Appendix 5. Review of Wireline-log Responses in Igneous and Igneous-derived Sedimentary Rocks
The selection of wireline-logging tools in Texaco Poersch #1 is essentially the same set as would be run in any recent wildcat drilled by a major oil company. Although the logging program was typical, there are special problems associated with the interpretation of the log responses of units penetrated by this well. Conventional logging tools are designed to measure properties sensitive to petroleum-reservoir parameters of porosity and fluid saturations. The most recent generation of tools is also useful in the discrimination of the common sedimentary lithologies of sandstone, limestone, and dolomite, and it is these rock types which are used in tool calibration of measurement scales. Igneous rocks are therefore "unusual," both with respect to the true meaning of responses normally keyed to sedimentary-rock properties and the necessity for novel interpretation techniques that can be linked with orthodox igneous petrography and petrogenesis. This information is also highly pertinent in the analysis of relatively immature clastic rocks which still bear the imprint of their igneous sources.
Information concerning the logging properties of igneous rocks is surprisingly meager, variable, and even paradoxical, as a consequence of the complex range of primary mineral associations, alteration products, and fracture-system geometries. The principal sources of data are drawn from logging operations by companies in search of metallic ores, evaluations of potential geothermal reservoirs, and oceanic crustal drilling by the Glomar "Challenger" and its successor, the "Resolution." In the following review, each tool will be considered in turn regarding its response within igneous sequences.
Lithodensity log
The modern density tool beams a stream of gamma rays into the formation from a radioactive source implanted on a pad pressed against the borehole wall. Two detectors measure the loss in gamma-ray flux caused by the formation lithology. The reduction is essentially a function of the aggregate electron density of the formation. The electron density is converted into an apparent density in units of grams per cubic centimeter, based on an assumed atomic number to atomic mass ratio (Z/A) of 0.5, and calibration with respect to standard sedimentary minerals. The apparent density generally tracks closely with the real density of common minerals, but appreciable differences can occur for minerals whose elemental Z/A ratio deviates significantly from the calibration ideal. Anomalous-density log values have sometimes been recorded in basic igneous rocks and, while not explained unequivocally, have been attributed to effects caused by electron densities higher than the medium used to calibrate the density-logging tool (West et al., 1975). However, the log densities of a range of igneous rocks reported from case studies summarized by Sanyal et al. (1980) generally shows close overall agreement with densities measured by conventional methods.
Long-term fluctuations of the density log within massive igneous rock units reflect overall changes in mineral assemblages and properties. These patterns may be indicative of either variations in primary crystalline phases and/or the products of hydrothermal alteration. The densities of igneous silicate minerals are sufficiently similar to make ambiguous the identification of specific mineral types. However, some useful diagnostic work is possible when density values are coordinated with other logs (notably the neutron and spectral gamma ray), or potential candidate minerals are known from petrographic observations of cuttings. A notable exception is provided by the presence of quantities of metallic minerals, such as magnetite or pyrite, which often cause a distinctive shift to higher densities. The genetic sequence of igneous-rock types from basic to acid compositions is matched by a trend from high densities to lower values (Keys, 1979).
The presence of fluid-filled vesicles or fractures causes a reduction in measured densities from their expected rock-matrix values. Because the density measurement is localized to the borehole wall in the vicinity of the source-detector pad, fractures or vugs may be missed as the contact device is drawn upwards through the borehole. However, diagnosis of possible fractures is aided by additional consideration of the caliper log and the count statistics of the two gamma-ray detectors. As a continuous measure of borehole diameter, the caliper registers borehole rugosity and highlights both washout zones and many fractures. The dual-detector system is designed to compensate both for effects of mud-cake thickness variation and minor borehole rugosity. Comparison of the two count rates yields a correction value which is recorded as a separate curve. When the borehole is relatively smooth (as indicated by the caliper), but the tool is located over a fluid-filled fracture, the short-spaced detector will be influenced more strongly than the long-spaced detector. Consequently, the combined examination of density, caliper, and density correction often provides useful indications of open fractures. However, it should always be remembered that both the simple caliper and density reading are localized, rather than full circumferential borehole measurements, so that fractures can be missed. Also, healed or mineralized fractures will not have the fluid component necessary for their recognition.
