CBM reservoir evaluation

This article discusses many aspects of reservoir evaluation specific to coalbed methane (CBM) wells, from initial data gathering through reserves estimation.

Core analyses

Core analyses are a critical part of analyzing CBM reservoirs to determine gas saturations. Coal cores must be placed in desorption canisters and heated to reservoir temperature. As the coal desorbs, gases are captured, and both their volume and composition are determined. Desorption continues for up to several months until the rate at which gas is being liberated from the coal becomes very small. At this point, the canisters are opened, and the cores can be described. The cores then are crushed in a mill that captures any remaining gas (residual gas), and the milled coal is mixed thoroughly to form a representative sample. Portions of this sample are used for:

• Sorption isotherm measurements
• Proximate analysis
• Ultimate analysis
• Vitrinite reflectance
• Maceral analysis
• Bulk density determination

Table 1 summarizes the various core analyses. An alternative to crushing the entire core is to first slab the core and crush one-half. The uncrushed half can be preserved for additional work including petrographic examination of the core and future coal analyses. The results of these core analyses are critical for both gas-in-place determinations and estimates of gas rates and recovery factors.

• 

  Table 1-Summary of the Various Core Analyses performed on Coal samples.

Log analyses

Because gas is sorbed to the walls of the coal micropores, openhole logs cannot calculate useful matrix porosity or gas saturation values. Nonetheless, logs are still useful for determining the location and thickness of coal seams and estimating their quality. Because of their low density, coals are identified most easily from a density log. They also can be recognized by a combination of other log responses including:

• High apparent neutron and sonic porosities
• High resistivities
• Low gamma ray values

Caliper logs also can be a useful coal indicator because coal intervals are often washed out by drilling operations. Mudlogging can detect coal seams through a combination of:

• Gas kicks
• Lithologic description
• Changes in the rate of drilling penetration

Mudlogging is recommended especially for exploratory or appraisal wells, which may contain unexpected coal seams or other gas-bearing lithologies.

Pure coals are characterized by low values of density and photoelectric effect, whereas ash-rich coals have much higher values. Micrologs can provide a qualitative indication of coal permeability based on the degree of separation between the micronormal and microinverse curves. Caliper logs also may indicate permeability by detecting a thicker mudcake across permeable coals. Newer wells may have access to logs that are more sophisticated, such as

• Geochemical
• Nuclear magnetic resonance
• Borehole imaging logs

These may have to be reprocessed with an emphasis on quantifying the location and properties of coals. These imaging logs are useful for identifying large fractures and thin higher-resistivity shales interbedded with the coals, but these tools do not have sufficient resolution to identify cleats. If a complete log suite is available, sophisticated computer models may be applied to estimate multiple coal characteristics.

Although logs are useful for identifying coal seams and estimating their gross character, coals typically are laminated at a much finer scale than can be resolved with logs. As a result, bright-banded coals with good cleating and high gas contents may be interbedded with ash-rich horizons, giving the appearance of a relatively homogeneous, poor-quality coal on the logs. This is similar to the thin-bedded-pay problem in clastic reservoirs and emphasizes the value of the use of coal cores for reservoir characterization. In addition, pure coals identified from logs may be composed of relatively gas-rich vitrinite macerals or gas-poor fusinite macerals. As a result, some low-density coals may be highly gas-productive while others may not.

Well testing

Buildup tests, injection/falloff tests, and slug tests each have been used successfully to determine critical reservoir and completion parameters in CBM reservoirs. In a buildup test, a well that is producing at a constant rate is shut in, and the downhole pressure is measured as it builds up. In an injection/falloff test, a well that is injecting at a constant rate is shut in, and the downhole pressure is measured as it falls off. In a slug test, a pressure differential is introduced instantaneously across the sandface, and the pressure response is measured. This typically is done by rapidly changing the fluid level in the well.

