Michael J. Pinto* and Michael McWilliams* ... -I .o. 0.0. Distanceji-om center of core (inches). FIG. 1. Variation in NRM intensity .... waves: W. H. Freeman & Co.
GEOPHYSICS,
VOL. 55, NO. 1 (JANUARY
1990);
P. 11l-l 15, 4 FIGS.
Short Note Drilling-induced
isothermal remanent magnetization
Michael J. Pinto* and Michael McWilliams* of this overprinting phenomenon for paleomagnetic core orientation.
INTRODUCTION
The recovery of core samples is important in petroleum exploration, mineral exploration, and scientific drilling projects; and often complete orientation of the samples (azimuth and plunge) is desirable. Recovered cores are usually not azimuthally oriented because of the costs associated with deployment and operation of downhole orientation tools. Inexpensive paleomagnetic orientation methods have been used with considerable success in the borehole environment (Van der Voo and Watts, 1978; Kodama, 1984; Bleakly et al., 1985a,b; Evans and Mailol, 1986;Layer et al., 1988; McWilliams and Pinto, 1988). In some cases, the technique has been hampered by secondary magnetizations associatedwith the drillstring and/or coring tool, magnetizations which have partially or completely overprinted the primary and secondary magnetizations used for orientation. We recently used paleomagnetic methods to orient granodioritic coresfrom the Cajon Pass Scientific Drilling Project (McWilliams and Pinto, 1988). In this study, we suspected that the samples had acquired an isothermal remanent magnetization (IRM) during coring. The remanent magnetization was strongest in specimens closest to the core-barrel wall and was approximately one order of magnitude weaker in those specimens near the center of the core (Figure 1). We suggest that this intensity variation is the result of the superposition of two magnetizations in the Cajon Pass samples. One is a Cretaceous primary thermoremanent magnetization (TRM) acquired by the sample after initial cooling and is constant throughout the core. The other is a secondary IRM, an overprint produced by the inner core barrel which varies with distance from the center of the core. To explain the overprinting phenomenon, we have developed a model for the magnetic field produced by a uniformly and axially magnetized core barrel. In the following sections we discussthe derivation of this model and the implications
THEORETICAL
BASIS
We begin by assuming that the hollow, cylindrical core barrel has a uniform and axial remanent magnetization M parallel to the axis of the drillstring. Since this is also parallel to the easy direction of magnetization for a long, thin rod or tube, the assumption is probably not far from correct. The magnetization is modeled by superposing two solid cylinders, each of length L but with opposing magnetizations and differing radii (Brown and Flax, 1964). The problem requires a complete solution for the field produced by a solid cylinder with a uniform axial magnetization. The magnetic field B created by a magnetized object can be generated by equivalent (Amperian) volume- and surfacecurrent distributions (Lorrain and Corson, 1970). The relationships among M, volume current density J, and surface current density K are J=VxM K=Mx&
where ii is the outward unit vector normal to the surface of the object. There is no contribution to the magnetic field from J, since V x M = 0 and, therefore, B depends only on K.
The Biot-Savart law is convenient for determining B because a high degree of symmetry is present. For surface currents, (r - r’)
(M
x f@x or
d.4’3
where dA’ is the element of surface area at the source, r is the vector from the origin to the point of measurement
Manuscriptreceived by the Editor April 3, 1989;revised manuscriptreceived July 18, 1989. *Department of Geophysics, School of Earth Sciences, Stanford University, Stanford, CA 943052215. 0 1990Society of Exploration Geophysicists. All rights reserved.
111
112
Pinto and McWilliams
(Figure 2), r’ is the vector from the origin to the sourcepoint, and k0 is the permeability of free space (coordinates of the source are primed, and coordinates of the point at which the field is measured are unprimed.) In cylindrical coordinates, VJ-Jc c B=4.,, JJ
vectors (I, 3, k) instead of cylindrical unit vectors (6, 4, 2) and (fi’, 4’. 2’): fi = cos +P + sin d$ 0 = cos $‘I + sin hf3 ^ +=-sin@+cos$j
(r - r’) (MS x 6’) x /1. a dz’ d+‘,
where a is the radius of the solid solenoid and M is the magnitude of M. For integration, we use Cartesian unit
6’ = -sin +‘P + cos $3 2 = i’ = 1;. The vector equations may now be expressed in terms of Cartesian unit vectors: MS X fi’ = Mb’ = -M
sin $‘i + M cos +‘j^
r = r cos +i + r sin & + zl;
f’ = a cos +‘i + a sin +‘j + z’k. .Ol
I
B does not depend on $, which is apparent from azimuthal symmetry. We may therefore fix r at C$= 0 without affecting the solution. It follows that r = rP + z,k and lr - r’l = [r2 -
2ur cos 4 + a’ + (z - z’)]‘~‘~.
