Journal of Natural Gas Science and Engineering 33 (2016) 988e1001

Contents lists available at ScienceDirect

Journal of Natural Gas Science and Engineering

journal homepage: www.elsevier.com/locate/jngse

Uncertainty in estimation of coalbed methane resources by geological

modelling
Fengde Zhou a, *, Zhenliang Guan b
a
School of Petroleum Engineering, University of New South Wales, NSW 2052, Australia
b
Key Laboratory of Tectonics and Petroleum Resources of Ministry of Education, China University of Geosciences, 430074, China

a r t i c l e i n f o a b s t r a c t

Article history: This paper presents an uncertainty analysis of coalbed methane resources estimation for a coal seam gas
Received 25 January 2016 field containing multiple coal seams. Firstly, logs from nine wells and laboratory data from nine coal
Received in revised form seams were used to predict the coal thickness, ash content and gas content at nine boreholes. Secondly,
31 March 2016
the structural models were determined using the well correlation, structural contour and coal seam
Accepted 4 April 2016
Available online 19 April 2016
thicknesses maps from coal mine data by the convergent gridder and sequential Gaussian simulation
methods. Then, distributions of coal density and ash content were generated in 3D by using sequential
Gaussian simulation based on the well log interpretations. Then, the distributions of gas content for each
Keywords:
Coalbed methane resources
cell were built by two methods; one is multivariable regression analysis in 3D and the other is sequential
Geological modelling Gaussian simulation based on the log interpreted gas content. Finally, coalbed methane resources were
Uncertainty analysis estimated based on the cell volume, coal density, net to gross ratio and gas content. In coalbed methane
resources estimation, four different density cutoffs were used to define the net to gross ratio of coal in 3D.
Results show that the gas contents decreases with increase in depth though with increases of vitrinite
reflectance ratio, fixed carbon content and pressure. Calculated gas content from linear multivariate
regression by using parameters of ash content, volatile matter content and fixed carbon content and
sample's burial depth matches well with laboratory measured values. The total coalbed methane re-
sources estimated are similar by the two geological modelling processes, the multivariable regression
analysis in 3D and the sequential Gaussian simulation. It has been found that the effects of coal density
cutoffs on coalbed methane resources for different coal seams are different.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction calculated based on the 3D distributions of coal quality and gas

Coalbed methane (CBM) resources Estimation is a complex
process (Jenkins, 2008) but it is important for planning and design
of coal seam gas production (Zhou et al., 2012). The resource esti-
mation is highly uncertain because it depends on the distributions
of coal thickness, coal quality, gas content, gas saturation, etc. But
the distribution predictions for those parameters are uncertain.
Hence, the uncertainty associated with the estimation of CBM
resource must be assessed (Zhou et al., 2012, 2015).
Three techniques, field analogs, volumetric methods and prob-
abilistic methods were used in CBM resources estimation (Jenkins,
2008). For the volumetric method, the CBM resources are
* Corresponding author. University of Queensland, School of Earth Sciences, QLD
4072, Australia.
E-mail address: f.zhou@uq.edu.au (F. Zhou).
content. Coal seam thickness, coal density, and gas content are
considered to be major parameters that introduce the uncertainty
in resource estimation (Wang et al., 1997; Zuo et al., 2009). Hence,
the uncertainty in calculating of CBM resources comes mainly from
two sources, well log interpretation and predicted distributions of
coal thickness, coal quality and gas content (Zhou et al., 2012). Note
that parameters related to 3D modelling, e.g. drilling density,
structural uncertainty (well correlation errors, fault geometry,
isopach/isochore issues) and gas saturation are also important for
uncertainty analysis but beyond the range of this paper.
Log interpretation by integrating laboratory and log data is
carried out firstly to estimate the coal quality and gas content at
boreholes. Next, the statistical analysis is carried out to characterise
coal quality and gas content in horizontal and vertical. Thus the
generated data are used in geological modelling. In laboratory, coal
gas content, weight percent of moisture, fixed carbon and volatile
matter content, desorption time and coal bulk density are normally

http://dx.doi.org/10.1016/j.jngse.2016.04.017
1875-5100/© 2016 Elsevier B.V. All rights reserved.
 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001 989

measured for each sample. Coal maceral compositions, e.g. vitrinite Age Coal Seams Lithology
Formation/Stages
reflectance ratio (VRO, %), vitrinite, liptinite and inertinite, are
measured in separate experiment. Coal ash and sodium content Early Cretaceous Dolerite dykes Mica
lamprophyre dykes and
(Heriawan and Koike, 2008a), volatile matter content (Hagelskamp Lohpiti sills
et al., 1988), sulfur content (Bancroft and Hobbs, 1986; Beretta et al.,

Upper
2010) need to be studied as part of the assessment of coal quality. Telmuchia Fine grained

Late Permian
The coal quality parameters were used to predict gas content (Zhou feldspathic

Raniganj
et al., 2012). Nolde and Spears (1998) calculated the gas content Jamdih sandstones, shales
using a linear regression equation in which the depth is the only with coal seams,
independent variable. According to Nolde and Spears, the gas maximum
thickness=725m

Lower
content increases with increase in depth. The authors presented a
Murlidih
model equation based on the desorption values for the 61 low
volatile bituminous coal samples. Fu et al. (2009) estimated the gas
content using multivariable regression analysis by relating data,

Middle Permian
Massive buff coloured

Baren Measure
Mahuda

Upper
such as burial depth, resistivity, sonic slowness and density log sandstones, grey shale,
(RHOB) with measured gas contents from 64 samples. Only one of Hariharpur carbonaceous shale
the five reported equations by Fu et al. (2009) shows that the gas with/without very thin
Petia lenses of coal, maximum

Lower
content decreases with increase in burial depth but the reason was
thickness=625m
not represented. Gas content was also estimated by relating with Shibabudih
the pressure, temperate, and the ratio of fixed carbon over volatile
matter content in wt% (Kim, 1977). XVIII
The log derived data is geostatistical analyzed to estimate dis-

Upper
tributions of coal thickness (Jakeman, 1980; Mastalerz and Phularitand XVII Buff coloured
Kenneth, 1994), coal quality (Cairncross and Cadle, 1988; medium and coarse
XVI grained sandstone,
Hagelskamp et al., 1988; Liu et al., 2005; Heriawan and Koike, grits, shale,
2008b; Beretta et al., 2010; Hindistan et al., 2010) and coal Early Permian
Brari XV carbonaceous shale,

Barakar
tonnage (Heriawan and Koike, 2008a). Sequential Gaussian simu- XIV siltstone and thick
XIII workable coal
lation (SGS) and ordinary Kriging methods are used to predict the Middle
XII
XI seams, maximum
distribution of coal thickness and coal quality (Beretta et al., 2010). X thickness=1250m
Zhou et al. (2012) presented an uncertainty analysis of CBM re- Nardkarki IX
sources using the stochastic reservoir modelling for a CBM field in VIII

southeast Qinshui Basin from China. They reported that the het- V/VI/VII
erogeneities in coal seam thickness and coal quality lead to IV Greenish shale, very
Lower

increasing uncertainty in estimating CBM resources. The authors Goladadih III fine grained sandstone,
II
I sandy shale,
also reported that the density variogram and coal seam roof surface
conglomerate and basal
contribute less to the uncertainty estimation of CBM resources. But Talchir tillite
sufficient data, sound log interpretation models and appropriate Archaean Metamorphics
geological methods can improve the reliability of resource
estimation. Fig. 1. A schematic of the stratigraphic succession showing coal seams. Coal seams
In this paper, the uncertainty in CBM resources estimation is numbers in rectangles are objectives of this study (After Saikia and Sarkar, 2013).

analyzed for a producing CBM field containing multiple coal seams.

