GEOCHEMICAL LOGS THOROUGHLY EVALUATE COALBEDS

David J. Johnston
Schlumberger Well Services
Denver

There are many advantages for using wire line logs in coal seams. Not only can they tell the presence and thickness of a coal seam, but a full coal seam analysis can be made if geochemical logs are used.

Specifically, geochemical logging data can yield a coal's mineral matter, fixed carbon, volatiles, macerals (organic constituents), cleating, relative flow rate, and various mechanical properties. All of these data improve correlating coal seams, reservoir analysis, and completion designs for coalbed-methane wells and coalmining operations.

With some knowledge of a basin's own unique coals, coal response parameters from wire line can be determined and used to give a proximate analysis and gas content for that basin.1 For coal, a proximate analysis determines the moisture, volatile material, ash, and fixed carbon content.

Wire line logs (traditionally density, neutron, and resistivity logs) have been used to identify and determine the thickness of coal seams. Coal responses are easily distinguished on these logs because:

The low density of coal will give a high density-porosity reading.
The high hydrogen content of coal will give a high neutron-porosity reading.

Fig. 1 shows the coal responses for these logs. In the depth track is a flag that identifies the coal seams.

The gamma ray, usually logged in combination with one of these logs, is presented in Track 1 as the solid curve. In coals, the gamma ray response usually is low unless there is a large amount of radioactive minerals in the coal.

A good example of a change in the volume of the radioactive minerals is the coal seam from 2,900 to 2,880 ft. Also presented in Track 1 is the caliper.

In Track 2, the resistivity is displayed. Notice that in all of the coals, the resistivity is greater than the surrounding beds. Track 3 shows density and neutron responses.

Some literature even describes how wire line logs (specifically microlog, digital sonic, and spontaneous potential measurements) actually respond to the permeability of a coal seam.2 3 However, this approach is at times inaccurate because of varying hole conditions (mud properties and hole size), drilling damage to coals at the borehole wall, and the elapsed time before logging.

Because coal permeability is a function of its cleats (a joint or system or joints along which a coal fractures), the best way to analyze coal permeability is with logs that identify the intensity of the cleats. This is one of many outputs gained by adding the geochemical log to the wire line suite used in coal examination.

With the addition of geochemical logging, with a geochemical logging tool or a spectral gamma ray logging tool,4 5 a more detailed analysis of a coal seam can be made. This analysis will be basin independent.

Such analyses yield a coal's mineral composition, rank, maceral identification, and creating information. From these data, flow estimates and mechanical properties can be calculated.

Also, the correlation of coal seams and their surrounding beds becomes easier. Finally, a gas content calculation can be determined from any equation using macerals, mineral matter, and coal rank as input.

DATA ACQUISITION

A geochemical logging tool measures multiple formation elements such as silicon, calcium, sulfur, iron, hydrogen, aluminum, potassium, thorium, and uranium. The tool also can measure carbon and oxygen.

In the case of Schlumberger's geochemical tool, this is done in a second operating mode.6-9

Because coal contains carbon, oxygen, hydrogen, nitrogen, and sulfur, the geochemical logging tool is a good choice for evaluating coal beds.10-13

Coals can be classified into seven ranks (Table 1)14 depending upon their carbon oxygen measurements.

As a coal's rank changes, the percentage of carbon increases while the percentage of oxygen, and hydrogen decrease.15

A coal's composition also is dependent upon its maceral makeup.

A coal's mineral matter (ash materials) can be determined by combining measurements from a geochemical log with those from lithology-density, neutron, and resistivity logs. If mineral matter is known, then the log data can be corrected to give a pure coal (mineral-free or dry ash-free, abbreviated as DAF) reading.

BOREHOLE EFFECTS

Data corrections often are necessary. For example, when using traditional wire line data to evaluate coal seams, the borehole effects on the log must be taken into consideration.16 Also, heavy muds with barite greatly affect the log response because barite causes the photoelectric cross section (Pef) measurement to be too high in invaded zones. This can occur when a coal is cleated (fractured, heavily jointed) or washed out.

Borehole effects also must be removed from geochemical logs.17 Data corresponding to hole rugosity or washouts greater than 4-6 in. should be examined closely to see if they are usable.

Borehole corrections also must be made to geochemical logs taken in air-filled holes. To date, however, the needed corrections have not been quantified, although research is promising.

THIN BEDS

If thin beds (less than 3-ft thick) are suspected, an inelastic station should be taken in addition to a continuous pass with a geochemical logging tool. High-resolution density logging (1.2 to 1.0-in. sample rate) provides good vertical resolution in a coal seam and shows the layering that could occur.

DATA ANALYSIS

As previously stated, coal seams have unique log responses that make it easy to identify a coal on the raw data.

Table 2 shows expected coal responses on wire line logs. Coals also have some unique mineral-matter log responses that can be used for identifying the mineral composition of ash, as listed in Table 3. After acquisition, geochemical logging data for coal evaluation are entered into an elemental analysis program for determining mineral matter, fixed carbon, volatiles, macerals, creating factor, and relative flow rate. Mechanical properties also are solved for.

