Professional Documents
Culture Documents
Chapter 6 - Resource Evaluation Methodologies - 2014 - Coal and Coalbed Gas
Chapter 6 - Resource Evaluation Methodologies - 2014 - Coal and Coalbed Gas
6
Resource Evaluation Methodologies
O U T L I N E
KEY POINTS
• Understanding coal resources and methods adsorption isotherm of gas to modify and
of assessments are important to planning improve the calculation.
coalbed gas projects. • An important parameter for GIP calculation
• Coal resource and reserve estimates are is the gas content of coal, which can be
based on geologic assurance determined by collected from wireline cores and cuttings.
the distance between observation points • However, gas content is measured from
where coal thickness is measured and desorption of gas from coal cores or cuttings
samples are collected for analysis. in sealed canisters by direct and indirect
• Increasing geologic assurance with methods.
closer distance between observation • The direct method is based on the analysis
points results in a progression of that gas released during desorption was
hypotheticalespeculative resources proportional to the square root of time.
(farthest spacing) to inferred, indicated,
• The indirect method is based on analysis of
and measured reserves (least spacing),
methane adsorption isotherms, which
respectively.
measure the optimum adsorbed gas at
• Geologic assurance-based methodology varying pressures and at constant
relies on the stratigraphic correlation of the temperature. Use of wireline or geophysical
same coal bed. logs (e.g. gamma ray, density, resistivity,
• Global assessments of coal resources and sonic, neutron) is another indirect method
reserves universally conform to the geologic of measuring gas content.
assurance method, but may be modified • Another specialized measurement of gas
from country to country according to depth, content of coal is the use of decline curve
distances between observation points, method.
thickness, and complexity of the geology
• Gas production data analysis based on
and structures of study areas.
estimate of ultimate recovery (EUR) and
• Coal resourceereserve methodology does spacing of coalbed gas wells has been
not directly correlate to the petroleum adopted from the petroleum industry for
industry’s method of estimating coalbed estimating reserves.
gas resourcesereserves.
• Case studies of Total Petroleum System
• Methods of calculation of coalbed gas (TPS) assessment of coalbed gas in the
resources and reserves adopted by the Powder River Basin, Wyoming and North
petroleum industry include: volumetric, Slope, Alaska, are reviewed for
material balance, production data analysis, production data analysis and analog
reservoir simulation, and analog methods. methods.
The volumetric and production data
• The Petroleum Resources Management
analysis are commonly used methods.
System guidelines have proposed
• Volumetric or gas-in-place (GIP) method is standardization of coalbed gas
commonly used and calculated by gas resourceereserve categories to contingent
drainage area, coal thickness, coal density, resources and proved, possible, and
gas content, and other parameters such as probable reserves based on well spacings.
coal quality, coal reservoir conditions, and Proved reserves are subcategorized into
THE METHODOLOGY CONUNDRUM 303
determine accurate coal and coalbed gas volu- undiscovered recoverable coalbed gas resources
metrics. Because coalbed reservoirs are internally are anticipated as a result of addition of more
heterogeneous and beds are highly variable in data (e.g. coalbed gas wells, gas content, gas pro-
lateral and vertical extent their thickness and dis- duction, stratigraphic control). Thus, as more
tribution can change rapidly over short distances. information becomes available and applied to
This fact prevents capture of the exact representa- assessment methodologies, it is expected that
tion of the geometry of the coal body and the pre- the estimates of undiscovered recoverable
cise establishment of the coal volume and proper coalbed gas will continue to vary through time.
application of coal and coalbed gas volumetric Overprinted on the local and regional varia-
calculations. Thus, coalbed gas volumetric esti- tions of coalbed gas calculations is the employ-
mates (e.g. total gas content in standard cubic ment of different assessment methodologies
meter) must account for rapid variations laterally utilized within a country and worldwide be-
and vertically within a coal reservoir in a well, tween countries. Thus, the magnitude of vari-
between wells, within a gas drainage area and ability of coalbed GIP estimates by a factor of
field, and within a basin (Stricker et al., 2006). three from various coal basins worldwide differs
The volumetric estimate of coalbed gas not due to a variety of assessment methodologies uti-
only depends on the assessment methodology lized from country to country. Boyer and Qingz-
but also on the quality, quantity, and recency of hao (1998) reported that the estimates of coalbed
the data. For example, estimates of the volume GIP or volumetric for major coal-bearing basins
of coalbed gas in the Powder River Basin in in the world ranges from 89 to 269 Tcm based
Wyoming and Montana, United States, have on methods developed in the United States.
changed several times from 1984 during prede- However, there are various methods of assess-
velopment to the 2004 during peak development. ments developed in the United States, which
Choate, Johnson, and McCord (1984) estimated vary from State to State, industry to Federal
the volume of coalbed gas from the Fort Union agencies, and Federal agency to agency. Thus,
coals in the Powder River Basin based on 1 trillion understanding the concepts behind specific
metric tons of coal resources calculated from oil assessment methodologies is critical to under-
and gas, geophysical logs, and average gas con- standing coalbed gas resources. According to
tent of 0.70 m3 per ton of coal as 0.85 Tcm GIP. Boyer and Qingzhao (1998) assessment method-
ARI (2002) indicated that estimates of coalbed ologies frequently exclude application of
gas resource in the Powder River Basin ranged resource estimation methods for conventional
by a factor of five between assessments of the hydrocarbons. The suggestion of applying
U.S. Department of Energy, U.S. Bureau of Land assessment methodology for conventional gas
Management, and the USGS. These federal by Boyer and Qingzhao (1998) to coalbed gas
agencies estimated undiscovered recoverable has been supplanted by reclassification of
coalbed gas from 0.22 to 1.10 Tcm, which is coalbed gas, shale gas, tight gas sand, and gas hy-
explained by different methodologies, geologic drates as unconventional hydrocarbons (Chapter
models, and resource assumptions (ARI, 2002). 2), which deserve a separate methodology of
The USGS estimated technically recoverable resource assessment (Schmoker, 2002, 2004;
coalbed gas using petroleum system methodol- Schmoker & Dyman, 1996).
ogy and estimate of ultimate recovery from pro- The assessment concepts advocated by Boyer
duction history of more than 1000 coalbed gas and Qingzhao (1998) significantly differ from
wells in 1999e2000 as about 0.35 Tcm for the that proposed by Schmoker and Dyman (1996)
Fort Union coals in the Powder River Basin and Schmoker (2002). Boyer and Qingzhao
(Flores, 2004). The differences in estimates of (1998) supported the used of volumetric and
THE METHODOLOGY CONUNDRUM 305
material or mass balance methods for assess- continuous gas reservoirs, in general, and
ments. The volumetric method or “macroassess- coalbed gas reservoir, in particular, as displayed
ment of the total GIP” employs the simple empirically by production wells and reservoir-
formula combining the volume of stored gas simulation models. In this method based on
per unit volume of coal reservoir with total reser- production characteristics in contrast to GIP or
voir volume (e.g. amount of coal resources). The volumetric, is the basis for forecasts of potential
material balance method utilizes the law of con- additions to reserves. This method is to some
servation of matter formulas combining reser- extent similar to the production-data-analysis
voir pressure conditions, gas adsorption/ method described by Clarkson and Barker
desorption analyses, and porosity/permeability (2011) that uses production type curves and
properties, which requires the knowledge and flowing material balance. Although Clarkson
understanding of the characteristics coalbed res- and Barker (2011) considered the geologic
ervoirs (King, 1990; Penuela, Ordonez, & assurance-based approach to estimate re-
Bejarno, 1998; Walsh, 1995). sources/reserves by the coal mining industry,
In contrast, the methodology proposed by the authors utilized the petroleum industry
Schmoker and Dyman (1996) and Schmoker well spacing approach with the premise of
(2002) is a knowledge-based assessment of Total continuous coal thickness within the gas
Petroleum System (TPS), which was adopted by drainage area. Rushing, Perego, and Blasingame
the USGS for world oil and gas resources and re- (2008) used reservoir simulation of constant
serves. The modification of the concept toward flowing pressure to determine the impact of
assessment of coalbed gas as unconventional coal reservoir properties on decline behavior of
gas system required recognition of coal beds as production wells.
a continuous accumulation, which occurs inde- Integrated with the various approaches is the
pendent of the water column and/or owe their probabilistic reserves assessment that uses for
accumulation to buoyancy of gas in water example a Monte Carlo method, which incorpo-
(Schmoker, 2002). In addition, as typified by con- rates estimates of well decline and recovery,
ventional hydrocarbon accumulation, unconven- volumetric, material balance, well pressure anal-
tional gas accumulation does not correspond to ysis caused by production, and other reservoir
a field or pool defined by a down-dip water con- parameters. The Monte Carlo simulation method
tact (Schmoker, 2002). As such, the customary for the reliability assessment of undiscovered en-
resource-assessment methods for conventional ergy resources and reserves is commonly used
gas accumulation cannot be used and a special- worldwide (Charpentier & Klett, 2000; PRMS,
ized method is required for the unconventional, 2011; Sarmiento & Steingrimsson, 2007; Stoltz &
continuous gas accumulations such as coalbed Jones, 1998; Swinkles, 2010; Yongguo, Yuhua,
gas, shale gas, tight gas, and gas hydrates. This Yong, & Qiuming, 2008). The probabilistic
chapter will focus on methods specific to coalbed method employs reservoir parameters to calcu-
gas. The approach suggested by Schmoker (2002) late the statistical uncertainty of GIP volumetric
requires an estimate of total GIP connected with a and recoverable resource volumes (Swinkles,
recovery factor to focus on the assessment from 2010). Stochastic method such as the Monte
total volumetric to prediction of additional undis- Carlo equations is applied to generate probabil-
covered gas reserves. The GIP method relies on ities, which lead to various quantitative
the measurement of the in situ gas content and risk analysis and decision-making methods
the amount of coal within an area. (Swinkles, 2010). Probability levels of the total
Another approach proposed by Schmoker recoverable gas volume can then be related to
(2002) is based on the production history of 1P, 2P, and 3P resource/reserve categories.
306 6. RESOURCE EVALUATION METHODOLOGIES
Barker (2008) applied the 1P, 2P, and 3P cate- coal is usually reported as original resources
gories for coalbed gas assessment. A case study that are in the ground prior to mining produc-
by Yongguo et al. (2008) shows that results of tion. The overburden thickness or rock interval
resource/reserve estimates using the Monte from the top of the coal to the surface is an
Carlo method have a narrower error range important factor in the economic extraction or
compared to the volumetric method. The uncer- mining of coal. However, coalbed gas extraction
tainty analysis is a key to lessening the high po- is not encumbered by depth depending on dril-
tential risks of economic assessment of coalbed ling/well costs as long as permeability exists in
gas. According to Moore (2012) the reserve cat- the coal reservoirs.
egories play an economic viability and commer-
ciality of the coalbed gas prospects. Coal Resource Classification System
The following methods have been adopted for
unconventional or continuous accumulations of A comprehensive coal resource classification
coalbed gas: (1) GIP or volumetric, (2) material includes identification of the areal coverage,
or mass balance, (3) production performance or spacing of data points, thickness of coal and
history, and (4) reservoir simulation. According overburden, density, rank and quality, and esti-
to Clarkson and Barker (2011) these methods mate of the quantity of coal. All these parameters
are applied during various stages of coalbed are important to coal development. However,
gas development. The application of these the most important parameters for coalbed gas
methods depends on the availability of data, development are the areal coverage, thickness,
which in turn, controls the accuracy of the density, and rank of coal. The coal resource clas-
assessment methodology. Generally, the GIP or sification system is intended to quantify total
volumetric method is applied during the early amounts of coal prior to and after mining. In
stage although it can be used throughout the addition, the classification system is intended
development period. The material balance, pro- to identify coal quantities that are (1) known,
duction history with decline curves, and reser- (2) undiscovered, (3) economically recoverable,
voir simulation can be applied when data on (4) potentially recoverable, and (5) subeconomic
production, permeability and porosity proper- resources at varying distances from data points
ties, and flowing and shut-in pressures are of measurements of coal thickness.
available. The key to the classification of coal resources
and reserves, which includes measured, indi-
cated, inferred, and hypothetical, is the distance
COAL RESOURCES VS GAS of data points from which coal thicknesses are
RESOURCES measured varying from close to far apart, respec-
tively. The distance of data or control points
Estimates of coalbed gas resources and re- demonstrates the perceptible continuity of a
serves are rooted in the understanding of coal bed between measurement points assuming
methods of assessment of coal resources and correlation of similar physical characteristics and
their classification. More specifically calculations stratigraphic position of the bed. Another factor
of the GIP or volumetrics of coalbed gas are that contributes to the reliability of resource
directly related to the mass of the coal. Coal re- and reserve classifications is the distribution of
sources are the total amount or tonnage (e.g. points. Often times coal resource and reserve es-
short or metric tons) of coal in the ground within timates are performed without regard of strati-
the limits of thickness, lateral continuity, and graphic correlation of the same coal and
overburden thickness or depth. The amount of enclosing beds. The degree of confidence in
COAL RESOURCES VS GAS RESOURCES 307
correlation increases with the knowledge and the compiled data and calculations of mapped
abundance of stratigraphic control points (e.g. and correlated individual coal beds. In the
oil and gas wells, coal drill holes, outcrops). absence of well-established correlations of coal
This is termed geologic assurance or the condi- beds, coal resources calculations are often per-
tion of certainty and confidence of the occur- formed in the form of bulk volume of coal, which
rence of a quantity of resources based on the is not an accurate representation of the quantity
data and spacing of control points from where of coal in an area.
the coal thickness is measured, described, and
sampled. However, the degree of geologic assur- Coal Resource Categories vs. Reliability
ance is controlled by the quantity and quality of
Estimates
the geologic data; that is, the more available data
the closer the spacing of data points and more Coal resources are classified according to the
reliable the coal resource estimates. Once the abundance and reliability of data used to perform
data points indicate that the stratigraphic corre- the resources estimates. The USGS provides the
lation and continuity of a coal bed and enclosing methods that are universally used for reliability
rocks is possible between data points within an estimates of coal resources, which often are used
area, then an estimate of the coal resources and as a guide (Averitt, 1974; Wood, Kehn, Carter, &
reserves of that coal bed can be calculated within Culbertson, 1983). Figure 6.1 demonstrates the
that area. Thus, a total coal resource and reserve areas of reliability using coal thickness measured
estimates of an area, coalfield, or a coal basin is from points in outcrops and well data.
Hypothethetical coal
(Area beyond
4.8-km radius)
Undiscovered
resources
Inferred coal
(Area between 1.2-
and 4.8-km radii)
(Area within
0.4-km radius)
Measured
Coal in subsurface coal
Identified (Area between
0.4- and 1.2-km radii)
resources Indicated coal
0 1 mile ured Drill
Meas al holes
0 1 km co
Bed
Ou Co
al Point of thickness of
of
tcr
op coal measurement
FIGURE 6.1 Diagram showing reliability categories using coal thickness at observations points (e.g. data) along the
outcroppings and subsurface. Radius originated from observations points. Source: Modified from Wood et al. (1983).
