CA1293991C - Nuclear magnetism logging tool using high-temperature superconducting squid detectors - Google Patents

Abstract

A B S T R A C T

NUCLEAR MAGNETISM LOGGING TOOL USING HIGH-TEMPERATURE
SUPERCONDUCTING SQUID DETECTORS

A logging device is disclosed for measuring the nuclear magnetism response of earth formations. The device employs a detector that is capable of detecting sinusoidal and slowly varying changes in magnetic field caused by precession of mobile nuclei about the earth's magnetic field. Such a detector may be one or more laser-pumped helium magnetometers, or one or more high-temperature Superconducting Quantum Interference Devices (SQUID).
The invention uses a micro-miniature Joule-Thomson refrigerator to maintain the high-temperature SQUIDs below their superconducting transition point. The SQUIDs are flux coupled in an axial gradiometer configuration to reduce motion-induced magnetic noise.

Description

NUCLEAR MAGNETISM LOGGING TOOL USING HIG~-TEMPERATURE
SUPERCONDUCTING SQUID DETECTORS

This invention relates to apparatus for use in a borehole, and more particularly, relates to nuclear magnetism logging apparatus for use in an earthen borehole.
The most accurate open borehole logging device for measuring 5 residual oil saturation and permeability in earth formations is the Nuclear Magnetism Log (NMI.). This logging tool uses the magnetic field from a solenoidal coil to polarize the protons in any fluids contained within the earth formation adjacent the tool. The solenoidal coil is then turned off and any mobile protons in-fluenced by the magnetic field of the solenoidal coil precess at their Larmor frequency about the earth's magnetic field. This precession may be measured as a damped sinusoidal voltage induced in a separate detection coil in the logging tool. The induced voltage decays rapidly, typically of the order of 20-50 milli-seconds, because of extremely short spin-spin or transverse relaxation times, T2~, of these fluids.
The initial amplitude of the nuclear magnetism signal is proportional to the "free-fluid" or producible fluids in the formation adjacent the tool. If Mn-EDTA is added to the drilling fluid and allowed to invade the formation, then the only residual nuclear magnetism signal will be from any oil phase in the formation ad~acent the tool. Thus, a NML log-inject-log procedure results in highly accurate measurements of residual oil saturation in the formation.
The spin-lattice or longitudinal relaxation time Tl of any fluid in an earth formation is related to the pore size(s) (and their distribution) containing the fluid; the pore sizes and distribution may then be related to the capillary pressure curve ~3~

which may then be related to the permeability of the Eormation in a well known way.
The NML attempts to measure Tl by repeated polarization cycles with successively longer polarization times. Since Tl is very much longer than the time T2* of the damped sinusoidal decay, Tl cannot be determined directly from the damped sinusoidal decay curve with only one measurement.
Although a highly useful logging tool, one of the main dis-advantages of the NML is its poor signal-to-noise ratio, which limits its accuracy to about 1 unit of porosity during continuous logging operation. Although this is adequate for measuring a free-fluid index, residual oil saturations are typically 1/3 or less of the total porosity, so that stationary NML operation is required in order to measure the residual oil saturation to sufficient accuracy for enhanced oil recovery requirements.
Typically, about 15 minutes of data is collected at a borehole location and then averaged to obtain residual oil saturation to better than 1 percent of saturation at that location. However, maintaining the tool stationary for this length of time slows data acquisition for the entire formation and increases the risk of sticking the tool in the borehole.
Another disadvantage of the NML is the shallow depth of in-vestigation into the formation. Again, due to poor signal-to-noise ratio, only signals from the initial few centimeters of the bore-hole wall can be detected.
Still another disadvantage is that repeated polarization cycles are required to obtain a discrete number of measurements of the Tl decay curve. This requires additional time, and since only a few measurements are obtained, the shape of the Tl decay curve so obtained is not very precise. This Tl decay curve then contains imprecise information on pore sizes and their distribution (and accordingly any capillary pressure or permeability determined therefrom is imprecise) in the formation, which is a severe disadvantage of the present NML tool.

