See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/327399840

Classification of Coal Resources Using Drill Hole Spacing Analysis (DHSA)

Article · August 2018

DOI: 10.17265/2328-2193/2018.04.003

CITATIONS READS

2 494

2 authors, including:

Waterman Sulistyana Bargawa

Universitas Pembangunan Nasional Veteran Yogyakarta
23 PUBLICATIONS   31 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Geostatistics research project View project

Geochemical characterization project View project

All content following this page was uploaded by Waterman Sulistyana Bargawa on 13 April 2020.

The user has requested enhancement of the downloaded file.

Journal of Geological Resource and Engineering 6 (2018) 151-159
doi:10.17265/2328-2193/2018.04.003
D DAVID PUBLISHING

Classification of Coal Resources Using Drill Hole

Spacing Analysis (DHSA)

Iskandar Zulkarnain and Waterman Sulistyana Bargawa

Master of Mining Engineering, UPN Veteran Yogyakarta, Jl. SWK 104 Yogyakarta 55283, Indonesia

Abstract: The classification of coal resources generally is based on geometric factors and the complexity of geological structures. The
classification has not considered coal quality factors such as ash content, sulphur content, caloric value. The development of
international classification standards has required a geostatistical analysis for the estimation and classification of coal resources. The
purpose of this research is to apply geostatistics method to determine optimal drill hole distance, and to analyze classification of coal
resource based on data of coal quality and quantity. Based on global estimation variance (GEV) approach from geostatistics, relative
error value was obtained. Drill hole spacing analysis (DHSA) results in optimal drill hole spacing on each coal seam for the coal
resources classification. Estimation using kriging block results in the value of kriging relative error. Coal resources classification was
based on relative error of 0-10% for measured resources, 10-20% for indicated resources and > 20% for inferred resources. Based on a
case study in a coal field consisting of three coal seams, the geostatistical approach produced the smallest distance on seam-3 as the
optimal borehole range in the research area. This classification yields a greater area of influence than the SNI standard on simple
geological complexity.

Key words: Geostatistics, bore hole spacing analysis, SNI (5015:2011), GEV, kriging relative error.

1. Introduction factors such as ash content, sulphur content, caloric

value.
Several international classification systems [1-4]
Geometry parameter, and coal quality aspects
have been developed in the past, the main ones are the
become an important aspect to determine the
American USGS Circular 831 (USGS, 1980) and SME
classification of coal resources [6]. Applying the
Guide (SME, 1999), the South-African SAMREC
approach to any coal basin with certain geological
Code (SAMREC, 2000), the Canadian CIM Guidelines
settings will have certain resource classification
(CIM, 2000) and National Instrument 43-101 (CSA,
parameters as well. The area of influence on each coal
2001), the European Code (EURO, 2002), the
basin will differ in different geological settings [7, 8].
Australasian JORC Code (JORC, 2012), and Indonesia
The development of resource classification standards
SNI (5015:2011). All these codes are broadly similar,
and coal reserves requires the use of geostatistical
although some differences in their definitions remain.
approaches. Coal resources classification research has
The JORC code is with little doubt the one that has
been widely developed using a geostatistical approach
found wider acceptance in countries that do not have
[9-13].
their own code. Generally, the basic classification of
resources and reserves for coal is a factor of quantity, 2. Objective
geometry, and the complexity of geological structures
The objective of this research is to make the
[5]. Limiting factors do not consider coal quality
classification of coal resources based on global
estimation variance (GEV) approach related to drill
Corresponding author: Waterman Sulistyana Bargawa,
Ph.D., Mr., research field: geostatistics and mining
hole spacing analysis. The results of the analysis will
environment. be compared with SNI 5015:2011 to evaluate the
152 Classification of Coal Resources Using Drill Hole Spacing Analysis (DHSA)

classification of resources.  E2 R    E2 r  / N (2)