308 6. RESOURCE EVALUATION METHODOLOGIES
Measured Coal Resources points for the indicated coal resources are farther
Estimation of measured coal resources is on apart or moderate degree of geologic assurance.
the basis of mappable and correlation of indi- As shown in Figure 6.1 the indicated coal re-
vidual coal beds whose thickness is measured sources of a bed are calculated from 0.4 to
from outcrops, coal drill holes, and oil and 1.2 km radius (0.8e2.4 km diameter) around a
gas wells. Measured coal resource estimates control point. Thus, the thickness and lateral
are based on closely spaced data control points extent of the coal bed is projected farther from
or high degree of geologic assurance the control point than that for the measured
(Figure 6.1). The measured coal resources of a coal resources. The distance between control
bed are calculated within 0.4 km radius points of thickness measurements for indicated
(0.8 km diameter) around a control point. coal resource is from 1.6 to 2.4 km apart to
Thus, the resources are calculated by projection assume coal bed continuity.
of the thickness, rank, and quality of the coal The zone of indicated coal resources as shown
beyond the point of measurement. The thick- in Figure 6.1 is about 2.4 km wide beyond the
ness of coal beds measured from the control zone of the measured coal resources. Indicated
points is determined by rank, which includes coal resources include anthracite and bitumi-
anthracite and bituminous coals 35 cm or nous coals 35 cm or more thick and lignite and
more thick and lignite and subbituminous subbituminous coals 75 cm or more thick.
coals 75 cm or more thick. These coal resources Much like the measured coal resources, the
extend up to 1800 m in depths. depth limit is up to 1800 m. The coal bed calcu-
The close spacing of control points demon- lated for indicated resources may be extended
strates better capability to trace the lateral con- beyond the indicated resource boundary for esti-
tinuity of the coal beds. Thus, geologic mates for inferred resources.
assurance is at a high degree. According to
Inferred Coal Resources
Averitt (1974), the close spacing of control
points permits close delineation of the coal Calculations of inferred coal resources are
body and the estimation of tonnage for based on presumed lateral extension or conti-
measured coal resources to be accurate within nuity of coal beds beyond the zones of
20%of the true tonnage. This estimated measured and indicated coal resources. Usu-
tonnage, however, does not represent the true ally the inferred coal resources are in inacces-
tonnage of the coal bed because where the sible areas with limited control points and a
spacing of control points are farther apart on low degree of geologic assurance. Thus, the
the same bed, resource calculations are lateral extent of coal beds is based on the
included as indicated and inferred with geologic character of the beds and enclosing
decreasing abundance of data. rocks. The zone of inferred coal resources
lies between 1.2 km and 4.8 km radius from
Indicated Coal Resources the control point of coal thickness measure-
Estimation of indicated coal resources is much ment. Estimates of the inferred coal resources
the same method as that used in measured coal are calculated by projection of the coal thick-
resources, which is based on mappable and cor- ness beyond the indicated coal resources
relation of individual coal bed(s). Data points area. Because the control points used in
where coal thickness is measured also are from the assessment of inferred resources are far
outcrops and subsurface observations. However, apart, the degree of geologic assurance is
unlike the measured coal resources, the control low.
UNIVERSAL GUIDELINES TO COAL RESOURCES ASSESSMENT 309
Undiscovered Coal Resources (Hypothetical UNIVERSAL GUIDELINES TO COAL
and Speculative) RESOURCES ASSESSMENT
The undiscovered coal resources are postu-
lated coal deposits either separate from or exten- Guides to coal resource assessment are very
sion of the measured, indicated, and inferred similar through out the world. The guidelines
coal resources. Based on the degrees of reliability, proposed the same resources categories with mi-
undiscovered resource is categorized as hypothet- nor variations in implementation given favorable
ical and speculative coal resources. Estimates of geological conditions. The broad guidelines sum-
hypothetical coal resources are based on a very marized from the United Nations (UN), USGS,
low degree of geologic assurance and usually Joint Ore Reserves Committee (JORC) (Coalfields
occur in unmapped and unexplored areas. Calcu- Geology Council of New South Wales and the
lations of tonnage of hypothetical coal resources Queensland Mining Council, 2001; UNFC, 2010;
are based on: (1) extrapolated coal beds beyond Wood et al., 1983) all have common recommenda-
the inferred coal resources beyond the radius of tion such as:
4.8 km from the control point and (2) geological 1. Maps for each coal bed or zone that contain
character of the coal. Much like the measured, coal resources,
indicated, and inferred coal resources, hypotheti- 2. Size of unit area or areal extent of coal
cal coal resources consist of anthracite and bitu- resources,
minous coals 35 cm or more thick and lignite 3. The resource category(s) on which the coal
and subbituminous coals 75 cm or more thick to resources estimate is based,
a depth of up to 1800 m. Speculative coal re- 4. The coal quality and rank,
sources have the lowest geologic assurance and 5. Thickness of the coal as measured from
include possible areas outside coal basins and stratigraphic control points or points of
coalfields deeper than 2000 m (Averitt, 1974). observation,
Usually, data of presumed occurrence of coal 6. The thickness of overburden to determine
below 2000 m are not collected because they are minability of coal beds overburden, and
uneconomic and beyond the technical capability 7. Economic considerations are emphasized
of being mined at the present time. such as (a) proposed mining methods, (b)
criteria used to limit, the reserves, and (c)
Other Countries mining recovery.
Other countries such as India consider the Of the three guidelines (USGS, JORC, UN),
depth of occurrence in the economic extraction, the USGS recommended a strategy for the esti-
scope for exploration, and coal resources catego- mation of resources in vicinities of where coal
rized in the depth ranges of: 0e300 m, 30e600 m, bed bifurcates into tongues (Wood et al., 1983).
and 600e1200 m. Germany, Poland and United This distinction is important because the main
Kingdom like India categorized coal resources requirement for coal resource assessment is the
at maximum depths of 1500, 1000, and 1200 m preparation of coal bed maps for each bed or
for hard coal, respectively, at specific thickness seam (see number 1 in the above list). The
(0.35e1.5 m), maximum sulfur content (2%), a USGS guideline recognizes the lateral variability
minimum calorific value (15e24 MJ/kg) (United and common occurrence of coal splitting, and
Nations, 2008) (Ersoy, 2005; Schmidt, Gerling, does not treat coal beds as a proverbial “layer
Thielemann, & Littke, 2007; Smakowski & cake” rock unit. The splitting of coal causes diffi-
Paszcza, 2010). culty in resource estimation, which requires the
310 6. RESOURCE EVALUATION METHODOLOGIES
delineation of the boundary between the main using these trends and features, on the nature
coal bed and the tongues. Thus, the areas of the of ancestral rivers, deltas, and coastal
main coal body and coal bed tongues are map- environments and related peat-forming
ped separately with the tongues being individu- swamps and/or bogs. In addition, directions
ally estimated for resources. Unfortunately this and distances of point source of sediments of
guideline has not been followed religiously. these environments are considered.
Moreover, failure to recognize this concept has 5. Expand stratigraphic work from known to
serious repercussions in the precise estimation unknown areas using surface (e.g. outcrops)
of coalbed gas reserves. and subsurface (geophysical logs of oil, gas,
According to Wood et al. (1983), when esti- and water wells) data points. Establish
mating coal resources in such a geologic condi- stratigraphic correlations and depositional
tion it is necessary to delineate the “coal split models of coal beds from these data
line” between the areas where the coal resources points.
are estimated separately. The “geologic condi- 6. Construct structural sections and contour
tion,” however, requires the sophistication abil- maps to use in estimating overburden
ity to recognize the split and interpret the thickness and rank of coal beds in both known
depositional environment of the interbedded and unknown areas.
rock units (see following sections). The “split 7. Project all data (e.g. isopach and isopleths
line” is defined where the coal bed tongues maps, depositional, stratigraphic, and
become thinner than the intervening parting of structural trends/features) of the above steps
the main coal bed. The parting is a non-coal layer from known to unknown areas.
in a coal bed, which does not exceed the thick- 8. Determine the average coal thickness per unit
ness of either the underlying or overlying parts area in each coal category (e.g. measure,
of the coal bed (Wood et al., 1983). All these indicated, inferred, and hypothetical).
points where the thickness of the main coal bed 9. Calculate and sum estimated tonnages for
and the tongues are measured are located in each category, which represent the best
the coal bed map. information for coal resources in the known
Wood et al. (1983) proposed the following area (e.g. identified resources) to unknown
approach of extrapolated coal bed map method area (e.g. undiscovered resources).
to estimate coal resources controlled by deposi-
tional environments.
Geologic Assurance
1. Collect all geologic data in the area adjacent to
the coal being estimated for resources. Fundamental to the resource assessment of
2. Collect all data on coal and overburden coal and coalbed gas resources and reserves is
thicknesses, coal quality, and coal rank. the concept of geologic assurance (Wood et al.,
3. Evaluate all data from (1) and (2). Construct 1983). Geologic assurance is controlled by
coal isopach map noting thickening and spacing, quantity, and quality of control points
thinning directions. Create coal quality (e.g. as affected by correlation and depositional envi-
calorific value, ash yield, trace elements) ronments of coal beds (Figure 6.2). That is, the
isopleths maps; the latter two parameters may more control points the closer the spacing be-
indicate direction and transport of sediments. tween data points and increase the certainty of
4. Distinguish depositional and erosional trends correlation of coal beds. Although this is an ideal
and features adjacent to and in the coal being situation, the standard operating procedure is
estimated for resources. Emphasis is placed, that initially resource assessments are normally
UNIVERSAL GUIDELINES TO COAL RESOURCES ASSESSMENT 311
Geologic assurance proceed, additional data points are collected and
the reliability of the stratigraphic correlations
and mappability of the coal beds are better
Distance and density of control points
established.
Spacing Guidelines
Correlation Depositional
environment Spacing or distance of data control points to
determine and ensure geologic assurance is a
Methods: universally accepted concept, which is dictated
Lithological Juxtaposed to peatlands:
by the geology of the coal, coalfield, and coal
Biological Seas, lagoons, lakes, bays,
Chemical Delta distributaries, basins. These factors are different worldwide,
Time-rock Crevasse splays, tidal which are reflected by variable guidelines
Roch sequence Creeks, rivers promulgated by different countries. Although
Strutures
spacing of control points for geologic assurance
FIGURE 6.2 Diagram showing a flow chart of the hier- are different worldwide, most countries create
archy of variables that determine geologic assurance. and modify their guidelines in conformity
with the UNFC (2010), USGS resources classifi-
performed in virgin areas with minimal or no cation (Wood et al., 1983), and Joint Ore Re-
stratigraphic data are available. Thus, correlation serves Committee code (JORC Code) and
of coal beds is governed by availability of strati- Guidelines (2004), Geological Survey of India
graphic data points, which are far apart with (2012), Indonesia Ministry of Energy & Mineral
the potential for miscorrelation of coal bed being Resources (2012). For example, Table 6.1 shows
very high. As exploration and development that the Indonesian standard for geologic
TABLE 6.1 The Geologic Assurance on Coal Resources Classification for Other Countries Such as
Indonesia is the Same Concept as Stated by Wood et al. (1983). However, it Also Depends
on Sedimentary and Tectonic Parameters, Which Lead to Classification of Geologic
Conditions into Simple, Moderate, and Complex
Geologic Conditions
1. Sedimentary
Coal thickness Little variation Some variation Large variation
TABLE 6.2 Indonesian Geologic Assurance Ranging from Simple to Complex Geologic Conditions Determines the
Distance between Observation Points for Measured, Indicated, Inferred, and Hypothetical
Coal Resources
assurance on coal resources classification is JORC Code of coal resource classification based
mainly controlled by the sedimentary and tec- on geologic assurance (Table 6.3) does not vary
tonic parameters, which dictate categorization much in distance between points of observation
of geologic conditions into simple, moderate, to that of the United States, Indonesia, and India
and complex. Sedimentary parameters are (Stoker, 2009).
distinguished according to the thickness vari-
ability, continuity, and splitting of coal beds to The Stratigraphic Factor
measure the complexity of the geologic condi- The stratigraphic factor is the foundation of
tions. In addition, tectonic parameters include geologic assurance toward estimation of coal
faults, folds, dipping beds, and igneous intru- and coalbed gas resources. Catuneanu (2006)
sions to characterize the various geologic condi- defined stratigraphy as the science of rock strata,
tions. Based on the range of simplicity and which encompasses their properties and attri-
complexity of the geologic conditions of the butes important to interpretation of the geologic
coal geology and tectonism, the distance be- origin and history. Although the science of stra-
tween control points is determined and the de- tigraphy covers a broad concept, it is the lateral
gree of geologic assurance is assigned to each and vertical characteristics of the rock strata
coal resource (e.g. measured, indicated, that are most applicable to coal resource assess-
inferred, and hypothetical) (Table 6.2). ments. More specifically stratigraphic correla-
The simple geologic condition with little vari- tion and depositional environments of the rock
ability in coal continuity and thickness, no coal strata greatly impact the geologic assurance.
splitting, limited or no faulting, folding, intru- The classic definition of stratigraphic correlation,
sions, and gently dipping beds can be assessed which is still applicable today, is the demonstra-
with control points that are far apart. In contrast, tion of equivalency of rock units (Krumbein &
complex geologic conditions with highly variable Sloss, 1963). However, stratigraphic equivalency
coal continuity and thickness, coal splits, faulting, of rock units in the more modern context of
folding, intrusions, and steeply dipping beds can sequence stratigraphy may be demonstrated by
be assessed with closely spaced control points. time-rock continuity despite changes in lithofa-
The concept of geologic assurance, which reflects cies beyond the scale of individual depositional
estimates of uncertainty, dictated by the degree of systems as suggested by Catuneanu (2006).
complexity of the geology and deformation is Nevertheless, in coal resource assessment in
exemplified by India where spacing from control which data collection of thickness of correlated
points is 200 m for proved (e.g. measured) re- individual coal beds is a requirement in the prep-
sources, 1000 m for indicated resources, and aration of coal bed maps, conventional stratig-
2000 m for inferred resources. The Australian raphy based on lithological or lithostratigraphy,
UNIVERSAL GUIDELINES TO COAL RESOURCES ASSESSMENT 313
TABLE 6.3 Evolution of the Concept of Geologic Assurance Based on Distance between Points of Observation, Which
Changed from Longer to Shorter Distances for Measured, Indicated and Inferred Coal Resources
Coal Resources
1986 <1 km between PO; <2 km between PO; extrapolate Presence of coal
extrapolate < 500 km <1 km from PO unambiguously determined
2001 <500 m to may be 1 km <1 km to may be 2 km <4 km between PO;
between PO; between PO; *extrapolate *extrapolate <2 km from PO
*extrapolate <500 m <1 km from PO
2003 <500 km between PO; <1 km between PO; should <4 km between PO; should not
should not be not be extrapolated more than be unreasonably extrapolated
extrapolated more than half the distance between PO beyond the last line of PO
half the distance
between PO
* Trends in coal thickness and quality should not be extrapolated more than recommended minimum kilometers from points of observation.
