS. 12-19 illustrate circuitry for detecting IQ amplitude v. time of the induced signals using downhole II controller/microprocessor 140 in conjunction with uphole 12 host computer/controller 27.

13 The purpose of the circuitry in FIGS. 12-19 is to 14 relate signal response of each of antenna group A 1

A

13 of FIGS. 2 and 2B and thereby obtain a series of overlapping 16 magnetic field measurements. In that way, lateral range 17 requirement (known to be a function of distance between the 18 sourea coil and antenna group A 1

A

13 can be decoupled 19 from system logging requirements. It should be further noted that all operations are paced by digital control 21 signals interacting between the host computer/controller 27 22 at the earth's surface and downhole controlle./micro- 23 processor 140 of FIG. 12.

24 CONTROLLER/MICROPROCESSOR 140 26 FIG. 12 illustrates controller/microprocessor 140 27 in more detail.

28 As shown, the controller/microprocessor 140 is 29 connected to host computer/controller 27 via power data bus 141 and includes a controller 142. The controller 142 3controls pre-amplification network 142, mixer network 143, 32 A/D convertor circuit 144 andi microprocessor 145 to detect 3 and digitize inducted signals at the antenna array 146.

33 34 36 'r'oI S88/09940 PCT/US88/01775 I> o 88/09940 *j 1 FIGS. 13A 13B illustrate how a pair of receiving 2 coils comprising a single anrtenna group A 1

-A

1 of FIG. 2B 3 can be formed into a three component array 150 wherein group 4 Al is multiplied by three so each pair has an axis of response 151, 152 or 153 orthogonal to the two remainder axes.

6 7 PRE-AMPLIFICATION CIRCUIT 141 8 FIG. 14 illustrates pre-amplification circuit 142 9 in detail.

As shown, incremental difference signal responses 11 from an antenna group A 1 are identified at the output 145 12 of amplifier 146 after the signals have passed through 13 blocking transformer 147 and switches 148. After amplifica- 14 tion at amplifier 147, the difference signal passes to mixer circuit 143.

16 The circuit 142 is designed for low noise, high 17 common-mode rejection and provides a high gain bandwidth .8 product to minimize temperature effects. Compensating 19 attenuator networks 150 before and after amplifier 146, are used to adjust the levels of neighboring antennas to achieve 21 good differencing of the mixed signals.

j 22 The second amplifrr 149 is used to boost the 23 signal so that further electronic amplification, attenua- 24 tion, or mixing does not degrade the signal-to-noise ratio (SNR) The SNR of each antenna is established prior to the 26 second stage of amplification at amplifier 149 and is the 27 highest SNR possible.

28 S29 MIXER CIRCUIT 143 S 30 FIGS. 15 and 15 illustrate mixer circuit 143 in 31 detail.

32 As shown in FIG. 15, ditect signals from the end 33 antennas group A 1 3 of FIG. 2B are first gain and phase 34 nulled via appropriate change in the values of variable resistor 200 and capacitor 201. In that way, response of the 36 antennas can be easily calibrated. Signal level control at 37 Vr 8 /01i77 b 1PEA/U U0 5

J

UU989 1 non-inverted inputs 202, 203 of amplifiers 204, 205 2 respectively, add additional range to the calibration 3 process. Filters 205 and 206 at the input and at the 4 output of amplifier 205 rid the signal of unwanted low and high frequency components as generated by the tool d.c.

6 power supply and the L-R-C circuit, 7 As shown in FIG. 16, mixed signals from antenna 8 group Ai...A 12 are similarly gain and phase nulled via 9 appropriate change in the values of variable resistors 210 and capacitors 211. After passing through switch board 11 212, additional range to the calibration process is 12 provided by having a normalization signal level appear only 13 at the non-inverted input 213 of amplifier 214. Filters 14 215 and 216 rid the signal of unwanted frequencies as previously discussed.

16 After passing through mix circuit 143, the 17 signals are digitized by A/D convertor 144 of FIG. 12. The 18 convertor 144 preferably has a dynamic range of at least 12 19 bits and digitization rate of 100 kaz to 1MHz. Once digitized, the signals are stored in on-board array 145A of 21 FIG. 12 and awaits the command from the host 22 computer/controller 27 to begin transmission to the 23 surface.