ELEMENTAL ANALYSIS

The required data inputs, from a geochemical logging tool, into the elemental analysis program for determining mineral matter, fixed carbon, and volatiles present are:

Density
Neutron measurements
Resistivity
Photoelectric factor, P,f
Potassium
Silicon (capture yield)
Calcium (capture yield)
Titanium (capture yield)
Iron (capture yield)
Sulfur (capture yield)
Aluminum (capture yield)
Carbon (inelastic yield)
Oxygen (inelastic yield)
Sigma.

Eight minerals as well as fixed carbon, volatiles. and porosity can be identified by the elemental analysis program for coal using data from the geochemical, spectral gamma ray, gamma-gamma density, compensated neutron, and induction logs. The identifiable minerals are quartz, calcite, siderite, pyrite, orthoclase, illite, kaolinite, and smectite.

Ideally, the program used will allow modification based upon the desired mineral, and will handle the simultaneous computation of multiple models.18 Figs. 2 and 3 illustrate outputs from an elemental analysis program for coals.

Fig. 2 is a solution for generic coal, the mineral content of the coals, and the minerals and fluids of the surrounding beds. There are two models presented. Model 1, next to the depth track, is used only for the surrounding beds.

Model 2 is the solution for the coal seams.

The type of minerals and fluids found for each model are listed.

Results from the coal model show the total volume of minerals found in each coal seam can be different and that type of minerals that are found in each coal can vary. The standard-deviation-results (SDR) curve, presented next to each model, is the confidence the program had in its solution. Excursions to the left are areas of less confidence.

Fig. 3 is the solution for the coal rank. It uses the outputs from Fig. 2 and solves for fixed carbon and volatiles. Model 1, next to the depth track, is used only for the surrounding beds and is identical to Model 1 in Fig. 2. Model 2 is the solution for the rank in the coal seams.

MACERALS

Elemental analysis to determine macerals also is conducted. Coal evolves from plant material partially decomposed in a swamp-like environment. Each part of the original plant material has different chemical compositions, which can result in a different chemical composition for coal seams otherwise having identical rank and ash.

Consequently, coals are subdivided into three types of macerals (organic content), based upon their reflectivity and form.

For example, vitrinite is the maceral that is composed primarily of woody tissues, bark, and leaves. Inertinite is composed of plant material that was partially oxidized before coalification, while liptinite forms from the waxy part (resins) of a plant. The chemical composition for each maceral is listed in Table 4 for different rank coals.19-21

The three maceral groups are solved for by first adjusting the density, neutron, carbon, and oxygen inputs for mineral matter content. These corrected data represent the dry, ash-free (DAF) coal responses and are used as input into an elemental analysis program modified for a maceral model.

Fig. 4 shows the results from a maceral model program run. In Track 1 the gamma ray and the dry ash free density of the coal are presented. In Track 2 is the combined model results of Fig. 3. In Track 3 is the maceral results. In this example the major maceral identified was vitrinite. Small amounts of inertinite were present, and no liptinite was found.

CLEATING

Mineral matter and maceral data from a given coal seam are used to determine a creating factor. The key to defining cleats is a combination of secondary minerals (kaolinite and calcite) and the macerals that are present.

It has been determined through previous work that macerals are important to the coal seam's ability to cleat. That is, if the "correct" type of macerals are not present, then the coal is unlikely to be cleated. It also has been determined that creating is unlikely if too much clay or mineral matter is present in a coal.

Cleating results are presented as types of cleat facies, as follows: very, partly, poor, high ash, high clay.

Fig. 5 is the result of a cleat-analysis program run. This figure is the same as Fig. 4 except for a fourth track that shows the cleat ratings. From this figure it can be determined that the "best" coal in terms of its cleats is from 2,900 to 2,880 ft. The "worst" coal is the deepest coal.

FLOW RATE

From the creating factor, it is possible to determine a relative flow rate. Flow rate can be calibrated if the pore pressure for each individual seam is known. This information can be obtained from drill stem tests, a formation-tester wire line tool, or any other means by which formation pressure can be measured.

Flow-rate logic includes a simple assumption. That is, as cleat frequency increases, the permeability of the coal increases; therefore, flow rate increases.

Currently, only three cleat ratings are used to determine flow rates, each rating contributing to the flow rate at a different level.

Fig. 6 displays the results of a flow rate program run. Track 3 presents the estimated flow rate.

The predicted flow rate for the lower two seams is about 330 MMcfd. The actual production from these two seams is approximately 250 MMcfd. For the upper two seams, the predicted production is 170 MMcfd. Total estimated flow from all of the coals is then 500 MMcfd. The upper seams have not been completed.

MECHANICAL PROPERTIES

To determine mechanical properties of a coal seam, sonic log data must be added to the mineral matter calculation derived from the geochemical logging data. Specifically, given mineralogical data from the geochemical analysis and sonic data, which yields compressional time, a shear time (m sec/ft) can be calculated by using published shear-to-compressional ratios (Table 5) for the minerals (or by directly measuring shear)13 and assuming a ratio value for fixed carbon and volatiles (estimated from historical core data).

