Coal Resource Classification System of the U.S. Geological Survey
By Gordon H. Wood, Jr., Thomas M. Kehn, M. Devereux Carter, and William C. Culbertson
GEOLOGICAL SURVEY CIRCULAR 891
GEOPHYSICAL LOGS AS A SOURCE OF COAL BED DATA
Geophysical well logs have been invaluable in the search for oil and gas because they provide rapid, economical, and detailed information on the thickness, lithologic characteristics, fluid content, correlation, structure, and depth of strata penetrated by a well. Such logs have been run in many thousands of oil and gas exploratory, discovery, and producing wells that have penetrated coal-bearing strata in many regions of the country.
Geophysical logs can be used to identify coal beds and to quantify their resources because coal has several unique physical properties including low natural radioactivity, low density, and high resistance to electrical currents; these properties contrast with those of most other rocks in the coal-bearing sequence. Thus, geophysical logs can provide information on the existence, continuity, thickness, and correlation of shallow to deeply buried coal beds in known coal-bearing areas that have not yet been fully explored and in the future may provide information in areas not previously thought to contain coal. Most of these logs are available to the public through commercial companies and may be studied at State geological surveys or at similar agency offices in some States.
Caution should be used in evaluating and interpreting the existence, thickness, and correlation of coal beds from the geophysical logs of oil and gas exploratory and production wells for two reasons. First, if only one type of log is available, other rock types may be misidentified as coal, for example, highly resistive limestone on a resistivity log or pure quartz sandstone on a natural radioactivity log. However, this problem can be mitigated by a thorough knowledge of the coal-bearing strata in the area under investigation and by an understanding of how these strata are recorded on a log. For example, if limestone beds are not present in the coal-bearing sequence and if all sandstone beds are more radioactive than the coal beds, problems of lithologic identification are nonexistent. If several log types are available, the coal generally can be identified with confidence despite other strata with similar log characteristics in the sequence. Second, some oil and gas logs are not suitable for identifying coal beds for reasons that are not always clear because they give readings that either mask coal beds, indicate coal beds where none are present, or are ambiguous. Here again, a knowledge of the stratigraphic positions of coal beds in coal-bearing sequences will aid identification of unsuitable logs. Because coal thicknesses interpreted from geophysical logs are considered as points of measurement for calculating coal resources, it is advisable to use only those coal thicknesses that are determined with confidence.
The geophysical log types generally used in coal bed recognition and stratigraphic identification and rank, quality, and thickness evaluations are the electrical, gamma ray, density, neutron, and acoustic velocity. The following discussion of log types is concerned principally with geophysical logs of oil and gas exploratory wells. It should be noted that coal exploration programs increasingly have become reliant on coal-oriented geophysical logs, which provide important data on thickness, depth, and correlation of coal beds, and locally on the composition of coal. A carefully chosen suite of coal-oriented logs can provide positive recognition of coal, identification of specific coal beds, and precise coal thickness measurements (Vaninetti and Thompson, 1982). Such suites are used currently to supplement information obtained from core holes, driller's logs, and examination of drill cuttings. They are useful especially where core or sample recovery of coal was incomplete. The ease and accuracy of recognizing, identifying, and evaluating coal beds with coal-oriented logs strongly contrasts with the difficulty of performing the same interpretations using the higher speed and less accurate geophysical logs of oil and gas wells, which generally are run at different instrument settings and with different equipment. Most oil and gas geophysical logs are recorded at a scale of 1 inch equals 50 feet, with selected sections recorded at the larger scale of 1 inch equals 20 feet, which is reduced for commercial sale to 1 inch equals 100 feet and 1 inch equals 40 feet, respectively.
By far the most common geophysical logs run in oil and gas exploratory wells are electric logs. Prior to the 1950's, conventional electrical logging surveys consisted of one measurement of the spontaneous potential (SP) and three measurements of apparent electrical resistivity of the rocks adjacent to the bore hole (fig. 26). These rock properties were measured only in the uncased part of a well that was filled with water or a water-based mud. The diameter of the well and the effect of adjacent rocks combined to give confusing curves on older electric logs. In solving this problem, a new family of resistivity curves, the focusing-electrode and the induction logs, came into use in the late 1950's. These logs provide better resolution of the coal beds than the older conventional logs and permit more accurate coal thickness measurements.