24 FLOWCHART 250 26 FIG. 17 illustrates operations of the detection 2 steps in detail via flowchart 250. Assume that a logging 2 station is at hand. Preamplification first occurs as 29 29 indicated at 251. Next, a mixing occurs at 252.

Thereafteri the signals are filtered, amplified and 31 digitized at 253 and 254 after which a conditional 32 S6tatament at 255 permits transfer of the data uphole if the 33 user does not Wish the data to be Fourier transformed 34 downhole. If Fourier transformation is desired, then the data is trnsferred to controller/micr-processor 80 via 36 3 instructions 256 and 257 for such purposes. Thereafter, 37 Sthe Fourier transforms are transformed to the host computer 38 via instructions 258 and 259.

SISTITUTE SHU

IPEA/US

WO 88/09940 PCT/US88/01775 ,W 81990 1 In addition, to the acquisition of the digitized 2 secondary field signa_... three other data signals must be 3 acquired. The first is a flux monitor signal via the flux 4 monitor 98 of FIG. 6. The output of the flux monitor 98 is proportional to the magnetic dipole moment of the transmis- 6 sion pulse, as previously mentioned. The signal from the 7 flux monitor 98 is amplified and filtered prior to digiti- 8 zation. This signal is not mixed but is used to verify 9 proper performance of the pulse power discharge circuit and to normalize the antenna signals to correct for pulse-to- 11 pulse differences.

12 A spontaneous potential (SP) measurement and/or 13 gamma ray measurement are also made. These are industry 14 standa-d measurements and are used to correlate the present log with other logging tools.

16 17 OPERATIONS OF THE LOGGING SOMOE 18 In the present invention, the operations are 19 carried out by a properly programmed digiual -logging code and the integrated use of the host computer/controller at 21 the earth's surface with the downhole control/micro- 22 processors 80 and 140.

23 24 CODE CHARACTERISTIC 26 DESTINATION

CODE

27 Lead To From 28 Radiation BIT MINE HIS HOST ME 29 Controller/Micro 1 1 2 0 1 Processor 31 32 Antenna 33 Controllefi/Micro 0 2 1 3 2 34 Processor 140 36 1 ~A4~ w 37 4uus 0 C, L; Z.

Qs 1 ADDRESS ID'S UART CONTROLLERS CODE 3 ReceJ.%iz Buffer Register A 4 Transmit Buffer Register A Control B 6 Status B 7 Modem Control C 8 Baud Select DATARDY 9

XMITRDY

i 1 i In this regard UART's are designated as universal 12 asynchtronous receiver transmitter which act as parallel-to- 13 serial convertors, while maintaining system format 14 integrity.

16 OPERATION 17 1. The tool is lowered into the borehole with 18 the wireline. During this period the tool status and 19 communications are checked and verified. Background data acquisitions are taken to verify the proper operation of the preamplifiers.

22 22 2. When the tool is ready to be drawn up the 23 borehole, the logging sequence is initiated. Wireline 2 operators strongly prefer that the tool be drawn without 2S stopping. When a tool is stopped in a borehole, it has a 26 high probability of becoming stuck. One major advantage of 27 a pulsed tool is that the data is unaffected by the draw 28 r28r 4 te so that the wireline operators constraints are 29 decoupled from the quality of the data.

3. The tool is set to pulse at predetermined 31 depths. The tool draw rate is selected to ensure that the 32 tool is ready to fire when hhe predetermined depth is 33 reached. Accurate depth information is provided 34 continuously by the wireline truck to the host computer in Sthe data truck.

36 37 38 v S'7 o' 1 1 t.

36 l' 1 4 When the fire commnd is given by the host 4 the data acquisitiion system i turned on; 38 9 the microprocessor in the command module 7 data acquisition terminates approximately initiates the recharging of the energy storage capacitors; 11 the microprocessor in the antenna module 12 directs the output of each antenna's RAM, in turn, down one 13 of two paths: through a Fast Fourier process in 14 firmware and then transmit the resulting data to the host computer on the surface at a 2400 baud data rate; (ii) 16 directly to the host computer at 19,200 baud.