The calculation of GEV values yields a relative
3. Methods and Material
error:
GEV yields global relative error values [14, 15].
Relative error = ±1.96∙ߪா ∙100%/mean (3)
The nomogram determines the value of the extension
The GEV obtained based on the nomogram is then
variance (  K ) dot against the square plane for the
2
used to estimate the relative error value. Then plotting
spherical model with the variance nugget being 0 and
between the relative error values against the drill hole
the sill value is 1 (see Fig. 1). Furthermore, point
spacing is to create a drill hole spacing analysis
variance estimates of the  K (r) field considering the
2

(DHSA) graph [6]. The DHSA graph analysis

value of the nugget variance and sill of each parameter
considers the value of drill spacing and relative error
[15]:
when it reaches 10%, 20% and 50% as the optimum
 E2 r   C 0  C *  e2  (1)
distance. Based on the DHSA graph and relative error
The estimation variance value divided by the values, the distance or area of influence for measured,
number of blocks (N) yields a GEV value: indicated, and inferred resources can be evaluated.

Fig. 1 Nomogram of extension variance/estimation [14, 15] values (࣌૛ࢋ ), point to square area for spherical models.
 Classification of Coal Resources Using Drill Hole Spacing Analysis (DHSA) 153

Geostatistical approach is generally used for Z*: estimated value

estimation of ore mineral grade [16-18]. Another  K2 : kriging variance
result of kriging is the use of kriging relative error of ߪ௄ : kriging standard deviation
the ordinary kriging estimation method for The calculation of the value of kriging relative error
classification of coal resources [19]. The classification is obtained from the standard deviation value of the
of coal resources uses the value of the kriging relative block unit divided by the value of the kriging estimate
error of the ordinary kriging estimation method. of the block unit, multiplied by the 95% confidence
Kriging is an estimation method of regionalized interval factor of 1.96 while the standard deviation
variable at point, or volume by using criteria to value is the square root of the estimated variance
minimize variance estimation. The kriging procedure value obtained from the ordinary kriging method.
needs to consider the following [20]: According to Ref. [21] the classification of
(a) estimated value using the equation: resources (Table 1) is based on confidence interval
n
(4)
Z* = w Z 
i 1
i i and tolerance of relative error.
(b) equation to calculate the weight, wi: Meanwhile, according to Ref. [24] the classification
n   n of resources and coal reserves is an effort to group
 w . (v, v)     (v,V )dan  w
i i 1 (5)
i 1 i 1 resources and coal reserves based on geological
(c) Kriging variance can be expressed by equation: confidence and economic feasibility. Table 2 shows
n  

   wi .  (v, V )   (V ,V )  
2
K (6) the classification of resources based on the distance of
i 1

Based on kriging variance we get the following the point of information according to geological
equation: conditions.
 The research area is located in East Kalimantan
 K2   (V , V )  w 1  M  (7)
Province Indonesia. Fig. 2 shows a map of drill hole
(d) kriging relative error is expressed by equation:
distribution in the study area.
K (8)
kriging relative error =  1.96  100 % Based on the geological model the research area
Z*
where: consists of 3 (three) coal seams.

Table 1 Resource classification [22, 23] based on drill hole distance and error tolerance.
Resource classification Max. extrapolation Max. spacing between information point Error tolerance
Measured 500 m +1 km < 500 m 0 - 10 %
Indicated 1,000 m +2 km < 1 km 10 -20 %
Infernal 2,000 m +4 km > 20%

Table 2 Resource classification [22] based on distance of information points according to geological condition.
Geological Resources
Criteria (m)
condition Inferred Indicated Measured
Simple Distance of information point 1,000 < x ≤ 1,500 500 < x ≤ 1,000 x ≤ 1,500
Moderate Distance of information point 500 < x ≤ 1,000 250 < x ≤ 500 x ≤ 250
Complex Distance of information point 200 < x ≤ 400 100 < x ≤ 400 x ≤ 100
154 Classification of Coal Resources Using Drill Hole Spacing Analysis (DHSA)

Fig. 2 Map of the distribution of drill hole in the study area.