Having derived compressional and shear data, Poisson's ratio (the ratio of transverse strain to longitudinal strain) can be calculated. The next step uses cleat facies identifications to determine if a coal is fractured. With this knowledge, a Poisson's ratio reduction can be calculated. Fig. 7 is a display of the results.

With an accurate Poisson's ratio, a thorough mechanical description of a given coal reservoir is possible. For example, a fracture height prediction can be calculated for each of the coal seams given Poisson's ratio, pore pressure, and fracture fluid weight. By knowing fracture height, an accurate stimulation design can be made.

FUTURE DEVELOPMENTS

In addition to developing the necessary borehole corrections that will make it possible to use geochemical logging in air-filled boreholes, work is ongoing in the area of coal gas-content calculation.

Reports already have been published indicating that different coal macerals generate different amounts of methane gas and have different capacities for sorption.4 22

The gas content for a coal seam is commonly calculated using mineral matter content, the maceral composition, and the hydrogen-to-carbon ratio for the dry, ash-free coal. (Levine's equation is one using these inputs to arrive at gas content.23

It is hoped that this type of gas content calculation will prove viable in any coal basin.

Further, it appears that gas saturation of a coal should be useful for determining gas adsorbed in a given seam. This, too, is being investigated, and hopefully will result in a gas saturation relationship that is measurable with wire line techniques.

ACKNOWLEDGMENT

The author thanks the Gas Research Institute for permission to present data from one of its funded wells.

REFERENCES

Howard, P., "Log Analysis of Coalbed Methane Wells in San Juan Basin," Schlumberger GFE paper, 1985.
Ludwick, L., and Bratton, T., "Application of Stoneley Permeability in the Four Corners Area," Schlumberger North Region Interpretation Symposium, 1989.
Mullen, M., "Log Evaluation in Wells Drilled for Coalbed Methane," Coalbed Methane, San Juan Basin, Rocky Mountain Association of Geologists, 1988.
Schlumberger publication, Geochemical Logging, SMP-9120, 1988.
Gilchrist, W., et al., "Application of Gamma Ray Spectroscopy for Formation Evaluation," SPWLA 21 st Annual Logging Symposium, July 6-9, 1982.
Hartzog, R., "Laboratory arid Field Evaluation of an Inelastic-Neutron-Scattering and Capture Gamma Ray Spectroscopy Tool," SPE Paper No. 7430, 1978.
Westaway, P., et al., "The Gamma Spectrometer Tool Inelastic and Capture Gamma-Ray Spectroscopy for Reservoir Analysis," SPE Paper No. 9461, 1978.
Gilchrist, W., et al., "Carbon/Oxygen Interpretation-A Theoretical Model," SPWLA 24th Annual Logging Symposium, June 27-30, 1983.
Roscoe, B., and Grau, J., "Response of the Carbon/Oxygen Measurement for an Inelastic Gamma Ray Spectroscopy Tool," SPE Paper No. 14460, 1985.
Chapman, S., et al., "The Emergence of Geochemical Well Logging," Nuclear Geochemistry, Vol. 36, No. 2, 1987, pp. 27-35.
Herron, M., and Herron, S., "Geological Applications for Geochemical Nell Logging," British Geological Society's meeting Geological Applications of Well Logs, June 1 2, 1988.
Herron, M.. "Future Applications of Elemental Concentrations from Geophysical Logging," Nuclear Geophysics, Vol. 1, No. 3, 1987, pp. 197-211.
Ellis, D., et al., "Mineral Logging Parameters: Nuclear and Acoustic," The Technical Review, Schlumberger, January 1988, pp. 38-52.
Practical Handbook of Physical Properties of Rocks and Minerals, Table 92, p. 104.
Practical Handbook of Physical Properties of Rocks and Minerals, pp. 102-9.
Log Interpretation Charts, Schlumberger, SMP-7006, 1989.
Schweitzer, J.S., et al., "Gamma Ray Spectroscopy Tool Environmental Effects," SPE 11144, 1982.
Quirein, J., et al., "A Coherent Framework for Developing and Applying Multiple Formation Evaluation Models," SPWLA 21st Annual Logging Symposium, June 913, 1986.
Dyrkacz, G.R., and Horwitz, P., "Separation of Coal Macerals," Fuel, Vol. 61, January 1982.
Dyrkacz, G.R., et al., "Fourier Transform Infrared Study of High-Purity Maceral Types," Fuel, Vol. 63, April 1984.
Practical Handbook of Physical Properties of Rocks and Minerals, Table 93, pp. 105-6.
Telle, W.R., and Thompson, D.A., "The Cahaba Coal Field: A Potential Coalbed Methane Development Area in Central Alabama," SOMED, The University of Alabama, Research Report 88-1, March 1988.
Levine, J., "Influence of Coal Composition on the Generation and Retention of Coalbed Natural Gas," Coalbed Methane Symposium, Tuscaloosa, Ala., Nov. 1619, 1987, pp. 15-18.