17 the data transmission to the surface is 18 completed; the storage capacitors are charged to the 6 prescribed voltage; 21 the tool automatically does a status check of 22 switch positions and communications links wth each 23 antepna; 24 the tool is ready for the next pulo when the prescribed depth is reached; 26 S5. Anter pulsing the tool a t selected intervals 7up the borehole thedata acquisition is complete. The ata 28 now resides in the host computer.

29 During these steps, the microprocessor/dontroller performs as follows: 1Control: operate switch on the supply charging circuit; 3 antenna 4 t--operate switch on the discharge circui t 33 S -operate switch to divert the current pulsi 34 37 Sup throughe borehole the data acquisition is complete. The datresistor; 352 now resides in the host computer During collect and transmit spontaneous poter/controial and gamma 1o data (used to correlate with other logging tool); Wti~ i S I WO 'j 9940 PCT/US88/01775 WO 994024' 1 Monitor:--tool electronics DC power levels; 2 charging voltage on the electrical power supply; 3 temperature in the command module electronics; 4 switch open or closed status: Tie microprocessor/controller 140 in the antenna 6 module performs the following functions: 7 Control:--turns on data acquisition channels; 8 sets mode (mixed or unmixed) of antenna 9 channels; I sets controllable gains in the signal condition- I ing boards; 12 collects and transmits antenna data; 13 collects and transmits flux monitor data; 14 sets the antenna relays for data acquisition, or background noise acquisition; 16 directs antenna and flux monitor data to the 17 surface directly or to digital signal processing 18 in the microprocessor.

19 FLOWCHARTS FIGS. 18 AND 19 21 In FIG. 18, the microprocessor/controllers are 22 initialized and a status check is completed at steps 260, 23 261. Therfafter, the particular devices are caused to be 24 called into operations as each serially answers the conditional queriet, of steps 262, 263, 264, 265 and 266. The 26 object of the queries is to determine an appropriate com- 27 mand. Performance is then built around execution of steps 28 267, 268 and 269. If performance is not possible, steps 269A 29 and 269B are executed.

In FIG. 19, operations of the antenna micro- 31 processor/controller 140 are set forth in some details.

32 As shown, a reset command at 270 resets the 33 add!:ess counter followed by the issuance of an acquire 34 command at 271. Conditional time command at 272 directs the transfer of either a or a bit at steps 27 275. The 36 address counter is then reset at step 274.

37 i {I ^A'f o] I

'A

O 88/09940 pCT/US88/01775 i \VO 88/09940 yo, 6 7 8 9 11 12 13 14 16 17 18 19 21 22 23 24 26 27 28 29 31 32 i33 34 36 37 INVERSION PROCESS FIG. 20 and 21 illustrate processing of the data in accordance with the present invention.

DATA PROCESSING The measurable data in a borehole are magnetic fields; a coil receiving antenna produces a voltage proportional to the time rate of change of magnetic flux threading the coil. Receiving coils may be oriented either vertically or horizontally; for the horizontally layered formation model it is sufficient to use vertical antennas; measuring formation dip angle of three-dimensional effects requires the use of horizontal antennas as well.

It may be shown by direct mathematical calculation that in order to determine formation electrical properties uniquely it is 'necessary to obtain magnetic data from a number of receiver antennas spaced vertically in the borehole. It is important to have the spatial information; measurements at different frequencies, which-' are readily available as a consequence of pulsed operation in the present invention, provide data more directly related to radial variations than to vertical layering. Vertical resolution of the present invention is directly related to the receiving antennas; the closer the vertical spacing the finer the resolution of the log.

There are practical limitations in antenna fabrication techniques and data acquisition electronics which limit available data and effect resolution and accuracy of the log. It is not sufficient to develop a mathematical relationship between borehole magnetic data and formation properties; the effects of electronic noise, finite antenna spacing, finite digital accuracy, and other practical problems must be taken into account. It may be demonstrated by direct numerical calculation that the mathematical zelation between formation properties and field data is sensitive to noise or data corruption of any variety, whether electrical, 'v f

I

c i i -Fw voltage waveform, 62 of FIG. 3A.