~3~9~

In the prior art the use of a Superconducting ~uantum Interference Device (SQUID) as a detector in a modified nuclear magnetism tool has been proposed, with the modification being the use oE two opposed superconducting magnets instead of the conventional polarizing solenoid to provide an increased depth of investigation into the borehole wall. ~lowever, the use of a SQUID
as a detector in a downhole device has not occurred because of the cryogenic and safety considerations of using liquid helium in the borehole environment. More specifically, liquid helium expands over 600 times on vaporizing and cannot be safely vented into the borehole. This problem, as well as that of providing an adequate amount of operating time downhole, in the high-temperature environment of a borehole, has so far prevented -the use of a SQUID
as a detector in a nuclear magnetism tool.
These and other limitations and disadvantages of the prior art are overcome by the present invention, however, and apparatus are provided for nuclear magnetism logging with a detector capable of detecting sinusoidal and slowly varying magnetic fields in an earthen borehole.
The present invention provides a logging device for measuring the nuclear magnetism response of earth formations using one or more detectors capable of essentially simultaneously detecting sinusoidal and slowly varying magnetic fields resulting from precession of mobile nuclei about the earth's magnetic field.
~ The apparatus according to the invention thereto comprises - means for providing a magnetic field that is capable of being energized or de-energized in response to preselected control signals, - detector means for detecting sinusoidal and slowly varying changes in magnetic field when said means for providing a magnetic field is de-energized and outputting signals indica-tive thereof, and - an electronics package for generating said preselected control signals and for receiving signals from said detector means.

~33'~

Preferably, the detectors in the apparatus according to the invention are high-temperature Superconducting Quantum Interference Device(s) (SQVID). Alternatively, one or more laser-pumped helium magnetometers may be used AS detectors. The presently preferred apparatus of the present invention uses a micro-miniature Joule-Thomson refrigerator, or alternatively, thermoelectri-cally-cooled Peltier modules, integrally mounted on the same wafer as the SQUID(s) to maintain any high-temperature SQUID(s) below its superconducting transition temperature Further, the SQUID(s) may be flux coupled in an axial gradiometer configuration to reduce any motion-induced magnetic field noise.
The preferred apparatus of the present invention provides a means for using a SQUID sensor in a nuclear magnetism logging tool without the need for any cryogenic liquid cooling. The method consists of fabricating one or more SQUIDs from a high-temperature superconducting material, such as for example, but not limlted to the rare earth-barium- copper oxide materials, such as Yttrium Barium Copper Oxide (YBa2Cu307), which is superconducting at temperatures exceeding 90 K. The SQUID(s) is epitaxially deposited on a high thermal conductivity substrate such as SrTiO3 or MgO.
The wafer substrate in turn is bonded to the cold stage of a micro-miniature Joule-Thomson refrigerator (or a Peltier module for higher temperature superconducting materials) which can cool down to about 80 K. A vacuum casing of non-metallic construction such as G-lO fiberglass surrounding the SQUID(s) and wafer provides thermal insulation to reduce the heat load and facilitate keeping the refrigerator cold stage at temperatures below the super-conducting transition temperature of the SQUID(s) material.
A SQUID detector and a laser-pumped helium magnetometer respond to all frequencies down to DC, while a resonant coil detector as used in the current state-of-the-art NML responds only to a narrow frequency range and has ~ero response at DC. Thus, a SQUID detector and a laser-pumped helium magnetometer may essentially simultaneously and directly measure both a T2* and T
decay in a single NML tool-with a single measurement.