Well logs, laboratory data and the data obtained from coal mining
difficult (Jeremic, 1985). The fault's opening ability affect the gas
are used to predict the coal structure, coal thickness, coal quality
saturation in coal and the dewatering process because the open
and gas content. Noteworthy is that the gas content distribution is
fault will be a conduit for gas and water flow.
estimated by using two different techniques: the multivariable
In the study area, logs are used to correlate coal seams with their
regression analysis in 3D and the SGS. The uncertainty is assessed
characteristics, namely stratigraphic sequence, thickness, variation
firstly by analyzing the effects of coal density cutoffs on resource
of log attributes etc. Fig. 2 shows two cross sections of the nine
estimation, secondly comparing the resource estimation based on
wells. Results show that the coal seams can be identified by typical
the distributed gas content by the multivariable regression analysis
values of bulk density (RHOB) and gamma ray (GR). The sandstone
in 3D and the SGS analysis and finally, analyzing the relationships of
sequences are characterized by lower GR response, however, higher
coal bulk density with ash, fixed carbon, and volatile matter con-
RHOB, which is similar to that of the mudstone; GR response in
tents. It is assumed that the statistics of gas content, proximate
mudstone is the highest of the three. Archaean gneiss is located at
analysis and logging data from the nine wells can be taken as
the south of the study area (Paul and Chatterjee, 2011; Singh et al.,
representative for the study area.
2013; Saikia and Sarkar, 2013) which was intersected by the well #F
(Fig. 2) at its deepest section. The Archaean gneiss is granite gneiss
2. Data
and characterized by high RHOB and low GR.
Structural map on the roofs of each coal seam and the thickness
2.1. Field data
contour maps of each coal seam were drawn by using coal mining
and coal seam gas boreholes. Fig. 3 shows the maps of structure and
Fig. 1 shows the litho-stratigraphic succession of coal seam XVIII
coal seam thickness for one of the coal seam No. IIIeIV. It shows
to coal seam I from top to bottom. These coal seams dip gently with
that there are 17 faults and 14 of them are normal faults. Three
dip angles ranging from 5
to 15
except in the vicinity of fault and
reverse faults are located in tectonic transition area. Coal seam No.
the seams dip towards the basin centre in general (Saikia and
IIIeIV is thicker in the middle and north part of the study area.
Sarkar, 2013). The mining methods for gently dip seam are the
same as in flat seams but mining conditions are a little more
990 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001

Fig. 2. Two well correlation sections, (a) Section #S1 and (b) Section #S2. MD in the first track is measured depth in m, RHOB in the second track is coal bulk density in g/cm3, and GR
in the third track is gamma ray in API. Distances between wells are not to scale.

2.2. Laboratory data
A summary of desorption and coal quality data obtained from 68
coal samples from five wells (wells No. A, B, C, D and E in Fig. 3) is
presented in Table 1. The desorption data include the total gas
content, natural desorption gas content, losing gas content, residual
gas content and desorption time. Coal quality tests include the
moisture content, ash content, fixed carbon content, volatile matter
content and coal density. Table 2 lists the correlation coefficient
matrix for the nine properties of the 68 samples which are listed in
Table 1. Fig. 4 shows the variation of raw gas content, moisture
content, ash content, volatile matter content, and fixed carbon with
depth. Results show that the raw gas content (Q) and volatile
matter content (VM) have negative relationships with depth; ash
content (ASHR) shows a negative proportionality with fixed carbon
(FC).
 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001 991

Fig. 3. Elevation on the roof of coal seam No. IIIeIV (a), and thickness contour map of coal seam No. IIIeIV (b).

3. Results rc is the measured coal bulk density in g/cm3.

3.1. Log interpretation
3.1.1. Coal quality
The proximate laboratory analysis provides the contents of ash,
moisture, fixed carbon and volatile matter as shown in Table 1.
Among these parameters, ash content (wt%) can be related to the
measured coal density. Fig. 5a shows the scatter plot of the
measured coal bulk density and the ash content. The correlation
coefficient (R2) has been found to be 0.74. The equation is:
ASHR ¼ 97:67$rc  114:13
where, ASHR is the calculated ash content in wt% as received basis,
In this study the lab measured density is close to the log derived
RHOB. Hence, we replace rc with RHOB in Eq. (1).
Fig. 5b shows the scatter-plot of fixed carbon content and ash
content and relationship derived from the regression analysis. The
model can be expressed as:
FC ¼ 0:902  ASHR þ 79:056 (2)
where, FC is the calculated fixed carbon content in wt% as received
basis.
Fig. 5c presents the relationship between volatile matter content
(1) with ash content. Results show that the relationship between them
is weak. This is because the volatile matter content is not only
related to ash content also to the depth and VRO (Amijaya and
992 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001

Table 1
Coal samples adsorption and coal quality industrial tests.

Sample number Mid-point depth, m Q1 (%) Q2 (%) Q3 (%) Moist (wt%) ASHR (wt%) rc (g/cc) VM (wt%) FC (wt%) VMdaf (wt%) FCdaf (wt%) Q (m3/t) Qdaf (m3/t) t (d)

