The practical application of Vectar Processed densities in proving the lateral continuity of coal Zones and Samples in the Ellisras Basin, South Africa in support of effective Mineral Resource adjudication

by

John Hendrey Sullivan

Submitted in fulfilment of the requirements for the degree of

Philosophiae Doctor in the

Faculty of Natural and Agricultural Sciences

University of Pretoria

Pretoria

2014

1

DECLARATION

I declare that this thesis, which I hereby submit for the degree Philosophiae Doctor at the

University of Pretoria, is my own work and has not previously been submitted by me for a degree at this or any other tertiary institution.

____________________________________

JH Sullivan

____________________________________

24 May 2014

2

ABSTRACT

The practical application of Vectar Processed densities in proving the lateral continuity of coal Zones and Samples in the Ellisras Basin, South Africa thus supporting a Competent Person to adjudicate Mineral Resources more effectively

by

JH Sullivan

Study Leader: Prof F. Camisani

Co-study Leader: Prof P.G. Eriksson

Department: Geology

PhD Degree:

Keywords: Ellisras Basin, Groenfontein, geophysical logging, cumulative distribution diagrams, lateral correlation, SANS data point, classification of Resources.

The Ellisras Basin, with huge coal resources, is fault-bounded along its southern and northern margins and is a graben-type deposit. The study area is situated in the south-western part of the

Limpopo Province of the Republic of South Africa and is geologically located in the Ellisras

Basin. In this area the basin is influenced by three major fault zones, the Eenzaamheid Fault delineating its southern limit, the Zoetfontein Fault near its northern limit and the Daarby Fault, with a down-throw of approximately 350 m towards the north-east. Sedimentological facies changes also influence the continuity of the coal zones, with deterioration in coal development.

The exploration project was a collaboration between two of the large role players in the South

African coal mining industry Sasol and Exxaro, for the purpose of identifying whether the coal

3

in the Ellisras Basin could be used for gasification purposes in the Sasol process, and that enough resources exist on the farms on which the two companies have the exploration rights..

The prospecting method used at the Project area, situated 50 kilometer west of the town of

Lephalale in the Limpopo Province of South Africa, comprises the drilling of cored exploration boreholes on a random spacing of ± 1 000 m x 1 000 m, together with infill percussion drilling.

The use of slimline geophysical methods to log lithologies is a technique which has been used extensively in the mining industry over a number of years. At the Project area the correlation between the measured densities derived from the traditional method of air and water measurement and those derived from Vectar processed derived densities from geophysical logging is better than 95%. As a method of “fingerprinting” the various coal zones and samples it was decided to calculate the distribution of relative densities in the chosen geological intersection. The data was then used to portray geophysically derived relative density cumulative distribution line diagrams (GDCDD) of the various lithotypes on either a sampleby-sample or zone-by-zone basis.

Using the classification method proposed, the various coal seams and zones can be correlated to a high degree and discrepancies easily identified. The lateral correlation between lithologies can be accurately described and substantiated, and this would convince a Competent Person that the method proposed is invaluable in classifying coal resources in the coal basins.

4

Acknowledgements

I want to recognize Sasol for allowing me to publish this dissertation and recognise individuals like Phil Grobler in the preparation of this document.

My wife (Deirdre) and children (David and Eliz-Mari) who suffered with a grumpy old man at home for the duration of this study. They have been a constant source of joy, blessing and inspiration.

Wikus van Deventer, Leon Roux and Neil Andersen for fruitful discussions and inspiration.

The SANS committee whose insight and discussions around the standard in general was most helpful.

The Lord who inspired and blessed me and woke me at odd hours with solutions implanted in my thoughts.

Klaus K ȕhne for an insightful review of this document as a whole

Professors’ Ferdi Camisani and Pat Eriksson who inspired me to keep going.

5

CONTENTS

1 INTRODUCTION .......................................................................................19

2 THE AIM OF THIS THESIS .....................................................................27

3 LITERATURE STUDY ..............................................................................29

3.1

SAMREC Code .......................................................................................................... 29

3.2

3.2.1

Applicability and accuracy of geophysical density logging of coal ........................... 33

Vectar Processing ..................................................................................................... 37

3.3

3.4

3.5

Distance only based drilling patterns ......................................................................... 40

Rules of conduct for a Competent Person .................................................................. 44

Comment on the literature study ................................................................................ 45

4 THE STUDY AREA ....................................................................................46

4.1

4.2

Regional structural setting .......................................................................................... 48

Local structural setting ............................................................................................... 50

4.3

4.4

4.5

Stratigraphy ................................................................................................................ 51