~3~

In addlkion ~o the S~UID element~, the SQUID electronics may also be thermally anchored to the refrigerator cold ætage.
This provldes an additional pexformance improvement by reduced noise.
One or more SQUID detectors may be flux coupled to a superconducting axial gradiometer detection coil. This coil may also be constructed of a hlgh-temperature superconductor material and may be thermally anchored to the refrigerator cold stage. The function of the detection loop is to incraase the flux sen~itlvity of the detector. The axial gradiometer configuration is preferred I to reduce flux noise from any motion of the logging tool in the earth's magnetic field while in the borehole. For this configuration, ona loop of the gradiometer is cen~ered in~ide the polarizing coil while the second counterwound loop extends outside the polarizlng coll. Thus, the inside loop detects the full signal from any polarized protons in the formation, whereas the outside loop detects sub~tantially no proton signal. A uniform field such as the Earth's magnetic field createæ an equal and opposite flux through each loop so that the Ear~h's magnetic field i8 effectively cancelled. Similarly, the laser-pumped helium magnetometers may be used in dif$erential pairs, with one inside the polarizing coil and the other fixed outside the polarizing coil.
Thus, the presently preferred apparatus of the present invention provides a SQUID detection module for use in a downhole logging tool that obviates the need for liquid cryogens, provides improved signal-to-noise for nuclear magnetism logging, and makes possible the direct measurement of the T1 decay curve of the earth formation.
According to another a~pect, the prevent invention provides a method for nuclear magnetism logging, comprising~
- generating a magnetic fleld in a borehole, - removing said magnetic field, and then - essentially simultaneously detecting the sinusoidal and slowly varying magnetic fields resulting from precession of mobile nuclei abouk the earth's magnetic field.

,, ~

It is therefore an object of this lnventlon to provide a nuclear magnetism logging device for obtainin~ signal-to-noiæe ratios several orders of mayni~ude above ~hat o~ the presen~ N~L
sta~e of the art.
It is another object of thls inven~lon to provide an NML
device khat may measure the entire Tl decay curve of an earth formation directly, wi~hout requiriny successive polariza~ion cycles.

- ' a Sa ,~

~3~

These and other advantages and objects of the present invention will become apparent from the following detailed description, wherein reference is made to the figures in the accompanying drawings, in which:
- Figure 1 shows a simplified block diagram of a nuclear magnetism logging tool, the magnetic field vectors associated with this nuclear magnetism logging tool and their relationship to the magnetic field of the Earth.
- Figure 2 shows a typical NML signal detected using a resonant coil tuned to the Larmor frequency.
- Figure 3 shows a typical NML signal detected using a SQUID as a detector instead of the resonant coil.
- Figures 4A and B show a SQUID detector and micro-miniature Joule-Thomson refrigerator integrally constructed on the same wafer substrate, and the details of the SQUID
detector, respectively.
- Figure 5 shows the axial gradiometer configùration of super-conducting coils flux-coupled to a SQUID that may be employed in the present invention.
- Figure 6 shows typical measuring electronics for a SQUID
detector.
Referring now to Figure 1, there may be seen a borehole 10 through an earth formation 20 and a nuclear magnetism logging tool 11 disposed Ln said borehole by cable 9. The logging tool 11 contains solenoidal coil 12 which generates a polarizing magnetic field vector 13 in the formation ad~acent the tool. The earth's magnetic field, He, 14 is shown at an angle ~ to the magnetization vector, Mo~ 15 representing the polarized proton spins. After the polarizing field is turned off, the proton spins 15 precess around the Earth's field 14 as shown. If the detector is a resonant coil 17 that is tuned to the Larmor frequency, it detects t'he precessing proton spins M . An instrumentation package 19 provides about 1 Kw of power to polarizing coil 12 and detects the voltage induced in resonant coil 17 by the precessing spins 15.

~3~

Figure 2 shows a typical signal 16 expected to be detected in resonant coil 17 by the precessing proton spins 15. The signal decays sinusoidally at the l.armor frequency with a time constant T2* which is much shorter than Tl. T2* represents the time constant for the decay of the envelope of the sinusoidal signal.
The Larmor frequency of precession is ~ x He, where ~ is 4.26 kHz/g for the proton. Note the Tl decay is not observed because the Tl decay is of the order of one second, which is far from the 2000 Hz Larmor frequency at which the resonant coil is tuned. Figure 2 also shows the measurement delay discussed later herein and indicates (by dotted line) the extension of the signal envelope back to time t indicative of ~F~ the porosity occupied by free fluids in the formation adjacent the tool.
SQUID detectors have been found to have advantages over conventional resonant coils for detecting NMR signals in the laboratory. SQUIDs detect the magnetic field (more precisely, total magnetic flux) linking the detection loop, whereas conven-tional NMR coils detect a voltage (rate of change of magnetic flux) induced in the loop. Thus, SQUIDs respond at all frequencies down to DC (i.e., have a very wide frequency response), whereas resonant coils respond only to a narrow frequency range (determined by the Q
of the coil) and have zero response at DC. The superiority of a SQUID as a detector over resonant coils is greatest for low ~R
frequencies, long spin lattice relaxation times Tl, and short transverse relaxation times T2*. Further, other non-SQUID type detectors capable of measuring DC or near DC and having a wide frequency response, such as laser-pumped helium magnetometers may also be employed as detectors to detect both the T2* sinusoidal decay and the slowly varying Tl decay.
Long Tl times and short T2* times are substantially the conditions that exist for downhole nuclear magnetism logging of earth formations. The Larmor frequency for protons in the Earth's magnetic field is about 2000 Hz, a very low frequency compared to the 10-500 M~lz used in laboratory NMR spectrometers. Secondly, the Tl relaxation time of earth formations is fairly long, typically ~3~