A-1 495.86 // // // 2.0 23.5 // 27.95 46.58 37.5 62.5 10.71 14.37 //
A-2 497.525 // // // 2.0 26.9 // 24.18 46.95 34.0 66.0 11.27 15.84 //
A-3 498.475 // // // 1.3 17.3 // 16.68 64.69 20.5 79.5 10.71 13.16 //
A-4 503.25 // // // 1.9 26.0 // 24.51 47.59 34.0 66.0 11.33 15.71 //
B-1 548.055 30 64 6 0.2 8.1 1.28 23.28 68.39 25.4 74.6 15.08 16.45 5.75
B-2 551.625 26 69 5 0.2 10.2 1.27 24.28 65.30 27.1 72.9 14.29 15.95 2.08
B-3 553.455 27 66 7 0.3 20.5 1.31 23.37 55.84 29.5 70.5 10.50 13.25 2.79
B-4 554.38 22 74 4 0.4 14.8 1.34 23.59 61.26 27.8 72.2 13.62 16.05 1.83
B-5 556.185 29 68 4 0.0 21.4 1.46 26.79 51.78 34.1 65.9 10.87 13.83 1.73
B-7 613.25 30 65 4 0.1 9.6 1.34 23.08 67.18 25.6 74.4 14.49 16.05 1.97
B-8 615.845 21 72 6 0.1 7.4 1.32 21.76 70.71 23.5 76.5 13.06 14.12 1.57
B-9 723.795 31 66 2 0.2 12.4 1.42 18.63 68.81 21.3 78.7 14.89 17.03 1.97
B-10 724.6 21 75 5 0.1 20.5 1.23 17.75 61.72 22.3 77.7 11.96 15.05 0.84
B-12 788.31 28 64 8 0.1 16.5 1.39 17.90 65.52 21.5 78.5 14.54 17.43 3.40
B-14 812.875 28 65 7 0.1 19.0 1.34 24.53 56.40 30.3 69.7 13.23 16.35 2.50
B-15 814.505 24 67 9 0.2 16.5 1.34 19.33 64.03 23.2 76.8 14.26 17.11 1.83
B-17 1001.805 24 64 11 0.4 24.3 1.39 17.38 57.93 23.1 76.9 12.68 16.95 3.08
B-18 1002.51 18 65 17 0.3 24.6 1.36 19.05 56.04 25.4 74.6 11.27 15.01 3.26
B-19 1013.59 21 72 7 0.2 12.8 1.40 16.27 70.78 18.7 81.3 12.10 13.90 1.64
B-22 1063.07 22 52 26 0.0 35.6 1.50 13.68 50.66 21.3 78.7 10.14 15.76 //
B-23 1066.845 9 83 8 0.3 48.1 1.64 11.05 40.53 21.4 78.6 8.79 17.04 2.62
B-24 1068.58 26 55 19 0.0 45.4 1.63 12.04 42.54 22.1 77.9 13.40 20.71
B-25 1072.215 13 79 8 0.0 45.2 1.58 17.94 36.80 32.8 67.2 10.66 19.47 2.71
B-26 1073.68 16 74 9 0.0 32.7 1.47 15.23 52.03 22.6 77.4 13.20 19.63 2.67
B-27 1120.125 28 66 6 0.1 35.9 1.46 22.46 41.56 35.1 64.9 11.78 18.40 6.83
B-28 1121.45 33 63 5 0.2 30.9 1.44 12.60 56.34 18.3 81.7 10.79 15.66 9.75
B-29 1124.685 3 54 42 0.0 45.2 1.58 17.94 36.80 32.8 67.2 8.52 15.56 //
B-30 1137.665 20 68 11 0.0 23.1 1.50 13.96 62.91 18.2 81.8 11.06 14.39 7.29
C-1 755.42 19 71 9 0.5 23.0 // 22.59 53.96 29.5 70.5 10.33 18.41 2.45
C-2 756.305 22 68 7 0.4 44.7 // 17.26 37.66 31.4 68.6 9.20 16.62 3.50
C-3 896.26 19 78 4 0.5 22.4 // 18.53 58.64 24.0 76.0 9.07 11.67 1.67
C-4 897.93 21 75 4 0.3 12.1 // 20.57 67.01 23.5 76.5 12.07 13.74 1.75
C-5 958.96 26 67 8 0.2 30.1 // 18.06 51.62 25.9 74.1 9.90 14.16 3.42
C-6 961.64 24 67 9 0.4 12.3 // 24.29 63.05 27.8 72.2 10.67 12.16 4.17
C-7 964.22 20 69 11 0.4 15.7 // 17.06 66.89 20.3 79.7 11.13 13.20 3.42
C-8 1019.8 14 76 11 0.7 23.0 // 18.75 57.59 24.6 75.4 9.72 12.62 3.21
C-9 1021.935 18 76 6 0.7 23.8 // 20.00 55.52 26.5 73.5 11.75 15.42 2.79
C-10 1025.225 19 81 0 0.7 10.0 // 19.59 69.72 21.9 78.1 12.27 13.64 2.08
C-11 1028.41 13 77 10 0.5 36.5 // 19.79 43.26 31.4 68.6 10.94 17.23 2.63
C-12 1030.07 14 18 6 0.3 9.9 // 20.67 69.15 23.0 77.0 11.89 13.20 2.29
C-13 1032.67 19 74 7 0.7 16.6 // 16.73 66.03 20.2 79.8 11.08 13.24 3.00
C-14 1089.68 13 63 24 0.4 16.7 // 18.83 64.03 22.7 77.3 9.80 11.77 7.83
C-15 1096.64 15 54 30 0.5 27.0 // 18.65 53.82 25.7 74.3 7.25 9.94 7.17
D-2 318.72 8 72 21 0.7 8.3 // 36.21 54.76 39.8 60.2 6.90 7.90 3.70
D-3 518.08 26 63 11 0.5 7.0 // 32.83 59.65 35.5 64.5 4.60 9.10 4.80
D-5 706.625 27 67 6 0.6 11.9 // 16.98 70.55 19.4 80.6 10.00 13.00 1.00
D-6 769.35 20 76 5 0.3 7.7 // 25.48 66.49 27.7 72.3 12.60 17.70 2.20
D-9 849.135 22 75 3 0.6 35.0 // 11.84 52.52 18.4 81.6 12.70 15.60 1.30
D-13 1057.555 27 62 11 0.5 2.1 // 20.64 76.73 21.2 78.8 11.50 16.60 0.86
D-16 1146.24 33 65 3 0.5 3.3 // 19.91 76.27 20.7 79.3 11.10 13.40 0.82
D-17 1147.46 31 64 5 0.5 13.7 // 17.92 67.84 20.9 79.1 12.90 15.70 0.82
D-18 1149.145 28 69 3 0.2 13.6 // 17.59 68.63 20.4 79.6 10.80 12.60 1.00
D-19 1214.95 30 64 6 0.9 11.1 // 12.50 75.55 14.2 85.8 12.60 16.60 4.94
E-1 1042.78 24 49 27 0.46 72.33 // 25.91 1.30 95.22 4.8 1.20 4.41 //
E-2 1116.97 15 80 4 2.5 22.75 // 3.59 71.16 4.80 95.2 5.06 6.77 //
E-3 1118.13 6 90 4 0.9 26.76 // 7.59 64.75 10.49 89.5 6.34 8.76 //
E-4 1130.47 11 85 4 0.82 14.75 // 17.69 66.74 20.95 79.0 8.57 10.15 //
E-5 1131.32 15 81 3 0.71 27.60 // 17.93 53.76 25.01 75.0 8.92 12.44 //
E-6 1183.895 7 90 4 0.86 39.13 // 10.53 49.48 17.55 82.5 7.59 12.65 //
E-7 1185.11 13 82 5 0.82 18.53 // 16.05 64.60 19.90 80.1 7.66 9.50 //
E-8 1409.99 9 32 59 0.8 19.28 // 10 69.92 12.51 87.5 3.37 4.22 //
E-9 1412.74 18 70 12 0.6 18.5 // 9.9 71.00 12.24 87.8 5.16 6.38 //
E-10 1417.19 22 34 43 0.90 12.80 // 15.4 70.90 17.84 82.2 4.44 5.14 //
E-11 1418.485 7 79 15 0.7 14 // 9.78 75.52 11.47 88.5 7.45 8.73 //
E-12 1420.835 12 66 21 0.8 18.77 // 9.89 70.54 12.30 87.7 4.68 5.82 //
E-13 1422.91 18 57 25 0.59 57.46 // 14.43 27.52 34.40 65.6 4.15 9.89 //
E-14 1459.31 19 52 29 0.9 9.83 // 8.6 80.67 9.63 90.4 5.06 5.67 //
E-15 1462.175 20 52 28 0.67 19.69 // 10.57 69.07 13.27 86.7 4.53 5.69 //

Note: Q1 ¼ naturally desorption gas content, Q2 ¼ lost gas content, Q3 ¼ residual gas content, Moist ¼ moisture content as received basis, ASHR ¼ ash content as received basis,
VM ¼ volatile matter content as received basis, FCN ¼ fixed carbon content as received basis, rc ¼ measured coal bulk density, daf ¼ dry-ash free basis, Q ¼ gas content as
received basis (Q ¼ Q1 þ Q2 þ Q3), Qdaf ¼ gas content as dry-ash free basis, t ¼ desorption time. wt% ¼ weight percent.