I

T1JfE SMET 1PEA/U3 PCT/U7 7 in 1989 1' mechanical, or digital. In addition to minimizing the sources of noise by novel antenna and circuit design, the present invention minimizes the effects of noise.

In a determinate data processing system the quantity of data acquired in the antenna module exactly equals the quantity of data specifying the layer module.

This type of system is attractive from a mathematical perspective, but it has substantial practical problems. In a noise-free situation a determinate system exactly and uniquely reproduces the formation properties of a layer model; if magnetic fields are calculated from an assuamed layered model to a very high degree of precision, and these values processed by the determinate system as though they were experimental data, then the formation parameters recovered by a determinate system agree exactly.

Unfortunately this system is particularl~y vulnerable to error. If the simulated data is corrupted by additive noise the resulting formation properties deviate from the true value and cause a false reading. The determinate system does niot provide satisfactory accuracy in 4tn inductive logging system designed for fiel~d usage.

The present invention utilizes an overdetermined, red'andant data processing system to minimize effects~ of noiso and to max'imize accuracy of the log. Magnetic data is acquired and digitized by the antenna module, with antenna locations as close as possible consistent with available data channels and, desired vertical resolution.

This data, consisting of a digitized representation of magnetic fields as a function of time, is processed by well known digital Fourier transform techniques to yield a representation of the field quantities in terms of their frequency spectrum. The number of frequency terms obtained in this process can be quite large, up to perhaps 2048, depending primarily on allocated computational resources and details of the pulse of the electromagnetic radiation source. The combination of magnetic field data at numerous antenna locations and numerous frequencies constitutes a matrix of data available far, Z,'33TITIJTE SHEE 1PEA/US

I,'

CCC PCT/US88/017 7 Si" ,WO "88/09940 1 redundant signal processing. This situation is to be corn- 2 pared with prier art induction logging systems, in which the 3 available data comprises one receiver antenna signal and one 4 frequency.

In the preferred embodiment of the invention data 6 is subdivided into several frequency regions in order to 7 provide information relating to radial variations in formag tion properties. Within each frequency region, say for o ex&aple 10 kHz to 15 kHz, the data from several frequencies and all antenna locations are collected and defined to be 1 the target data matrix. It is the objective of the signal 12 processing to determine formation properties that best match 13 the target data, in the sense of minimizing the average 14 error between calculated and target data.

Formation properties are determined in the present 16 invention by systematically varying formation model parame- 17 ters to find optimum values that minimize the mean square 18 error between target magnetic field data and calculated 19 model data. A first estimate of model parameters is made, based ca, an approximate solution to the field equations, and 21 subsequently the values .are improved by means of an 22 iterative method until optimum model parameters have been 93 determined. Fig. 20 is a flowchart of the preferred inver- 24 sion process. As shown, digital data from pulsed operation of the tool arrives in matrix form. The radiating antenna 26 and each of the receiving antennas is associated with a 27 number of frequency components derived from the Fourier 28 transform signal process. Each frequency component is a 29 complex digital number representing amplitude and phase normalized with respect to the source.

31 Formation properties are represented by a matrix 32 of ,1lues. For example, at 301 an initial estimate of forma- 33 tion properties is generated. In the preferred embodiment 34 formation properties are represented by bed thickness, bed location, and bed conductivity; these properties may be 36 augmented by dip angle and azimuthal variation. The dimeri- 37 1 d 1 32 PCI^S dA/0i?7§i I PEA/US 05 JUL 1989 1 sion of the formation matrix the total number of 2 parameters to be determined) must be less than or equal to 3 the dimension of the receiver antenna data matrix of 300.

4 In the preferred embodiment, there are 13 receiver antenna locations and two frequencies, corresponding to 26 complex 6 numbers, or a total dimension of 52 real numbers in the 7 receiver data matrix. There are 16 layers, each with a 8 variable location, for a total of 32 real numbers in the 9 formation property matrix. The inversion calculation is preformed once for each pair of frequencies; different 11 calculations at different frequencies determine how the 12 formation varies laterally from the borehole. In the 13 preferred embodiment calculations are performed at 14 frequencies of 5.0, 5.2, 10.0, 10.4, 20.0, and 20.8 kHz.