about 1.0 second, while the T2* rel~ation time is short, typically about 20-50 milliseconds. This high Tl/T2* ratio found in downhole nuclear magnetism logging is the situation where SQUIDs or similarly sensitive, wide frequency response detectors sre most advantageous over conventional resonant coil detectors.
A measure of the improvement in signal-to-noise (S/N) ratio of SQUID vs. resonant coil detection is given in the paper "New Technique For Improved Low-Temperature SQUID NMR Measurements" by R. A. Webb (Rev. Sci. Instum., Vol. 48, No. 12, pp. 1585-1594, December, 1977). For typical laboratory values of SQUID noise figure, coil Q, etc.
SQUID S/N 2.2 x 10 (Tl/T2)1/2 COIL S/N (~ )1/2 Using typical earth formation values of Tl = 1000 ms, T2* - 20 ms, and ~ D 2*~*(2 x 10 ) = 1.26 x 10 , the SQUID S/N is a factor of 139 greater than the coil S/N.
Moreover, for NML logging, the improvement in S/N for a SQUID
detector compared to resonant coil detector should be even greater because the conventional NML has a deadtime of about 25 milli-seconds after the polarizing field is turned off. This measurement delay or deadtime is due to coupling and ringing between the polarizing coils and its circuitry, and the detection coils and its circuitry; this delay is illustrated in Figure 2. Because the delay or deadtime is an appreciable part of the T2* decay, A
majority of the signal has decayed before the NML begins recording.
For example, if T2* ~ 25 milliseconds and the NML deadtime is 25 milliseconds, then l/e of the signal has decayed before detection even begins.
A SQUID or similar detector does not have this problem.
Unlike the resonant coil, which is sensitive only to the rapid transverse dephasing of the nuclear spins (T2*) at the Larmor frequency and has no sensitivity at DC, the SQUID or similar detector measures the z-component of the magnetization at fre-quencies down to DC. Thus, even instrument deadtimes of the order of 25 milliseconds will not significantly reduce the signal level 3~

with SQUID or similar detectors. Such a similar detector that is a non-SQUID type detector capable of measuring DC or near DC, is a laser-pumped helium magnetometer; such magnetometers may also be employed as detectors to essentially simultaneously detect both the T2* sinusoidal decay and the slowly varying Tl decay.
In generall an optically-pumped magnetometer uses a lamp or laser containing an alkali metal or helium whose light is passed through a cell containing the vapor of the same element and impinging on a photodetector on the other side of the vapor cell.
An optically-pumped magnetometer utilizes a population of electrons in this alkali vapor gas or in metastable helium to obtain a continuous measurement of the magnetic field intensity. The output of the photodetector is amplified and fed to a coil surrounding the vapor cell. This electro-optic system is an oscillator whose frequency is directly proportional to the magnetic field intensity.
The magnetometer takes advantage of optical pumping to cause atomic or electron spin precession similar that employed by an NML to cause proton precession. Resonance absorption and reradiation of energy of various or specific resonant lines are a function of magnetic field intensity.
For downhole nuclear magnetism logging, another very important advantage of SQUIDs or similar detectors is that they may directly measure the Tl decay curve. It is not necessary to apply repeated polarizing cycles to obtain the Tl decay curve needed for pore size and permeability measurements. In order to measure the Tl decay curve directly, a detector must respond at very low frequencies below l Hz, because Tl is typically of the order of l second.
Thus, SQUID or similar detectors are sensitive not only to the transverse dephasing of the spins but also to the thermal relaxation Tl. Since Tl is typically 50 times greater than T2 in earth formations, SQUID or similar detectors also may detect a signal for the much longer spin lattice relaxation time Tl.
Figure 3 shows a typical signal expected to be detected if resonant coil 17 of Figure t is replaced by a SQUID detector 18 and ; 35 consists of a sinusoidally-decaying signal at the Larmor frequency : L2~3~