Coal types and distribution ......................................................................................... 53

Sub-outcrop maps of the various coal zones .............................................................. 61

5 TECHNICAL INFORMATION RELATED TO THE

PREPARATION OF THIS THESIS .................................................66

5.1

5.2

5.2.1

Prospecting methods used at Groenfontein ................................................................ 66

Drilling techniques, planning, control and material recovery .................................... 66

Cored drilling ........................................................................................................... 66

5.2.2

Percussion drilling .................................................................................................... 66

5.3

Geophysical density logging performed at the Study area ......................................... 67

6 ACCURACY OF THE GEOPHYSICAL DATA IN

PREDICTING THE RELATIVE DENSITY OF THE

COAL SEAMS ON THE STUDY AREA .........................................71

6.1

6.2

Statistical evaluation of the measured relative densities of cored material................ 73

Reliability of measurements of sample masses .......................................................... 73

7 STATISTICAL DEPICTION OF THE GEOPHYSICAL DATA

AND ASSOCIATED CALCULATIONS ..........................................76

6

7.1

Using the geophysical log to ‘fingerprint’ the various coal zones and specific parts (samples) of the coal seams. ............................................................... 76

8 DISCUSSION ON THE USE OF STATISTICALLY DERIVED

GDCDD WHEN EVALUATING SAMPLES OVER THE

STUDY AREA......................................................................................86

8.1

Method for discussing the GDCDD of samples ......................................................... 89

8.1.1

Graphs ...................................................................................................................... 89

8.1.2

8.2

8.2.1

Tables ....................................................................................................................... 89

Specific example from the study area ........................................................................ 93

Sample 01A .............................................................................................................. 93

8.3

8.4

8.5

8.6

Relationship between GDCDD and yield calculated from proximate data and the relationship between cored and percussion boreholes’ GDCDD curves...................................................................................................................... 100

Comparison between analytically derived yield and the GDCDD curves ............... 104

Anonymous check of information from neighbours ................................................ 106

Summary on the use of statistically derived GDCDD curves when evaluating samples over the Study area ................................................................. 108

9 DISCUSSION ON THE USE OF STATISTICALLY DERIVED

GDCDD WHEN EVALUATING COAL ZONES .........................114

9.1

9.2

Methodology used .................................................................................................... 116

Coal Zone 11 ............................................................................................................ 117

9.3

9.4

The importance of the standard deviation when evaluating zones: ......................... 125

Discussion on the use of GDCCD when evaluating the coal zones of the

Study Area. ............................................................................................................. 134

10 CONCLUSION ..........................................................................................136

11 WORKS CITED ........................................................................................141

7

LIST OF FIGURES

Figure 1 The relationships between the various categories of reporting of Resources and Reserves according to the SAMREC Code ......................................................... 21

Figure 2 Multiple seam deposit type Coal resource (<50%ash) and thick interbedded seam deposit type on right, SANS (2004) .............................................. 22

Figure 3 Drilling densities required for a multiple seam deposit .............................................. 23

Figure 4 Drilling densities required for a thick interbedded seam deposit ............................... 24

Figure 5 Distribution of boreholes to get to eight boreholes per 100 hectare........................... 25

Figure 6 Statistics for coal density measurement by various means illustrating that accurate densities can be measured with geophysics. *Density = induced density, Ryder (2002). ................................................................................................ 35

Figure 7 Dual density sonde (from Weatherford promotional material) showing various operating parameters and physical measurements ......................................... 36

Figure 8 Example of a geophysical density log. The graph on left is natural gamma and calliper log and the one on the right is the various density logs .......................... 39

Figure 9 Map illustrating a resource classification scheme using only distance criteria. The solid stars represent the location of drill holes. The filled circles represent the area covered by the search parameters surrounding a hole. The filled circles may represent the portion of a deposit, potentially classified as Indicated Resources, Pincock Perspectives (2009) ................................ 40

Figure 10 Rectangles drawn around data points (boreholes) as modelled in the geological modelling package - Minex ...................................................................... 41

Figure 11 Circles drawn around data points (boreholes) at the study area, with

ArcGIS, an ESRI product ........................................................................................... 42

Figure 12 Resource classification based on drilling density, Pincock Perspectives

(2009) ......................................................................................................................... 43

Figure 13 Location of the Study area after Sambo ( 2008)....................................................... 46

Figure 14 Location of the nine farms and geographic areas in the Study area ......................... 47

Figure 15 The Ellisras Basin and other Karoo basins in southern Africa, Mtimkulu

(2009) ......................................................................................................................... 48