Electromagnetic fields at the above frequencies, are 16 calculated at 302. Then the results of such calculations 17 are evaluated at 303 such that the final formation values 18 should be accurate within a tolerance range of five per 19 cent as determined at 304. If the final formation values are not within specification, the formation 21 properties are re-evaluated at 305 using changed elements 22 of a Gauss-Newton Matrix to yield updated estimate of 23 formation properties at 306. The process is then repeated 2 through loop 307. If the formation values are within specification limits as determined at 304, the in-range 26 results are stored 308.

2"7 27 A forward solution of the electromagnetic 2 equation is used to calculate the antenna matrix on the 29 basis of specified formation properties. The preferred 3 forward solution is a new analytical solution to the 31 31 electromagnetic equations. Alternatively a layering 32 3 technique based on Helmholtz potential theory may be used as in the manner of W.D. Kennedy, H.F. Morrison, S.M.

34 Curry, S.P. Gill, "Induction Log Response in Deviated Boreholes"; Transactions of the SPWLA Annual Logging 36 3 Symposium, Society of Professional Well Log Analysts, Houston, Texas; Vol II, paper FFF, June 9-13, 1986 38 /Je h'I1rTU r J L 4 V cv n~.1 i_5ese elements.

37 38 -vi PCT/US88/01775 WO 88/09940 33 1 The preferred solution represents an exact ana- 2 lytical solution to Maxwell's electromagnetic equations in a 3 formation whose electrical properties vary in an arbitrary 4 manner, continuous or discontinuous, in a specified direction. In the case of a non dipping formation, the proper- 6 ties vary in the vertical direction only. The solution is for the case of a vertical source antenna and vertical b receiver antennas.

9 In formation with specified electrical parameters, define the following complex functiona, which in general are 11 discontinuous functions of vertical position, changing 12 abruptly at bed boundaries: 13 14 16 17 19 k 21 6 formation conductivity 22 formation dielectric value 23 formation magnetic permeability 24 s real integration variable radian frequency 26 27 These discont-inous functions are recursively 28 integrated along the vertical axis to generate three con- 29 tinuous functions with spec-ified initial conditions:, 31 S32 Z d2 33 (2) 36 37 1 37 AL-u coii, leading to peak power 38L ,TUTE

SHEET

ILAUS

4"-0 the N 'I U' p WO 88/09940 31- PCT/US88/01775 'i -7 Zk V- cc) /9 t- 00) d2' -Al x 1 18 19 21 22 23 24 26 27 28 29 31 32 33 34 36 In these equations 5 the vertical Position z is equal. to zero at the dipole source and increases in the Positive direction toward the receiver antenna array located below the source.

The magnetic field at specified receiver location along the vertical axis is given by the follow~ing integral; BZ

I

Zrrj W exe&e-4 Iwo 8/0940 5~*PCT/US88/017 7 Equation 5 represents the exact solution for the 2 measured magnetic field, normalized with respect to the 3 :magnetic moment of the dipole source, for a formation with 4an arbitrary continuous or discontinuous variations in all electric parameters. The solution may be extended to 6 include dip angle and azimuthal variations.

7 More specifically, in the preferred induction 8 logging system the formation is assumed to have constant 9 magnetic permeability and negligible dielectric effects, and formation parameters are given by bed., thickness, bed 11 location, and bed conductivity. Since conductivity is 12 assumed to be constant within a bed, Equations I through 4 13 are integrated to provide simpler step by step recursion 14 relations: 16 17 e 1 19r 23 24 262 28 32 34 Ir 36 37 -yo WO 88/0940 PCT/US88/01775 -i WO 8809940 *4 i In this simplified model, the formation properties 2y are averaged above the source, and processed to vary only in 3; the region sensed by the antenna array. In the preferred 4 embodiment, there are 16 beds, which may be located anywhere below the source; the inversion process locates the bads and 6 defines the bed conductivities.