with time constant T2* which is proportional to MOSin 0, as wel]. as a non-sinusoidal component decaying wi~h time constant Tl which i~
proportional to MoCos 9. The sinusoidal T2* decay is due to the component of polari~ed nuclear spins perpendicular to the earth's field He, while the Tl component is due to the component of nuclear spin aligned along the earth's Eield He. The Larmor sinusoidal T2*
decay can be separated from the near-DC Tl decay using either an electronic low-pass filter, a copper eddy-current container as a shield around the detection coils, or a digital filter applied to the recorded data by realtime or postprocessing software. The longer Tl decay will not be measured by coil 17 while SQUID 18 measures essentially simultaneously both the sinusoidal T2* decay as well as the Tl decay. However, other non-SQUID type detectors capable of measuring DC or near DC, such as laser-pumped helium magnetometers may also be employed as detectors to essentially simultaneously detect both the T2* sinusoidal decay and the slowly varying Tl decay.
Figure 4A shows a SQUID detector 20 and a cooling module 21, which is preEerably a micro-miniature Joule-Thomson refrigerator, on a substrate 30. Although only one SQUID detector 20 is shown in Figure 4A, more than one such SQUID detector may be located on the substrate 30. The substrate 30 is itself used as the cold stage of the Joule-Thomson refrigerator. This provides excellent thermal contact between the SQUID and cold stage. Such thermal contact is required to provide the lowest possible temperature for the SQUIDs, so that their noise figures will be optimal. The SQUIDs are fabricated onto the substrate as close to the Joule-Thomson reservoir as possible to ensure maximum cooling. The SQUID may be of either the RF-biased or DC-biased type, although only the DC-biased type is discussed herein. The SQUID is made from high temperature superconducting material, such as, for example, but not limited to rare earth-barium-copper oxide materials. The term "high temperature superconducting material" is used herein to mean a material whose superconducting transition temperature is above the boiling point of liquid nitrogen (i.e., - 77 K).

~35~
- 11. -As shown in Figure 4B, the DC ~QUID consists of two granular weak links formed by epitaxial growth oE a rare earth, barium, copper oxide material, such as YBa2Cu307 on an appropriate substrate. The granular weak links are about 15 microns long and 1 micron thick and are defined by insulating gaps 47, 48; the weak links are said to be granular in that the weak link is actually between grains of the superconductor that are weakly Josephson-coupled. If the weak links are too narrow ( 15 microns) the current around central insulating portion 46 may become too high and quench the superconductor material causing the SQUID not to operate Although, a square 46 is shown in Figure 4B, other geometries may be employed. Similarly, if the weak links are too wide, too much current fl.ows so that the "staircase" operating characteristics of a SQUID are not achieved and the SQUID does not operate properly.
Although YBa2Cu307 is discussed herein as the preferred material for making the SQUID, any of the high-temperature super-conductors, such as those with oxygen-deEicient Perovskite structure may be used. For YBa2Cu3O7, the substrate should have a molecular structure similar to YBa2Cu3O7 (for any other material employed for the SQUID, the substrate should have a molecular structure similar to that material) to achieve epitaxial growth, and should also be a good thermal conductor near liquid nitrogen temperatures to achieve good thermal connection to cooling module 21. Examples of such substrates may be, but are not limited to SrTiO3 or MgO. For example SrTiO3 has an "a" cell dimension of 3.9 Angstroms while YBa2Cu3O7 has a corresponding cell dimension of 3.83 Angstroms. Epitaxial growth o:E oriented crystals is important in YBa2Cu307 superconductors for achieving the highest possible critical fields and currents. Since it is preEerable that the SQUIDs remain supe.rconducting during the polarizing cycle (when they will be subjected to magnetic fields in excess oE 1 kilogauss), the SQUIDs should preferably have a critical field in excess of the strength of any anticipated polarizing magnetic Eield.