Littke, 2006). Amijaya and Littke (2006) also reported that the volatile matter content decreases by about 20e30 wt% (daf) for
 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001 993

Table 2
Correlation coefficient matrix for the various properties of 68 samples in study area.

Depth (m) Moist (wt%) ASHR (wt%) VM (wt%) FCN (wt%) VMdaf (wt%) FCdaf (wt%) Q (m3/t) Qdaf (m3/t)

Depth (m) 1.00

Moist (wt%) 0.03 1.00
ASHR (wt%) 0.21 0.05 1.00
VM (wt%) 0.74 0.14 0.20 1.00
FC (wt%) 0.12 0.08 0.90 0.23 1.00
VMdaf (wt%) 0.37 0.11 0.49 0.65 0.77 1.00
FCdaf (wt%) 0.37 0.11 0.49 0.65 0.77 1.00 1.00
Q (m3/t) 0.50 0.36 0.31 0.29 0.19 0.10 0.10 1.00
Qdaf (m3/t) 0.42 0.39 0.09 0.25 0.18 0.06 0.06 0.86 1.00

Q (m3/t) Moist (wt%) ASHR (wt%) VM (wt%) FC (wt%)

0 10 20 0 2 4 0 50 100 0 20 40 0 50 100
0
(a) (b) (c) (d) (e)

200

400

600
Depth (m)

800

1000

1200

1400

1600

Fig. 4. Variation of (a) raw gas content, (b) moisture content, (c) ash content, (d) volatile matter content, and (e) fixed carbon with burial depth.

each 1% VRO increase in the range between 0.5 and 2.0% VRO. Hence
we first interpreted the moisture content and then calculated the
volatile matter content. Fig. 6 shows the plot of the measured ash
content and moisture content which is quite scatter and therefore
the relationship between two parameters is weak. The measured
average, minimum and maximum moisture contents are 0.51%,
0.02% and 2.5% respectively for the 68 samples presented in Table 1.
The equation of moisture content is:
MOIN ¼ 0:0043  ASHR þ 0:5967
Then we have calculated the volatile matter content based on
material balance using the equation:
VM ¼ 100  ASHR  MOIN  FC
where, VM is the calculated volatile matter content in wt% and
MOIN is the moisture content in wt% as received basis.
Fig. 4a shows the gas content with depth. It can be seen from the
figure that the gas content decreases with increase in depth. This is
due to the fact that the temperature, which has negative effect on
gas content, increases with increase in depth. Although VRO in-
creases with increase in depth and has a positive effect, its overall
effect on gas content is low (see Fig. 8a; Laxminarayana and
Crosdale, 1999). Faiz et al. (2007) reported that for a depth up to
about 600 m the average in-situ gas content increases. With further
increase in depth (up to 900 m) the gas content tends to plateau or
(3) even decreases which is consistent with the combined effects of
temperature and pressure on the gas sorption capacity during the
geological history. Fig. 8 also shows the changing of vitrinite and
liptinite with depth in which data are from wells #A, B, C, D and E.
(4) Results indicate that the coal rank in the study area is medium to
high volatile bituminous coal.
In this study, a regression equation is used to calculate gas
content as a function of ash content, volatile matter content and
fixed carbon content and sample's burial depth. The equation is:

3.1.2. Gas content

Q ¼ 0:0067  D þ 2:4728  ASHR þ 2:5364  VM þ 2:5468
68 samples were used to estimate the gas content for those nine
wells. The average, minimum and maximum gas contents are 10.2%,  FC  234:72
1.2% and 15.1%, respectively (see Fig. 7). The average gas contents in (5)
wells #A, B, C, D and E are 11.01 m3/t (4 samples), 12.00 m3/t (24
samples), 10.05 m3/t (15 samples), 10.74 m3/t (10 samples) and where, Q is the calculated raw gas content in m3/t as received basis,
5.61 m3/t (15 samples), respectively. D is the burial depth in m.
994 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001

60 100
y = 97.67x - 114.13 (a) 90 (b)
50 80
R² = 0.7418
70
40

FC (wt%)
ASHR (wt%)
60
30 50
40
20 30
20 y = -0.9088x + 79.241
10
10 R² = 0.8188
0 0
1 1.2 1.4 1.6 1.8 0 10 20 30 40 50 60 70 80 90 100
ρc (m3/t) ASHR (wt%)

40 16
35 (c) (d)
14
30 12
VM (wt%)

Q (wt%)
25 10
20 8
15 6
10 4
y = -0.0892x + 20.184 y = -0.0728x + 11.7
5 2
R² = 0.041 R² = 0.0938
0 0
0 20 40 60 80 0 20 40 60 80
ASHR (wt%) ASHR (wt%)

Fig. 5. Relationships between (a) ash content and coal bulk density, (b) ash content and fixed carbon content, (c) ash content and volatile matter content and (d) ash content and
raw gas content. The relationship between ash content and volatile matter content is weak because the volatile matter content is not only related with ash content but depth and
VRO. The gas content is not only related with ash content but depth, reservoir pressure, temperature, etc.

3 16
Histogram of measured gas content min=1.20
14 max=15.08
2.5 Normal distribu on
average=10.21
12
std.devia on=3.10
2
MOIN (wt.%)

10 kurtosis=0.23
Percentage (%)

1.5 8

6
1
4
0.5 2

0 0
3-4
4-5
0-1
1-2
2-3

5-6
6-7
7-8
8-9
9-10

>20
14-15
10-11
11-12
12-13
13-14

15-16
16-17
17-18
18-19
19-20
0 20 40 60 80
Ash content (wt%)
Gas content (m3/t)
Fig. 6. Relationship between the measured ash content and moisture content. Samples
Fig. 7. Histogram of the measured gas content on as received basis and the best fitting
were collected from wells #A, B, C, D, and E.
normal distribution.