Figure 16 Sketch map of the Ellisras Basin showing major structural elements ...................... 49

8

Figure 17 Coal-bearing formations within the stratigraphy of the Ellisras basin-fill,

Bordy et al (2010). Coal zones 11-4 are of the thick interbedded coal type with Coal zones 3-1 being multiple seam deposits, according to SANS

(2004) ......................................................................................................................... 50

Figure 18 Local structural geology and borehole positions at the Study Area ......................... 51

Figure 19 The Grootegeluk Formation coals at Groenfontein, adapted from Bordy et al (2010) ..................................................................................................................... 53

Figure 20 Photograph of typical cores from Grootegeluk Formation coals. The yellow pencil marks are lithology boundaries with associated measured depths below surface. White pencil writing denotes sample numbers and the amount of plastic bags needed to place the material in weighing purposes ...................................................................................................................... 54

Figure 21 Thickness (general) statistics of the Grootegeluk Formation at the Study area ............................................................................................................................. 54

Figure 22 Typical upwards coarsening cycles on which sample boundaries are based. CO9 is coal Zone 09 and the other units are depth in meters.......................... 55

Figure 23 Thickness statistics (MINEX) of the thickness of the coal zones in the

Swartrant Formation ................................................................................................... 56

Figure 24 Variation in lithology of the Swartrant Formation coals seams. Red lines denote zone boundaries and the illegible writing on the left side of the lithology log and to the right of the geophysical log is the depth in one meter intervals ............................................................................................................ 57

Figure 25 Swartrant coal seams with laminated siltstone ......................................................... 58

Figure 26 Swartrant coal seams, showing dull coal interbedded with sandstone

(coarse, white rock) and mudrock (dark grey laminated rocks). The yellow pencil marks are lithology boundaries with associated measured depths below surface. White pencil writing denotes sample numbers ................................. 58

Figure 27 Sub-dividing the various coals into coal zones and samples in the Ellisras

Basin. Study area on left and Grootegeluk mine on right .......................................... 59

Figure 28 zones and mining benches as defined at Grootegeluk mine ..................................... 60

Figure 29 Geographical areas identified in the Study area ....................................................... 61

9

Figure 30 Northern Area coal zones. The top part is a plan showing the sub-outcrop positions of the coal zones and the bottom parts illustrates sections through the stratigraphy at various positions to illustrate faults and the dip of the rocks. Vertical exaggeration is ten times .................................................................... 62

Figure 31 Central Area coal zones. The top part is a plan showing the sub-outcrop positions of the coal zones and the bottom parts illustrates sections through the stratigraphy at various positions to illustrate faults and the dip of the rocks. Vertical exaggeration is ten times .................................................................... 63

Figure 32 Southern Area top coal zones. The top part is a plan showing the suboutcrop positions of the coal zones and the bottom parts illustrates sections through the stratigraphy at various positions to illustrate faults and the dip of the rocks. Vertical exaggeration is ten times ......................................................... 64

Figure 33 North-south section through all farms along line A-A’ in Figure 29.

Surface topography is not shown ............................................................................... 65

Figure 34 Standard setup of geophysical logging equipment in the field................................. 68

Figure 35 Diagram of setup of geophysical logging equipment in the field ............................ 69

Figure 36 Plot of lithology and relative densities determined by geophysical logging. Three curves are long-spaced density (off-set for clarity), shortspaced density (serrated line) and Vectar processed density (smoothed line) ............................................................................................................................. 71

Figure 37 Photo of the local geology and a superimposed geophysical log of a borehole in a bulk sample pit illustrating the correlation between the geophysical logging and the lithotypes ...................................................................... 72

Figure 38 Regression plot of geophysically derived Vectar processed densities versus weighed densities ............................................................................................ 74

Figure 39 Regression plot of geophysically derived compensated densities versus weighed densities ....................................................................................................... 74

Figure 40 Regression plot of geophysically derived long-spaced densities versus weighed densities ....................................................................................................... 75

Figure 41 Regression plot of geophysically derived short-spaced density versus weighed density .......................................................................................................... 75

10

Figure 42 Detail illustration of gradual increase in Rd near the mudstone/coal contact as the probe moves upwards in the borehole. “Deduced” data is created in the process ................................................................................................. 77

Figure 43 Correlation between boreholes using the geophysical profile .................................. 78

Figure 44 Detail illustration of how an identified sample would be classified using the various Rd intervals and counting the number of times a specific Rd value would fall in a specific class interval. The data of the graph is reported per centimeter and everything to the left of the vertical R 1,35 line would report to the 1,35 interval on the distribution diagram and so on ................... 80