7 The forward solution program steps are described 8 in FIG. 21. A standard numerical quadrature integration 9 routine, for example Laguerre integration, is used for the i'ntegration of the field strength in terms of the integra- 11 tifon: variables S via execution of step 311. Each value of S, 12. from step 311 is then combined with the input material 1:r properties of step 311 to calculate layer functions at 312, 14 and' the results in turn are used to recursively evaluate the auxiliary functions defined in Eqs. 6 and 7 at step 313.

16 After incremental' integration of t 4 e B-field (Eq. 5) has Sbeen completed via execution of steps 314 and 315, the 1.8& output of the forward solution at 316 are thus normalized 19 receiver antenna signals field strength per unit magnetic moment of the source) in the same data format as 21 the experimentally measured data. Storage then occurs 317 22 as elements of a normalized receiver data matrix.

23 FIG. 22. illustrates the format of normalized 24 receiver data matrix 320 in detail The matrix 320 comprises a frequency domain trans- 26 formation of the original (amplitude v time/depth) detected 27 secondary signals normalized to a logging station 28 1, 2, as shown. It comprises a series of elements 29 having an ar litude value A,A I3,B' C,C associated in Srows and columns identified with common antenna depth 31 location and cr.mMon. component of frequency. That is to 32 say, each element B,B C,C represents a component Sof the amplitude of the detected signal at each logging Sstation and each also corresponds to a particular common frequency-ahtenna location. For example, for logging station 36 1 for common column the elements are identified as 37 o

I

PCT/US88/01775 WO 88/09940 7 1 Afo; 2,fo; A 3,fo; An3fo assuming the antenna S 3,fo,fo 2 array comprises thirteen (13) separate antenna groups where 3 the row subscripts 1, corresponds to each of such 4 groups and the common column frequency subscript is fo, fl..fn- Similarly for a particular common antenna depth 6 number 1, or 13, at logging station the elements 7 are identified as Aifo; A'l,f1; A",f 2 where the row 8 subscripts corresponding to the common antenna group 9 and the column subscripts corresponds to each of the frequency components fo, as shown.

11 "Frequency slicing" can occur about one or more 12 columns of each matrix 320, say fl-f 3 Moreover, such tech- 13 nique is employed in the present invention for investigati- I, on purposes.

ITERATIVE ERROR MINIMIZATION The preferred method for inverting the forward solution has been previously discussed in conjunction with tCG. 20. Recall that such method, used to determine formation property matrix from the antenna data matrix, is an iterative method that provides a best fit, in the least square error sense, to the measured data. This method best handles complications caused by non-ideal data that contains additive noise or perturbations from non-ideal formations.

Additive noise arises from thermal effects in the antenna preamplifier circuitry and finite digitization in the analog-to-digital convertor circuitry.

Note that initial estimate of formation properties occurs at 301 in FIG. 21. This estimate may be a constant matrix or an approximate analytical estimate and is used with the forward solution to generate antenna signal data.

The calculated antenna matrix is then compared with the measured data, and a mean square rror evaluated, as described before in matrix format as indicated in FIG. 22.

i1 WO 88/09940 i PCT/US88/01775 37 The mean square error is an analytic function of the formation property matrix by virtue of the analytic forward solution expressed in Eqs. 1 to 5 and the Jacobean erivative can be evaluated. The Jacobean represents the derivative of the antenna signal matrix with respect to the formatio parameter matrix, on a coefficient by coefficient basis. It may be calculated explicitly in forms of the analytic solution expressed in Eqs. 1 to 7 or it may be evaluated numerically. In the preferred embodiment the Jacobean differential coefficients are evaluated numerically by determining the ratio of antenna matrix variation to a small variation of formation properties.

The iteration procedure for obtaining an inverse solution in the preferred embodiment is the well known Gauss-Newton technique. The technique may be described by placing the contents of the measured antenna data matrix in a linear vector T, the calculated antenna matrix in a linear vector C, and the formation property matrix in a linear vector X. In the preferred embodiment T and C-consist of 52 numbers.