PO, points of observation.
Source: Modified from Joint Ore Reserves Committee code (JORC Code) and Guidelines (2004).
lithofacies association, and biological or biostra- maps of coal bed or coal zone, which are known
tigraphy related to individual depositional sys- to contain coal resources, as a part of document-
tems is still the time-tested approach ing the estimates of resources. Correlation is the
(Figure 6.2). Rock units may be traditionally underpinning of coal bed or coal zone maps,
recognized by their physical (e.g. color, texture, which record the following information presum-
thickness, internal structures), biological (e.g. ably of the same coal bed: (1) trace of the outcrop
plant and animal fossils), and chemical (e.g. chem- of the coal bed, (2) control points where thickness
ical composition) properties. However, in coal is measured from the same bed at surface expo-
physical and biological attributes (e.g. black and sures, and (3) control wells, drill holes, and sub-
plant-rich) are more a detriment than a benefit surface points where coal thickness of the same
following the axiom of they all look alike. coal was measured. The problem is as difficult
Stratigraphic correlations of coal beds and where the coal beds are steeply dipping and/or
associated rock strata revolve around their depo- highly deformed (e.g. faulted).
sitional environments and related sedimentologic In coal-bearing rocks, correlation is a demon-
processes by which they form. Conventional stration of perceptible continuity of a coal bed
stratigraphic correlation of rock strata requires between control points by showing correspon-
knowledge and interpretation of depositional en- dence in character and stratigraphic position
vironments. Simply put correlation of coal beds (Wood et al., 1983). The traditional method of
and associated rock strata form a base for studies correlation is to trace and map a coal bed with
of depositional environments (Ferm & Horne, distinctive lithologic, biologic, and chemical fea-
1979; Fielding, 1984; Flores, 1986; Flores, Spear, tures along surface exposures from one locality
Kinney, Purchase, & Gallagher, 2010; Flores, to another locality (Figure 6.2). In addition to
Spear, Purchase, & Gallagher, 2010; McCabe, these features, recognizing the contacts between
1984; Ryer, 1984). However, according to Wood the beds and adjoining units may aid in the map-
et al. (1983) it is a means to an end to establish ping of a coal bed. Also, mapping a coal bed is
314 6. RESOURCE EVALUATION METHODOLOGIES
practiced in the subsurface using geophysical well as structural features (folds, faults, igneous
logs and/or lithological samples (e.g. rock cut- intrusions, etc.) may be of value in establishing
tings and cores) from well to well (Flores, Spear, correlations (Figure 6.2) (see Catuneanu, 2006
Kinney, et al., 2010; Flores, Spear, Purchase, and references therein).
et al., 2010; Wood et al., 1983). This may be Thus, coal bed correlation is based on under-
possible in areas where closely spaced wells standing the stratigraphy, however, it demands
permit mapping, however, this method may be more sophistication in interpreting the strati-
restricted in areas where lithology changes occur graphic data of the coal and enclosing rock units
rapidly laterally. In both surface and subsurface to develop depositional correlation-based
conditions, the density of control points (e.g. out- models. This deposition correlation approach
crops, wells) and rate of lithological variation in- is complicated by the areal extent of coal beds
fluence the accuracy of coal bed correlations. limited by erosion and/or nondeposition. This
Thus, other methods may be applied in support problem is exacerbated by lateral changes of
of coal bed correlation such as biological (e.g. the coal beds into other lithological units with
maceral composition, palynology) and chemical differing properties sometimes called lithofa-
(e.g. inorganic and trace elements) attributes of cies change (Reading, 1996). Figure 6.3 demon-
the rock unit (Ayers, 1986; Crowley, Ruppert, strates the lateral and vertical traceability of
Stanton, & Belki, 1994; Flores, Spear, Kinney, coal bed successions, which are restricted by
et al., 2010; Flores, Spear, Purchase, et al., 2010; lithofacies change in the Cretaceous Fruitland
Gochioco, 1992; Nichols, 1999; Scholes & Formation in the San Juan Basin (Fassett,
Johnston, 1993; Schweinfurth, 2002; Stanton, 2010). The Fruitland coal beds are separated
Moore, Warwick, Crowley, & Flores, 1989). above and below as well as laterally from other
Other methods such as sequential position of rock units by their depositional continuity and
the rock unit with respect to underlying and extent, which are controlled by surrounding
overlying rock units and unconformities as depositional environments.
Seaward (basinward)
Fruitland
Formation
Pictured Cliffs Sandstone
coals
100 m
10 km
Huerfanito bentonite bed (datum)
FIGURE 6.3 The lateral and vertical variabilities of coal beds in the Cretaceous Fruitland Formation in the San Juan Basin,
New Mexico, United States. Source: Adopted from Fassett (2010).
UNIVERSAL GUIDELINES TO COAL RESOURCES ASSESSMENT 315
The Depositional Connection another in an en echelon pattern and vertically
Geologic assurance, stratigraphy, and dis- stacking was controlled by relative sea-level
tance of control points to estimate coal resource fluctuations. Continuity of coal beds was deter-
categories are all directly connected to deposi- mined by Fassett (1989) based on a time-rock
tional environments. However, the depositional unit (Huerfanito bentonite bed), which is volca-
factor has played minor and secondary roles in nic in origin. The conventional approach would
coal and coalbed gas resources assessment for have used the Pictured Cliffs Sandstone as a
the simple reason that resource specialists have time-rock unit, which would have correlated
not expended their time identifying the deposi- the coal bed immediately on top of the sand-
tional component of coal deposits. Although stone as a single, very continuous bed (e.g. layer
the principle of coal deposition should be the un- cake) beyond its limited depositional site
derpinning test of geologic assurance, resource controlled by seaward regression.
analysts generally lack the experience or time
for this type of investigation. All too often Coal bed Variability
resource assessment is on a short fuse and based Definition of the areal extent and thickness
on sparse to readily available data, which are not of coal beds is the primary objective of resource
amenable to exhaustive depositional study. assessment, which can only be delineated with
A depositional environment is defined as a the use of stratigraphic control points and by
site where sediments (e.g. detrital, chemical) proper correlation as a result of their deposi-
accumulated, governed by physical, biological, tional environments. The rate of lateral and
and chemical processes related to modern and vertical variability of coal beds depends on
applied to ancient environments, and lithified their depositional environments and changes
into sedimentary rock units. The areal extent in depositional trends with respect to related
and peat thickness of peatlands are controlled, peat-forming environments. Thus, the degree
in part, by the areal distributions and processes of certainty of geologic assurance of coal
of surrounding depositional environments. In resource assessment is as variable as the area
addition, because peatlands are genetically in which peat/coal accumulated. The
linked and grade into these contemporaneous, consensus is that size or extent of the area
physically active environments, the peatlands (e.g. swamp, mire, bog) of peat/coal accumula-
respond to the changing depositional trends. tion is controlled by the types of depositional
For example, a shift in the paleoshoreline in the environments (e.g. delta plain vs alluvial)
basin of deposition controls the seaward or land- (Ferm & Cavaroc, 1968; Ferm & Staub, 1984;
ward position of the peatlands and resulting Ferm et al., 1979; Flint, Aitken, & Hampson,
thickness of coal deposits. 1995; Flores, 1986; Stanton et al., 1989; McCabe
Figure 6.3 demonstrates this phenomenon & Parrish, 1992; Moore 2012). However, Field-
where base level lowering produced seaward ing (1987) suggested that depositional control
stratal stacking of marine sandstones and juxta- is overshadowed by subsidence and sediment
posed coal beds formed on back-barrier and supply.
delta plain peatlands (Flores & Erpenbeck, Research by Ayers (2002), Ferm et al. (1979),
1978; Fassett, 1989). Along this setting Creta- Greb, Eble, Hower (1999a,b), Horne, Ferm,
ceous Fruitland coals are continuous but thin Carrucio, and Baganz (1978), McCabe (1984),
and thicken as well as split with a 20-km dis- Ryer (1981), and Warwick and Stanton (1988)
tance perpendicular to depositional strike. demonstrated small- and large-scale variability
More importantly the stratigraphic architecture of the areal extent, thickness, and shape of coal
of the coal-sandstone is overlapping one beds. Many of these workers have advocated
316 6. RESOURCE EVALUATION METHODOLOGIES
the need for more stratigraphic data and addi- which the detailed stratigraphy of a Pennsylva-
tional closely spaced control points in order to nian (Hartshorne) coal beds within a
demonstrate apparent continuity of coal beds. 100,699 km2 area in the Arkoma Basin reveals
They suggest that the most informative cross sec- the shape, anatomy, and pattern of develop-
tions of coal-bearing rocks are those in which ment of the deposit in a fluvial setting. The
control points are closely spaced in which there isopach map (Figure 6.5) shows a coal body
can be no doubt about lateral continuity of coal about 31e62 km wide and 187 km long with
beds. The next best thing in which no interpreta- two linear coal subbodies as much as 1.8 m
tion is required between control points to trace thick trending east to west (Figure 6.5). The
coal bed continuity along a continuous road linear coal subbodies are flanked on the south-
cut or other large-scale excavation. Demonstra- west and east by no coal. Between these linear
tion of the scale of lateral continuity, thickness coal subbodies is an area of about 0.6e1.8 m
variability, and areal extent of coal beds from thick coal separated by a east-west trending,
ancient deltaic to fluvial coal-bearing deposits ribbon-like area of no coal or sandstone
in relationship to stratigraphic control during (yellow color).
the past 30 years is represented by the works of A stratigraphic cross section (Figure 6.4)
Ferm et al. (1979), Flores, Spear, Kinney, et al. showing detailed lithofacies descriptions and
(2010), Flores, Spear, Purchase, et al. (2010), variations provides a clue to the depositional
and Greb et al. (1999a, b). settings of the coal body and laterally equiva-
lent sedimentary rock units. The 1.8-m thick
Small-Scale Variability coal subbodies along the margins results
Milici and Houseknecht (2009) demon- from the presence of an upper coal bench,
strated this concept in Figures 6.4 and 6.5 in which splits from the lower (main) coal bench
and thins toward the center or toward the
fluvial channel (Figures 6.4 and 6.5). The nar-
ARKANSAS row east-west belt of fluvial channel sandstone
–
OUTBOARD – OKLAHOMA – INBOARD (yellow color) with no coal represents tempo-
ral deposition with the merged lower and up-
McAL.
McAL.
Coal Split
per Hartshorne coal beds formed in the
adjoining mires (Figure 6.5). The ribbon-like
LOWER UPPER
HART.
Scale U.S. e
lin
T .9 N.
ntic
na
5 0 5 10 15.5 21 26
km
lto
Mi
Arkoma Basin T .8 N. Backbone
N anticline
T .7 N.
T .6 N.
T .5 N.
R. 15 E. R. 16 E. R. 17 E. R. 18 E. R. 19 E. R. 20 E. R. 21 E. R. 22 E. R. 23 E. R. 24 E. R. 25 E. R. 26 E. R. 27 E.
University of Oklahoma
Channel Sandstones, UPPER HARTSHORNE COAL ISOPACH
U L Coal 1.2–1.8 m thick
Hartshorne coals missing Fault ARKOMA BASIN
Coal 0.6 m thick or less PLATE 2 Coal >1.8 m
John Gossling (1994) 12/22/93
Coal 0.6–1.2 m thick (Digital compilation by R.C. Milici, U.S. Geological Survey, 2008) Geologic structures - no coal
FIGURE 6.5 Map showing the areal distribution and thickness of the Upper Hartshorne coal bed in the Arkoma Basin,
Arkansas and Oklahoma, United States. The areal distribution of the Upper Hartshorne coal bed depicts the split by a fluvial
channel sandstone (see Fig. 8.4). U ¼ upper; L ¼ lower. Source: Adopted by Milici and Houseknecht (2009).
and Pedlow (1977) in delta plain depositional the rocks surrounding the coal bodies are
environment show more variable shapes. The important in mapping the shape and margins
Allegheny coal beds in the Appalachian region of the coal beds. In addition, mapping of split
in western Pennsylvania in a marginal delta lines of the coal along channel margins pro-
plain occur in five elongate subbodies ranging vide a better and more precise coal bed map
from 3.2 to 6.4 km wide and 9.6e16.1 km of merged single versus split beds for separate
long within 434-km2 area. The cross-sectional resource assessment. Finally, larger data sets
shape of the coal subbodies is tabular averaging 1.6 km apart made for better delin-
decreasing in thickness rapidly toward the eation of the small, elongate coal subbodies.
mire edges from 0.6 m to less than 15 cm and Greb et al. (1999a, b) demonstrated the value
no coal. Ferm et al. (1979) interpreted the of large data sets (3800 locations within
coal subbodies formed in interfluve areas 1860 km2 area) in mapping Middle Pennsylva-
and the sites of no coal as tidal creeks infilled nian (Duckmantian) coal bed in the central
by siltstones and shales. In this setting the Appalachian region (Figure 6.6). Like in the
lateral continuity of coal beds may be extrapo- western Pennsylvania Alleghenian coal bed,
lated up to 3.2 km between control points. In the Duckmantian coal bed shows similar strat-
these case studies by Ferm et al. (1979) and igraphic vertical and lateral variations
Pedlow (1977), the general characteristics of (Figure 6.7). That is benches of the coal bed
318 6. RESOURCE EVALUATION METHODOLOGIES
Appalachian
Basin
Central
Appalachian
Basin
0 5 10 km
0 5 miles
FIGURE 6.6 A map showing the density of observations points of a Duckmantian coal bed in the central Appalachian
Basin in eastern Kentucky, United States. Source: Adopted from Greb et al. (1999a, b).
SW NE
50 cm
1 km
CU CU
X X
CL X
Location
NE
SW
FIGURE 6.7 A cross section of the Pennsylvanian fire clay coal beds showing the upper (CU) and lower (CL) beds separated
by clay parting in eastern Kentucky, United States. Source: Adopted from Greb et al. (1999a, b).
pinch out into adjoining detrital rocks. Howev- the coal subbodies of the Alleghenian coal.
er, the closely spaced data points permitted Figure 6.8 shows the isopach map of the lower
isopaching of the coal benches into elongate bench as an eastewest oriented lenticular coal
to lenticular shape much more complex than sub-body (38.6 km long and 23 km wide)
UNIVERSAL GUIDELINES TO COAL RESOURCES ASSESSMENT 319
5 mi
10 km
?
?
?