The preceding remarks are true for both the older compensated formation density tool and the more recent lithodensity device which was run in the Poersch #1 hole. In addition to an improved measurement of formation density, the lithodensity log incorporates a curve of the photoelectric absorption index (Pe). The photoelectric effect corresponds to the absorption of low-energy gamma rays and is a direct function of the atomic number, Z, of elements within the formation. The index is measured in units of barns per electron and may be converted to a volumetric absorption index, U, through multiplication by the density, and has units of barns per cubic centimeter. Values of Pe and U are tabulated for common sedimentary minerals in logging-service company manuals (e.g., Schlumberger, 1988). Consequently, the photoelectric curve is a valuable new means to assess mineralogy, when used in conjunction with density and neutron-log responses. The photoelectric absorption of most igneous mineral assemblages has not yet been addressed in the literature. However, the direct dependence between atomic number and photoelectric absorption index is given by the equation:
Z = 10 Pe0.218
Transformation of the photoelectric curve to a computed log of aggregate atomic number is therefore a useful means to relate logged measurements with their expectations calculated from elemental analyses of igneous-rock types.
Neutron log
The neutron tool bombards the formation with high-energy neutrons which experience reductions in energy through collision with formation atomic nuclei. The major loss in energy occurs in collisions with hydrogen nuclei, so that the neutron log is primarily an index of relative hydrogen concentration. The measurement is sensitive to hydrogen in any form, regardless of its presence in a fluid such as water or oil, or as a mineral constituent such as a hydroxyl radical. Estimation must be linked with a matrix calibration standard of known neutron cross section capture, and the scales are commonly expressed in units of percent volume water related to a matrix of either calcite, quartz, or dolomite. In Poersch #1, the neutron log was recorded with respect to an assumed quartz matrix.
The "neutron porosity" of the basic igneous rocks in Poersch #1 is fairly high and averages 19.5%. This figure shows good concordance with neutron responses of Tertiary basalts logged in Nevada (Sethi and Fertl, 1979), which averaged 20% and oceanic-crustal basalts logged by the Glomar Challenger in the equatorial Pacific, which ranged between 15 and 25% (Anderson et al., 1984). In cases of extensive fracturing or fluid-filled vesicles, the elevated neutron response can be attributed directly to water content. However, the major cause appears to be the influence of bound water within the matrix minerals. Fresh basalt would be expected to have a low neutron response, since the only significant hydrous mineral are the amphiboles, and hornblende is reported to have an equivalent neutron porosity of 8% (Edmundson and Raymer, 1979). Hydrothermal alteration of basalts is a widespread phenomenon and generates minerals such as chlorite, serpentine, epidote, and zeolites, all of which have a significant hydroxyl component. These cause an accentuation of the neutron response so that, for example, chlorite has an equivalent neutron porosity of 52%. Anderson et al. (1984) found that differencing of the neutron and density porosity logs in oceanic basalts resulted in a "hydroxyl mineral content" log which faithfully reproduced the degree of alteration observed in continuous cores. In practice, use of the neutron log in qualitative estimation of hydrothermal alteration must be tempered by considerations of the wide variability of hydrous-mineral assemblages as well as the unknown cross section capture characteristics of the non-hydrous phase.
Sonic log
The borehole-compensated sonic tool has a dual transmitter-dual receiver system designed to measure the speed of ultrasonic sound through the formation of the borehole wall. The transmitters alternate the triggering of acoustic pulses whose first arrival times are recorded at the receivers. The geometry of the tool design compensates for variable tool angle, transmission through the borehole mud, and moderate variations in borehole rugosity. The log measurement is continuously recorded as transit time in units of microseconds per foot and represents an average of the formation over the vertical distance between the receivers (the span).
In sedimentary-reservoir lithologies, the recorded transit time is primarily a function of the intergranular and intercrystalline pore volume, since the velocity of sound through fluids (and gases) is substantially slower than that of minerals in the rock framework. The explicit matrix transit times of igneous-rock assemblages are largely unknown. Field transit times for igneous rocks summarized from case studies of sonic logging runs by Sanyal and others (1980, p. 225-256) show wide ranges and overlaps that do not allow distinction by igneous type or crystal size. They are probably controlled by factors of rigidity and mechanical strength determined by cooling history, hydrothermal alteration, and fracture development which are both complex and poorly understood.
In igneous-rock sequences marked by significant open-fracture systems, the sonic log is often useful in the recognition of fracture zones, when considered in conjunction with density and/or neutron logs. While the density and neutron logs are sensitive to all kinds of porosity, the sonic log is usually "blind" to the porosity component in open vertical fractures. This difference occurs because the sonic tool records the first (and fastest) arrival time, which will follow a route in the borehole wall which misses the fracture(s). By contrast, the neutron and density tools will respond to all porosities and will include fractures, provided they are in the vicinity of the contact pad. Note that if a fracture is healed, the necessary fluid component will not be present for its recognition by any of these tools.
