ABSTRACT

Proper selection and application of calibrated well logs provide valuable information for exploration, drilling, and reservoir engineering. Log-derived interpretive information defines subsurface pressure and temperature, type of depositional environment, and evaluates the production potential of clastic and carbonate reservoir rocks.

This study also focuses on overpressure detection, pore pressure evaluation and its correlation with the distribution of commercial hydrocarbon resources in clastic Gulf Coast sediments. Additional emphasis is placed on recent experiences with gamma ray spectral logging techniques to locate permeable and fractured intervals in the Cretaceous carbonate trend, to determine the source rock potential of shales, and the amount and type of clay minerals present.

Field case examples illustrate the above mentioned techniques.

SUBSURFACE TEMPERATURE AND PRESSURE ENVIRONMENT

Subsurface formation temperature available from open hole well logs listed on the log heading is always lower than the true, or static, formation temperature. Due to the cooling of formations by circulating drilling mud (conditioning a well prior to logging, for example) the recorded bottom-hole temperature (BHT) may be 20°F to 80°F lower than the actual formation temperature.

Since true or static formation temperature is an important parameter in exploration, drilling, logging, well completion, and reservoir engineering there is a method which permits the determination of static formation temperature from maximum recording thermometer (BHT) data recorded during all routine logging operations. The recommended technique requires the use of BHT data on each logging run, including information as to mud circulating time and time that the logging device was last at the bottom of the wellbore (Fertl and Timko, 1972; Dowdle and Cobb, 1975).

The basic concept is the straight-line relationship on semilogarithmic paper of BHT in °F (from well log heading) vs. the ratio of t/ (t + t), where t = time in hours after circulation stopped; t = circulating time in hours for well conditioning. Then, exptrapolation of this straight line to a ratio of t/ (t + t) = 1.0 will determine the true static formation temperature. Figure 1 illustrates the extrapolation technique for true static formation temperature in two wells which have been drilled in quite different geothermal regimes. Well no. 1 is a high temperature well located in the South China Sea; well no. 2 is a deep hole drilled onshore Texas (Fertl, 1976).

Abnormal formation pressures can be caused by many factors. In some areas, a combination of these factors prevails. To place the possible causes of abnormal formation pressures in proper perspective, it is necessary to understand the importance of petrophysical and geochemical parameters and their relationship to the stratigraphy and structural and tectonic history of a given area or basin.

Crossplots of formation pressure gradient and temperature from Gulf Coast wells (fig. 2) yield important findings (Timko

Figure 1. Extrapolation technique for true formation temperature. 1 = High-temperature well. Four logs were run to 6518 ft (1987 m). First log, 4 hours after mud circulation stopped; recorded 218°F (103°C). Log 2, 6 hours later, measured 264°F (128°C). Straight-line extrapolation to infinte time (log 1.0) indicates a BHT of 281°F (138°C). 2 = Deep Texas onshore well. Three logs were run to 12,548 ft (3826 m). First-recorded temperature was 272°F (133°C), whereas actual stabilized BHT is 284°f (140°C). t = circulating time, in hours; t = time after circulation stopped, in hours. (Fertl, 1976)

End_Page 64-------------------------

and Fertl, 1971). A temperature window for 215°F to 290°F coincides with the range of highest pressure gradients in hydrocarbon reservoirs. Note that this temperature range lies in the bulk part of second-stage clay dehydration as proposed by Burst (1969), i.e. it lies in the zone of maximum fluid distribution.

A generalized correlation has been developed between a "typical" Gulf Coast shale-resistivity profile (fig. 3) and the distribution of oil and gas field in this area (Timko and Fertl, 1971). Application of these statistical study results indicates whether it is possible for commercial production to exist below the depth to which the well has already been drilled and logged, and whether it is economically attractive to continue drilling a borehole below a given depth in sand/shale sequences. Such limits on potential hydrocarbon targets in super-pressure environments are caused by geologic factors and severe production problems. Recently, similar findings have been reported for the exploration activities in the Mackenzie delta, Canada, by Evans et al. (1975).