Also fixing r at 4 = 0 leads to a correspondencebetween the unprimed cylindrical unit vectors and the Cartesian unit vectors:
I
I
.Ol :
Expanding (MP’
Field
meaSUrCme”t
x
6’)
x
(r - r’), we get
_
point
-2.0 Distanceji-om
-I .o
0.0
center of core (inches)
FIG. 1. Variation in NRM intensity with distance from the center of the core. (a) NRM intensity variations from a suite of samples from core 17 (4432 ft-4451 ft) in the Cajon Pass borehole. Points are plotted at the center of each 0.86 in specimen. (b) NRM intensity variations from a single plug from a depth of 4435.4 ft. The outermost specimen has been cut into three 0.25 in subspecimens. Note the progressive decreasein NRM intensity with distance from the sidewall in the subspecimens; the trend correlates well with the theoretical predictions (solid line; scaled to minimum value of NRM intensity).
I I
---r--L FIG. 2. Cartesian and cylindrical coordinate systems with a model core barrel.
114
Pinto and McWilliams
whether the observed NRM intensities could be created by the field present at the end of the core barrel. Each specimen was first demagnetized in peak alternating fields of 70 mT. A dc magnetic field was then applied to the specimen, and the IRM produced was measured. IRM was imparted using higher dc fields until the specimens had reached saturation; the laboratory-induced IRM intensity increases approximately linearly over the range of field strengths observed inside the core barrel. Our IRM studies indicate that an IRM equal in magnitude to the observed NRM values could be produced in most specimens in fields less than 5 Gauss (S mT) and that the NRM intensity of specimens that were in close contact with the core-barrel wall could be produced isothermally in fields of less than 20 Gauss (2.0 mT). An effect which warrants discussion is drilling-induced remanence (DIR), described by Burmester (1977). During drilling and sawing of samples of granite, Burmester found that a large DIR was imparted to the samples, a DIR which had a coercive force in excess of 92.5 mT, even though it was acquired in a field of less than 0.2 mT (the ambient field near the rock cutting tools). The outer 0.4 mm of a sample was
then removed by etching in hydrochloric acid, removing the DIR. Burmester concluded that the DIR resided in the coarse magnetite grains at the surface of the specimen that had been cut during drilling, suggesting that the DIR was a stress-aided viscous remanent magnetization. Although this effect has been observed in other studies (Jackson and Van der Voo. 1985). it does not appear to be responsible for the overprint seen in our study or in other recent studies. Gzdemir et al. (1988) have documented a DIR in granodioritic cores which they ascribe not to a stress-aided, or piezoremanent. viscous remanent magnetization but to a high-temperature IRM/VRM; and Audunsson and Levi (1989) have documented a DIR in basalt cores which they ascribe primarily to an IRM. In samples from Cajon Pass, the overprint magnetization resides not only in the samples in contact with the coring bit, but also in samples from the interior of the core which were not in contact with the bit. If the overprint magnetization were a stress-aided viscous remanent magnetization. then either (I) only the outermost specimens should be affected (if the DIR were from initial rotary coring) or (2) all the specimens should be overprinted
FIG. 3. Theoretical magnetic fields at the end of a long, uniformly magnetized cylinder (shaded region depicts the cylinder wall). The cylinder terminates at Z = 0.0, is magnetized in the -Z direction, and extends in the -Z direction. The cylinder axis lies at R = 0.0. (a) Field trajectories at the end of the cylinder (arrow length does not correspond with field intensity). (b) Axial component (arbitrary units). The axial component is discontinuous at the boundary of the cylinder. Contour interval 0.1 (heavy lines) and 0.05 (lighter lines). (c) Radial component (arbitrary units). The radial component is continuous at the boundary of the cylinder. Contour interval as in Figure 2b. (d) Total intensity (arbitrary units). Contour interval as in Figure 2b.