3~g~

Epitaxial growth of the SQUID material on the SrTiO3 substrate may be achieved by several techniques well known in the art, such as molecular beam epitaxy, electron beam evaporation, sputterine from single and multiple targets, pulsed excimer laser ablation of single targets, sol gel, and plasma oxidation. For example, in the molecular beam epitaxy method, a thin layer (- 20 microns) of Y-Ba-Cu is first vapor deposited onto the SrTi03 substrate at the correct stoichiometry and then oxidized to YBa2Cu3O7. Next a thin layer of gold is deposited on top of the YBa2Cu307 layer. Finally, a photoresist layer is placed on the gold layer for photolithography patterning. The photoresist and gold are then ion milled in the SQUID pattern, i.e. removing the resist and gold covering central portion 46 and gaps 47, 48 shown in Figure 4B;
that is, the gold and resist are left above the portion of the layer designed to be superconducting. An oxygen ion beam of 0.3-3 MeV is then used to implant oxygen into the exposed YBa2Cu307 of central portion 46 and gaps 47, 4S, causing these areas to become an insulator, while the gold protects the portion of the pattern designed to be superconducting. Finally, the remaining gold is removed by ion milling, leaving a perfectly planar structure.
A plurality of SQUID sensors may be so constructed on the substrate. The advantage of having a plurality oE sensors is that the SQUID noise of each sensor will be random and uncorrelated, and therefore, summing the SQUID outputs will result in further improvement in signal-to-noise ratio.
The micro-miniature refrigerator operates by Joule-Thomson expansion of a working gas, such as nitrogen. There are no cryogenic fluids or moving parts, which results in very low vibration and correspondingly low noise. M:Lcro-miniature Joule-Thomson refrigerators are commercially available, such as for example, the System I refrigerator manufactured by MMR Technolo-gies, Inc. of Mountain View, California. However, these MMR
Technologies refrigerators are manufactured from glass or silicon and can not be used as an epitaxial substrate ~ YBa2Cu3O7 SQUIDs.
As shown in Figure 4A, a microminiature Joule-Thomson refrigerator 9~

40 has three main parts. Thcse parts are a he.at exchanger 41, an expansion capillary 42, and a liquid reservoir 43. The refrige-rator may be constructed from a SrTiO3 wafer 30 using the photo-fabrication technique. For example, in the photofabrication technique, a mask for the refrigerator portion such as that shown in Figure 4A is first prepared. The wafer 30 may then b~ coated with a thick, water solution of gelatine activated with ammonium bichromate. The solution on the wafer 30 is then dried, covered by a mask, exposed to ultraviolet light through openings in the mask and then developed in hot water. This resist, when dried, forms a tough resilient coating which can withstand an abrasive etch with Al2O3 powder. The unexposed gelatine washes off in hot water leaving a pattern of unprotected SrTiO3.
An abrasive etching is performed by exposing the wafer partially covered with resist to a blast of A1203 or other powder of sufficient hardness, in sizes ranging from 10-30 microns; this blast is scanned across the surface of the waEer. Channels of precisely controlled depth in the range of 2-100 microns and with nearly vertical side walls can be etched in the wafer 30 in this manner. The remaining resist is removed by chemical solution of the gelatine.
To complete the refrigerator portion, a cover plate is bonded to the etched wafer, using a variety of possible adhesives, such as low temperature epoxy. Some care must be taken to assure that the adhesive does not flow into the micron size channels.
The refrigerant working gas is contained in a pressurized tank, typically at pressures of 140 bar, and allowed to expand through the Joule-Thomson refrigerator into a larger low pressure collection tank (not shown) within the logging tool. The collection tank is many times (typically at least 10 times) larger than the pressurized tank so that the expanded gas does not reach a high pressure and degrade refrigerator operation. Typical refrigerator operating times are about 12 hours with 1 liter of nitrogen gas at 140 bar pressure. Alternatively, a closed cycle compressor could be used for the gas, although such moving parts ~3~