GAMMA RAY SPECTRAL LOGGING APPLICATIONS

Gamma rays are the radiations originating within an atomic nucleus. A nucleus gives off excessive energy (gamma rays) as the results of radioactive decay or an induced nuclear reaction. Radioactive decay consists of the emission or capture of elementary or composite particles with the consequent transformations into daughter nuclei characterized by different atomic numbers and, in some cases, by different mass numbers. Both uranium and thorium are characterized by specific decay series. The only unstable isotope of potassium is the nuclide potassium 40, the major contributor, which emits a single, easily identifiable gamma ray at 1.46 MeV.

In addition to total gamma ray counts, the Spectralog measures the gamma rays emitted by potassium 40 at 1.46 MeV, the uranium series nuclide bismuth 214 emanating gamma rays at 1.764 MeV, and the thorium series nuclide thallium 208 emanating gamma rays at 2.614 MeV. Pertinent instrument specifications for the Spectralog, the schematic of instrumentation, and the energy window calibration have been reported in the logging literature previously (Wichmann et al., 1975; Fertl, 1979).

Under oxidizing and slightly alkaline conditions, the uranyl ion (UO₂⁺⁺) is soluble in subsurface waters containing carbonte, bicarbonate, or hydroxyl ions. Some of this dissolved uranium may be absorbed on ferric hydroxide, which in turn often coprecipitates with calcium. Furthermore, radioactive salts are usually found to coprecipitate with barium sulfate (inasmuch as barium is a chemical analog to radium), which is the least water-soluble salt. Depending on variations in temperature, pressure, flow, and geochemical conditions, this radioactive salt precipitates in a non-reversible manner on the cement annulus or the pipe (i.e., casing) itself. In other words, micro-crystals of water-insoluble "radiobarite" [Ba (Ra) SO₄] formed suspended in colloidal solutions may, under dynamic flow conditions, be transported through permeable reservoir rock, such as during primary production or waterflood operations, until final precipitation occurs at perforated or around unperforated cased wellbores which penetrate the subject formation.

The well-known application of gamma ray logging techniques for locating oil-depleted, water-flooded strata and, hence, bypassed oil has been discussed by several investigators (King and Bradley, 1977; Doering and Smith, 1974; Fertl et al., 1978). Use of conventional gamma ray logs to successfully locate behind-the-pipe saltwater invasion in six Texas oilfields was recently discussed (King and Bradley, 1977). Here again, the Spectralog has unique application in cased wellbores to assist in locating watered-out or bypassed oil stringers and simultaneously provides a reliable shaliness estimate.

Figure 2. Formation pressure gradient vs formation temperature in 60 overpressured Gulf Coast wells.

Figure 3. Typical Gulf Coast shale resistivity profile based on the short normal curve, correlated to distribution of gas-oil reservoirs. Profile is based on hundreds of commercially productive wells.

End_Page 65-------------------------

Clastic Reservoir Rocks

Clean clastic formations exhibit low values for K, Th, and U. However, presence of highly radioactive minerals (monozite, zircon, mica, etc.) may substantially increase the radioactivity of clean clastic rocks. Both K and Th concentrations provide reliable shaliness estimates based on either linear or exponential correlations.

V_SH1 = (A-A_MIN)/(A_MAX-A_MIN)

A = value from Spectralog (K in %, or Th in ppm) in zone of interest

A_MIN = minimum log value (K in %, or Th in ppm) in nearby clean, shale-free formations

A_MAX = maximum log value (K in %, or Th in ppm) in nearby essentially pure shale zone

This linear mathematical relationship can be applied by simply scaling the proper logging curve from 0 to 100%. In some areas a value of V_sh = 0.5V_SH1 has been found to be a good approximation. More refined relationships include:

V_SH = 0.33 (2^2VSH1-1.0) - Highly Consolidated and Mesozoic Rocks

V_SH = 0.083 (2^3.7VSH1-1.0) - Tertiary Clastics

Potassium, Thorium, and Uranium Distribution in Clay Minerals and Shales

Generally speaking, most shales exhibit high K- and Th-content, but low U-values. Kaolinite, a K-deficient clay mineral, is a notable exception. Bentonites, frequently used as important stratigraphic time markers over large areal extent, show excessively high Th-content. Clay typing based on K- and Th-values is shown in figure 4.