4. Results Fig. 4 is an example of DHSA chart analysis to

determine the optimum distance in Seam-3 with
Table 3 shows descriptive statistics on coal seam.
classification of measured resources being 750 m,
Variogram model is spherical used for all data of coal
1,100 m for indicated resources and 2,150 m for
quality (ash, total sulphur, relative density, calorific
inferred resources. Fig. 5 shows a histogram of
value) and thickness for the three seams of coal.
comparison of measured, indicated, and inferred
Table 4 shows the variographic results for each coal
resource classification based on the borehole distance
quality parameter. (area of influence). Relative error values are obtained
Table 5 shows the calculation of relative error values from the results of reading DHSA graphics for Seam 1,
by the method of GEV. The research area consisted of Seam 2, and Seam 3. Drill hole distance based on the
3 (three) coal seams. DHSA and GEV methods calculation of relative error values showed higher
produce optimal borehole spacing at each seam. The values compared with the standard of SNI.
classification of measured resources in Seam-1 is 950 Based on the comparison of measured, indicated,
m at relative error of 10%, 1,650 m for indicated and inferred resource classifications in Fig. 5, the
resources at relative error of 20%, and 2,650 m for optimal distance for the research area using the result
inferred resources at relative error of 50%. At Seam-2 of the relative error value analysis on Seam 3: distance
the classification of measured resources is 1,150 m at measured resource classification is 750 m at ER 10%,
relative error of 10%, 1,850 m for indicated resources indicated is 1,100 m at ER 20%, and inferred is 2,150
at relative error of 20%, and 4,300 m for inferred m at ER 50%. Based on Fig.5 the classification of
resources at relative error 50% while the classification measured resources in this study produces the same
of measured resources at seam-3 is 750 m at relative distance or area of influence as Saraji, but is higher
error of 10%, 1,100 m for indicated resources at than the Peak Down Bowen Basin, Coal Guideline and
relative error 20% and 2,150 m for inferred resources at SNI. For the classification of indicated resources, this
relative error of 50%. Fig. 4 shows a drill hole spacing research is more conservative than Saraji and Peak
analysis (DHSA) graph for Seam-3. Down Bowen Basin, but the distance is higher than the
 Classification of Coal Resources Using Drill Hole Spacing Analysis (DHSA) 155

Table 3 Descriptive statistics for Seam-1, Seam-2, and Seam-3.

Seam 1 Seam 2 Seam 3
No Parameter
Thick CV TS RD Ash Thick CV TS RD Ash Thick CV TS RD Ash
1 Mean 9 4.408 0.3 1.3 2.5 17 4.742 0.2 1.3 2.1 2.2 4.712 0.9 1.3 4.1
2 Median 10.5 4.320 0.3 1.3 2.5 19.9 4.573 0.2 1.3 2.1 2.3 4.575 0.2 1.3 3.7
3 Mode 11 4.218 0.2 1.3 2.7 26 5.470 0.1 1.3 2.2 2.7 4.393 0.2 1.3 3.1
4 Variance 11.9 61.763 0 0 0.2 65.1 176.226 0 0 0.3 0.5 188.591 1.5 0 1.9
Standard
5 3.5 248.5 0.2 0 0.5 8.1 419.8 0.2 0 0.6 0.7 434.3 1.2 0 1.4
Deviation
6 Minimum 0.7 4,110 0.1 1.3 1.2 2 4.252 0.1 1.3 1.2 0.3 4.073 0.1 1.3 2
7 Maximum 16 5,170 1.1 1.3 4.1 28.8 5.593 0.8 1.3 3.7 5 5.762 4 1.4 9.5
8 Count 213 36 36 36 36 134 67 67 67 67 330 49 49 49 49
Coefficient of
9 0.4 0.1 0.5 0 0.2 0.5 0.1 0.7 0 0.3 0.3 0.1 1.4 0 0.3
variation
10 Weighted avg 4.400 0.3 1.3 2.5 4.597 0.2 1.3 2.1 4.675 0.7 1.3 40
sign. weighted
11 0 0 0 0 0 -0.2 0 0 0 -0.2 0 0
Average

Table 4 Variogram parameter using spherical model.