Slug tests are relatively simple to run and are inexpensive compared with other types of well tests. However, slug tests can be used only in underpressured reservoirs and may not investigate a large reservoir volume. This is an important consideration because CBM reservoirs typically are very heterogeneous, requiring a large radius of investigation to characterize them adequately. Slug test results may be used to design other single or multiple well tests for determining additional reservoir parameters.

Conventional drillstem or buildup tests can be run in CBM wells, but, in many cases, the reservoir pressure will be too low to lift produced water to the surface. This limits the ability to obtain a large radius of investigation with these tests in low-permeability reservoirs. However, reliable test data and results can be obtained if the test is run long enough to reach infinite-acting radial flow. For example, drillstem tests conducted in low-permeability coal seams in the Ordos basin of China were followed by injection/falloff tests that provided similar results.

In an injection/falloff test, it is important to establish communication with all the coal layers before testing. This can be achieved by breaking down the perforations with a small ballout treatment followed by a spinner survey to confirm communication. After allowing the water level in the well to stabilize, injection should commence at a low, constant rate to avoid changing the wellbore stress. This rate should be below the formation-parting pressure to avoid long periods of linear flow that could mask the infinite-acting radial flow regime. The formation-parting pressure can be determined by a step-rate test before the injection falloff test. The maximum acceptable injection pressure should be less than 80% of the estimated parting pressure. After the injection period, the well is shut in and the bottomhole pressure is monitored for a period of time that is usually approximately twice the injection time. A downhole shut-in device may be used to minimize storage effects and reduce the test time. A downhole shut-in device is critical in underpressured reservoirs to avoid problems of falling liquid levels in the wellbore during the falloff period.

Injection/falloff tests are more expensive than slug tests and buildup tests, but they have at least three advantages. First, injection/falloff tests do not require reservoir flow; therefore, they can be run in underpressured as well as normal and overpressured reservoirs. It usually is easier to measure injection rates in an injection/falloff test than it is to estimate flow rates in a drillstem test when fluids are not produced to the surface. Second, injection/falloff tests usually are not affected by complications resulting from gas desorption because the reservoir pressure does not fall below the initial pressure during the test. The reservoir pressure will fall below the initial saturation pressure in the drawdown portion of a drillstem test and may result in gas desorption near the wellbore. Third, injection/falloff tests typically investigate larger reservoir volumes than slug tests or buildup tests, especially in underpressured reservoirs, which cannot flow fluids to the surface.

Well-test permeability is a critical parameter for estimating CBM production rates and ultimate recovery. Fig. 1 illustrates a plot of effective permeability vs. expected ultimate recovery.

• 

  Fig. 1-Plot of effective permeability (from well testing) vs. expected ultimate recovery (from material balance and simulation) for individual wells from a CBM lease in the San Juan basin of New Mexico.

It is important to obtain good permeability estimates from well testing early in the life of each well, preferably before hydraulic fracture stimulation and production. If the well is not tested before fracture stimulation, it may be difficult to run a test long enough to reach infinite-acting radial flow and determine the average value of permeability. Once the permeability is known from a prefracture test, post-fracture tests can be used to determine fracture properties. Wells should be tested before production to avoid two-phase flow during the tests. Slug tests and injection/falloff tests performed before production are most likely to result in a single-phase flow of water. The data from these tests can be analyzed with conventional methods. Although it is easier to analyze data from tests with single-phase flow, it is often important to test wells with two-phase flow. Tests with two-phase flow may be required to track permeability changes over the life of a field because permeability can vary significantly as a function of pressure and gas desorption as a CBM reservoir is produced.

Effective permeability will change during the productive life of a CBM reservoir because of changes in relative permeability as fluid saturations change. Effective permeability also may vary because of changes in absolute permeability as the reservoir is produced. During early production, the coal matrix expands as pore pressure is reduced, resulting in a decrease in absolute permeability. With continued production, the matrix contracts as gas desorbs from the coal, resulting in an increase in absolute permeability. These changes can be tracked over time with pressure-transient testing. The absolute permeability can be calculated from a two-phase test with the use of relative permeability curves derived from simulation, analogous fields, or the published literature. The coal degasification pseudopressure function developed by Kamal and Six can be used to analyze CBM well tests with two phases flowing. The method incorporates sorption isotherm and relative permeability relations. Mavor describes an alternative method to analyze tests with two-phase flow.