FIGURE 6.8 An isopach map of the upper part of the fire clay coal indicating a prominent elongate, eastewest trend.
Source: Adopted from Greb et al. (1999a, b).
surrounded by small elongate subbodies (412 km and rider coal benches with high ash yields,
long and 4 km wide). The larger lenticular sub- which occasionally merged with the upper
body has a very serpentine boundary suggesting coal bench, draped lateral lithofacies. These
more detail interpretation resulting from closely depositional and tectonic controls may explain
spaced control points. Figure 6.9 shows the the right angle reorientation of the directions
isopach map of the upper bench as a south- of elongation between the lower and upper
eastenortheast lenticular body along the thickest benches. The fine-tuning of the shape and
part (from 35 to >105 cm) with branching changes in orientation of coal subbodies are
smaller lenticular coal subbodies to the north- critical to certainty of geologic assurance in
west; both subbodies delineated by also very coal resource assessment, which can only be
serpentine boundaries. documented by more data sets and close
Greb et al. (1999a, b) interpreted the vegeta- spacing of control points as demonstrated by
tion during accumulation of the upper bench the works of Ferm et al. (1979) and Greb
of the coal bed as shifted from planar to et al. (1999a, b).
domed mires. Furthermore, Greb et al.
(1999a, b) suggested that faulting also Large-Scale Variability
controlled depositional environments resulting Ferm et al. (1979) and Greb et al. (1999a, b)
in channel sandstones with angular changes in have demonstrated the variability of a single
orientation. In addition, channel sandstones coal bed within peat/coal depositional
320 6. RESOURCE EVALUATION METHODOLOGIES
Thickness 0 0 to 35 35 to 70
(cm)
70 to 105 >105
FIGURE 6.9 An isopach map of the lower part of the fire clay coal indicating a prominent lenticular, southwestenortheast
trend. Source: Adopted from Greb et al. (1999a, b).
environments with sizes or areal extents from (Figure 6.10). The stratigraphic data sets are
450 to 1860 km2 areas with 250 to 3800 locations about four times greater but cover four times
where coal thicknesses were measured, respec- more area than that of Greb et al. (1999a, b).
tively. The density of control points in these Figures 6.11 and 6.12 show a merged
areas varies from 0.5 to 2.0 per km2 area. Control WyodakeAnderson coal bed, which either
point densities are within spacings required for splits or pinch out along its western margin.
estimation of measured and indicated coal re- The WyodakeAnderson coal bed splits into
sources categories. A larger scale peatecoal as many as five coal beds, which in turn split,
depositional environment of a single coal is merge, and pinch out (Flores, Spear, Kinney,
described by Flores, Spear, Kinney, et al. (2010) et al., 2010). Splitting caused the merged
in the southern part of the Powder River Basin WyodakeAnderson coal bed to expand into
in Wyoming. This area is about 7770 km2 under- an interval of as much as 305 m thick, which
lain by the WyodakeAnderson coal bed, which contains thin to thick coal beds interbedded
is 6e46 m thick, and overlain by 61e366 m thick with thick mudstone, siltstone, and sandstone
overburden (rock interval above the coal to the beds (Flores, Spear, Kinney, et al., 2010). This
surface). There are 14,000 coalbed gas, oil and single WyodakeAnderson coal bed is one of
gas wells, and coal drill holes with density of the largest coal deposit accumulated in a
control points of about 1.8 per km2 area but peat-forming depositional environment inter-
only 10,000 geophysical logs of these wells preted as a domed or raised bog (Flores, Spear,
were used to measure the coal thickness Kinney, et al., 2010). However, the coal deposit
COAL RESOURCE ASSESSMENT METHODOLOGY 321
North group
of coal mines
Gillette
Middle group
of coal mines
Merged Wyodak-Anderson coal
Split line
Wright
South group
of coal mines
Explaination
15 mi
FIGURE 6.10 A map of the merged Wyodak-Anderson coal zone of the Paleocene Fort Union Formation in the south and east of
the Powder River Basin, Wyoming. See index map for location in the United States. Source: Modified from Flores, Spear, Kinney, et al.
(2010), Flores, Spear, Purchase, et al. (2010).
322
4500
(Felix)
Formation Formation
Wasatch
(Smith)
(Swartz)
4000
Fort Union
(Anderson)
Lower Wyodak
(Canyon)
Lower Wyodak ? ?
?
?
3000
?
?
107˚ 105˚
NORTH
DAKOTA 2500
SOUTH
EXPLANATION
DAKOTA
Sandstone
MONTANA Decker Quietus Coal
45˚
WYOMING
Sheridan Powder River Coalbed methane
Basin
well
West East
Wyodak cropline Gillette Oil and gas well 2000
0 40 miles
0 40 km
43˚
NEBRASKA
FIGURE 6.11 An east to west cross section across the central part of the Powder River Basin in Wyoming, United States (see index map for location). The
merged Wyodak coal zone splitting on the easternmost part of the cross section. The reader is referred to Flores et al. (2010) for the references cited in the
cross-section. Source: Modified from Flores, Spear, Kinney, et al. (2010), Flores, Spear, Purchase, et al. (2010).
COAL RESOURCE ASSESSMENT METHODOLOGY
FIGURE 6.12 A north to south cross section across the central part of the Powder River Basin in Wyoming, United States (see index map for location).
The merged Wyodak coal zone splitting northward. The reader is referred to Flores et al. (2010) for the references cited in the cross-section. Source: Modified
323
from Flores, Spear, Kinney, et al. (2010), Flores, Spear, Purchase, et al. (2010).
324 6. RESOURCE EVALUATION METHODOLOGIES
TABLE 6.4 Average Specific Gravity and Weight of Various Rank Coals in the United States
The 9.1 m threshold was selected because values. The coal bed thickness model was con-
the nearest coal bed above and below the verted to a fixed raster data set, or GRID file
WyodakeAnderson coal bed is about format, with a 1/4-acre (0.10 ha) cell size. The
10.7e18.3 m. Filtering for wells that are dupli- thickness GRID was used as input in ESRI’s
cated and for WyodakeAnderson coal bed Spatial Analyst (ESRI, 2009) extension to create
greater than 49 m due to splits netted 6630 wells a raster tonnage data set using Wood and
from the original 14,000 wells. Additional others’ (1983) and Mercier and others’ (2010)
filtering was performed on this data by buff- formula in Raster Calculator (Scott Kinney
ering wells used along the cross sections and Tracey Mercier, Personal Commun., 2010):
reducing the total to 2989 wells, which were
used to isopach by simple Kriging method Thickness GRID (in ft or m) 1770 (weight of coal
allowing weighted averaging of the thickness per acre ft or per hectare) 0.25 (area of each cell, in
of the WyodakeAnderson coal bed. The isopach acres or hectares)
map (Figure 6.13) was used to calculate the coal or
resources of the merged WyodakeAnderson
coal bed in the southern Powder River Basin. Thickness GRID (in ft or m) 1605 (weight of coal
per acre ft or per hectare) 0.25 (area of each cell, in
acres or hectare)
Cell-Based Coal Resource Calculations
The calculation methodology used for this The resultant GRID contained tonnage
assessment was a cell-based modeling adaptation values per cell can be converted to interna-
of Wood and others (1983). Initially, a hierarchy of tional units used to summarize resources by
data points was established to create a coal thick- any given zone. By using a cell-based instead
ness isopach. The order of importance, or confi- of a vector-based methodology, coal thickness
dence, was (1) cross-section wells, (2) drill holes and tonnage estimates can be generated for
outside a 1.2-km buffer zone of cross-section each 1/4-acre (0.10 ha) cell in the GRIDs as
wells, (3) coalbed gas wells (e.g. drilled holes, opposed to larger polygonal areas. This can
cored drill holes reported in the WOGCC data- provide a more efficient mechanism to report
base) outside a 1.2-km buffer zone of drill holes summary statistics as any polygonal area can
and the previously established cross-section be overlain on the thickness or tonnage GRIDs
buffer zone, and (4) additional wells outside a to generate summary statistics summarized.
1.2-km buffer zone of State wells and the two pre- Examples of other polygonal zones that
viously established buffer zones. The wells were can be summarized include (1) counties, (2)
filtered for greater than 45 m thick of the merged 7.5-min quadrangles or topographic maps, (3)
WyodakeAnderson coal bed along the split measured, indicated, and inferred buffer zones,
boundary because of over thickening cause by and (4) overburden categories (Figure 6.14)
introduction of interburden (for example, sand- (Wood et al., 1983).
stone, siltstone, and mudstone). This hierarchy Coal resources were calculated for the
resulted in a decrease of approximately 9000 wells entire merged Wyodak-Anderson coal bed
used for resource calculations and created a delineated by split boundary after removing
method to assess confidence in the data. mined-out and clinker areas. After resources
The wells point data set was modeled in were calculated, a cell-based Euclidean alloca-
ESRI’s GeoStatistical Analyst (ESRI, 2009) tion method and a vector-based Voronoi
extension using simple Kriging to interpolate method (De Berg, van Kreveld, Overmars, &
and extrapolate weighted-average thickness Schwarzkopf, 2000) were used to validate
COAL RESOURCE ASSESSMENT METHODOLOGY 327
106°00’ 105°30' 105°00’
WE
EXPLANATION
NG
DI
STO
R
- HA Clinker areas
EP
N-H
S LE
T EN
AT
Coal mines
MONTANA
I Single coal
CR
UR
SO
EEK
S 25–50
MI
L E
STUDY AREA
LI TT H 50–75
NC
(represented)
WYOMING
RA
44°30'
Gillette
North group Inferred split coal area
of coal mines
Wright EN 76–100
I NG
POWDER RIVER 100
5050
S PR 101–125
50
BASIN
50
75
E T
R OZ
126–150
75
Gillette Lineament
50
50
10 50 75
50
0
75
50
50 0 5 10 20 miles
75
75
50
10 75
75
0 75
75
75 75 0 10 20 km
100
75
K
75
EE
0
10 75 Thickness contour interval is
CR
7575
75
75
50
Middle group E 25 ft (7.6 m)
of coal mines OT
Y
75
CO
H
100 75 UT
SO
44°00’
100
75 50 75
75
K
E EE
100
50
SAG CR
100
-O
75
SU SSEX ER
DL
75 Wright
FID
75
75
10
0
125
100
12
100
5
75
50
100
South group
of coal mines
75
T ON
50
-UP
IN TO
A RM 75
75
43°30'
BU
FF
A
FIGURE 6.13 An isopach map of the merged Wyodak-Anderson coal zone of the Fort Union Formation in the
southern part of the Powder River Basin in Wyoming. The thickness of the merged Wyodak-Anderson coal is un-
affected by northwest-southeast and northeast-southwest basement lineaments. Source: From Flores, Spear, Kinney, et al.
(2010), Flores, Spear, Purchase, et al. (2010).
328 6. RESOURCE EVALUATION METHODOLOGIES
EXPLANATION
Coal mines
< 5:1
6:1–10:1
STUDY AREA
(represented) WYOMING
Gillette
11:1–15:1
44°30'
25
25
5
26:1–30:1
25
31:1–35:1
45
Gillette
35
36:1–40:1
40
5
41:1–45:1
5
25
46:1–50:1
5 5
35
40
25
20
35
45
35
35
50
of coal mines
30
0 5 10 20 miles
5
5
5
44°00’ 0 10 20 km
20
5
20
15
20
Wright
15
South group
of coal mines
15
15
10
43°30'
FIGURE 6.14 A ratio map of the overburden rock per thickness of the merged Wyodak-Anderson coal in the southern part
of the Powder River Basin in Wyoming. The overburden-coal thickness ratio increases toward the northwest.
DATA COLLECTION 329
the calculated totals. Both methods generated and desorbing gas over a period of time is neces-
a less than one percent difference in resource sary to measure and estimate the volume of gas
and reserve totals than the cell-based Kriging charge in a coal bed reservoir. Coal samples for
method. Wood et al., (1983) prepared guide- gas desorption can be from drill cuttings and
lines for standardized reporting categories in core (Montgomery & Barker, 2003). However, ac-
order to have rational and uniformity of pro- curate measurements of gas content are best ob-
cedures to follow as well as for accurate in- tained from coal cores. Because of the cost of
ventory of resources. The purpose of these drilling and coring, time constraints from pro-
guidelines is to create coal resource estimates longed coal desorption and laboratory analyses,
prepared by various workers, which are loss of gas during transport from the field to the
meaningful comparison world-wide, and for laboratory, and demand by companies for speedy
a nation, state, county, region, basin, town- initial appraisal, the conventional procedure of
ship and range and quadrangle. Federal coal gas desorption from coal cores is being replaced
and surface ownership are categories also by new technology of in situ gas measurement
included because the Powder River Basin in the wellbore (Lamarre & Pope, 2007; Carlson,
contains split-state ownership and large Hartman, Obluda, & Miller, 2008). Lamarre and
amount of subsurface mineral ownership. Pope (2007) described this new technology as us-
The guidelines also require reporting accord- ing a downhole chemical-sensing tool called
ing to overburden (e.g. from 0e100 ft or Raman spectrometer, which is lowered by wire-
0e30 m to >500 ft or >150 m), and net line into a 12.7 cm-diameter wellbore, developed
coal thickness (e.g. from 2e5 ft or 0.6e1.5 m specifically for coalbed gas measurement. The
to >40 ft or 12 m). Thus, the merged Wyodake Raman spectrometer provides the critical desorp-
Anderson coal body, excluding the “tongues,” tion pressure of gas dissolved in coal bed reser-
in the southern Powder River Basin was voir fluid (e.g. water) in the wellbore.
estimated to contain measured resources of Despite the new technology for in situ gas
53.5 billion metric tons, indicated of 95.2 billion measurement conventional coring of coal for
metric tons, inferred resources of 35.4 billion gas desorption at the well site still provides the
metric tons, and hypothetical resources of most reliable gas content data. However, reliable
748.4 million metric tons or a total resources collection of data depends on the methods fol-
of 184.9 billion metric tons. lowed at the well site, which include use of
proper equipment (e.g. drill rig, core barrels, lab-
oratory trailer, canisters), techniques (e.g. prepa-
DATA COLLECTION ration and handling of cores samples), procedure
of measurement of gas and management of data,
Precision of data collection dictates the degree and deployment of appropriate number of field
of accuracy of resource assessment. Coalbed gas personnel. When a coalbed gas drilling and cor-
resource assessment of GIP or gas concentration ing project is properly implemented, it is a major
involves complicated data collection and ana- undertaking as typified by the cooperative pro-
lyses. This process mainly includes collection of gram between the USGS, Wyoming Bureau of
data to estimate gas content and reservoir proper- Land Management (BLM-WY), and 29 gas oper-
ties to determine gas saturation. These estimates ators in the Powder River Basin from 1999 to
rely on data from acquisition of coal samples 2009 (Flores et al., 2006). The cooperative field-
from coalbed gas exploration wells for gas laboratory project, with the author as project
desorption and isotherm analysis. The procedure chief, was created to minimize cost and provide
involves the collection of samples of fresh coal real time data from coal cores donated by the
330 6. RESOURCE EVALUATION METHODOLOGIES
coalbed gas companies and operators collected, similar to that used to drill conventional water
analyzed, managed, and interpreted by the wells. Elsewhere for deeper coal beds drill rigs
USGS for policy decision makers in the Wyom- used are much like those in conventional oil
ing Reservoir Management Group of the and gas development.