Seam Parameter Co Sill Range CV

Ash 0 0.24 950 0.2

Calorific value 0 61763 1,600 0.06
1 Relative density 0 0.00033 2,550 0.01
Thickness 3.4 8 1,800 0.38
Total sulphur 0 0.038 1,300 0.47
Ash 0 0.3046 850 0.26
Calorific value 0 215000 3,000 0.09
2 Relative density 0.000225 0.000125 735 0.01
Thickness 11 50 1,800 0.48
Total sulphur 0 0.0228 2,950 0.7
Ash 0 1.9 1,650 0.34
Calorific value 0 210000 3,010 0.09
3 Relative density 0.00006 0.000339 1,300 0.02
Thickness 0.14 0.47 1,700 0.33
Total sulphur 0 0.02111 2,820 1.43
Co: nugget variance, CV: coefficient of variation
156 Classification of Coal Resources Using Drill Hole Spacing Analysis (DHSA)

Table 5 Relative error value calculation using GEV method.

Parameter Mean h l _X _Y N a Co C h/a l/a Varians σ2Er σ2ER σER Error
9 100 100 54 49 2,646 1,800 3.4 8 0.056 0.056 0.020 3.560 0,001 0.037 0.80%
Thickness 9 200 200 27 25 6,75 1,800 3.4 8 0.111 0.111 0.040 3.720 0,006 0.074 1.62%
9 300 300 18 17 306 1,800 3.4 8 0.167 0.167 0.056 3.848 0,013 0.112 2.44%
2.5 100 100 54 49 2,646 950 0 0.24 0.105 0.105 0.030 0.007 0,000 0.002 0.13%
Ash 2.5 200 200 27 25 675 950 0 0.24 0.211 0.211 0.070 0.017 0,000 0.005 0.40%
2.5 300 300 18 17 306 950 0 0.24 0.316 0.316 0.110 0.026 0,000 0.009 0.74%
4.408 100 100 54 49 2,646 1,600 0 61.763 0.063 0.063 0.023 1389.668 0,525 0.725 0.03%
CV 4.408 200 200 27 25 675 1,600 0 61.763 0.125 0.125 0.045 2779.335 4,118 2.029 0.09%
4.408 300 300 18 17 306 1,600 0 61.763 0.188 0.188 0.065 4014.595 13,120 3.622 0.16%
1.3 100 100 54 49 2,646 2,550 0 0.00033 0.039 0.039 0.011 0.000 0,000 0.000 0.01%
RD 1.3 200 200 27 25 675 2,550 0 0.00033 0.078 0.078 0.021 0.000 0,000 0.000 0.02%
1.3 300 300 18 17 306 2,550 0 0.00033 0.118 0.118 0.042 0.000 0,000 0.000 0.03%
0.3 100 100 54 49 2,646 1,300 0 0.038 0.077 0.077 0.025 0.001 0,000 0.001 0.36%
TS 0.3 200 200 27 25 675 1,300 0 0.038 0.154 0.154 0.050 0.002 0,000 0.002 1.00%
0.3 300 300 18 17 306 1,300 0 0.038 0.231 0.231 0.080 0.003 0,000 0.003 1.88%

DHSA : SEAM-3
100%

10%
ERROR RELATIVE (%)

Thickness
1% Ash
CV
RD
0% TS

0%
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
2500
2600
2700
2800
2900
3000
100
200
300
400
500
600
700
800
900

DRILL HOLE SPACING

Fig. 3 Drill hole spacing analysis (DHSA) graph of Seam-3.