Fig. 9 shows the relationship between the measured and

calculated gas content as received basis. Results show that the
calculated gas content by regression matches well with the
measured gas content. But it also shows that at low gas content the
measured values of some points are lower than the calculated and
at high GC, the opposite.
3.1.3. Coal bulk density cutoff
The coal bulk density, which is used to identify coal, typically
referred to the cutoff density. The density cutoff values affect coal
seam thickness, average seam density and gas in place calculation
(Seidle, 2011). Seidle (2011) also reported that the standard coal
density cutoff is 2.0 g/cm3 according to the coal seam gas industry.
However, a low density cutoff value of 1.75 g/cm3 was also used to
define the net coal seam thickness by Lamarre (2003). In the study
area, we analyzed the uncertainty of gas resources by using four
different density cutoffs, 1.65 g/cm3, 1.75 g/cm3, 1.85 g/cm3 and
2.0 g/cm3, respectively. From Eq. (1), the calculated ash contents are
57 wt%, 67 wt% and 81 wt% for coal bulk densities of 1.75 g/cm3,
1.85 g/cm3 and 2.0 g/cm3, respectively.
3.2. Reservoir modelling
3.2.1. Structural modelling
Two well cross sections are used to correlate the well tops of the
nine coal seams (see Fig. 2). Three wells, E, J and G intersected
different faults. Fault points are identified at each well intersection.
The well #E missed the coal seams, IX and VIII due to the fault.
Again well #G missed the coal seam XII for the same reason. Well #J
intersected only a part of the coal seam IX.
After well correlation, the study area is gridded into
25 m  25 m cells on the horizontal plane with a total number of
cells of 387  254 ¼ 98,298. At the vertical direction, the nine coal
seams are layered into 0.5 m and the 8 interlayers are not layered
 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001 995

VRO (%) Vitrinite (%) Lip nite (%)

0 1 2 3 0 50 100 0 50 100
0
(a) (b) (c)

200

400

600
Depth (m)

800

1000

1200

1400

1600

Fig. 8. Variation of (a) vitrinite reflectance (VRO), (b) vitrinite and (c) liptinite with depth.

20 3.2.2. Distribution of coal thickness

The contour maps on coal seam thickness (Fig. 3b shows one of
y = 0.6256x + 3.8252
18 the coal seams) were digitized then imported into the geological
R² = 0.6257
modelling software. Coal thicknesses are distributed by using the
16 moving average method (Petrel, 2014) based on the well tops and
the contour maps. In the final model, the maximum average coal
14
seam net thickness is 16 m in coal seam IIIeIV followed by coal
Calculated GC (m3/t)

12 seams VeVII (13 m), XV(11 m), XIVA(7 m), XIV(6 m), XIII(6 m),
XII(5 m), XI(5 m) and IX(3 m) as listed in Table 3.
10 With those digitized coal thickness data, the direction of the
major and minor distance was determined by generating 2D hori-
8
zonal variogram maps (Petrel, 2014). The minor distance direction
6 is perpendicular to the major direction (Petrel, 2014). After known
the major and minor directions, an exponential model is chosen to
4 fit the sample's experimental variogram along major and minor
directions for each coal seam. Fig. 11 shows the analysed variogram
2
range and orientation of coal thickness in the nine seams. The
0 figure shows that for coal seams, IIIeIV, VeVII, IX, XI, XII, XIII, XIV,
XIVA and XV the major directions are
0 2 4 6 8 10 12 14 16 18 20
15
, 5
, 45
, 65
, 45
, 45
, 5
, 45
and 45
; the major
Measured GC (m3/t) distances are 10,000 m, 2686 m, 2354 m, 5100 m, 7756 m, 9843 m,
6932 m, 10,000 m and 10,000 m; and the minor distances are
Fig. 9. Relationship between the measured and calculated gas content on as received
basis. Dotted and solid lines present diagonal and fitted lines, respectively.
3408 m, 2357 m, 1888 m, 1643 m, 2140 m, 3358 m, 2928 m, 2953 m
and 2629 m, respectively.

vertically. The reservoir structure model was built (see Fig. 10) using
3.2.3. Distribution of coal quality
the well tops, structural contour maps and coal seam thickness
There are only nine wells with logs hence it is impossible to
maps which were sourced from coal mining data as shown as un-
analysis the semi-variance of the logs and coal properties derived
filled dot points in Fig. 3. In the study area, there are 17 faults which
from logs. However, Saikia and Sarkar (2013) reported that coal ash
were included in reservoir structural modelling. Convergent
content and volatile matter has a fair relationship with coal thick-
gridder method is used to generate the structural model. The well
ness. Hence, we built the distributions of coal quality properties,
data (points of intersection with seam top) is constrained in
e.g. ash content, and coal density, using the variogram of coal
segment with an influence radius of 600 m. Fig. 10a shows the well
thickness. Sills at both major and minor directions are 1.0; nugget is
locations.
0.1 and vertical range is 10 m.
996 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001

Fig. 10. Structural contour map on the top of coal seam No. XV (a), and a vertical section shows the faults and coal distributions (b). One trust fault is shown in left of the section
map.

Based on the above analysis, the coal density was modelled density and ash content in 3D.
using the SGS method with the simply Kriging (SK) algorithm. The Table 3 lists the statistics for the net coal thickness, coal density
SGS method honors borehole data, input distributions, variograms and ash content. Results show that the average bulk density is
and trends. In this study, we also built the distributions of coal 1.85 g/cm3 for coal seam XII followed by coal seams IIIeIV(1.64 g/
 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001 997

Table 3
Statistics for variables of the nine coal seams in their models.

Statistics parameters Minimum Maximum Mean Standard deviation Variance

Coal seam No. XV

Net Thickness, m 0.12 25.43 10.63 4.73 22.34
RHOB, g/cm3 1.24 2.70 1.63 0.43 0.18
ASHR, wt% 6.92 72.73 27.21 13.82 191.00
Coal seam No. XIVA
Net Thickness, m 0.08 15.69 6.94 3.44 11.81
RHOB, g/cm3 1.31 2.47 1.44 0.15 0.02
ASHR, wt% 13.50 79.70 30.02 13.67 186.80
Coal seam No. XIV
Net Thickness, m 0.09 11.99 5.83 2.46 6.04
RHOB, g/cm3 1.27 2.28 1.61 0.16 0.03
ASHR, wt% 9.44 71.86 34.91 12.57 157.92
Coal seam No. XIII
Net Thickness, m 0.07 18.04 5.50 3.40 11.54
RHOB, g/cm3 1.22 2.61 1.48 0.22 0.05
ASHR, wt% 5.33 73.80 27.33 14.07 197.85
Coal seam No. XII
Net Thickness, m 0.08 11.24 5.34 2.64 6.95
RHOB, g/cm3 1.44 2.66 1.85 0.31 0.10
ASHR, wt% 26.17 75.85 53.50 16.56 274.23
Coal seam No. XI
Net Thickness, m 0.08 11.67 4.68 2.69 7.22
RHOB, g/cm3 1.30 2.28 1.48 0.11 0.01
ASHR, wt% 12.69 55.10 32.81 8.44 71.29
Coal seam No. IX
Net Thickness, m 0.12 10.78 3.01 1.97 3.90
RHOB, g/cm3 1.32 2.77 1.57 0.16 0.03
ASHR, wt% 14.49 74.66 34.09 8.20 67.29
Coal seam No. VeVII
Net Thickness, m 0.10 35.10 13.47 9.10 82.89
RHOB, g/cm3 1.18 2.55 1.63 0.17 0.03
ASHR, wt% 2.01 79.65 41.64 10.85 117.78
Coal seam No. IIIeIV
Net Thickness, m 0.11 37.97 16.04 7.45 55.52
RHOB, g/cm3 1.20 2.51 1.64 0.14 0.02
ASHR, wt% 3.08 75.37 46.39 11.22 125.99