BLM-WY.
Core-Recovery Systems
Core-recovery systems for collecting coal
Coal Sampling and Handling Strategies
samples from wellbores include conventional,
Critical to coal sampling from the wellbore is wireline, sidewall, and pressure cores as well
the collection of fresh and continuous core sam- as drill cuttings. Each of these systems has ad-
ples for gas desorption. More importantly, in or- vantages and disadvantages in terms of cost,
der to minimize lost gas the core retrieval must quality, quantity, reliability, and time con-
be performed with the least amount of time of straints (Table 6.5). Waechter, Hampton, and
travel to the surface. These sampling issues Shipps (2004) suggested that each of these core
depend on the core-recovery systems, core barrel recovery systems may be utilized under
equipment, experience of operator on drilling different sets of geological conditions. Conven-
and coring through coal beds, and physical char- tional coring method, which requires numerous
acteristics of the coal bed (Figure 6.15). The expe- trip times out of the wellbore, are amenable for
rience of the author in the Powder River Basin use in shallow wells with less than two coal
indicates that operators with experience in dril- beds. In order to retrieve the core from a con-
ling for coal mines have a better “feel of the cor- ventional core drill, the core barrel is detached
ing rpms” of the rock than others and highly from the hole, which is time consuming process
fractured coals are difficult to retrieve contin- because each rod is removed one at a time
uous samples. In the Powder River Basin (Figure 6.16). Wireline coring method is
because of the shallow depths (e.g. <1000 m) of amenable for use in deeper wellbores with
coal beds the rigs for drilling coalbed methane more numerous coal beds. The mechanism of
in most cases are usually moderate size very wireline coring is such that after coring the
Conventional core Less expensive; potentially large and Potential lost of gas; numerous
continuous samples; possible evaluation trip times
of discrete coal benches for cleats,
fractures, and petrology; minimal
contamination
Wireline core Better percent recovery; continuous Expensive; minimal trip times
samples; coring is fast; minor trip times;
better sample collection from deep
wellbores
Sidewall core Reduce cost; minimal trip time; minimize Small size samples; need more
lost gas; sampling units can be accurately samples for accurate
picked from wireline logs; minimal lost representation; limited
gas; amenable to sampling deep potential for additional
wellbores analyses
core barrel is left at the bottom of the hole while near bottomhole pressure, preventing gas expan-
the inner tube with the rocks is detached and sion and fluid loss (Eaton, Redford, Segrest, &
pulled up rapidly to the surface. This process Christensen, 1991). A gel displaces drilling
permits repeated coring through thick intervals mud from the core barrel to keep in situ pressure
of coal and rock units and core retrieval at a constant and inner barrel frozen for transport
very fast rate (e.g. in minutes depending on to the laboratory. Owen and Sharer (1992) and
the depth of the wellbore). Rotary cores Diamond and Schatzel (1998) suggested that
collected from the side of the wellbore after pressure coring is an alternative to direct mea-
completion of wireline or geophysical logging, surement method for obtaining coal sample is
as the name implies, cuts sidewall core. This reservoir pressure, eliminating lost gas. Drill cut-
method is more appropriate for in deeper well- tings are small granular particles of crushed coal
bore but is constrained in washout areas of the produced by surface rotary-drilled wellbores.
sidewall. Cuttings have shorter distance to the center of
Pressure coring is a self-contained system, the coal fragment, which promotes rapid rate
which captures and maintains the core at or of gas diffusion. Xue et al. (2011) devised new
332 6. RESOURCE EVALUATION METHODOLOGIES
(A) (B)
FIGURE 6.16 Conventional core barrel assembly (A) containing a split inner tube (B) for collecting continuous coal core in
the Powder River Basin, Wyoming.
drill cuttings sampling-while-drilling system into 0.6 m sections (Stricker et al., 2006). Long
based on double tubing drill holes and reversed canisters were used specifically for the
circulation of pressurized air. The drill cuttings USGS-BLM cooperative project in the Powder
sampling system is recommended for soft coal, River Basin, because of the thick coal beds. Using
which may be used in direct measurement of the long canisters permit more desorption of the
gas content. Coal samples are difficult to obtain thick coals. Standard procedure suggests 0.3 m
in soft coal due to wellbore deformation and long canisters to be used for desorption and
collapse. description of the cores but this author suggests
Prior to collection of conventional cores, each that the sooner the coal core is inserted in the
borehole will be drilled near the top of the coal canister and sealed the least gas is lost and coal
bed and will run a typically 15.2 cm steel pipe oxidized. Each coal section is inserted into a
casing. The coring will attempt recovery of a airtight canister especially designed and con-
continuous core, from bottom to top, of the structed by the USGS for gas desorption
coal bed by using a 7.6 cm diameter, 4.5 m long (Figure 6.17) (Stricker et al. 2006). The canister
split core barrel. Coring runs will be performed is sealed and purged with helium (inert gas) to
depending of the thickness of the coal. That is, make sure that there is no gas leak from the
if the thickness of the coal is 4.5 m, one run is canister. Gas leakage is tested by submerging
only attempted. If the thickness of the coal is the bottom and top of canisters into the
30.5 m, seven runs will be attempted and so on. water bath. The canister is immersed in
As part of the calculation of the total gas content, 76.2 91.4 cm-diameter plastic tanks filled with
during drilling the depths, time when coring water (water baths) heated by thermostatically
started, time when the core barrel started up controlled by water heaters and maintained to
the hole after coring, and time when the core bar- constant temperature to simulate reservoir tem-
rel arrived on the surface are recorded for use in perature of the coal bed in the ground. The water
calculation of lost gas. Once on the surface, the baths conduct more even heat to the plastic can-
split core barrel is opened and the coal is divided isters than the aluminum canisters.
DATA COLLECTION 333
PVC Pipe
68.6 x 11.4 x 0.6 cm
wall thickness
11.4 cm coupling
10.2 cm
plug
FIGURE 6.17 Design and dimensions of PVC pipe gas canister utilized by the USGS (Stricker et al. 2006) and Bureau of Land
Management coalbed gas cooperative project in the Powder River Basin, Wyoming, United States. Source: Flores et al., 2006.
BOX 6.1
canister ahead of time so the solvents used in aluminum is its high heat conduction leading
construction have evaporated to reduce or elimi- to faster temperature equilibration in the water
nate interference with measurements of gas tank. A major disadvantage is that aluminum
composition (wetness). is reactive metal, easily corroded especially by
Aluminum canisters are relatively heavy the low ph water associated with coal. The
and bulky. The aluminum canisters are not low pH water may react and produce
easily fabricated in the field and most designs hydrogen gas and alter gas composition and
require welding. A significant advantage of isotopic results (Barker et al., 2002).
FIGURE 6.19 Spreadsheet used for data to be recorded, collected, and managed, which are summarized in a table for the
lost, measured or desorbed, and residual gas as well as other information during desorption of a subbituminous coal in the
Powder River Basin (Stricker et al., 2006).
DATA COLLECTION 337
and cannot be transported on passenger planes Measurement of coal density of a low-ash bitu-
as baggage by US law regulation, (2) not minous coal using the water-filled headspace
readily available, (3) only low pressure helium method is exemplified according to Barker
can be used with safety in plastic-plugged PVC et al. (2002) of the following:
canisters, getting a vigorous flow of gas the
Measured by water addition: canister
valves of the canisters.
A disadvantage to water-filled headspace is headspace ¼ 1948g
that water promotes biological growth and Total canister volume ¼ Empty canister filled
chemically reactive. For example, it was known with water canister weight ¼ 2762 g
early in the testing for gas in the Powder River
Therefore, within this canister, coal volume is
Basin that operators add water in the canisters
equal to (2762 g 1948 g) (1 cc/g) ¼ 814 cc
to minimize oxidation. Our experiments tested
this process and found that during measure- Canister þ coal mass ¼ 2122 g;
ments of gas for a period of time there was a
spike in gas volume (Stricker & Flores, 2002). Canister mass ¼ 1028 g
This was interpreted to be a methanogenic
Coal mass ¼ 2122 g 1028 g ¼ 1094 g;
spike, which was determined as a significant
part of methanogenic pathway of generation Coal density ¼ 1094 g/814 cc ¼ 1.34 g/cc
of biogenic gas in the Powder River Basin
(Flores, McGarry, Stricker, 2005; Chapter 4). The above measurements plus the weight of the
Thus, as water in the canisters promotes metha- empty canisters are used to calculate the head-
nogenesis in low rank coal, carefully filling the space volume and density of the coal in each
headspace with water replacing empty space canister. These measurements plus the desorp-
in the coal and canister is still a beneficial proce- tion data collected on-site and in the Denver lab-
dure to measure water mass and headspace as oratory, in turn, will be used in the final
long as the procedure is done at the end and af- calculations of the gas content in cc/g or scf
ter the gas measurements when water is with- per ton of coal.
drawn as rapidly as it is filled. No lasting The calculations will utilize the Direct Method
effects to the coal quality are manifested by determination of gas content, which was devel-
this procedure. With just using distilled water oped originally to estimate the gas content of
for headspace measurement, it is possible to coals to be mined (Kissell, McCulloch, & Elder,
calculate easily the apparent sample density, 1973; Diamond, Schatzel, Garcia, & Ulery, 2001;
once the headspace volume in the canister is Australian Standard, 1999). The USBM Direct
measured. Method determines the most accurate estimate
Filling the headspace with water and of lost gas volume (gas lost from the coal sample
measuring the mass of the water added is a com- during the elapsed time from the start of desorp-
mon procedure. However, according to Barker tion to the time that the canister is sealed)
et al. (2002), the headspace volume is assumed compared to other methods. Thus, the total gas
equal to the weight of the water added to the content for each coal in the canister will be esti-
canister converted to volume by assuming that mated using the Direct Method lost gas proce-
1 gram (g) of water ¼ 1 cubic centimeter (cc) of dure plus the measured gas content (gas
headspace volume. These workers suggested volume released from the coal sample during
purging with helium reduces microbial activity desorption measurements divided by the coal
and that adding water is only required if mass) and residual gas content (volume of gas
the headspace is not purged by helium. remaining in the coal sample at the end of
338 6. RESOURCE EVALUATION METHODOLOGIES
canister desorption measurements divided by present more comprehensive analyses and meth-
the coal mass). The USGS has determined that odologies of these types of gas measurements.
the residual gas content for all the Powder River
Basin coals is approximately 2.5% of the total gas Lost Gas
content, although other analyses contain more The lost gas is volume of gas that cannot be
residual gas depending on the various labora- directly measured because it is desorbed during
tories employed. Using coal bed reservoir GIP drilling, coring, travel to the surface, and prep-
analysis for a specific purpose of gas resource aration of the coal sample before sealing the
assessment follows these calculations. canister. Lost gas makes a part of the total gas
content that escaped from the cores or cuttings
(e.g. small particles) caused by the gas diffu-
Types of Gas Measurements
sion, hydrostatic pressure variability, the time
The commonly used analytical methods for of retrieval, and depth of the borehole.
measurement of total gas content include deter- Waechter et al. (2004) suggested that sample
mination of lost, measured or desorbed, and re- preparation and drilling/coring methodology
sidual gases. Each of these measurement types also are important factors to lost gas. Lost gas
is measured separately and summed up estimate is extrapolated from the rate of gas desorption
the total gas content of the coal. The works of of the coal in the pressurized canister during
Mavor and Pratt (1996), Mavor, Pratt, Britton the early stages of desorption (e.g. first 9 h).
(1994), and McLennan, Schafer, and Pratt (1995) Extrapolation is made from the beginning of
BOX 6.2
FIGURE 6.20 Direct method graph showing cumulative gas measurements during desorption through time of a subbi-
tuminous coal in the Powder River Basin (Stricker et al., 2006). Straight line is fitted to the gas desorption data collected during
the first 9 h for lost gas. Gas continuous to be desorbed until it is depleted as measured gas. sm3/mt, standard cubic meter per
metric ton.
340 6. RESOURCE EVALUATION METHODOLOGIES
Stainless steel
(A) Drum (15.2 ×
20.3 cm)
(B)
FIGURE 6.21 Equipment used to measure residual gas from crushed coal varying in size from particles to fragments. Coal
is crushed by steel balls in a stainless steel drum on a tumbler (A) from which desorbed gas is measured by digital pressure
equipment (B). Source: Photographs courtesy of Gary Stricker.
recovery and sealing of the coal samples in the The direct method was developed by the
container to minimize lost gas, which Bertard U.S. Bureau of Mines (USBM) in the early
et al. (1970) considered suitable as long as it is 1970s to measure gas content in coal mines
less than 20% of the total gas content. After for general safety applications (Diamond
determination of the lost and measured gas, the et al., 2001). The USBM direct method was
coal sample was crushed to measure the remain- modified from Bertard et al. (1970) methodol-
ing gas volume using a graduated cylinder sus- ogy but adopted for core coal samples. In addi-
pended over a pan of water into which gas tion, the coal sample was collected in standard
released by crushing was collected. The lost gas aluminum canister (0.3-m long and 10.2 cm
plus measured gas plus residual gas is the diameter) for gas desorption. Gas volume was
total gas content of the coal expressed in cc/g measured by periodically releasing the pres-
or scf/t. sure resulting from accumulated gas in the
342 6. RESOURCE EVALUATION METHODOLOGIES
canister into a water-filled graduated cylinder. Schatzel (1998) expressed by the following
The lost gas, measured or desorbed gas, and equation:
remaining or residual gas were added to
make up the total gas content (Diamond et al. Qt ¼ ðQl þ Qm Þ=Mt þ Qr =Mc
1986). According to Diamond et al. (1986) the where:
lost gas volume (Ql) was estimated from Ql ¼ lost gas volume.
several direct measurements of the early gas Qm ¼ measured or desorbed gas volume.
desorption immediately after the coal was Qr ¼ remaining or residual gas volume.
sealed in the canister, which were made every Mt ¼ total air-dried mass (weight) of the coal
15 min for up to 9 h. The gas content data are sample.
plotted on a graph of cumulative desorbed Mc ¼ air-dried mass (weight) of the coal sam-
gas against the square root of time of desorp- ple crushed to a powder in the ball mill.