The determination of fractures from logging operations is still largely an art form and is usually based on a combination of "fracture indicators" as discussed in the main part of this report. The systematic evaluation of fracture systems calls for a specialized logging program such as the full waveform sonic tool (utilizing both compressional- and shear-wave characterization), borehole televiewer, and formation microscanner. However, these tools were not run in the Poersch #1 borehole, so that recognition of potential open fractures was drawn from diagnostic criteria of several tools, such as the density-correction curve (mentioned earlier), and the spectral gamma-ray and resistivity logs (described in the following sections).
Spectral gamma-ray log
The conventional gamma-ray tool contains a scintillation crystal which records the total gamma-ray emission of the formation in the borehole wall. The measurement scale follows an arbitrary but standard convention of API units, which are anchored to an instrument zero reading and a value of 100 units matched with an "average mid-continent shale" calibrator in the test pit at the University of Houston. The isotopic origin of gamma rays is given by their energy level. The three principal sources of gamma rays in nature are the potassium-40 isotope, and radioactive isotopes of the uranium and thorium series. Through the use of selected energy windows and sophisticated computer processing, the spectral gamma-ray tool records logs of thorium and uranium in parts per million, and potassium in percent.
The major uses of the spectral gamma-ray log in sedimentary sections have been for more accurate determinations of shale content, the typing of clay-mineral species, and the recognition of fracture zones (Serra et al., 1980). Since the tool has been deployed commercially for little over 10 years, case studies of its application in igneous-rock sequences is limited. However, the geochemical nature of its measurement holds great promise for the interpretation of igneous petrography and petrogenesis. The genetic sequence of magma types is matched by a progressive increase in potassium, uranium, and thorium ranging from basic, through intermediate to acid members. The gabbro/diabase/basalt clan is characterized by very low levels of all three sources. They are strongly contrasted with granite/felsite/rhyolite rocks and aplite/ pegmatite veins, which have significantly enhanced contents derived from potassium feldspars and uranium and thorium minerals associated with later stages of fractionation. The fink between these isotopes and aspects of hydrothermal alteration has still to be established but may best be understood in the context of orthodox geochemical models.
The spectral gamma-ray log has found great success in the location of fractures in sedimentary reservoirs (e.g., Fertl et al., 1980). These are often (but not invariably) marked by anomalously high and localized levels of uranium, caused by the precipitation of uranium salts at fracture margins from migrating solutions. The same phenomenon was observed by West and Laughlin (1976) in igneous rocks. The use of the thorium to uranium ratio (Th/U) is often a valuable aid to accentuate relative uranium enrichment, since it tends to differentiate between the soluble hexavalent uranium ion and the relatively insoluble tetravalent uranium and thorium ions (Zelt, 1985).
The bulk of the sedimentary rocks in the Poersch #1 borehole are arkoses and subarkoses, which are relatively immature products of erosion from granitic sources. As such, the differentiation of arkose and "granite wash" from granite, felsite, and pegmatite can often be ambiguous in petrographic examination of cuttings. The spectral gamma-ray log measurements provide additional evidence to clarify these distinctions. The weathering of granitic material in a hydrous, oxidizing environment results in the alteration of potassium feldspars to kaolinite and other clay minerals, while uranium is oxidized to its mobile hexavalent state and forms soluble uranyl salts. Both potassium and uranium pass into solution with a resultant depletion in these elements in the maturing sedimentary product. Thorium has a low solubility so that its behavior is more difficult to forecast, particularly as its content in clay minerals is generally high, apparently as a result of absorption between platelets (Hassan et al., 1976).
Resistivity logs
The resistivity of most rock-forming minerals is extremely high and generally exceeds the measurement limits (about 2,000 ohm-meters) of most commercial resistivity tools. Resistivity logs record the influence of conductive components whose nature is well understood, although the mathematical description of their contribution to resistivity remains approximate and empirical even after decades of study. Shales have low resistivity caused by cation-exchange mechanisms in the wet clay-mineral phase. The conductivity of shale-free sedimentary lithologies is a function of the volume of pore water, water salinity, temperature, and the tortuosity of the pore geometry which channels electrical current through the rock. The universally used empirical equation proposed by Archie (1942) takes the form:
F = Ro/Rw = 1/φm
This relationship states that the formation factor (F) is the ratio of the resistivity of brine-saturated rock (Ro) to the resistivity of the brine (Rw), and is inversely proportional to the pore volume (φ) powered by a "cementation factor" (m). The cementation factor is controlled by the tortuosity of the pore network, which is an expression of the current path length through the rock.