will cause vibrations that may degrade signal-to-noise (S/N~.
Alternatively, a Peltier module may be used as a cooling module 21 when superconducting materials whose transition temperature is above about 180 K are used Eor SQUIDs, since the temperature down to which current Peltier modules may cool is about 180 K.
The refrigerator cold stage and SQUID may be surrounded by a vacuum can (not shown). The vacuum can provides a vacuum space necessary for reducing the heat leakage into the cold stage and the SQUID sensor. The vacuum can may be constructed of any non-metallic and non-magnetic material with low thermal conductivity at liquid nitrogen temperatures. Thin strips of superinsulation (i.e., aluminized mylar) can be added within the vacuum space to improve ther~al isolation of the SQUID. However, the strips must be cut vertically to reduce eddy currents in the superinsulation.
Figure 5 shows the axial gradiometer coils that may be used in the practice of this invention. The function of these coils is to improve flux sensitivity and reduce magnetic noise from tool motion in the Earth's magnetic field. The pickup coils 51 and 52 are coaxial and counterwound so that a uniform magnetic field through the coils is exactly cancelled. One of the picXup coils 51 is centered inside the polarizing coil 12, while the other coil 52 is substantially outside the polarizing coil. The pickup coil 51 centered in the polarizing coil 12 will detect the maximum signal from the earth formation 20, while the pickup coil 52 outside the polarizing coil will detect substantially no signal from the formation. These coils may be flux coupled to one or more SQUIDs, or each SQUID may have its own set of coils.
However, the logging tool may also be operated with the SQUID
itself as the flux sensor and with the tool clamped against the borehole wall to reduce vibration. If flux-coupled coils are used, the SQUID sensor is shielded from magnetic fields by a super-conducting shield of YBa2Cu3O7 surrounding the SQUID. The super-conducting leads to the coil enter through small holes in the superconducting shield.

The gradiometer coils may be thin epitaxial Eilms deposited on a SrTiO3 substrate, similar to the SQUID fabrication. Alter-natively, the coils may be made from short lengths of super-conducting YBa2Cu307 wire. This wire may be made from a gold or silver tube packed with a compressed powder of YBa2Cu307. The tube i9 sealed, swaged to smaller diameter, then refired above 900C and allowed to slowly cool (~ 20 hours). Although wire made in this manner may not have the highest obtainable critical currents, values in e~cess of 1000 A/cm may be obtained at zero magnetic field, and 200 A/cm above 1000 gauss.
Similarly, laser-pumped helium magnetometer detectors may be employed in pairs, with one magnetometer inside the polarizing coil and one outside the polarizing coil. These two magnetometers are again differentially coupled to eliminate signals from motion in the earth's field.
Although a DC granular weak link thin-film type of SQUID has been described hereinabove, other types of SQUIDS may also be fabricated according to the teachings of this invention. Among these are tunnel junction SQUIDS (superconductor-insulator-super-conductor, superconductor-normal metal-superconductor, and super-conductor-semiconductor-superconductor), thin film weak links such as those of Dayem and Grimes, point-contact SQUIDs, and various hybrid combinations of thin Eilm and point-contact SQUIDs.
Similarly, the SQUIDs may be either single junctlon RF-biased type or double junction DC SQUIDs.
Figure 6 shows a block diagram of the electronics required for operating a DC SQUID 620, as is well known in the art. The electronics must consist of a blas current source 610, an amplification means 600, 601, 602, a modulation means 603, 604, 605, a feedbflck means 607, 608, a filter means 606, a monitor means 609, and a zero offset 612. At the present time, it is preferred that a DC SQUID be operated in a feedback mode so that the SQUID is maintained at a particularly sharp point on the SQUID's staircase current versus voltage operating characteristics, as selected by the bias current. The modulation and feedback may be at any ~z53~3~r~L