Multiple well tests can be used to indicate the degree of communication between wells and to determine permeability anisotropy. Coals typically demonstrate greater permeability in the face cleat direction because these fractures are more continuous and have wider apertures than butt cleats. Directional permeability ratios as high as 17:1 have been reported because of this anisotropy. It is important to understand both the direction and magnitude of permeability anisotropy early in the project life because it can have a significant impact on the choice of well-pattern geometry and orientation and well spacing.

Commercial software or conventional pressure-transient equations can be used to design CBM tests properly. If the permeability range is unknown, the test can be designed for the lowest acceptable permeability that would result in a viable project, usually 1 to 5 md, depending on other factors such as coal thickness, gas content, and initial saturation state. Although coal has a dual porosity nature, most CBM tests can be analyzed with a homogeneous model because all the Darcy flow occurs within the cleat system. Conventional well-test analyses generally are preferred for analyzing data from CBM well tests because they are relatively straightforward. In some cases, however, reservoir and flow conditions do not follow the assumptions on which conventional well-test analysis methods are based. Numerical simulators are useful for history matching well-test data when conventional analyses are inadequate.

Pilot projects

Multiwell pilots are a key element in appraising the potential of a CBM reservoir. A typical pilot consists of several closely spaced wells that are produced for a sufficient period to understand the potential of the reservoir and determine if it can be developed commercially. The key objectives of a pilot are to:

• Quantify variations in key reservoir parameters
  • Net thickness
  • Gas content
  • Gas saturation
  • Permeability
• Assess the ability to dewater the reservoir as indicated by decreasing water rates and reservoir pressures
• Determine gas productivity and the potential for commercial gas rates
• Test completion options
  • Hydraulic fracture stimulation
  • Cavitation
  • Artificial-lift methods
• Evaluate full-field development issues
  • Well spacing
  • Pattern geometry

Choosing the size of the pilot is a critical consideration. The pilot should be large enough to evaluate a representative part of the reservoir but small enough to achieve definitive results in a short period. Generally, pilot wells will need to produce for a minimum of 6 to 12 months at a well spacing of less than 40 acres. Numerical simulation should be used to optimize these values and predict the performance of the pilot on the basis of individual reservoir characteristics. A pattern containing an isolated center well, such as a five-spot, is preferable and can be implemented with a successful appraisal well as part of the pattern. Once the wells are drilled and completed, it is critical to collect high-quality surveillance data on a regular basis including:

• Individual-well water and gas rates
• Flowing bottomhole pressures
• Shut-in bottomhole pressures

A successful pilot will show

• Increasing gas rates
• Decreasing water rates
• Decreasing reservoir pressures with time

If the pilot gas rates are approaching an economic level, the pilot can be expanded to development-scale spacing. If the gas rates are increasing but clearly subeconomic, the pilot can be expanded at the current well spacing to a nine-spot or other configuration to minimize water influx and assess whether economic rates will be achievable. If water and gas rates are low, the initial pilot wells can be produced for a longer period, or the project can be terminated. Before making any decision, it is critical to reconcile pilot well performance with core, log, and well-test data, preferably through reservoir simulation. This work will ensure, for example, that a good CBM reservoir is not being abandoned because of poor well completions.