tion. The estimated lost gas volume is then The USBM modified the direct method in the
extrapolated from the initial linear portion of early 1980s as a result of assessment of low vol-
the desorption curve through the point on the umes of gas adsorbed within the organic matter
x-axis (see Figure 6.20). Ulery and Hyman associated with oil shale. The original direct
(1991) suggested improvement of the USBM method was not a sufficiently precise procedure
direct method by making notes on gas temper- to measure of the low gas content of oil shale
ature and ambient pressure to correct gas mea- such that a new technology was developed to
surements to standard temperature and more accurately measure the gas. Also, the new
pressure (STP). direct method increased the accuracy for estima-
The measured or desorbed gas (Qm) is calcu- tion of gas content of coal with measurement of
lated periodically using the water displacement differential pressure buildup between readings
procedure with the water-filled graduated cyl- and by the ideal gas law to calculate the volume
inder. After an extended period of time gas of gas released into the desorption canister
desorption reaches very slow rate with longer (Diamond et al., 1986). The new methodology
periods of time of measurement. Recommenda- abandoned the water displacement procedure
tion for termination of gas desorption varies and relies on an apparatus with pressure trans-
from (1) daily rate of less than 0.05 cm3/g by ducers and digital readouts. This eliminates
McCulloch et al. (1975), (2) average rate of problems of water-soluble gas and headspace
10 cm3/g daily for one week period by Dia- discrepancies.
mond and Levine (1981) to 0.05e10 cm3/g per The versatility of the direct method is that gas
day for one week period by Ulery and Hyman content of coal samples can be measured using
(1991). After calculations of measured or des- whole coal samples from wireline conventional
orbed gas are discontinued, the residual gas cores and cuttings or crushed samples. Whole
(Qr) is analyzed by crushing as much of the des- or core coal samples are used for prolonged
orbed coal sample to a powder (200 mesh) to desorption technique and crushed samples are
release the remaining in situ gas (Diamond & used for quick desorption technique. The pro-
Levine, 1981). The volume of gas released is longed desorption technique is mostly utilized
measured by the water displacement method in the United States to determine coalmine
after the sample cools to ambient temperature. methane to predict outburst, coalbed gas
The gas volume released by crushing is attrib- resource assessment (e.g. GIP), and basic
uted only to the portion of the sample crushed research on reservoir properties (Diamond &
to a powder. The total gas content Qt of the sam- Schatzel, 1998). The whole or core coal samples
ple is then estimated according to Diamond and are usually obtained from boreholes drilled on
DATA COLLECTION 343
the surface or through the coal mine face. The Smith and Williams (1981) devised a direct
prolonged desorption method allows the release method procedure to resolve the problem of
of gas to a very low rate of desorption, which accurately measuring the total gas content of
often takes from weeks to months (Stricker coal cuttings from rotary-drilled boreholes. The
et al., 2006). Thus, results of gas content mea- lost gas component of the total gas content was
surements are often long in coming, which is central to the water displacement measurement
not amenable to the quick results demanded by procedure. Smith and Williams (1981) assumed
the industry. However, Stricker et al. (2006) that no gas remains in the coal cuttings at the
devised a technique by which in two week pre- end of desorption period; thus, residual gas is
liminary results are provided as soon as the gas not determined. Due to the speed of gas desorp-
content (lost plus part of measured) curve is tion from coal cuttings, final results are generally
plotted and extended by a predictive formula. obtainable in one week.
Stricker et al. (2006) recommended that the cutoff
rate of desorption is 0.003 cc/g (0.09 scf/t). After Indirect Methods
prolonged gas desorption, the coal sample is Kim (1977) indirectly estimated gas contents
then crushed into small granules in order to from adsorption isotherm data. Adsorption iso-
release rapidly residual gas. therms measure the optimum adsorbed
Instead of waiting for an extended period of methane at varying pressures and at constant
time to obtain results of gas content measure- temperature (Chapter 4). Thus, an adsorption
ments, the coal samples (e.g. core, cuttings) isotherm is a measure of the maximum methane
are crushed for rapid analysis, which is sorption or idealized storage capacity of the
commonly used in Australia and elsewhere coal. Mavor and Nelson (1997) suggested if
(Saghafi & Williams, 1998; Saghafi, Williams & the in situ storage capacity is equal to the gas
Battino, 1998; Moore, 2012). The measured gas content if the coal is saturated with gas. Howev-
content and estimated total gas content is avail- er, most coal reservoirs are undersaturated with
able within days. Thus, the quick-crushing gas; thus, the gas content will be overestimated
method is suitable for coal mine applications in relationship to the adsorption isotherm. This
and sometimes coalbed gas resource assessment assumption is particularly correct in coal beds
by the industry. Sawyer et al. (1987) indicated at shallow depths. Nevertheless, Mavor and
that only 63% of the total gas content is desorbed Nelson (1997) and Mavor, Pratt & Nelson
with the quick-crush method because the lost, (1995) suggested that the adsorption isotherms
measured, and residual gas are difficult to deter- should serve as useful checks for gas contents,
mine completely. Saghafi and Williams (1998) which should be much lower than the
and Saghafi et al. (1998) indicated the accuracy adsorbed gas.
of measurement of gas content of coal using Another indirect method of estimating gas
rapid crushing technique. The rapid crushing contents includes using wireline logs (Diamond
technique accelerates the rate of gas desorption & Schatzel, 1998; Mullen, 1988, 1991; Hawkins
from coal, which allows measurement of gas et al., 1992). Scholes and Johnston (1993)
content in hours instead of days or weeks as defined wireline logs as measurements of one
necessary for the traditional method of desorp- or more physical properties of rocks as a func-
tion (Saghafi et al., 1998). The quick crushing tion of depths. The metal cable or wireline con-
technique is used when gas content estimates tains electrical conductors in the core that
are needed for either prevention of mine powers the downhole tool and communicates
outburst or immediate calculation of coalbed the toll signature back to the surface (Scholes
gas reserves. & Johnston, 1993). The signatures are displayed
344 6. RESOURCE EVALUATION METHODOLOGIES
as geophysical logs such as resistivity, gamma gas desorption and adsorption measurements
ray, acoustic or sonic, neutron, and density in the laboratory can be correlated to low bulk
(Scholes & Johnston (1993). The tool does not density permitting an estimate of both adsorbed
measure the gas content of the coal beds in the and desorbed gas volumes in coal. The relation-
wellbore but rather the logs are utilized as ship of maceral composition (e.g. vitrain-rich or
empirical estimation curves. Diamond, Murrie, woody, fusain, attritus) versus clay partings in
and McCulloch (1976) and McFall et al. (1986) a subbituminous coal in the Powder River Basin
determined gas contents from empirical estima- to a density log may yield more sensitive re-
tion curves of measured gas content results sponses than originally thought (Figure 6.22). A
plotted against depth and apparent rank of frequency diagram of measured coal densities
coal beds. These workers used algorithms of 590 coal cores of the subbituminous coal in
relating gas content to coal rank derived from the Power River Basin is shown in Figure 6.23.
actual total gas content and to proximate anal- This shows that the coal density mostly occurs
ysis of coal cores collected in the study area. within a relatively narrow range between 1.25
The log-derived coal property data are then and 1.3 g/cc.
used to estimate gas content in relation to coal Fu, Qin, Wang, and Rudolph (2009) use multi-
rank and depth. variate regression to establish The relationships
Recent workers have proposed an indirect between measured gas content and depth, as
method of estimating gas content using geophys- well as density, natural gamma, resistivity and
ical logs, which was reviewed by Saghafi, sonic logs were established by Hou, Hong and
Hatherly, and Pinetown (2011). The premise of Hao (2002) and Fu et al. (2009) conducted in
this concept is that coal consists of organic and various coalfields in China. Results were mixed
inorganic composition and that the gas content from high correlation coefficients (R2 ¼ 0.9) to
is related to the abundance of these components. negative correlation. Bhanja and Srivastava
Saghafi et al. (2011) suggested that the vertical (2008) estimated gas content for Jharia coalfield
and lateral changes of the inorganic component in India from sonic, density, photoelectric ab-
of the coal might respond to geophysical logs. sorption index (Pe), and gamma ray logs. These
Thus, the changes of the log responses might pro- authors asserted that the estimates of gas content
vide an estimate of the gas content of the coal. The of bituminous and lignite coals determined from
log responses might be typified by the density log the logs were calibrated with gas data from the
but Saghafi et al. (2011) indicated that using den- studied coalfield. Bhanja and Srivastava, 2008
sity log alone not in conjunction with other logs had proposed the following empirical equation
(e.g. gamma ray, sonic or induction, photoelectric to estimate the methane gas content of coal using
absorption), which also respond to variations in sonic, density, gamma ray and Pe logs for Jharia
the inorganic content of the coal, is an ineffective coalfield:
method. Saghafi et al. (2011) suggested that more
experimental and theoretical work is required V ¼ 0:767 fDt=ðrc GR PeÞg þ 10:67
before the relationships between the various where:
geophysical logs and gas content can be estab- V ¼ Volume of gas (cc/gm).
lished (Chapter 7). Dt ¼ Sonic log response for coal ¼ 120 ms/ft
Bowler and Tedesco (2004) proposed that (for Jharia coalfield).
bituminous coals have a low bulk density and rc ¼ Density log value for coal seam (g/cc).
are easily identified. Based on high volatile B GR ¼ Gamma ray response for coal (cps).
and A bituminous coal cores in the U.S. Illinois Pe ¼ Photo-electric absorption index value for
Basin, Illinois, and Forest City Basin, Kansas coal (1.8 b/electron for Jharia coalfield).
DATA COLLECTION 345
m m FIGURE 6.22 Continuous core section
D D of the Wyodak-Anderson subbituminous
F 50.0 51.0
F Clay/ coal in the Power River Basin, Wyoming
Attritus with a correlative density log. Coal mac-
50.2 eral and inorganic partings control the
Clay parting behavior of the log. D, density.
50.3
51.1
A
50.4 A
Massive Woody
F
50.5
51.2
Massive Woody
50.6
F
F
50.7
F
51.3
50.8
Coal parting
50.9
A = Attritus
51.0 51.5 F = Fusain
Scale in meters
1.3
1.4
1.5
1.15
1.25
1.35
1.45
of Boyer and Qingzhao (1998). These workers George coals as resource and reserve units
suggested that the prospect size area is impor- and recognized the coal zones as “plays.”
tant to gas producers and the basinal or regional Magoon and Dow (1994) defined a play is
area is significant to national and international one or more prospects and a prospect is a
institutions. possible trap to be investigated for commercial
Although the estimates of coal resources and accumulations of hydrocarbons in the context
reserves are keyed to the coal areas as defined of Petroleum Systems. Elsewhere, in the Raton
by Wood et al. (1983), coalbed gas resources Basin, Colorado and New Mexico, United
and reserves are often standardized using con- States the USGS assessed the Petroleum System
ventional oil and gas, and unconventional gas in the province as combined conventional and
assessment guidelines in terms of Petroleum unconventional assessment units (Higley,
Systems. For example, assessment of coalbed 2007). Thus, despite the sincere attempt to stan-
gas resources and reserves in the Powder River dardize methodology and terminology for
Basin in Wyoming and Montana, United States, resource and reserve assessments of coalbed
which is identified as a geologic-based prov- gas in relationship to conventional oil and
ince, assessed various Tertiary and Cretaceous gas, the process remains fluid. Boyer and
coal-bearing and associated sandstone units Qingzhao (1998) correctly noted that the
as unconventional and conventional assess- distinct characteristics of coalbed gas resources
ment units (Flores, 2004). In contrast, the gas and reserves prevent direct application of
industry in the Powder River Basin empha- established assessment guidelines for conven-
sized coal zones such as the Wyodak and Big tional oil and gas.
BOX 6.3
Coalbed gas
Bitumen
Shale gas
gas particularly estimation of economically and significantly across the assessment province
technically recoverable quantity of coalbed gas (Attanasi & Rice, 1995). The AU was evaluated
(Attanasi & Rice, 1995; Rice, Young, & Paul, by the (tested) cells, which are determined by
1995; Schmoker & Dyman, 1996; Schmoker & the spacing of coalbed gas wells permitted by
Klett, 2002, 2007). According to Attanasi and the State. The TPS was analyzed in the context
Rice (1995) the coalbed gas assessment is based of probabilistic methods by Schmoker and Klett
on the geologic elements of the TPS assessment (2002), which is best applied in risk analysis for
distinct in the province, which include the petroleum exploration ventures by Rose (2001)
geologic and stratigraphic frameworks, source (Figure 6.26).
and reservoir rock types, hydrocarbon genera-
tion, migration, trap and seal formations, and Case Study: TPS and Coalbed Gas
timing. The assessment is prepared to assess
Assessment
coalbed gas resources within an assessment
unit (AU) consisting of a volume of rocks with The USGS has assessed coalbed gas TPS in
similar characteristics and including discovered conjunction with the 2000 National Oil and Gas
and undiscovered coalbed gas resources. The Assessment program in various coal basins in
AU is organized to estimate continuous accumu- the conterminous United States and Alaska
lation (e.g. coalbed gas) and/or conventional (Flores, 2004; Higley, 2007; Roberts, 2008). For
accumulation (e.g. sandstones as structural and example, the coalbed gas TPS in the Powder River
stratigraphic traps) in which EUR, and gas and Basin in Wyoming and Montana was assessed in
water production profiles of wells vary 2000 consisting of the Tertiary-Upper Cretaceous
Coproduct ratios
(E.G. Gas composition; wet vs dry gas;
methane vs heavy hydrocarbons)
FIGURE 6.26 Flow chart showing the sequence of investigation using the TPS toward assessment of coalbed gas resources
and reserves. Source: Modified from Schmoker and Klett (2002).
ASSESSMENT OF COALBED GAS AS A PETROLEUM SYSTEM 355
coalbed gas TPS consisting of four AUs (Flores, groups of thirds are shown in Figure 6.28
2004). The AUs are based on the thickness and dis- (Cook, 2004). Based on this analysis the median
tribution of coal beds and the characteristics of total recovery per cell is 5.4 million m3 for the
associated fluvial channel sandstones. In three of first one-third, 6.8 million m3 for the second
the AUs the subbituminous coal beds served as one-third, and 5.9 million m3 for the third one-
source and reservoir rocks and in the fourth AU third (Flores, 2004). The increase in EURs is
the coal bed served as source for gas, which mainly explained by extensive draining of gas
migrated to the sandstones. from neighboring undrilled leases and subordi-
The coalbed gas development, which started nately by the use improved technology for
in the 1980s (Hobbs, 1978), was just maturing well completions. The primary life span of
when the USGS assessed the Powder River Ba- coalbed gas wells in the eastern part of the basin
sin in the year 2000 (Flores, 2004). Presently from 2000 to 2004 is about 7e8 years after which
most of the coalbed gas production (e.g. 70%) the well life was extended by booster pumps
is mainly from two coal zones or plays (e.g. (Figure 6.29).