 Classification of Coal Resources Using Drill Hole Spacing Analysis (DHSA) 157

Distance of area of influence

based on relative error
5000

4500

4000 4300
Distance of area of influence (m)

3500

3000

2500 2650
2000 2150
1500 1850
1650
1000 1150 1100
950 1000
500 750
250 500
0
MEASURED INDICATED INFERRED

Seam 1 Seam 2 Seam 3 SNI

Fig. 4 Area of influence (or distance) on coal resource classification.

Fig. 5 Various studies: Saraji, Bowen Basin (Bertolli 2013), coal guideline, and SNI 5015:2011 to compare the optimum drill
range for resource classification based on relative error.
158 Classification of Coal Resources Using Drill Hole Spacing Analysis (DHSA)

Table 6 Example results of estimation and resource classification on CV parameters of Seam-1.

CV (Calorific value)
X Y
Estimate Variance Stdv Relative error Classification
105688 208144 4614.56 75608.6 274.9702 11.67915 Indicated
105938 208144 4551.02 72280.6 268.8505 11.57866 Indicated
105438 208394 4516.63 69172 263.0057 11.41318 Indicated
105688 208394 4542.44 65188 255.3194 11.01668 Indicated
105938 208394 4523.78 71376 267.1629 11.57526 Indicated
105188 208644 4456.67 74910.8 273.6984 12.03699 Indicated
105438 208644 4479.73 67627.5 260.0529 11.37800 Indicated
105688 208644 4471.24 61656.7 248.3077 10.88474 Indicated
105938 208644 4423.43 65807 256.5288 11.36666 Indicated
105688 208894 4203.44 72018.3 268.3623 12.51332 Indicated
105938 208894 4175.06 64312.4 253.5989 11.90531 Indicated

Coal Guideline and SNI. As for the classification of influence or distance is higher than SNI, but it is still
inferred resources, this research is more conservative within range of other methods.
than other methods. (c) High value of kriging variance will cause high
Here is a discussion of the value of kriging relative relative error value. The high relative error values in
error for classification of coal resources. Based on coal quality in the study area were totally sulphur (>
Eq. (3) calculation of kriging value relative error 100%) and ash (> 50%) included in inferred resource
is obtained from standard deviation value of unit block classification. Geologically high sulphur and ash
with 95% confidence interval. Table 6 shows the contents are associated with sediments deposited in the
results of resource classification based on the relative brackish-water environment.
error kriging value.
References
High value of kriging variance will cause high
[1] Emery, X., Ortiz, J. M., Juan, J., and Rodríguez, J. J. 2006.
relative error value. The highest relative error values in
“Quantifying Uncertainty in Mineral Resources by Use of
coal quality parameters are total sulphur > 100% and Classification Schemes and Conditional Simulations.”
ash > 50% included in inferred resource classification. Mathematical Geology 38: 445-64.
Geologically high sulphur and ash contents are [2] Goh, E., and Effendi, S. 2017. “Overview of an Effective
Governance Policy for Mineral Resource Sustainability in
associated with sediments deposited in
Malaysia.” Resources Policy 52: 1-6.
marine-brackish water environments. Fe element in the [3] Hatton, W., and Fardell. 2012. “New Discoveries of Coal
marine-brackish water environment is present in large in Mozambique—Development of the Coal Resource
numbers, whereas bacterial activity plays a major role Estimation Methodology for International Resource
Reporting Standards.” International Journal of Coal
in the formation of high sulphur.
Geology 89: 2-12.
[4] Gandhi, S. M., and Sarkar, B. C. 2016. “Chapter 13:
4. Conclusion Mineral Resources Classification, Book Chapter.”
Essentials of Mineral Exploration and Evaluation,
(a) Based on comparison of measured, indicated, and
309-20.
inferred resource classification at the most optimum [5] Li, G. 2016. “Coal Reservoir Characteristics and Their
distance at Seam-3 with distance of 750 m measured Controlling Factors in the Eastern Ordos Basin in China.”
resource classification at 10% relative error, indicated International Journal of Mining Science and Technology
26: 1051-8.
1,100 m at 20% relative error, and inferred 2,150 m at
[6] Bertoli, O., Paul, A., Casley, Z., and Dunn, D. 2013.
ER 50%. “Geostatistical Drill Hole Spacing Analysis for Coal
(b) The results of this study indicate the area of Resource Classification in the Bowen Basin, Queensland.”
 Classification of Coal Resources Using Drill Hole Spacing Analysis (DHSA) 159