The Eagle Ford shale of Creataceous age, located within the Cretaceous carbonate trend in Texas, is sandwiched between the Austin Chalk and Buda Limestone. The Eagle Ford formation is the potential source rock for the oil in the carbonate reservoirs. It exhibits a wide range of lithology, varying from a typically dark, organic-rich shale to a limey, chalky, dolomitic formation, or shales sometimes interbedded with siltstone. Porosity in the Eagle Ford formation is usually less than 5%. Even though by itself not to be considered the sole target for exploratory drilling, the formation is frequently fractured and then becomes a potential completion target of more tha just localized interest.

Several of these organic-rich shales frequently owe their production potential i.e., permeability, to natural fracture systems in an otherwise essentially impermeable rock. Natural fracture systems are concentrated in brittle, usually calcareous or silty zones. The Spectralog easily pinpoints the calcareous or silty zones, both being characterized by low potassium and low thorium, but extremely high uranium concentration. Based on this concept, the Spectralog has already assisted in many successful completion or recompletion attempts.

Whereas figure 5 shows the Eagle Ford shale as a tight source rock, figure 6 illustrates an oil-productive Eagle ford interval in Frio County, Texas. For comparison, figure 7 presents the log response over the entire Eagle Ford shale-Buda Limestone-Del Rio shale section in the Cretaceous carbonate trend, south Texas.

Source Rock Potential in Argillaceous Sediments

As early as 1944, it was stated (Beers and Goodman, 1944) that "use of radioactive elements as tracers in the study of sedimentation may aid in defining the source beds of

Figure 4. Thorium/potassium ratio (TH/K) variations in potasium-rich feldspars, glauconite, muscovite, illite, mixed layere, kaolinite, chlorite, and bauxite minerals. (After Hassan et al., 1976).

Figure 5. Spectralog response over Austin Chalk, Eagle Ford shale and Buda Limestone in Cretaceous carbonate trend, Caldwell Co., Texas. (After Fertl et al., 1978). The relatively thin Eagle Ford shale in the subject well is characterized by very high potassium and thorium values, and excessively high uranium concentration. These Spectralog response characteristics are typical for an organic rich shale.

End_Page 66-------------------------

oil." Laboratory data for Paleozoic black shales have shown linear correlations between the carbon content (in %) and the uranium content, and the thorium/uranium ratio. In 1945, the radioactivity and organic content of 510 samples of sedimentary rocks were studied and "a marked relation between certain types of organic content and radioactivity" was observed (Russel, 1945). In 1978, a patent was issued for the in situ evaluation of the source rock potential of earth formations based on downhole gamma ray spectral logging application (Supernew et al., 1978).

The Spectralog, hence, allows a continuous monitoring of the source rock potential of shales in both open and cased wellbores. The potential of the Spectralog to hydrocarbon exploration thus becomes obvious (fig. 8). New and old wells can be logged to determine source rock potential variations both versus depth and on a regional basis using the appropriate mapping techniques.

Cretaceous Carbonate Trend

Pure carbonates (limestone, chalk) are characterized by very low K-, Th-, and U-contents. However, presence of alkalines (authigenic feldspar, argillaceous material, etc.) may increase significantly the K-content.

Whereas precipitation of uranium is inhibited by the presence of carbonate ions, which is reflected in the generally low uranium concentration in limestones and dolomites, U-precipitation from aqueous solutions (subsurface waters) easily takes place in reducing environments in presence of carbonaceous amterial and sulfides. Such conditions, combined with precipitation during ground-water movement over geologic time through zones of high fluid transmissibility

Figure 6. This Spectralog is shown over the Eagle Ford shale formation in the Pearsall area, Frio County, Texas. The total counts (gamma ray) response appears to be completely unrelated to the "shaliness" of the formation. The K curve indicates a relatively "clean," calcareous Eagle Ford formation, particularly between 5450 ft and 5500 ft. Note the frequent correlation between decreasing K and increasing U concentration. (After Fertl, 1979). Selective perforating in such zones often yields commercial production without any massive well stimulation.