Wyodak and Big George plays) in the upper The total area of the Upper Fort Union
part of the Paleocene Fort Union Formation AU in the Powder River Basin varies from
(Flores, Spear, Kinney, et al., 2010, Flores, Spear, a minimum of 8,447,000 acres (3,418,379 ha), a
Purchase, et al., 2010)). The Wyodak coal zone median of 8,892,000 acres (3,598,464 ha), and a
(Wyodak-Anderson) is shallow coalbed gas maximum of 9,337,000 acres (3,778,549 ha).
play along the eastern margin (see Figures Estimated coalbed gas drainage area per cell
6.10, 6.11 and 6.12) and the Big George (Wyo- having potential for additions to reserves in
dak-Anderson equivalent) is a deep coalbed the next 30 years is a minimum of 40 acres or
gas play in the central part of the Powder River 16 ha (16 wells per square mile), median of
Basin. The number of evaluated (tested) cells as 80 acres or 32 ha (8 wells per square mile),
of April 2000 in the upper part of Fort Union and a maximum of 640 acres or 258 ha (one
Formation AU was 3090 of which 2449 have well per square mile) as adopted from
EUR (Figure 6.27). The EURs from these cells permitted well spacing by the Wyoming Oil
are from 1179 producing coalbed gas wells, and Gas Conservation Commission. However,
which were completed at the end of 1999; how- because of increased issues of management of
ever, only 638 wells at the time had sufficient co-produced water the Wyoming Oil and Gas
production history to calculate EURs (Cook, Conservation Commission later changed the
2004; Flores, 2004). The 638 coalbed gas wells well spacing from 40 to 80 acres.
were completed mainly from the Wyodak coal
zone play along the shallow eastern margin of
Analog Method
the basin (Figure 6.27). The 638 wells were orga-
nized into groups of thirds with the first group Many coalfields and coal regions of the world
of one-third comprising the oldest and longest contain coal resources data but rarely coalbed
producing (1980e99) coalbed gas wells. The sec- gas data. Moore et al. (2009) and Moore and
ond group of one-third coalbed gas-producing Friedrich (2010) described the uncertainty that
wells was developed from 1998 to January comes with estimation of coal and coal bed
1999. The third group of one-third comprised resources with limited or no data. Traditionally
the newest and shortest producing (01/ and historically, in a given coalfield or coal re-
1999e11/1999) coalbed gas wells. The EURs of gion in the world, coal resources data are more
all the coalbed gas wells greater than the mini- available having been investigated longer than
mum size (0.56 million m3) analyzed in the coalbed gas data. Thus, estimation of coalbed
356 6. RESOURCE EVALUATION METHODOLOGIES
Montana
Wyoming
SHERIDAN
CROOK
Coal mines
WESTON
U.S.A.
JOHNSON CAMPBELL
Wyodak–Anderson
Boundary
NIOBRARA
NATRONA
Explanation
CONVERSE
Coalbed gas well
km
0 10 20 30 40 50
FIGURE 6.27 Map showing coalbed gas wells in the eastern part or shallow margin of the Powder River Basin, Wyoming
at the end of 1999. Source: Data source from Wyoming Oil and Gas Conservation Commission website. Blue dots are coalbed gas wells.
Red areas are coal mine sites.
gas resources contains more uncertainty than from lignite to anthracite rank coals in different
estimation of coal resources. Although coalbed geologic and tectonic settings from Pennsylva-
gas exploration and development have been nian to Miocene ages.
instituted for the past 20e30 years, data collec- The analog method is used when there is
tion have been concentrated in coalfields in the no or limited coalbed gas data (e.g. produc-
United States, Australia, and Canada. During tion, adsorption isotherm, gas content, gas
the last decade additional coalbed gas data composition, etc.), which typifies virgin, unex-
have been collected in China, Colombia, India, plored, and undeveloped areas. These coalbed
Indonesia, Japan, and some European countries. gas data may be used in virgin areas with the
Coalbed gas data (e.g. adsorption isotherms, gas same coal properties such as thickness, rank,
content, gas production, etc.) were collected composition, quality, thermal maturity, and
ASSESSMENT OF COALBED GAS AS A PETROLEUM SYSTEM 357
3
First one third
Second one third
Third one third
0.3
EUR (BCM)
Decline curves
0.03
0.003
0.0003
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Percentage of producing coalbed methane wells that has
EUR greater than minimum (0.0006 BCM) that were analyzed.
FIGURE 6.28 Distribution of estimate ultimate recovery for coalbed gas wells in the first, second, and third on-thirds
analyzed in the upper part of the Fort Union Formation in the Powder River Basin, Wyoming. EUR, estimate ultimate re-
covery; MMCF, million cubic feet. Source: Adopted from Flores (2004).
similar geologic and tectonic settings. For overlying Eocene Wasatch Formation AU, which
example, Flores et al. (2004) assessed the poten- was assessed by Flores (2004) as a hypothetical
tial of coalbed GIP resources of Alaska with AU. The majority of the test coalbed gas wells
emphasis on the Cook Inlet Basin using the Pow- in the Fort Union Formation drill though the
der River Basin coal bed analogs. The two basins overlying Wasatch Formation. That is, a test
are similar in coal rank, composition, and quality well is one that has penetrated and tested the
but slightly different in geologic and tectonic set- Upper Fort Union AU. Because there is only
tings. What is known of the Cook Inlet Basin is very limited or no coalbed gas wells that tested
that the sandstones contain biogenic gas, which the Wasatch Formation AU at the end of 1999,
migrated from the coal beds much like the Pow- the assessment used coalbed gas production
der River Basin (Magoon & Anders, 1990; Flores data as an analog from the tested coal beds in
et al. 2004; Warwick & Flores, 2008; Dawson the Upper Fort Union Formation AU. The
et al., 2012). In addition, limited information of coal beds included the Anderson coal bed of
gas content is available (Smith, 1995), which the Wyodak coal zone, which is stratigraphi-
can be used in the GIP assessment. This informa- cally closest bed, similar rank, coal quality,
tion was complimented by gas content and and thickness to the Wasatch coal beds. The
adsorption isotherms of comparable Powder Wasatch Formation AU had to have a mini-
River Basin coal beds in terms of rank and vitri- mum EUR per cell of 0.56 million m3 for assess-
nite reflectance. ment; however, 164 cells were estimated to
Even in the Powder River Basin where have EURs greater than 0.56 million m3. The
coalbed gas data abound in the Fort Union For- percentage of total assessment unit area in the
mation there is a general lack of data for the Wasatch Formation AU that has potential for
358 6. RESOURCE EVALUATION METHODOLOGIES
5,663,369
5,097,032
4,530,695
Production enhanced
3,964,359 by vacuum
pumps
3,398,022
Cubic meters
2,831,685
2,265,348
1,699,011
1,132,674
566,337
0
1995 1996 1997 1998 1999 2000 2001 2002
Years
FIGURE 6.29 Distribution of annual production of 30 coalbed gas wells in the eastern part of the Powder River Basin from
1995 to 2002. During the 7 year period the coalbed gas wells peaked and declined production in 2000 after which the wells
were connected to booster pumps to prolong their life. Source: Data source from Wyoming Oil and Gas Conservation Commission
website.
additions to reserves in the next 30 years was to be derived from producing wells from known
estimated at a minimum of 1%, a median of continuous accumulations with analogous
13%, and a maximum of 24%. This is based geologic characteristics, which Roberts (2008)
on the occurrence of net coal thickness of described as analog areas. The method of TPS
15 m in the west-central part of the Powder assessment is similar to that described in
River Basin in Wyoming. The total recovery the Powder River Basin by Flores (2004) in which
per cell for untested cells having potential for each continuous accumulation is assigned to
addition to reserves was estimated a minimum four coal-bearing formations as AUs. Roberts
of 0.56 million m3, median of 5.1 million m3, (2008) indicated that data necessary for the
and a maximum of 85 million m3. assessment of coalbed gas consist of: (1) area per
Roberts (2008) assessed the undiscovered, cell of untested cells, (2) percentage of the untested
technically recoverable coalbed gas resources in AU area, and (3) total recovery of petroleum per
the Cretaceous and Tertiary rocks of the North cell for untested cells having the potential for addi-
Slope and adjacent State waters where limited tions to reserves in 30 years. Comparison of vitri-
or no coalbed gas production is occurring. The nite reflectance (R0) of Tertiary lignite and
lack of coalbed gas production from reservoir subbituminous coal beds in the Sagavanirktok
accumulation necessitated cell drainage areas Formation ranges 0.33e0.52% R0 and is analogous
ASSESSMENT OF COALBED GAS AS A PETROLEUM SYSTEM 359
to the Tertiary Fort Union coal beds in the Powder follows: 1P (proved) and 2P (proved plus prob-
River Basin, which ranges from 0.31 to 0.47% R0 able) reserves, and 3P (proved plus probable
(Flores, 2004). Thus, Roberts (2008) selected the plus possible) resources (Figure 6.32). Barker
producing coalbed gas wells from Fort Union For- (2008) and Clarkson and Barker (2011) defined
mation as analogs for estimation of drainage area reserves in the context of PRMS (2011) as: (1)
and total recovery per cell of untested cells. The demonstrated commercial production poten-
Powder River Basin coal bed analogs were limited tial; (2) marketable and commercial gas compo-
to the Anderson or Canyon coal beds, which sition, gas content, and coal thickness; (3)
belong to the Wyodak coal zone (Flores, 2004). depth within accepted economic limits of a
The estimates of total recovery per cell in the Saga- coalbed gas fairway; (4) feasible development
vanirktok Formation coalbed gas AU were based plan that is economically viable under existing
on Powder River Basin analog wells with the market; and (5) development within a realistic
following: (1) a minimum EUR of 0.56 million m3 time period.
and (2) a median EUR of 5.1 million m3 was Barker (2008) and Clarkson and Barker (2011)
applied to untested cells in the AU. Roberts have discussed the current practices of classifica-
(2008) assumed that much like in the Powder River tion and estimation of coalbed gas resources and
Basin, gas was generated from coal beds by micro- reserves with adoption of the incremental
bial activity (Flores et al., 2005). The Powder River approach to delineation and development used
Basin coal bed analogs were also used in the Creta- by the coal mining industry and well spacing
ceous coal-bearing AUs in which the gas volumes used by the petroleum industry. The coal mining
per unit of acreage and ultimate recovery for indi- industry approach is the same as the Wood et al.
vidual wells might be slightly comparable. Rob- (1983) approach of coal resources and reserves
erts (2008) maximum estimate of 42.5 million m3 with increasing uncertainty as the geologic
is consistent with maximum EURs from coalbed assurance decreases or as the distance of control
gas well analogs in the Powder River Basin points (e.g. data) increases consequently the pro-
(Figure 6.30). gressive classification of proved, probable, and
possible reserves. According to Clarkson and
Barker (2011) the coal geologic assurance-based
Petroleum Industry PRMS
method applied to the petroleum well spacing
The concepts of petroleum resources classifi- method provides no precise assessment of the
cation were discussed and defined by Ross range of uncertainty in recovery efficiency,
(2011) and for additional information the which resulted in large reserves and different
reader is referred to the petroleum industry’s reserve growths. Thus, in order to comply with
guidelines for application of the PRMS the PRMS (2011) principles, Barker (2008) and
(PRMS, 2011). Ross (2011) defined petroleum- Clarkson and Barker (2011) offered the recent
related projects such as a coalbed gas accord- practices to define resource/reserve areas.
ing to its chances of commerciality by using: According to Barker (2008) and Clarkson
(1) reserves, (2) contingent resources, and (3) and Barker (2011) this definition presumes
prospective resources (Figure 6.31). Thus, for that the group of wells is established in the
a coalbed gas project that satisfies the require- coalbed gas fairway that is composed of a
ments for commerciality as set out in Ross laterally continuous coal thickness with satis-
(2011) and PRMS (2011), reserves may be factory gas content and permeability. Whether
assigned by three estimates according to recov- the lateral continuity of coal thickness is within
erable sales quantities and adopted by Barker the same correlated coal bed(s) cannot be ascer-
(2008) and Carlson and Barker (2011) as tained but discussed in the following section.
360 6. RESOURCE EVALUATION METHODOLOGIES
(A) 0.3
0.03
EUR (BCM)
0.004 BCM
0.003
0.0003
0.00003
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Percentage Sample
(B) 0.3
0.03
EUR (BCM)
0.0003
0.00003
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Percentage Sample
FIGURE 6.30 Distribution of estimate ultimate recovery (EUR) for the coalbed gas wells in the Anderson (A) and Canyon
(B) coal beds in the upper part of the Fort Union Formation, Powder River Basin, Wyoming. The EURs were used as analogs
for the Tertiary and Cretaceous coals in the North Slope, Alaska. Median EUR in each graph represented by the red dot. 0.13
BCFG ¼ 3.7 million m3. Source: Adopted from Roberts (2008).
ASSESSMENT OF COALBED GAS AS A PETROLEUM SYSTEM 361
Project maturity
Production sub-classes
On production
Commercial
Approved for
Reserves development
Development pending
Sub-commercial
Contingent Development unclarified
resources or on hold
Development
not viable
Unrecoverable
Prospect
Undiscovered
Prospective
Lead
PIIP
resources
Play
Unrecoverable
Range of uncertainty
FIGURE 6.31 The Petroleum Resource Management System classification is based on the range of uncertainty and
increasing chance of commerciality. Source: Adopted from Ross (2011).
These resources/reserve areas are well may be extended to farther distances as sup-
demonstrated in Figure 6.32 which starts at ported by coal geology and quality. The small-
the largest area or 3P with contingent resources est reserve area or 1P (Figure 6.32) contains
determined from drilling, logging, testing, and proved reserves, which include undeveloped,
sampling coalbed gas content and thickness to developed nonproducing, and developed pro-
establish enough presence of potential gas and ducing. Proved undeveloped is a small drainage
gas rate, which may be undemonstrated and area for producing and nonproducing wells
uneconomic (Barker, 2008). The 3P area contains proven to contain commercial and recoverable
mainly possible resources, wherein according to coalbed gas (Barker, 2008; Clarkson & Barker
well spacing rule the distance from probable lo- 2011). Well spacing varies from 80 to 320 acres
cations is two-well spacings. However, these (e.g. 1 acre ¼ 0.40 ha) in the United States and
can be extended to farther distances as sup- as much as 550 acres in Australia (Jenkins and
ported by coal and quality. The next smaller Boyer, 2008; King, 2008). Moore (2012) dis-
reserve area (Figure 6.32) contains probable cussed the certification of reserves of a coalbed
and proved reserves in which probable reserves gas project based on the PRMS (2011) methodol-
according to well spacing rules are two-well ogy. In order to achieve certification at the
spacings between proved locations. These also reserve level, commercial flow rates must be
362 6. RESOURCE EVALUATION METHODOLOGIES
Permit boundary
3P / or Outside
10
contingent fairway
00
resource
m
200
2P
m
FIGURE 6.32 Classification of resource and reserve areas (1P, 2P, and 3P) for coalbed gas with the depth limits of
commercial production ranging from 200 to 1000 m. 1P, proven reserves; 2P, proven plus probable reserves; 3P, proven plus
probable plus possible. Source: Adopted from Barker (2008; 2011).
demonstrated and be supported by the wells recently by Jenkins (2008), Moore et al., (2009)
containing gas data throughout the permit and Moore and Friedrich (2010). These workers
area, measured permeability tests, and demon- and others have emphasized more often than
strable areal extent of the reservoir (Moore, not that the sources of inaccuracies of resource
2012). Moore (2012) noted that the value of and reserve estimates begin from the inherent
most coalbed gas projects is based on the 2P re- heterogeneity of the coal beds and the vari-
serves in contrast to 1P for conventional gas ability of associated noncoal rocks. Jenkins
projects. (2008) suggested the parameters (e.g. areal
extent, thickness, gas, content, coal density,
coal rank, adsorption isotherms, etc.), which
METHODOLOGIES: ROOM FOR serve as input to determine gas resources and
IMPROVEMENT reserves using probabilistic techniques are
very variable physically and compositionally.