The internal structure of igneous rocks differs drastically from sedimentary reservoir lithologies. In cooling from a parent magma, the component minerals form an interlocking mesh of crystals which almost invariably precludes the development of connective pores. Possible exceptions could occur in highly vesicular volcanic rocks, where both volume of pore water and connectivity would be required for the transmission of electrical current. In ancient volcanics, the vesicles are usually occluded by amygdaloidal minerals, so that conductivity from this source is unlikely.
The resistivity of most igneous minerals is extremely high, so that the general expectation for massive igneous rocks are resistivities which exceed the measurement scale of commercial resistivity logs. Exceptions to this generalization are provided by metallic minerals such as magnetite and pyrite. However, their role as significant contributors to conductivity is controlled by their relative amount and degree of connectivity within the rock framework. Detailed studies of this phenomenon with respect to pyrite were described by Clavier et al. (1976).
Appreciable conductivity in igneous rock which is matched by measurable resistivity features are most commonly attributed to open-fracture systems (Sanyal et al., 1980). In the Poersch #1 borehole, three resistivity logs were run in combination. Their coordinated behavior can be related to expectations of responses likely to be caused by fractures. The spherically focused log (SFL) is most strongly influenced by rock close to the borehole wall; the medium-induction log (ILM) responds to rock further away; the deep-induction log (ILD) records resistivity at greater distances from the borehole. The SFL system develops approximately constant potential shells around the current-emitting electrodes. Measurement of current at return-electrodes is a function of the conductivity within the shallow zone of investigation. The two induction tools induce eddy currents within the formation which causes magnetic fields. The magnetic fields, in turn, induce currents in receiver coils. All three tools are conductivity devices whose operating range is best in moderately conductive formations. In highly resistive rocks, aberrations are not uncommon as a consequence of short-circuit mechanisms. Under these conditions, the borehole mud column is (by contrast) relatively conductive, and its "borehole signal" can dominate the measured resistivity in complex and unpredictable ways.
The occurrence of open fractures introduces planar conductive features, whose effect is often registered on the resistivity-log combination as a distinctive response pattern. Under normal conditions, the resistivity logs track as sub-parallel curves in a set order. The ILD reads the lowest, the ILM an intermediate, and the SFL as the highest resistivity. This ordering is a consequence of the radial change in dominant fluid type from relatively fresh mud filtrate to more saline connate waters. However, fluid-filled fractures in the vicinity of the borehole can influence the shallow investigation SFL tool, with a significantly reduced resistivity reading. This may be less than that recorded by the other tools. The effect is also accentuated by the physical design principle of the induction tools which tends to cause them to be insensitive to vertical fractures (Sanyal et al., 1980).
Spontaneous potential (SP) log
The SP logging tool is a "passive" device which responds to natural electrical potentials that occur within the borehole. The log is recorded in units of millivolts. Events which occur on the SP log are caused by natural battery effects between the borehole and formations in the wall. Three types of potential mechanism have been recognized: the liquid junction, membrane, and electrokinetic potentials. The first two species are electrochemical and dominate the SP response in sedimentary sequences of interbedded shales and porous, permeable reservoir lithologies. Their magnitude is controlled by the contrast in salinities of invasion-mud filtrate and saline-formation waters, through the flow of ions from the more concentrated to the more dilute solution. The movement of solute ions causes current flow and the initiation of a small voltage potential which is measured by the SP log.
The electrochemical potentials are usually minimal in massive igneous rocks, where both mud-filtrate invasion and the existence of interparticle connate water are largely precluded. However, localized features on the SP log have been observed commonly at fracture zones and are usually attributed to "streaming" or electrokinetic potential of ion movement within the fractures (Kintzinger et al., 1977).
In practice, SP logs of igneous sections are characteristically torpid, featureless curves which appear to wander aimlessly and show little concordance with features observed on other logs. Under these conditions, the susceptibility of the SP tool to other, extraneous sources of minor voltages make them suspect at best. However, a strong potential feature at the location of possible fracture zones indicated by other logs is useful supporting evidence.
Kansas Geological Survey, Texaco Poersch #1 Report
Placed on web March 5, 2010; originally published June 1988.
Comments to webadmin@kgs.ku.edu
The URL for this page is http://www.kgs.ku.edu/Publications/OFR/1988/OFR88_22/11_app5.html