cm3), XV(1.63 g/cm3), VeVII(1.63 g/cm3), XIV(1.61 g/cm3), IX(1.57 g/
cm3), XI(1.48 g/cm3), XIII(1.48 g/cm3), XIVA(1.44 g/cm3). While the
maximum average ash content is 53.50 wt% for coal seam XII fol-
lowed by coal seams IIIeIV(46 wt%), VeVII(42 wt%), XIV(35 wt%),
IX(34 wt%), XI(33 wt%), XIVA(30 wt%), XIII(27 wt%) and XV(27 wt%).
3.3. Estimation of CBM resources
In calculating the CBM resources, we have used two processes
which are shown in Fig. 12. In Process 1, we have calculated the gas
content distribution in 3D based on the distribution of coal density,
ash content, fixed carbon, burial depth and volatile matter content
using Eq. (5). In Process 2, however, we calculated the gas content at
borehole first and then built its distribution in 3D using SGS. In both
processes, the gas amount in each cell is calculated using the
following equation:
CBMR ¼ Vb $RHOB$Q $NTG
where, CBMR is the cells' gas amount in m3, Vb is the cells' bulk
volume in m3, RHOB is the cells' coal density in g/cm3, and Q is the
cells' gas content in m3/t, NTG is the cells' net-to-gross ratio of coal
thickness in fraction.
4. Discussion
cm3, 1.75 g/cm3, 1.85 g/cm3 and 2.0 g/cm3, are used to define the
NTG based on the distribution of coal density in 3D. The estimated
resources for the nine coal seams with different coal density cutoffs
for the two processes are presented in Table 4, Figs. 13 and 14. As
can be seen from Table 4 and Figs. 12 and 13 that the estimated total
resources for the four density cutoff for the nine seams based on
Process 1 are 17.4  109 m3, 19.7  109 m3, 21.0  109 m3 and
22.3  109 m3 and based on process 2 are 16.4  109 m3,
18.7  109 m3, 20.1  109 m3, 21.6  109 m3 respectively. It can be
observed that the estimated resources for most coal seams in
Process 2 are lower than those estimated in Process 1. This is
because in the study area, the depth has a negative effect on gas
content as shown in Fig. 4. The effect of depth is not considered in
calculation of resource estimation in Process 1 when generating the
distribution of the gas content. In Process 2, however, the gas
content distribution was generated by using SGS which considered
the decreasing trend of gas content with the increase in depth.
(6) Note that the structural uncertainties and spatial variations of
depositional environment are also important for CBM resources
calculation. In this study, hand-drawn structure and coal thickness
distribution maps from mining companies are used in geological
modelling. Though wells G, J and F are located away from wells A, B,
C, D, H and I, the estimated gas data from wells A, B, C, D, H and I will
dominant the resources calculation results because their close
locations.

4.1. Effect of geological modelling processes on CBM resources 4.2. Effect of coal density cutoffs on CBM resources

As mentioned above, two processes are used to estimate the When estimating the incremental changes of resource estima-
resources. In each process four different density cutoffs of 1.65 g/ tion by using the Process 1 for the nine coal seams, starting from
998 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001

θ=15° θ=-5° θ=-45°

Lmax=10000m Lmax=2686m Lmax=2354m
Lmin=3408m Lmin=2357m Lmin=1888m

III-IV V-VII IX
θ=-65° θ=-45° θ=-45°
Lmax=5100m Lmax=7756m Lmax=9843m
Lmin=1643m Lmin=2140m Lmin=3358m

XI XII XIII

XIVA
XV
θ=-5° θ=-45° θ=-45°
Lmax=6932m Lmax=10000m Lmax=10000m
Lmin=2928m XIV Lmin=2953m Lmin=2629m

Fig. 11. Analyzed horizontal variogram of the digitized thickness data from nine thickness maps. q ¼ major showing the direction, Lmax ¼ major distance, and Lmin ¼ minor distance.

Log: RHOB Laboratory data

Eq. 1

BH: ASHR Step 1:

Log
interpretation
Eqs. 2 and 3
Eq.4
BH: FC, VM BH: Q

Structural
3D: RHOB 3D: Q
Model
Eq. 1
Density
ASHR Cutoff
Eq. 2 Eq. 3 Step 2:
Reservoir
FC MOIN NTG Vb D
Modelling

Eq. 4
VM

Eq.5
3D: Q
Eq.6 Eq.6
Step 3:
CBMR CBMR Calculation
Process No.1 Process No.2

Fig. 12. Diagram showing the two different processes in calculating CBM resources (BH ¼ borehole, 3D ¼ three dimensions, other abbreviations are described in the text; after Zhou
et al., 2015).
 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001 999

Table 4
Estimated resources for the nine coal seams with four different density cutoffs and two different processes.

Coal seam No CBMR by process 1 (109m3) CBMR by process 2 (109m3)

1.65 (g/cm3) 1.75 (g/cm3) 1.85 (g/cm3) 2.00 (g/cm3) 1.65 (g/cm3) 1.75 (g/cm3) 1.85 (g/cm3) 2.00 (g/cm3)

XV 3.1 3.2 3.3 3.5 2.9 3.1 3.2 3.4

XIVA 2.4 2.4 2.5 2.5 2.3 2.3 2.3 2.4
XIV 1.5 1.8 2.0 2.2 1.5 1.8 2.1 2.3
XIII 1.8 1.9 2.0 2.0 1.8 1.9 1.9 2.0
XII 0.5 0.7 1.0 1.5 0.5 0.6 0.9 1.5
XI 1.5 1.6 1.6 1.6 1.5 1.6 1.6 1.6
IX 0.9 1.0 1.1 1.1 0.9 1.0 1.1 1.1
VeVII 2.8 3.3 3.6 3.8 2.4 2.9 3.2 3.4
IIIeIV 2.9 3.8 4.0 4.1 2.7 3.5 3.7 3.9
Total 17.4 19.7 21.0 22.3 16.4 18.7 20.1 21.6

7 nine coal seams, from high to low, can be ordered as XII, IIIeIV, XIV,
≤1.65 (g/cm3) VeVII, XV, IX, XIII, XIVA and XI (Fig. 15).
6 ≤1.75 (g/cm3)
≤1.85 (g/cm3)
5 ≤2.00 (g/cm3)
4.3. Well log derived gas content
CBMR (109m3)