Drilling-induced
equally (if the DIR were from plugging and sawing the samples in the laboratory). In addition, the overprint magnetization observed in the Cajon Pass samples has a much lower coercive force; the force is frequently removed in peak alternating fields of 10 mT and is always eliminated after treatment in 25 mT. To further confirm that DIR was not present, we cut some of the outermost specimensinto three, equally sized subspecimens. As expected, subspecimens closestto the core barrel had substantially larger NRM than subspecimensfarther away, and Figure lb clearly showsthat the intensity changes cannot be a surface effect. IMPLICATIONS
FOR PALEOMAGNETIC ORIENTATION
CORE
Experimental results have shown that drill-string overprints can partially or completely replace the remanent magnetization needed for paleomagnetic core orientation. The requisite magnetization may be a primary magnetization acquired at approximately the time the rock formed or a secondarymagnetization acquired later. Of particular importance is the effect of IRM overprinting on secondary viscous remanent magnetization (VRM). VRM is acquired in natural samplesover periods up to IO’ s in the presenceof the recent geomagnetic field; it is therefore parallel to the present-day geomagneticfield at the sampling site. Unoriented cores can easily be oriented if the VRM of a sample can be isolated and aligned with the present magnetic north direction by rotation
-
TheoreticalField Measured Field NRM intensity
n
.
-3
-2
-1
0
1
2
3
Distancefrom centerof core (inches)
FIG. 4. Comparison of measured NRM intensities, field values predicted by theory, and actual field values measured at the tip of the core barrel. Fields and magnetization intensities are each scaled to the minimum value,since each has arbitrary units. The size of each measured field svmbol representsuncertainty in fluxgate sensor position relaEiveto the core barrel.
115
IRM
about the borehole axis. Using VRM for orientation is potentially more accurate than using the primary magnetization, because the VRM is not affected by structural complications (use of the primary magnetization assumesthere has been no relative rotation between the core sampling location at depth and the surface exposures). However, VRM usually has a lower coercive force than primary TRM or CRM (chemical remanent magnetization), and thus it is more susceptible to overprinting by drilling-induced IRM. Our experimental results and theoretical calculations suggest three practical guidelines: (1) nonmagnetic core barrels should be used whenever possible (stainless steel is sufficiently nonmagnetic); (2) ferromagnetic materials near the end of the drill string should be demagnetized before coring begins (a degaussingcoil should be sufficient); and (3) it is best to avoid specimens from the outside of the core for paleomagnetic orientation. ACKNOWLEDGMENTS
We thank R. Blakely, D. Dunlop, and J. Rosenbaum for critical reviews. J. Kirschvink loaned us a fluxgate magnetometer at the Cajon Passwell site. Supported in part by the National Science Foundation through DOSECC, Inc. (McWilliams) and by Stanford University through the Undergraduate Research Opportunities program (Pinto). REFERENCES \udunsson, H., and Levi. S., 1989, Drilling induced remanent magnetization in basalt drillcores: Geophysical J., in press. Bleakly, D. C.. Van Alstine. D. R., and Packer, D. R., 1985a, Core orientation: Controlling errors minimizes risk and cost in core orientation: Oil and Gas J., Dec. 2, 103-109. 1985b, Core orientation: How to evaluate orientation data, quality control: Oil and Gas J., Dec. 9, 46-52. Brown, G. V., and Flax. L., 1964, Superposition of semi-infinite solenoids for calculating magnetic fields of thick solenoids: J. Appl. Phys., 35, 1764-1767. Burmester, R. F., 1977, Origin and stability of drilling induced remanence: Geophys. J. Roy. Astr. Sot., 48, I-14. Evans. M. E.. and Mailol, J. M., 1986, A paleomagnetic investigation of a Permian redbed sequence from a mining drill core: Geophys. J. Roy. Astr. Sot., 14, 667687. Jackson,M., and Van der Voo, R., 1985, Drilling-induced remanence in carbonate rocks: occurrence. stability, and grain-size dependence: Geophys. J. Roy. Astr. Sot., 81, 75-87. Kodama. K. P., 1984, Paleomagnetism of granitic intrusives from the Precambrian basement under eastern Kansas: orienting drillcores using secondary magnetization components: Geophys. J. Roy. AsIr. Sue., 76, 273-287. Layer, P. W., KrBner, A., McWilliams, M., and Clauer, N., 1988, Regional magnetic overprinting of Witwatersrand Supergroup sediments, South Africa: J. Geophys. Res., 93, 2191-2200. Lorrain, P.. and Corson, D. R., 1970, Electromagnetic fields and waves: W. H. Freeman & Co. McWilliams. M.. and Pinto. M. J.. 1988. Paleomagnetic results from granitic basement rocks in the Cajon Pass scientific drillhole: Geophys. Res. Lett.. 15, 1069-1072. 8zdemir. 0.. Dunlop, D. J., Reid, B., and Hyodo, H., 1988, An early Proterozoic VGP from an oriented drillcore into the Precambrian basement of southern Alberta: Geophys. J. Roy. Astr. Sot., 95, 69-78. Van der Voo, R., and Watts, D. R., 1978, Paleomagnetic results from igneous and sedimentary rocks from the Michigan Basin borehole: J. Geophys. Res., 83, 5844-5848.