The Jacobean is defined to be the matrix of differentials of the calculated value C with respect to the parameter values X:

I

ax The iteration proceeds from step n to step n+1 as follows: Xnt (Xr)JT (T Ca WO 88/09940 31PCT/US88/0 1775 1 The iteration rproceeds until the change in succes- 2 sive values diff. ;s by a p'fedetermined error value (typi- 3 cally1% 4 To those skilled in the art to which this inventiJon relates, ma~ny i~sin construction and widely dif- 6 fering embodim.,ntz anut aiiplications of the invention will 7 suggest: themselves without departing from the spirit and 8 scope of the invention. The disclosures and the descrip- 9tions herein4 are purely illustrative and are not intended to 1,be in any ranse limiting.

IL What is claimed is: 12 13 14 16 19 21 22 23 24 26 27 28 29 31.

32 33 34 36 37

Claims (14)

1. A method for systematically logging an earth 2 formation around a borehole using a logging sonde to generate 3 a primary magnetic field in the formation at depth and 4 detecting an induced secondary magnetic field indicative of a formation electrical parameter, comprising the steps of: 6 g eratig a series of time spaced primary 7 or pet i ver magnetic field pulse's a a series of stations along the 8 borehole; 9 (ii) detecting response data relative to the secondary magnetic field induced in said formation by each 11 said primary magnetic field pulse; and 12 (iii) processing said response data to provide an 13 indication of a formation electrical parameter around the 14 borehole by modeling response of the adjacent formation 16 using a series of assumed conductivity and bed thickness 17 values along with actual parametners of the pulsed primary Sfield, and cross-checking a forward solution of the model 1 of step against measured detected components of the secondary magnetic field at depth. 21 22

2. The method of claim 1 wherein the peak power of 23 said primary field pulse is at least one megawatt. 24

3. The method of Claim 1 wherein the peak power of 26 said primary field pulse is about 50 megawatts. 27 28 4. The method of Claim 1 wherein tho peak power of 29 said primary field pulse is in a range of 1-1000 megawatts. 31 5. The method of Claim 1 wherein the step (ii) of 32 detecting the secondary magnetic field includes the substeps 33 using a series of antennas, each of which independently 34 detect components of said secondary field associated with different formation regions and digitizing each detected 36 coiponent independently. 37 38 SC-SSTtYE SuET IPEA/US so -uunter is then reset at step 274. 37 31 i. 4 4 PCT/US 0 17 7 0 5 JUL1989

6. The method of Claim 1 wherein the forward solution and the cross-checking of that solution against 3detected components of the secondary magnetic field at depth 4 are carried out in the frequency domain.

7. The method of Claim 6 in which the step of 7 detecting the secondary magnetic field includes using a series 8 of antennas in the sonde and wherein frequency transforms 9 related to the instant depth of said antennas during each said pulse are used in the forward solution and cross checking 11 12 steps. 13

8. The method of Claim 7 in which the frequency 14 transforms can be frequency sliced to concentrate response 13 about a frequency range R. 16 17 18 9. The method of Claim 8 in which R is about 1-40 18 Skilohertz thereby yielding acerate conductivity values. 1 10. The method of Claim 8 in which R is about 1-30 21 22 megahertz thereby yielding accurate dielectric values. 23 24 11. The method of Claim 7 wherein the frequency transforms are in mat ix format in which row and column 26 designations are antenna depth and frequency. 27 28 12. The method of 'Claim 7 including the steps of 29 mixing signals from adjacent antennas in said seric-' of antennas in the sonde so that a digitized signal is produced 31 which represents the second derivative or curvature of the 32 magnetic fieli. 33 34 13. An apparatus for systematically logging an earth formation around a borehole using a logging sonde by 36 generating a primary magnetic field in the formation at depth 37 and detecting an induced secondary magnetic field indicative 38 of a formation electrical parameter, comprising: I P di! US >r 1 P PCT/US /01 'cru R 17 7_ 1JUL 981 means for generating a series of time spaced 2 primary magnetic field pulses of high peak power at a series of stations along the borehole, said generating means 4 including a radiation coil and a power source network in selective operative contact therewith, said power source 6 network including a magnetohydrodynamic (MHD) source and a 7 switch network; 8 (ii) means for detecting response data relative to 9 Sa secondary magnetic field induced in said formation by each Ssaid primary magnetic field pulse, and 12 (iii) means for processing said response data to 1provide an indication of at least one electrical parameter of the formation around the borehole. 14