frequency whi.ch allows the electronics to follow any changes in magnetic field in the logging tool. Figure 6 shows a 100 kHz modulation oscillator 604.
More particularly, the detected signal is flux coupled into the SQUID 620 via the input coil 621. If the SQUID 620 is the detector itself, then input coil 621 is not present and the detected flux is the flux passing through the insulating, central portion of the SQUID (i.e., item 46 of Figure 4B). The SQUID is modulated by modulation oscillator 604 at a frequency sufficient to allow the electronics to follow changes in the magnetic field by modulation and feedback coil 622. Advantageously said frequency is 100 kHz. A phaseshifter 603 is connected with said oscillator. As the SQUID senses a change in magnetic field, the SQUID trys to move up or down the vertical portion of the selected operating point (selected by its bias current) on its staircase operating characteristic. The amplifier means 600, 601, 602 amplify this change which is then filtered by filter means 606 to provide unmodulated feedback, via feedback means 607, 608 and modulation and feedback coil 622, to the SQUID 620. In particular the amplifier means comprise an amplifier 600 with amplification xlO0, a tuned amplifier 601 with amplification xlO0 and quality Q ~ 3, and a broadband amplifier 602 with amplification x0-300. The broadband amplifier 602 is also connected with a multiplier 611. As a further detail it is noted that the filter means 606 contains lO0 and 200 kHz traps. The feedback keeps the SQUID at its selected operating point. The amount and type of feedback is detected by monitor means buffer amplifier 609 and provided as an output to indicate changes in the magnetic field sensed by the SQUID.
Many other variations and modifications may be made in the apparatus and techniques hereinbefore described, by those having experience in this technology, without departing from the concepts of the present invention. Accordingly, it should be clearly understood that the apparatus and methods depicted in the accompanying drawings and referred to in the foregoing description ~2~3~g3l are illustrative only and are not intended as limitations on the scope of the invention.

.

,, ' , '~ :
. , ' "'~ ~ ' , ;
. ' .
~; , ' ' .

Claims (15)

1. A nuclear magnetism logging tool, comprising:
- means for providing a magnetic field that is capable of being energized or de-energized in response to preselected control signals, - detector means for detecting sinusoidal and slowly varying changes in magnetic field when said means for providing a magnetic field is de-energized and outputting signals indicative thereof, and - an electronics package for generating said preselected control signals and for receiving signals from said detector means.

2. A nuclear magnetism logging tool as described in claim 1, wherein said detector means is a laser pumped helium magnetometer.

3. A nuclear magnetism logging tool as described in claim 1, wherein said detector means is a superconducting quantum interference device (SQUID).

4. A nuclear magnetism logging tool as described in claim 3, further comprising means for cooling said SQUID detector means for maintaining it at a temperature below the superconducting transition temperature of its superconducting material.

5. A nuclear magnetism logging tool as described in claim 1, wherein said means for providing a magnetic field is a solenoidal coil.

6. A nuclear magnetism logging tool as described in claim 3, wherein said SQUID employs a high temperature superconducting material.

7. A nuclear magnetism logging tool as described in claim 6, wherein said means for cooling employs a Joule-Thomson refrigeration cycle.

8. A nuclear magnetism logging tool as described in claim 6, wherein said high-temperature superconducting material is substantially an oxygen-deficient perovskite.

9. A nuclear magnetism logging tool as described in claim 8, wherein said high temperature superconducting material is substantially YBa2Cu3O7.

10. A nuclear magnetism logging tool as described in claim 3 further comprising a pair of superconducting axial gradiometer detector coils flux coupled to said SQUID, and wherein one of said coils is disposed within the means for providing a magnetic field and the other coil is disposed outside the means for providing a magnetic field.

11. A nuclear magnetism logging tool as described in claim 10, further comprising a superconducting shield disposed around said SQUID.

12. A nuclear magnetism logging tool as described in claim 4, wherein said SQUID and means for cooling are mounted adjacent each other on a thermally conductive substrate.

13., A nuclear magnetism logging tool as described in claim 12, wherein said substrate is SrTiO3.

14. A nuclear magnetism logging tool as described in claim 12, wherein said substrate is MgO.

15. A method for nuclear magnetism logging, comprising:
- generating a magnetic field in a borehole, - removing said magnetic field, and then - essentially simultaneously detecting the sinusoidal and slowly varying magnetic fields resulting from precession of mobile nuclei about the earth's magnetic field.