4
In this study, the gas content is calculated based on the esti-
3 mated ash, fixed carbon and volatile matter contents and burial
depth. The theoretical relationships of bulk density with ash, fixed
2
carbon, volatile matter and moisture contents were established by
1 varying individual parameters for 5000 times randomly. This was
achieved by varying individual parameters randomly, such as the
0 ash content from 1 to 40 v.%, moisture content from 0 to 3 v.%, fixed
XV XIVA XIV XIII XII XI IX V-VII III-IV carbon contents from 6 to 98 v.% and volatile matter from 0 to
Coal seam No. 54 v.%.
Measured values of fixed carbon content and ash content for 68
Fig. 13. Calculated CBM resources for different coal seams with different coal density
cutoffs via Process # 1. The effect of variation of density cutoffs on CBM resources of the
samples were plotted in Fig. 16. From this plot, two correlations
coal seams, from high to low, sequenced as No. XII, XIV, IIIeIV, VeVII, IX. XV, XIII, XI, bounding all the data points (upper and lower limit) were estab-
and XIVA. lished. These two limits were used to generate fixed carbon con-
tent. Data points for the volatile matter were generated using the
material balance therefore, subtracting the sum of moisture con-
7 tent, ash content and fixed carbon content from 100 v.%. According
≤1.65 (g/cm3)
to Anderson et al. (2003) the densities for the ash, fixed carbon,
6 ≤1.75 (g/cm3)
volatile matter and moisture are 2.75 g/cm3, 1.35 g/cm3, 0.90 g/cm3
≤1.85 (g/cm3)
5 and 1.00 g/cm3 respectively which were used to calculate bulk
≤2.00 (g/cm3)
density. Following equations are used to convert unit from wt% to
CBMR (10 9 m 3 )

4 v.%:

3
60
2 Increase density cutoff
Incremental percetage in CBM resources, %

from 1.65-1.75g/cm3
1 50 Increase density cutoff
0 from 1.75-1.85g/cm3
XV XIVA XIV XIII XII XI IX V-VII III-IV 40 Increase density cutoff
Coal seam No. from 1.85-2.00g/cm3

Fig. 14. Calculated CBM resources for different coal seams with different coal density 30
cutoffs via Process # 2. The effect of variation of density cutoffs on CBM resources of
the coal seams, from high to low, sequenced as No. XII, XIV, IIIeIV, VeVII, IX, XV, XIII,
XI, and XIVA.
20

the top XV, XIVA, XIV, XIII, XII, XI, IX, VeVII and IIIeIV, it was found 10
that with change of density cutoff from 1.65 g/cm3 to 1.75 g/cm3 are
4.9%, 1.7%, 23.4%, 5.4%, 27.5%, 3.3%, 9.7%, 18.6%, and 30.8%; from
1.75 g/cm3 to 1.85 g/cm3 are 2.2%, 0.8%, 13.4%, 1.0%, 47.7%, 1.9%, 4.9%, 0
9.0% and 5.8%; and from 1.85 g/cm3 to 2.00 g/cm3 are 5.5%, 2.0%, XV XIVA XIV XIII XII XI IX V-VII III-IV
7.9%, 2.6%, 56.3%, 0.6%, 0%, 5.2%, and 2.5% respectively. In general,
the effect of change in density cutoffs on resource estimation of the Fig. 15. Incremental percentage of CBM resource with increasing density cutoff for the
nine coal seams.
1000 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001

100 50
90 Measured y = 69.206x - 83.165
45
y=50-1.1x R² = 0.8776
80 40
y=100-1.5x Random points
70 35
Well #B
60
FC (v.%)

30 Linear (Random points)

ASHR (v.%)
50 25 Linear (Well #B)
40
20
30
15
20
10
10 y = 61.339x - 74.033
5 R² = 0.7665
0
0 20 40 60 80 100 0
ASHR (v.%) 1 1.2 1.4 1.6 1.8
ρc (g/cm3)
Fig. 16. Relationship between ash content and fixed carbon content in volume percent.
The volume percentages are calculated with the densities of ash, fixed carbon, volatile Fig. 17. Comparison of the measured and theoretical relationships between coal bulk
matter and moisture as 2.75 g/cm3, 1.35 g/cm3, 0.90 g/cm3 and 1.00 g/cm3, respectively. density and ash content. The theoretical calculated coal bulk density use the random
v.% ¼ volumetric percent. Hatched lines are boundary lines showing the ranges of generated volume percentage of ash, fixed carbon and moisture content and their
ASHR and FC. density. The used densities of ash, fixed carbon, volatile matter and moisture are
2.75 g/cm3, 1.35 g/cm3, 0.90 g/cm3 and 1.00 g/cm3, respectively.

100
ASHRðwt%Þ
ASHRðv:%Þ ¼ (7) Theori cal
V$rash 90
80 Measured
70
FCðwt%Þ
FCðv:%Þ ¼
FC (v.%)

(8) 60
V$rfc
50
40
VMðwt%Þ 30
VMðv:%Þ ¼ (9)
V$rvm 20
10
ASHRðwt%Þ FCðwt%Þ VMðwt%Þ Moistðwt%Þ 0
V¼ þ þ þ (10) 1 1.2 1.4 1.6 1.8 2
rash rfc rvm rmoist
ρc (g/cm3)
where rash, rash, rash, rash are density of ash, fixed carbon, volatile Fig. 18. Comparison of the measured and theoretical relationships between the coal
matter and moisture, respectively. bulk density and fixed carbon content.
The calculated bulk densities were plotted as a function of ash
content, fixed carbon content and volatile matter content in
Figs. 17e19. From Fig. 17 it can be seen that only one of the 24 100
measured points is outside the calculated values (points). The slope Theori cal
90 (b)
of the best fit of the measured points (well #B) is slightly lower than
that of the best fit of the calculated points. Results show that the 80 Measured
uncertainty of ash content is within ±6 v.%. Figs. 18 and 19 present 70
the bulk density against volatile matter content and fixed carbon
VM (v.%)

60
content. These figures show that all the measured data points are
50
within the area defined by the calculated data points. The re-
lationships of the coal bulk density with fixed carbon and volatile 40
matter content are, however, very weak (data points are scattered). 30
In theory, the volatile matter content increases with increasing coal 20
bulk density and the fixed carbon content decreases with
increasing coal bulk density. Results show that the resource esti- 10
mation based on the relationships of bulk density with the fixed 0
carbon and volatile matter contents can yield high uncertainty. 1 1.2 1.4 1.6 1.8 2
Table 3 lists the summary statistics of the thickness, bulk density ρc (g/cm3)
and ash content of the nine coal seams obtained from 3D reservoir
model. The statistics can be improved with more well data. How- Fig. 19. Comparison of the measured and theoretical relationships between the coal
bulk density and volatile matter content.
ever, the effects of density cutoffs and different calculation pro-
cesses on the estimation of CBM resources will be similar as this
study.
 F. Zhou, Z. Guan / Journal of Natural Gas Science and Engineering 33 (2016) 988e1001 1001