14. The apparatus of Claim 13 in which the MHD 16 1 source generates said pulsed primary field by selectively 18 driving a radiation coil with a pulsed oscillating current wherein the primary field energy per logging cycle is substantially constant. 21 22 15. The apparatus of Claim 14 in which the peak 23 power of said primary field is in a 'range of 10-1000 24 megawatts. 26 16. An apparatus for systematically logging an 27 earth formation around a borehole using a logging sonde by 28 generating a primary magnetic field in the formation at depth 29 and detecting an induced secondary magnetic field indicative 'of a formation electrical parameter, comprising: 31 means for gen rating a series of time spaced porm ht f' k oAver 32 primary magnetic field pulse sat a series of stations along 33 the borehole; (ii) means for detecting response data relative 34 to a secondary magnetic field induced in said formation by each said primary magnetic field pulse, and (iii) means for 36 processing said response data to provide an indication of at 37 least one electrical parameter of the formation around the 38 borehole; wherein said processing means includes means for tSu"ST1T1IUTE SHEET IPEA/US e? t us 8 /01775 5 JUL 989 modeling the response of the adjacent formation using a series 2 2of assumed conductivity and bed thicknes, values along with 3 actual parametners of the pulsed primary field, and means 4for cross-checking a forward solution of against detected of the secondary magnetic field. 6 7 17. The apparatus of Claim 16 in which peak power 8of said primary field is at least one megawatt.

18. The apparatus of Claim 16 in which the peak Spower of said primary field is about 50 megawatts. 12 13

19. The apparatus of Claim 16 in which the peak 14 power of said primary field is in the range of 1-100 Smegawatts. 16 17 The apparatus of claim 16 in which said means 18 for detecting the secondary magnetic field includes a 19 series of antennas, each of which independently detect 21 components of said secondary field associated with different 22 formation regions, and (ii) means connected to said series of 23 antennas for digitizing each detected component independently. 24

21. The apparatus of Claim 16 in which said 26 modeling means that provides the forward solution of and 27 said cross-checking means of are carried out in the 28 frequency domain. 29

22. The apparatus of Claim 21 in which said 31 processing means provides frequency transforms that can be 32 frequency sliced to concentrate response about a frequency 33 range R. 34

23. The apparatus of Claim 22 in which R is about 36 1-40 kilohertz thereby yielding conductivity values. 37 38 SUBSTITUTE SHEET 41 USI C S 0O S S .me C 44

24. The apparatus of claim 22 in which Ris about 1 f to 10 megahertz thereby yielding dielectric values. The apparatus of claim 22 in which the frequency-depth transforms are in matrix format in which rows and columns designations are antenna depth and frequency.

26. A method for systematically logging an earth formation around a borehole using a logging sonde to generate a primary magnetic field in the formation at depth and detecting an induced secondary magnetic field indicative of a formation electrical parameter, comprising the steps of: generating a series of time spaced primary magnetic field pulses of high peak power at a series of stations along the borehole by selectively driving a radiation coil with a pulsed current, the selective driving of said radiation coil being controlled by a digital logging code, which is generated by a host computer/controller at the earth's surface in association with at least one Z. microprocessor/controller on-board the logging sonde; (ii) detecting response data relative to a secondary magnetic field induced in said formation by each said primary magnetic field pulse; and (iii) processing said response data to provide an indication of a formation electrical parameter around the borehole. "I 27. A method for systematically logging an eartt formation around a borehole substantially as herein described with reference to the accompanying drawings. 9 28. An apparatus for systematically logging an earth formation around a borehole substantially as herein described with reference to the accompanying drawings. S.. DATED this 22nd day of January 1991 MPI By their Patent Attorneys GRIFFITH HACK CO. s/EM I.- ci Sr *0 S i.