Figure 7. Spectralog shows drastically different response in the organic-rich (source-rock), partly calcareous, Eagle Ford shale (>K, >> U,>Th) down to 4122 ft, over the relatively clean and tight Buda Limestone (< K, < U, < Th) from 4122 ft to 4218 ft, and illustrates the typical shale response in the Del Rio shale (>>K, > U, >>Th) below 4218 ft (After Fertl, 1979).

End_Page 67-------------------------

(i.e., high permeability intervals, natural fissure and fracture systems, fault zones, etc.), may be the cause of U-enrichment in excess of 20 ppm.

Hence, in carbonates, the Spectralog assists in the description of rock types, detailed zone correlation, and reliable shaliness estimates and frequently assists in locating those intervals which: 1) provide high production rates from natural fracture systems; and 2) pinpoint target zones for well stimulation such as fracturing or acidizing. Increased oil production based on recompletions using Spectralog is illustrated in figure 9; an important correlation between mud log shows and Spectralog response is presented in figure 10.

CLAY MINERALS IN RESERVOIR ROCKS

Significant shaliness (clay content) of potential reservoir rocks can impose drilling, completion, and production problems. To a varying degree shaliness of reservoir rocks affects all electric, radioactive, and nuclear logging techniques. Furthermore, such shaley and/or silty reservoir rocks may exhibit log-derived water saturation values as high as 60 to 70%, but are still capable of producing waterfree oil.

Three basic clay distribution models of "shaliness" in potential reservoir rocks are considered in formation evaluation using well logs. These include: 1) clay dispersion; 2) clay laminations; and 3) structural clay. A schematic presentation of these three clay models and, most importantly, their drastically different impact on effective porosity and reservoir permeability is illustrated in figure 11.

Dispersed clays may occur in the pore space as: a) discrete, i.e. not intergrown, particles; b) coating of sand grains by intergrown clay crystal linings; and c) clay crystal bridging across the pore space (fig. 12) (Neasham, 1977). As illustrated previously, a direct relationship exists between increasing clay content and decreasing effective reservoir porosity. Furthermore, the three categories of dispersed clay distribution affect reservoir permeability, significantly differently.

Discrete clay particle distribution is typical for sandstones with "patchy" kaolinite, which is frequently arranged at random within the rock. Not only is the effective porosity reduced, but this kaolinite often behaves as migrating fines (Pittman and Thomas, 1978). Coating and pore-lining clay minerals are illite, chlorite, and montmorillonites. They frequently intergrow and then form continuous clay "layers" containing ample micro-porespace with pore diameters up to two microns. The "water holding" capacity of clay layers, particularly of the smectite-type, to retain large amounts of "non-movable" pore waters is apparent and so should be the effect on water saturation calculations versus production behavior.

Pore-briding of clay minerals also occurs with illite, chlorite, and smectites (montmorillonites). Often extensive intergrowth of clay crystals will be present (i.e., "bridging" will frequently occur together).

Shaliness of reservoir rocks to varying degrees affects all electric, radioactive and nuclear logging techniques. In addition, effects due to presence of hydrocarbons are superimposed. The combined effect of shaliness and hydrocarbon effects does complicate any quantitative log interpretation, but -- surprisingly at first glimpse -- does offer several benefits for qualitative, quick-look type log evaluation.

Figure 8. Organic carbon content (%) versus uranium/potassium ratio (left) and uranium content (right) in New Albany Shale. (After Supernaw et al., 1978)

Figure 9. The Spectralog was run over the Austin Chalk section in a well located in Bexar Co., Texas (After Fertl et al., 1978). Initially perforated from 1400 ft in the "cleanest" chalk, the well tested at 3 BOD. Based on a subsequent Spectralog run the operator also perforated the two intervals of low potassium but excessively high uranium values at 1360 ft to 1366 ft and from 1380 ft to 1384 ft. Note that on a conventional gamma ray curve these zones would look "dirty," whereas the Spectralog clearly shows the Austin Chalk to be relatively shale free. The recompletion attempt based on the Spectralog proved successful, with production increasing sixfold to 18 BOD.