The spontaneous potential (SP) log measures the difference in electrical potential between rock types, and the resulting curve is recorded on the left-hand side of the log as a single trace. This curve generally reflects the invasion of drilling fluid into the rocks, so a permeable sandstone bed tends to record as a large deflection to the left of the log response for shale (fig. 27 and fig. 28). There are many exceptions to this generalization as shown by the deflections caused by high-porosity coal beds (see fig. 29, SP curve). There are also many wells where the SP curve is nearly featureless in a coal-bearing section and the porosity is recorded the same as shale (fig. 28A and fig. 28B).
Three types of resistivity curves are recorded on the right side of a geophysical log. These are the 16-inch normal (short normal); the 64-inch normal (long normal); and the 18-foot, 8-inch or 24-foot lateral (lateral); the names referring respectively to the spacing and to the configuration of the electrodes in the probe. These curves record the resistance of rock types to the flow of an electric current. Because most coal beds are highly resistant to the flow of an electric current compared with most adjacent rocks, resistivity curves generally show a large deflection opposite a coal bed. In the short and long (16-inch and 64-inch, respectively) normal curves, however, coal beds thinner than the electrode spacing show a "reverse" (low resistivity) curve bounded by two small peaks (fig. 28A and fig. 28B, No. 6 coal). The lateral curve shows a large deflection opposite thin coal beds and a low deflection below the bed. The lateral curve is of little value in the measurement of the thickness of coal beds because it is asymmetric and generally offset from the coal bed. Nevertheless, this curve can be useful in correlating coal beds (fig. 28A and fig. 28B).
FOCUSING-ELECTRODE AND INDUCTION LOGS
Focusing-electrode logs (for example, lateral logs) use special electrodes to send a narrow focused electric current horizontally into adjacent rocks. This results in a resistivity curve that has good resolution of thin resistive beds such as coal (fig. 29 and fig. 30). These focusing and lateral logs measure the conductivity (inverse of resistivity) and the resistivity of rocks by means of induced alternating currents. Commonly, an induction log is run in conjunction with a SP and 16-inch normal log (fig. 27). The induction log is recorded simultaneously as two curves, conductivity and resistivity. The conductivity curve is hyperbolic, which compresses the parts of the curve characterized by low conductivity. The resistivity curve, however, is not compressed, so it can be compared directly to the short normal curve and can be used for measurement of the thickness of a coal bed. Combinations of induction and focusing-electrode logs are also common.
With the exception of some high-moisture-content lignites, most coal beds are responsible for high-resistivity deflections on resistivity curves. However, some other rock types such as limestone or resistive sandstones also show high-resistivity deflections and may be mistaken for coal. Limestone is indistinguishable from coal on most electric logs. Fortunately, limestone beds are absent in many regions. In the Mid-continent region, however, limestone is abundant in the coal-bearing sequence and generally can be differentiated from coal only by use of supplementary logs (see fig. 31) or by examination of closely spaced samples of drill cuttings.
Resistive sandstone beds that are permeable generally can be differentiated from coal beds because of their large deflections on a SP log (figs. 26, 27, and 28). Where a SP log is featureless or ambiguous, a knowledge of the lithology of a coal-bearing sequence can help differentiate coal from sandstone. For example, in the Powder River basin of Montana and Wyoming, many sandstone beds are gradational with adjacent low resistivity shale beds, whereas the shale beds are in sharp contrast with adjacent high-resistivity coal beds. As a result, the resistivity curves delineating coal beds are more nearly parallel than the curves representing the contacts of sandstone beds.
The gamma ray log records the natural gamma radiation from rocks adjacent to a drill hole. Coal generally has low natural radioactivity as compared with other rocks, particularly shale, in a coal-bearing sequence (figs. 29, 30, and 31). In some coal-bearing regions, limestone and sandstone may have similar low natural radioactivity (for example, Midcontinent or Appalachian regions) so that in those regions supplemental logs such as density or acoustic velocity are needed to identify coal. In other regions, gamma ray logs alone are sufficient to identify coal beds as, for example, in the Fruitland Formation (Fassett and Hinds, 1971) and in the Northern Great Plains where no other rock types in the Tertiary coal-bearing sequence, including limestone, are known to have as low a natural radioactivity as coal. Even in areas where a gamma ray log is diagnostic of coal, a few oil and gas well gamma ray logs are useless because their time constant is so long and their sensitivity is so low that a coal bed either cannot be detected or its boundaries are obscured. Locally some coal beds are uraniferous; consequently, a high radioactivity is recorded on gamma ray logs. These uranium-bearing coal beds usually can be identified by using other logging methods.