5. Conclusions References

In this study 68 laboratory data sets and logs from nine wells Amijaya, H., Littke, R., 2006. Properties of thermally metamorphosed coal from
Tanjung Enim Area, South Sumatra Basin, Indonesia with special reference to
and a vast of mining data (structural maps and coal thickness the coalification path of macerals. Int. J. Coal Geol. 66, 271e295.
contour maps) are used in well correlation, log interpretation, Anderson, J., Basinski, P., Beaton, A., Boyer, C., Bulat, D., Ray, S., Reinheimer, D.,
variogram analysis, structure modelling and stochastic modelling Schlachter, G., Colson, L., Olsen, T., John, Z., Khan, R., Low, N., Ryan, B.,
Schoderbek, D., 2003. Producing natural gas from coal. Oilfield Rev. 15 (3),
for estimation of CBM resources. CBM resources were estimated by 8e30.
using two different methods for four different density cutoffs. From Bancroft, B.A., Hobbs, G.R., 1986. Distribution of kriging error and stationarity of the
the results of this study, the following conclusions can be drawn: variogram in a coal property. Math. Geol. 18, 635e652.
Beretta, F.S., Costa, J.F., Koppe, J.C., 2010. Reducing coal quality attributes variability
using properly designed blending piles helped by geostatistical simulation. Int.
1) With increase of depth the raw gas content decreases even J. Coal Geol. 84, 83e93.
though vitrinite reflectance ratio and fixed carbon content in- Cairncross, B., Cadle, A.B., 1988. Paleoenvironmental control on coal formation,
distribution and quality in the Permian Vryheid formation, East Witbank
crease. A possible explanation for this is that the temperature
Coalfield, South Africa. Int. J. Coal Geol. 9, 343e370.
has a negative effect on gas content. As the depth increases Faiz, M., Saghafi, A., Sherwood, N., Wang, I., 2007. The influence of petrological
temperature increases and therefore, the gas content decreases. properties and burial history on coal seam methane reservoir characterisation,
Sydney Basin, Australia. Int. J. Coal Geol. 70, 193e208.
2) It has been also found that the multivariate regression by using
Fu, X., Qin, Y., Wang, G.G.X., Rudolph, V., 2009. Evaluation of gas content of coalbed
ash content, fixed carbon, volatile matter content and burial methane reservoirs with the aid of geophysical logging technology. Fuel 88,
depth as inputs is an effective method for estimation of the raw 2269e2277.
gas content. Hagelskamp, H.H., Eriksson, P.G., Snyman, C.P., 1988. The effect of depositional
environment on coal distribution and quality parameters in a portion of the
3) The effect of density cutoffs on CBM resource estimation has Highveld coalfield, South Africa. Int. J. Coal Geol. 10, 51e77.
been found different in the different coal seams because of the Heriawan, M.N., Koike, K., 2008a. Identifying spatial heterogeneity of coal resource
heterogeneity of coal density that exists throughout the nine quality in identifying spatial heterogeneity of coal resource quality in a multi-
layer coal deposit by multivariate geostatistics. Int. J. Coal Geol. 73, 307e330.
coal seams. Heriawan, M.N., Koike, K., 2008b. Uncertainty assessment of coal tonnage by spatial
4) The estimated total coalbed methane resources based on the modeling of SEAM distribution and coal quality. Int. J. Coal Geol. 76, 217e226.
SGS analysis (Process 2) shows 5% lower than that using Hindistan, A.M., Tercan, E.A., Ünver, B., 2010. Geostatistical coal quality control in
Longwall mining. Int. J. Coal Geol. 81, 139e150.
multivariate regressed analysis (Process 1) because the depth Jakeman, L.B., 1980. The relationship between formation structure and thickness in
has a negative effect on gas content which was not considered in the Permo-Triassic succession of the Southern coalfield, Sydney Basin, New
generating the gas content distribution in Process 1. In Process 2, South Wales, Australia. Math. Geol. 12, 185e212.
Jenkins, C., 2008. Practices and pitfalls in estimating coalbed methane resources
however, the gas content distribution was generated by using
and reserves. In: Search and Discovery Article #80011 from Oral Presentation at
SGS which used the decreasing trend of gas content with the AAPG Annual Convention, San Antonio, TX.
increase in depth. Jeremic, M.L., 1985. Strata Mechanics in Coal Mining. A.A. Balkema Publishers,
Netherlands, pp. 19e20.
5) In theory, the relationship between the coal bulk density and the
Kim, A.G., 1977. Estimating Methane Content of Bituminous Coalbeds From
ash content is dependent on the relationship between the ash Adsorption Data. Report No. RI 8245.
content and the fixed carbon content. In this study, the fixed Lamarre, R.A., 2003. Hydrodynamic and stratigraphic controls for a large coalbed
carbon content has a fair relationship with the ash content. methane accumulation in Ferron coals of east-central Utah. Int. J. Coal Geol. 56,
97e110.
Naturally, ash and fixed carbon content are two dominant pa- Laxminarayana, C., Crosdale, P.J., 1999. Role of coal type and rank on methane
rameters for coal bulk density because for their high density sorption characteristics of Bowen Basin, Australia coals. Int. J. Coal Geol. 40,
values. 309e325.
Liu, G., Zheng, L., Gao, L., Zhang, H., Peng, Z., 2005. The characterization of coal
quality from the Jining Coalfield. Energy 30, 1903e1914.
Mastalerz, M., Kenneth, R.W., 1994. Variations in seam thickness, coal type and coal
quality in the Namurian succession of the intrasudetic basin (southwestern
Poland). Palaeogeogr. Palaeoclimatol. Palaeoecol. 106, 157e169.
Nomenclature Nolde, J.E., Spears, D., 1998. A preliminary assessment of in place coalbed methane
resources in the Virginia portion of the central Appalachian Basin. Int. J. Coal
ASHR ash content (as received basis) Geol. 38, 115e136.
Paul, S., Chatterjee, R., 2011. Determination of in-situ stress direction from cleat
CBMR coalbed methane resources
orientation mapping for coal bed methane exploration in south-eastern part of
D burial depth, m Jharia coalfield, India. Int. J. Coal Geol. 87, 87e96.
FC fixed carbon content (as received basis) Petrel, 2014. Petrel Help Center. Schlumberger.
Saikia, K., Sarkar, B.C., 2013. Coal exploration modelling using geostatistics in Jharia
GR Gamma, API
coalfield, India. Int. J. Coal Geol. 112 (1), 36e52.
Moist moisture content (as received basis) Seidle, J., 2011. Fundamentals of Coalbed Methane Reservoir Engineering. PennWell
NTG net-to-gross ratio Corporation, Tulsa, Oklahoma, USA.
R correlation coefficient Singh, A.K., Sharma, M., Singh, M.P., 2013. SEM and reflected light petrography: a
case study on natural cokes from seam XIV, Jharia coalfield, India. Fuel 112,
R2 correlation of determination 502e512.
RHOB logging coal bulk density, g/cm3 Wang, H., Li, Y., Wang, E., Zhao, Z., 1997. Strategic ground water management for the
Q raw gas content (as received basis), m3/t reduction of karst land collapse hazard in Tangshan, China. Eng. Geol. 48,
135e148.
Q1 naturally desorption gas content, m3/t Zhou, F., Allison, G., Wang, J., Sun, Q., Xiong, D., Cinar, Y., 2012. Stochastic modeling
Q2 lost gas content, m3/t of coalbed methane resources: a case Study in southeast Qinshui Basin, China.
Q3 residual gas content, m3/t Int. J. Coal Geol. 99, 16e26.
Zhou, F., Yao, G., Tyson, S., 2015. Impact of geological modeling processes on spatial
SGS sequential Gaussian simulation coalbed methane resource estimation. Int. J. Coal Geol. 146, 14e27.
VM volatile matter content (as received basis) Zuo, J., Peng, S., Yongjun, L., Chen, Z., Xie, H., 2009. Investigation of karst collapse
VRO vitrinite reflectance ratio based on 3-D seismic technique and DDA method at Xieqiao coal mine, China.
Int. J. Coal Geol. 78, 276e287.
rc measured coal bulk density, g/cm3
t desorption time, days