A staged piloting approach is the best way to minimize the time and cost of evaluating a CBM reservoir. In a frontier area, multiple pilots may be needed to prove up a large enough area to declare commerciality and obtain gas sales contracts. The number of appraisal wells that should be drilled and offset by additional wells to create pilots must be determined. If several widely spaced appraisal wells indicate similar reservoir properties, a single pilot may be sufficient to decision a large area. However, if properties vary dramatically, multiple pilots may be needed. A good approach is to drill the first pilot wells around the most prospective appraisal well. If this pilot is unable to produce gas at economic rates, then it becomes unlikely that additional pilots will be successful, leading to an early exit from the project. Alternatively, if the first pilot is successful, management will be enthusiastic about expansion and additional investment. Unfortunately, management often loses interest in a CBM prospect because of the multiyear time commitment, the money required to reach a decision point, and because the technical staff often does not have a clearly defined evaluation and exit strategy.

Numerical simulation studies

Because of their layered, fractured, and heterogeneous nature, CBM reservoirs are very complex. Reservoir properties can vary rapidly, and many variations are difficult to quantify. Some of these properties, such as porosity and gas saturation, must be determined from sources such as cores, analogous reservoirs, and correlations rather than from wireline logs. Other properties, such as compressibility and gas storage capacity, are difficult to measure in the lab and can range over several orders of magnitude. Additional complications include fluid contributions from noncoal layers and the likelihood of strong directional permeability trends.

The impacts of these reservoir complexities are best resolved with numerical reservoir simulation. The advantages of numerical simulation include

• The ability to integrate widely different data types
  • Reservoir data
  • Completion data
  • Well performance data
• Help resolve data discrepancies and provide key insights into production mechanisms
• Incorporate unique components
  • Gas storage
  • Diffusion mechanisms
• Understand and revise the geologic model including estimates of aquifer size and strength
• Evaluate development options
  • Well spacing
  • Well pattern
  • Fracture design
• Provide a reasonable basis for rate and reserve estimates

However, because of their complexity, rate and reserve forecasts for CBM reservoirs are generally less certain than the forecasts for conventional oil and gas reservoirs. Perhaps the most valuable use of CBM simulation is to evaluate the effects of variations in key parameters. Because of the difficulty in quantifying the areal and vertical variation of every CBM parameter, simulation can be used to test the impact of various parameter combinations on overall reservoir performance.

In addition to the usual data types required for numerical simulation, CBM simulation requires gas content values at initial reservoir conditions, sorption isotherms, the diffusion coefficient, and parameters to estimate changes in absolute permeability as a function of pore-pressure depletion and gas desorption. Because these properties may vary significantly, it is critical to have representative core and well-test data from each coal seam. To obtain a meaningful history match, high-quality surveillance data must be obtained from producing wells at regular time intervals. Water and gas production data can be obtained easily, and flowing bottomhole pressures can be estimated from fluid levels in a pumping well. Numerical simulation can be used to determine how often shut-in pressures should be obtained, and these often can be measured in conjunction with well work or other planned shut-in periods.

Gas-in-place determination

Gas in place in a CBM reservoir consists of free gas residing in the cleat system plus the gas that is sorbed onto the surface of the coal.

G=A•h•43,560φcl(1-Swi)Bgi + 1.36Gcρc(1-fa-fw), where

G	=	gas in place, Mscf,
A	=	areal extent, acres,
h	=	net coal thickness, ft,
φcl	=	cleat porosity, fraction,
Swi	=	initial water saturation fraction in the cleats, fraction,
Bgi	=	initial gas formation volume factor, Mscf/ft3,
Gc	=	gas content (dry ash-free basis), scf/ton,
ρc	=	coal density (dry ash-free basis), lbm/ft3,
fa	=	ash weight fraction, lbm ash/lbm coal, and
fw	=	water weight fraction, lbm water/lbm coal.

Generally, coal thickness is estimated by counting those intervals with a bulk density of less than 1.75 g/cm³. Larger density cutoff values are sometimes used, but this requires that a lower average gas content be used as well. Coal density values can be determined from a density log or from core measurements.