End_Page 68-------------------------

Figure 10. Spectralog and mud log information correlate in this prolific oil well, located in Frio Co., Texas, which also offsets other significant oil producers. Close study of this data shows an interesting correlation between the shale-free, U-rich intervals and significant hydrocarbon shows recorded on the mud log over this Austin Chalk section. (After Fertl et al., 1978).

End_Page 69-------------------------

For clastic reservoir rocks all three porosity-type log responses in gauge wellbores can be expressed as shown in figure 13a and 13 b. Overlays of comparable scaled logs quite often locate hydrocarbon and especially gas-bearing zones (the same log combinations and concepts are also used in crossplot techniques). It allows a quick look at the extent of the pay and assists in locating the proper intervals to be tested and/or perforated.

Presence of gas is indicated by the difference in the porosity from acoustic and neutron logs. The acoustic indicates high apparent porosity whereas the neutron, when affected by gas, shows low apparent porosity. Since both logs respond to porosity and shaliness in similar fashion, such overlays work in clean and shaly sands.

Presence of gas may also be detected through comparison of density and neutron. This overlay shows the greatest separation in clean sands, whereas shaliness tends to mask hydrocarbon effect. Acoustic-density overlays show the least efficiency, especially in shaly, uncompacted formations. Under such conditions, acoustic readings are affected by lithology, compaction, porosity, and hydrocarbon content in the pore space.

The "q-factor" can be used successfully to estimate if shaly reservoir rocks are too tight for commercial production (Alger et al., 1963; Fertl and Watt, 1977). This q-factor provides a basis for determination of cut-off permeability in pay sands. One expects this permeability cut-off to be dependent on both shaliness and effective porosity of the reservoir rock. Past attempts suggested a tentative permeability cut-off when values for the q-factor exceeded twice the value of effective porosity or when q was larger than 0.4. However, field experience with this cut-off, particularly in sands of moderate porosity (15 to 25%), has not always been satisfactory.

Figure 14 illustrates a proposed permeability cut-off over the entire range of reservoir porosities encountered. The chart is based on field data from the U.S. Gulf Coast area, New Mexico, California, and Wyoming (Fertl and Watt, 1977). Basically, shaly reservoirs are classified in three

Figure 11. Clay distribution models and effect on effective reservoir porosity.

Figure 12. Cross plot of porosity and air permeability (Klinkenberg corrected) for selected sandstone samples which have contrasting dispersed clay morphology.

Figure 13. (a) Generalized response of porosity logs.

End_Page 70-------------------------

categories, producible, non-producible, and zones which require stimulation. Note that in the range of low effective reservoir porosity the region of possible production is separated from the essentially tight formations by a narrow band. Formations which plot within this band frequently require stimulation to obtain production.

SPONTANEOUS POTENTIAL (SP), GAMMA RAY AND SPECTRALOG CURVES DEFINE DEPOSITIONAL ENVIRONMENTS

Extensive literature describes the use of well logs in hydrocarbon exploration. The present review focuses on the qualitative use of well logs applied to stratigraphic correlations and interpretations of depositional environments. However, it cannot be overemphasized that in such studies one must use all available data and be thoroughly familiar with the response of and all possible effects on the specific logging curves used.

Log-derived patterns provide basic clues for depositional variations in clastic sediment sequences. Generally speaking, clays, shales, and silts are deposited in a low-energy environment, whereas sands and coarser materials are deposited in a high-energy environment.

Typical SP-curve patterns relating to depositional environments and sand features are shown in figures 15 and 16. The Gamma Ray and Spectralog record the natural radioactivity of a formation. Radioactivity usually increases with increasing shaliness or, in other words, with decreasing grain size. Hence, similar to the SP-curve, these radioactive logs can be used as grain size profile indicators. Figure 17 summarizes the four major natural radioactivity curve motifs.