A gamma ray log is the most versatile of the geophysical logs for the following reasons: (1) it does not require fluid in the hole; (2) it is not sensitive to small variations in hole diameter; and (3) it can be used to detect coal beds through well casing. In fact, near-surface gamma ray logs of cased oil and gas wells are a prime source of data for identifying and measuring the thickness of shallow coal beds in the Northern Great Plains region.
The gamma ray logs can detect shale partings in a coal bed, but generally the thickness of thin partings is exaggerated. In the Powder River basin of Montana, there is an example where the gamma ray log records a 0.3-foot shale parting at the same thickness as a 2-foot parting.
A gamma-gamma density log records the bulk density of rocks adjacent to a drill hole by measuring the induced gamma rays emitted by the rocks after bombardment by a gamma ray source encased in a probe and lowered into the drill hole. The denser the adjacent rocks, the more gamma rays are absorbed and not returned to a detector in the probe where they are measured in grams/cm3. Most ranks of coal are low in density (about 0.7 to 1.8 grams/cm3) compared to adjacent rocks; therefore, a density log is an excellent tool for coal-bed evaluation. The density log is capable of identifying detailed variations in the density of rocks. Density logs from oil and gas wells are commonly run at 30 feet per minute, are recorded at 1 inch equals 20 feet, and then are reduced to 1 inch equals 100 feet. At this scale and speed, thin coal beds and partings can be detected and their thicknesses accurately measured (fig. 30). Unfortunately, density logs must be run in uncased holes and are strongly affected by differences in hole diameter; thus, the curve recorded for a caved or enlarged part of a well may simulate the curve of a coal bed. A caliper log, which measures the diameter of a well, is generally run in conjunction with these logs so that the proper interpretation and correction can be made. Density logs should be used in conjunction with other logs to avoid mistaking spurious low-density readings for coal. These logs are generally recorded simultaneously with a gamma ray log, which resolves most uncertainties of rock identification.
The neutron log, which is similar to the density log, records the varying intensity of gamma rays resulting from inducing neutrons into rocks adjacent to a drill hole by a probe containing a radioactive source. In general, the number of gamma rays and neutrons detected as they return to a detector in the probe are inversely proportional to the hydrogen ion content of a particular rock type. The curve recorded by the detector of returning gamma rays and neutrons can be interpreted to indicate the relative fluid content of the rocks and therefore their porosity and permeability. This log is commonly called the porosity index. The curve of the neutron log records high readings adjacent to permeable fluid-filled rocks because of their high hydrogen contents, but it also records high adjacent to a coal bed because of its high carbon content (fig. 31). A clay with high-moisture content will also record high reading on this curve. Therefore, a high-moisture day adjacent to a coal bed can obscure the contact between the day and coal and record a spurious thickness of coal. Other types of logs such as the density or gamma ray will allow a more accurate measurement of the thickness of coal. Neutron logs are also strongly affected by caved or over- sized diameter holes; therefore, neutron logs should be used in conjunction with caliper logs.
Acoustic velocity logs record the velocity of pulsed sonic waves generated in a probe, transmitted into rocks surrounding the probe in a drill hole, and reflected to receivers in the probe. The results are recorded as the inverse of velocity--that is, the time in microseconds for the waves to travel 1 foot (interval transit time). The velocity of the sonic wave depends both upon the lithology and the porosity of the rock type being penetrated. If the rock type is known, therefore, the acoustic velocity log can be used as a measure of porosity. A decrease in velocity (increase in interval transit time) can be interpreted to be the result of an increase in porosity. Coal generally has a longer interval transit time (lower velocity) than adjacent rocks. Because an acoustic velocity log can record velocity changes in great detail, it is commonly recorded on an expanded scale (1 inch = 20 feet) along with a gamma ray and caliper log. However, its value as a tool for identifying coal beds is dependent on the nature of the coal-bearing sequence. An acoustic velocity log is an excellent tool in deeply buried rocks such as the Tertiary rocks of southwest Wyoming. Acoustic velocity logs can be used to delineate coal beds with precision (see fig. 29). They are of lesser value where the same rocks are near the surface, are poorly consolidated, or are fractured because the many spurious low-velocity deflections are recorded. Such logs are everywhere useful in distinguishing coal from limestone, which has a much shorter interval transit time.