Gas content values are obtained from coal core-desorption measurements corrected for lost gas and residual gas. Values of several hundred scf/ton are quite common, although values can range from less than 10 to more than 1,000 scf/ton in high-rank coals. The ash fraction is derived from proximate analysis of desorption-canister samples and ranges from a value of zero in pure coals to a value of one in mudstones. The water weight fraction is also obtained from proximate analysis and ranges from less than 0.05 in medium-volatile bituminous coals to more than 0.5 in subbituminous coals. Cleat porosities, which are difficult to measure, typically are assigned values ranging from 0.01 to 0.05. Initial water saturation in the cleats is generally assumed to be 1 unless the cleats contain free gas.

Determining accurate coalbed reservoir gas-in-place parameters is often a difficult, time-consuming process, and the resulting estimates must be revised many times as additional wells are drilled and more information becomes available. Because few wells are cored and coal recoveries generally will be less than 100%, the existing coal samples may not be representative of such a heterogeneous reservoir. Lost-gas corrections, especially if they are large, often yield erroneous gas content values. Density logs do not capture the fine-scale variability of coal seams and typically are only a rough approximation of the net coal thickness that is actually contributing gas. Several references exist to guide this work including a comprehensive publication by the Gas Research Institute (GRI).

Reserves determination

Estimates of remaining CBM reserves are commonly made throughout the life of a project. These estimates begin with qualitative values generated before drilling appraisal wells and extend to quantitative reserve numbers based on the production history of development wells. During the initial screening stage, parametric studies provide a means to relate values of key reservoir parameters generically to recovery factors. As the first wells are drilled, data from analogous producing fields can be used to estimate the potential gas reserves of a new asset if their reservoir characteristics are similar.

Once coal cores have been cut and analyzed, the measured gas content and sorption isotherm data can be used to estimate a recovery factor and reserves. The gas content indicated by the isotherm at an assumed abandonment pressure is subtracted from the total gas content of the core at current reservoir pressure. This value, divided by the total gas content of the core, provides an estimated recovery factor (Fig. 1). This calculation assumes that the permeability and coal-seam continuity are sufficient to achieve the expected abandonment pressure in an economic time frame. This is a critical assumption because the shape of the sorption isotherm curves dictates that most of the gas is produced at low-pressure values. A more sophisticated tool for estimating reserves at this stage is numerical reservoir simulation, which can be used to determine whether the assumed abandonment pressure is realistic for the expected range of permeabilities.

• 

  Fig. 1—A sorption isotherm curve showing how a recovery factor is calculated.

Estimates of gas in place and the recovery factor may be improved after multiple wells have been completed with the use of a modified material-balance technique. This approach requires:

• Substantial production and reservoir pressure data
• Estimates of gas saturation
• Estimates of effective porosity
• Estimates of formation compressibility
• Estimates of water influx

The technique uses a conventional material-balance equation modified to account for gas desorption from the coal seams.

Decline-curve analysis can be used in the mid-to-late stages of the field’s producing life. This method is dependent on production data and implicitly assumes decreasing gas production in its forecast. The increasing gas rates that characterize early CBM reservoir behavior preclude the application of decline-curve analysis during this period. However, if the reservoir data, completion type, and early production character of a new well are similar to a mature well in the area, then the profile from this mature well may be used to estimate the performance of the new well. Some operators use mature-well data to create a series of type curves for estimating the performance of new wells.

Nomenclature

G	=	gas in place, Mscf,
A	=	areal extent, acres,
h	=	net coal thickness, ft,
φcl	=	cleat porosity, fraction,
Swi	=	initial water saturation fraction in the cleats, fraction,
Bgi	=	initial gas formation volume factor, Mscf/ft3,
Gc	=	gas content (dry ash-free basis), scf/ton,
ρc	=	coal density (dry ash-free basis), lbm/ft3,
fa	=	ash weight fraction, lbm ash/lbm coal, and
fw	=	water weight fraction, lbm water/lbm coal.

Noteworthy papers in OnePetro

Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read

External links

Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro

See also

Coalbed methane

CBM reservoir fundamentals

CBM basin assessment

CBM case studies

PEH:Coalbed_Methane

Page champions

George J. Koperna, Jr.