Improvement of methodologies for estima- Coal bed heterogeneity, in turn, depends on
tion of coalbed gas resources and reserves has depositional environments; hence the impor-
been suggested by many workers more tance of understanding non-coal lithologies
METHODOLOGIES: ROOM FOR IMPROVEMENT 363
associated with the coal beds. Moore et al. the GIP equation, use of coal resources values
(2009) indicated that the heterogeneities of is suggested for these two parameters. The A
these parameters and other input data intro- and t parameters of the GIP equations have
duce several levels of uncertainties, which af- been measured for coal resources in many
fects the accuracy of estimations of coalbed coalfields and coal regions throughout the
gas resources such as GIP. According to Moore world. During this process, the degree of un-
et al. (2009) the sources of parameter uncer- certainty of the lateral continuity and thickness
tainty include the following: (1) various mea- of coal beds relies on the distance between
surement errors, (2) simplification of a observed points of the same coal bed (Wood
complex geological model, (3) insufficient et al. 1983). That is the less the distance
data to distinguish the complex geological (<0.5 m) between observation points for the
model, and (4) uncertainty of the geological thickness and occurrence of the same coal
model itself. The ideal situation is to provide bed the more likely (“best” of Moore et al.,
a measure of value such as “worst,” “bad,” 2009) is the coal bed to be identified as coal re-
and “best” case scenario to express the uncer- serves. The more the distance (>0.5e1.25 m)
tainty of each parameter (Moore et al., 2009). between observation points of the same coal
For example, when the heterogeneity of a bed the less certain (“bad to worst” of Moore
parameter such as coal thickness is low (e.g. et al., 2009) the coal bed occurrence; thus the
nonsplitting, nonmerging, nonpinching later- coal should only calculated as probable re-
ally), the range of value from “best” to “worst” serves (“bad”) or hypothetical (“worst”) re-
is low. Thus, assigning a measure of value of sources. Thus, the degree of uncertainty of
uncertainty of parameters is a good start of the estimates is expected to decrease with
improvement of methodology. closer observation points (e.g. data). Therefore,
The diversity of GIP (see previous section) is the geologic assurance-based coal resource and
indicated that the methodology continues to be reserve estimates should be adopted for calcu-
improved to achieve better estimate of the lations of the coalbed GIP resources and re-
coalbed gas resources. However, recent im- serves where no well production data is
provements toward the estimation of coalbed available.
GIP are focused on the internal properties of
the coal reservoir such as adsorption isotherms,
Post Mortem with Hindsight
chemistry, reservoir pressure, burial, etc. (Chen
et al., 2011; Dallegge & Barker, 2000; Simpson, More than three decades have passed in the
2008). Although these parameters are important evaluation and extraction of coalbed gas and
to the GIP equations, one parameter on the we seem to be no closer to understanding a
other side of the equation, which has been scru- multidiscipline assessment of coal and coalbed
tinized toward its application to GIP estimation, gas resources and reserves. The concept that
is the A or study area or drainage area. This coalbed gas is an integral part of the natural
parameter is defined as the areal coverage and gas resource is valid; thus, the commodity
extent of a coal bed to be drained for gas. This must be assessed using the methodologies
definition assumes a high degree of lateral con- applied to petroleum system, however, the
tinuity and uniformity of thickness (t in the methodologies may be misplaced without
equation) of coal bed in a three-dimensional knowledge of the coal system (Warwick,
framework. 2005). The last decades of coalbed gas assess-
In order to avoid introducing more error ment should pave the way to improve the
than necessary to the A and t parameters of methodology in terms of lessons learned from
364 6. RESOURCE EVALUATION METHODOLOGIES
coalbed gas producing basins. The mature most important factor in the variability of esti-
coalbed-gas producing basins in the United mates of coal and coalbed GIP resources and
States, Canada, and Australia, provide lessons reserves. The accuracy of gas content applied
in exploration and development (Chapters to rapidly thickening and thinning coal beds
7e9). For example, the Powder River Basin will cause overestimation or underestimation
has been producing coalbed gas since early of GIP resources and successive calculations
1980s or about 30 years. Coalbed gas resource of recoverable resources as well as proved
assessments of the basin have been performed and unproved reserves, which in turn, are pro-
twice during the past 18 years (Flores, 2004; vided for certification (Moore, 2012). The strat-
Rice & Finn, 1996). The difference between igraphic distribution of coal beds controls the
Rice and Finn (1996) and Flores (2004) coalbed gas drainage area and gas volume, which plays
gas resources assessment is the amount of a major role in calculation of GIP and produc-
EURs of producing wells and more advanced tion. It is suggested that for greater accuracy
analysis of the coal stratigraphy and resources to resource estimations to separate coal bodies
and coal reservoir data, which was performed of split beds from merge coal bodies and to
by the National Coal Resources Project, and make separate calculations after correlations
the USGSeBureau of Land Management Coop- are properly performed using depositional en-
eration Project (Fort Union Coal Assessment vironments as a guide.
Team, 1999; Flores et al., 2006). Based on production history the coalbed gas
The coalbed gas assessment by Flores (2004) model in the Powder River Basin is expected to
and succeeding development in the Powder add reserves into the next 30 years (Flores,
River Basin has provided insights for improve- 2004). This expectation is deemed an insuffi-
ment of methodologies. The most critical cient assessment without additional informa-
parameter to understand at the outset of any tion on reservoir properties from the work of
assessment, exploration, and development is Flores et al. (2006). For example, the Flores
the coal geology and stratigraphy. Even (2004) assessment of the TPS Upper Fort Union
though, the basin contains abundant coal AU, which drew production data for EURs
resource and stratigraphic data, the informa- of wells concentrated in the eastern margin of
tion was not fully examined before use in the basin has not taken into consideration of
terms of coal reservoir continuity and charac- rapid development, which saw the increased
terization. The failure to define controls and infill drilling around producing wells in the
limits of coal beds led to inconsistent reservoir area from 2000 to 2009 adding more than
correlation and nomenclature, which in turn, 19,000 wells in 9 years. The work of Stricker
led to misidentification and confused reporting et al. (2006) showed that Wyodak coal zone tar-
of gas production. Furthermore, the problem geted for production the shallow eastern
created difficulties in gas drainage and protec- margin of the basin consisting of subbitumi-
tion of correlative rights, which are interest of nous C coal beds in contrast to the deeper cen-
different ownerships to develop the same reser- tral part of the basin, which contains
voirs in adjoining leases. The issue of coal bed subbituminous B and A coal beds. Coalbed
correlation or miscorrelation also leads to gas content values have been positively veri-
miscalculation of GIP resources and reserves, fied by workers (Stricker et al., 2006; Higley,
which is exacerbated by the merging, splitting, 2007) showing that higher rank coals contain
and pinching out of coal beds to form thick to more adsorbed gas. Thus TPS assessment unit
thin beds, respectively. Coal bed correlation based mainly on the shallow, lower rank coals
influenced by availability of data is a single to estimate addition to reserves for the next
METHODOLOGIES: ROOM FOR IMPROVEMENT 365
30 years with the higher rank and deeper coals high-pressure methane adsorption isotherms
(e.g. Big George coal zone) that still remain to of the same age and rank coals of the Powder
be developed may have pitfall. In addition ac- River Basin and those in Colombia, Indonesia,
cording to Flores (2004) the subbituminous and the Philippines show significant differences
coal in the Powder River Basin is typically un- in the gas storage capacity. Figures 6.33, 6.34
dersaturated. The subbituminous C and B are and 6.35 show the differences in adsorption iso-
2e3 times more undersaturated at 66% and therms of the subbituminous C, B, and A coals
57%, respectively, than subbituminous A coal in the Powder River Basin and Tertiary coal ba-
at 23%. Thus, the saturation of coal beds plays sins in Colombia, Indonesia, and the
an important role in determining a more real- Philippines. The gas storage capacity of subbi-
istic coalbed gas resources/reserves either in tuminous coals based on the isotherms in
the form of gas-in-place or production history Colombia, Indonesia, and the Philippines is
models. about twice as much as the same coal rank in
The use of the coalbed gas assessment in the the Powder River Basin. The major controlling
Powder River Basin as an analog to similar factor of the difference in adsorption isotherms
rank coal beds in the United States and world- is the vitrinite composition, which average
wide should be tempered with moderation. about 10e15% more in Colombia, Indonesia,
Recent studies of Flores, Spear, Kinney, et al. and the Philippines (Flores & Stricker, 2012).
(2010), Flores, Spear, Purchase, et al. (2010) Finally, based on adsorption isotherms of all
and Flores and Stricker (2012) comparing the coal ranks from these countries, the coals in
FIGURE 6.33 High-pressure methane adsorption isotherms of Tertiary subbituminous C coals of the Philippines and
Powder River Basin in the United States. Arbitrary comparison of the adsorbed gas at 8 MPa shows the Philippine coal store
about twice more gas than the Powder River Basin coals. MPa, megapascal; cc/gm, cubic centimeter per gram; sub, subbi-
tuminous. Source: Adopted from Flores and Stricker (2012).
366 6. RESOURCE EVALUATION METHODOLOGIES
FIGURE 6.34 High-pressure methane adsorption isotherms of Tertiary subbituminous B coals of Indonesia and Powder
River Basin in the United States. Arbitrary comparison of the adsorbed gas at 8 MPa shows the Indonesian coal store about
twice more gas than the Powder River Basin coals. MPa, megapascal; cc/gm, cubic centimeter per gram; sub, subbituminous.
Source: Adopted from Flores and Stricker (2012).
FIGURE 6.35 High-pressure methane adsorption isotherms of Tertiary subbituminous A coals of Colombia and Powder
River Basin in the United States. Arbitrary comparison of the adsorbed gas at 8 MPa (megapascal) shows the Colombia coal
store about twice more gas than the Powder River Basin coals. MPa, megapascal; cc/gm, cubic centimeter per gram; sub,
subbituminous. Source: Adopted from Flores and Stricker (2012).
SUMMARY 367
FIGURE 6.36 Relationships of adsorbed methane and apparent rank from lignite to anthracite of Pennsylvanian
to Miocene coals in the United States, Colombia, Indonesia, and the Philippines. Adsorbed methane in Indonesian and
Philippine coals at the boundary of subbituminous A to high volatile C bituminous rank is about 7.5 Scm/Mt more than the
United States coals. Source: Adopted from Flores and Stricker (2012).
Colombia, Indonesia, and the Philippines hold resources). The closer the distance between
more gas than coals in the United States data points the higher the certainty of existence
(Figure 6.36). of coal resources (e.g. inferred, indicated, and
measured reserves). Geologic assurance method-
ology is based on accurate stratigraphic correla-
SUMMARY tion for measurement of coal continuity,
thickness, and coal quality of the same coal
The key to implementing plans for coalbed bed. Assessments of coal resources and reserves
gas exploration and development projects is worldwide conform to the geologic assurance
the knowledge of coal resources and methods principle, which is modified from country to
of assessments. Understanding coal classifica- country according to depth, distance between
tion and categories will assist in the estimation observation points, thickness, and complexity
of coalbed gas resources. The critical feature of of the coal geology and structures of the study
coal resource and reserve estimates is the use areas.
of geologic assurance approach to impart cer- The coal resourceereserve methodology
tainty. Geologic assurance reflects the spacing does not directly translate to the petroleum
between observation points where coal thickness industry’s method of estimating coalbed gas
is measured and samples are collected for coal resourcesereserves because of lack of a range
quality analysis. The degree of assurance in- of uncertainty of gas recovery. The methods of
creases with more geologic and stratigraphic estimation of coalbed gas resources and re-
data, closer data points, and better quality serves adopted by the petroleum industry
data. The farther the distance between data include: volumetric, material balance, produc-
points the lower the certainty of existence of tion data analysis, reservoir simulation, and
coal resources (e.g. hypotheticalespeculative analog methods. The volumetric and
368 6. RESOURCE EVALUATION METHODOLOGIES
production data analysis are the most ray, density, resistivity, sonic, neutron) is
commonly used methods. another indirect method of measuring gas con-
The volumetric or GIP method is calculated tent. Other specialized measurements of gas con-
using gas drainage area, coal thickness, coal den- tent of coal utilize decline curve methods.
sity, and gas content plus other parameters to Gas production data analysis based on esti-
improve the estimates such as coal quality (e.g. mate of ultimate recovery (EUR) and spacing
ash and moisture content) and coal reservoir of coalbed gas wells have been adopted for esti-
conditions (e.g. pressure, water saturation, frac- mating reserves from the petroleum industry.
ture porosity), and adsorption isotherm or gas The methodology is demonstrated in case
storage capacity. studies of TPS assessments of coalbed gas in
An important parameter of the GIP method is the Powder River Basin, Wyoming and North
the gas content of the coal, which can be Slope, Alaska, which represent the production
collected and measured from wireline cores data analysis and analog methods, respectively.
and cuttings that are sealed in PVC or aluminum The PRMS methodology proposed standardiza-
canisters. tion of coalbed gas resource-reserve categories
The gas content is measured by desorption of to contingent resources and proved, possible,
gas from the coal cores or cuttings using the and probable reserves based on well spacings.
direct and indirect methods. The direct method The proved reserves are, in turn, subcategorized
is based on the desorbed gas from coal cores into undeveloped, and developed and proved
and cuttings and assumption that gas released reserves based on well spacing. The various
during desorption was proportional to the methodologies to estimate coalbed gas re-
square root of time. The indirect method is based sources and reserves are not perfect and have
on analysis of methane adsorption isotherms, room for improvement. The post mortem of
which measure the optimum adsorbed gas at the methodologies with hindsight of lessons
varying pressures and at constant temperature. learned from previous assessments is a good
Use of wireline or geophysical logs (e.g. gamma start for this process.