International Journal of Coal Geology 112: 107-13. Practical Approach. Netherlands: Springer.
[7] Eble, C. F., and Greb, S. F. 2018. “Geochemical, [15] Michel, D. 1977. Geostatistical Ore Reserve Estimation.
Petrographic and Palynologic Characteristics of Two Late New York: Oxford.
Middle Pennsylvanian (Asturian) Coal-to-Shale [16] Bargawa, W. S., and Amri, N. A. 2016. “Mineral
Sequences in the Eastern Interior Basin, USA.” Resources Estimation Based on Block Modeling.”
International Journal of Coal Geology 190: 99-125. Presented at Progress on applied mathematics in science
[8] O'Keefe, J. M. K., Bechtel, A., Christanis, K., Dai, S., and and engineering, AIP Conf. Proc. 1705, 020001.
Hower, J. C. 2013. “On the Fundamental Difference [17] Bargawa, W. S., Rauf, A., and Amri, N. A. 2016. “Mineral
between Coal Rank and Coal Type.” International Journal Resource Modeling Using Pod Indicator Kriging.”
of Coal Geology 118: 58-87. Presented at Progress on Applied Mathematics in Science
[9] Srivastava, R. M. 2013. “Geostatistics: A Toolkit for Data and Engineering, AIP Conf. Proc. 1705, 020025.
Analysis, Spatial Prediction and Risk Management in the [18] Bargawa, W. S. 2016. “Mineral Resource Estimation
Coal Industry.” International Journal of Coal Geology Using Weighted Jackknife Kriging.” AIP Conf. Proc. 1755,
(Special Issue on Geostatistics). 120001.
[10] Sabourin, R. 1983. “Geostatistics as a Tool to Define [19] Blackwell, G. 1999. “Relative Kriging Errors—A Basis
Various Categories of Resources.” Journal of the for Mineral Resource Classification.” Exploration and
International Association for Mathematical Geology 15: Mining Geology 7: 99-105.
131-43. [20] Armstrong, M. 1998. Theory of Kriging Chapter of Basic
[11] Cornah, A., Vann, J., and Driver, I. 2013. “Comparison of Linear Geostatistics. Berlin Heidelberg: Springer, 83-102.
Three Geostatistical Approaches to Quantify the Impact of [21] Diehl, P., and David, M. 1982. “Classification of Ore
Drill Spacing on Resource Confidence for a Coal Seam Reserves/Resources Based on Geostatictical Methods.”
(with a Case Example from Moranbah North, Queensland, CIM. Bull. 75 (838): 127-35.
Australia).” International Journal of Coal Geology 112: [22] JORC. 2012. “Reporting for Exploration Results, Mineral
114-24. Resources and Ore Reserve.”
[12] Silva, D. S. F., and Boisvert, J. B. 2014). “Mineral [23] Souza, L. E., Costa, J. F. C. L., and Koppe, J. C. 2010.
Resource Classification: A Comparison of New and “Comparative Analysis of the Resource Classification
Existing Techniques.” The Journal of the Southern Techniques: Case Study of the Conceicao Mine, Brazil.”
African Institute of Mining and Metallurgy 144: 3. Applied Earth Science (Trans. Inst. Min. Metall. B) 119:
[13] Jain, P. K. 2017. “Valuation of Mineral Resources with 166-75.
Special Reference to India.” Mineral Economics, 1-9. [24] Standard Nasional Indonesia. 2011. Pedoman pelaporan
[14] Annels, A. E. 1991. Mineral Deposit Evaluation: A sumberdaya dan cadangan batubara, SNI 5015:2011.

View publication stats