MEASUREMENT OF THE COAL BED THICKNESS
The measurement of thickness of coal beds on geophysical logs requires the identification of the top and base of the coal beds by either of the following methods: (1) point of inflection method (the points where curves change direction), or (2) midpoints of inflection method (arbitrary points located midway between the points of inflection). The points of inflection method is used for measuring the thickness of coal on gamma ray logs, on resistivity logs, and for thin beds on density, neutron, and acoustic velocity logs. It is important to understand that the thickness of coal measured on the short normal and long normal resistivity curves is less than the true thickness of the coal bed by the amount of the electrode spacing; that is, 16 inches less for the short normal curve, and 64 inches less for the long normal curve. The midpoints of inflection method is used to identify the top and base of thick coal beds on the density, neutron, and acoustic velocity logs (see figs. 26, 27, 29, and 30 for examples). In general, the SP, long normal resistivity, lateral resistivity, and induction curves are not suitable for accurate measurement of the thickness of coal beds.
The precision and accuracy of coal bed thickness measurements on geophysical logs are dependent on several factors, such as the speed of logging, the scale at which the log is recorded, the type of log, the type of equipment, and the instrument settings as well as the ability of the user to pick the correct top and base of a coal bed.
Density, neutron, gamma ray, and acoustic velocity logs in oil and gas wells are recorded at speeds of 30 to 60 feet per minute and at a scale of 1 inch equals 20 feet. Such logs generally permit the measurement of the thickness of coal beds within an error of +/-1 foot and allow identification of coal beds as thin as about 2 feet. A suite of logs run in a coal exploratory drill hole at slow speeds of 15 feet or less per minute and recorded at 1 inch equals 10 feet or less, with a standard mineral probe and recording instruments set for coal identification, can measure coal beds as thin as +/-0.5 foot. The use of special equipment and slower logging speeds can improve the precision of thickness measurements.
Resistivity logs of oil and gas wells generally are run at higher speeds of about 100 feet per minute and are recorded at 1 inch equals 50 feet, with the result that coal beds thinner than about 2 feet cannot be identified or measured as to thickness. Some of the focusing electrode logs recorded at 1 inch equals 20 feet may permit greater precision. In contrast with oil and gas logs, logs of the "single point" resistivity type in coal exploration drill holes can provide precision similar to those of radiation-type logs.
Geophysical logs can also provide information on the composition and rank of coal. High-ash or shaly zones (partings) in coal generally are recorded on logs by curves indicating that they are more radioactive, more dense, and less resistive than purer coal. High-rank coal is more dense and has a shorter interval travel time in sonic logs than low-rank coal. Efforts have been made to quantify compositional factors of coal, such as moisture and ash content, using geophysical logs. Bond and others (1971) report that in the Illinois basin an experimental program using computer processed data from logs of coal exploratory drill holes produced excellent determinations of moisture and ash in coal. However, attempts to quantify coal composition using geophysical logs from wells in the Northern Great Plains have had erratic results.
Geophysical logs that have been properly related to a known stratigraphic section can be used to correlate coal beds and determine their structure. In many areas a coal bed and adjacent rocks are recorded on geophysical logs as a unique curve, or sets of curves, that can be recognized on logs of drill holes throughout a large area (see fig. 28, No. 9 coal bed). Recognition of "signature" curves permits correlation of stratigraphic units across any area where the lithology remains reasonably constant and in some places allows correlation throughout a coal basin. Properly correlated logs can provide the necessary data to construct coal bed maps that show structure, thickness of coal, and thickness of overburden. Geophysical logs are also useful in studying the ancestoral sedimentary environments of coal beds. In the Appalachian coal region, Wanless and others (1963) used electric logs of oil and gas wells augmented by drill-hole and outcrop data to identify and map sedimentary environments of some coal-bearing rocks of Pennsylvanian age. Daniels and Scott (1980; 1982) in Kentucky and in Montana used geophysical logs of coal drill holes to refine their interpretations of drillers' lithologic logs in a study of ancestoral sedimentary swamp environments of Pennsylvanian and Tertiary rocks.
Geophysical logs of oil and gas wells are the source of abundant data on coal beds, but the value of any of these logs in any particular coal area or basin depends in large part on a knowledge of the coal-bearing rocks and of how the data are recorded on the curves of the geophysical logs. Consequently, it is advisable for those planning to use geophysical logs for coal interpretation to study the logs of drill holes in an area where the lithology is known to determine the reaction of logging equipment to various rock types. Use of a suite of several types of logs is recommended for the best results. It is important to recognize and discard poor quality or ambiguous logs so that erroneous data is not used in calculating coal resources.
Created by the EERT WWW Staff.
[an error occurred while processing this directive]