Figure 45 Geophysically derived cumulative distribution (GDCDD) line diagram for various samples in borehole G2500038 on Groenfontein. The horizontal axis depicts the relative density and the vertical axis the cumulative value of the density data per sample ........................................................ 82

Figure 46 Geophysically derived cumulative distribution line diagram of Sample

01B in 18 boreholes . One line represents a cumulative distribution graph for a sample of a specific borehole ............................................................................. 83

Figure 47 Anomalous curve for G250019 Sample 08A ........................................................... 84

Figure 48 Proximate information for the yield at various fractions for borehole

G250019 Sample 08A. ............................................................................................... 85

Figure 49 lithological/geophysical log for G250019 sample 08A ............................................ 85

Figure 50 Four farms (Geelbekpan, Groenfontein, Gannavlakte and Ringbult) chosen to determine whether the method could be applied to more than one adjacent area ............................................................................................................... 87

Figure 51 Geophysical profile along A-A’ illustrating varying geology over the

Study area. Coloured labels denote the bottom of coal zones. Red lines denote major stratigraphic boundaries ....................................................................... 88

Figure 52 Standard deviation calculations at specific Rd's (1,60 and 2,0) - green blocks.......................................................................................................................... 92

Figure 53 Combined GDCDD of Sample 01A (Spread at and Rd of 2.0 indicated with light blue dotted line) ......................................................................................... 94

Figure 54 Graphical representation of Sample 01A and Sample 01B on the Study area illustrating the varying lithology of this sample over the Study area ................. 98

11

Figure 55 Thickness of Sample 01A over the Study area showing thickest development of Sample 01A in the in the south (blue area) ...................................... 99

Figure 56 Plan of section line (B - B’) on Groenfontein ........................................................ 100

Figure 57 Profile along section B - B’ on Groenfontein, Different color represents different Samples ...................................................................................................... 101

Figure 58 Sample 01A, GDCDD (on Groenfontein) at the top and yield curves along a selected profile at the bottom. Yellow vertical lines are discussed in point 8.4 on p. 104 ................................................................................................ 102

Figure 59 Local lithological variation, based on geophysical logging, from north to south in Sample 01A, ending on Ringbult (on the right of the figure).

Colored labels denote the bottom of individual coal zones. Red line is the correlation of Zone 11 .............................................................................................. 103

Figure 60 Combined actual yield and geophysically derived data for Welgelegen showing divergence starting at about an Rd of 1.75. Darker lines are usually actual analytical densities and lighter colored lines the GDCDD curves. Horizontal axis is relative density and vertical axis is percentage ............... 105

Figure 61 Anonymous data for Sample 32A illustrating faulting in the borehole with the green curve. Jumps in lines are a function of the data being collected in 10 cm intervals ...................................................................................... 106

Figure 62 Anonymous data for Sample 02A illustrating effect of weathering in the area ........................................................................................................................... 107

Figure 63 Standard deviation line graph for the combined average standard deviation of Rd 1.6 and 2.0, for all the samples in the Ellisras basin found in the

Study area (per farm) and the associated trend lines ................................................ 113

Figure 64 Grootegeluk subdivision of coal zones into mining benches, Dreyer

(1994) ....................................................................................................................... 114

Figure 65 Study area subdivision of coal samples into coal zones ......................................... 115

Figure 66 Trend of the GDCDD curves along profile A-A’ for Zone 11 in the boreholes, on Groenfontein (borehole number listed below the graph).

Each curve represents the GDCDD in one borehole. The average trend curve is also indicated (thick red line) for all the curves combined. The

12

lowest (borehole G250005), highest (G250725) and mean (G250025) curves are indicated on the graph ............................................................................. 116

Figure 67 General statistics at a Rd of 2 and 1.6 of the GDCDD’s of coal Zone 11 at

Groenfontein, plus and minus two standard deviations (2σ) in yellow blocks........................................................................................................................ 117

Figure 68 Coal Zone 11 geological/geophysical profile across Groenfontein. Note that the lithology in G250726 was inadequately logged, but the geophysical log has highlighted this fact ................................................................. 119

Figure 69 Combined GDCDD of Coal Zone 11 on Welgelegen illustrating the effect of erosion on coal Zone 11 ....................................................................................... 120

Figure 70 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of coal Zone 11 at

Welgelegen. Plus and minus two standard deviations in yellow blocks .................. 121

Figure 71 Coal Zone 11 profile, illustrating why low, high and average values occur in the GDCDD diagram on Welgelegen as a result of local geological conditions ................................................................................................ 122

Figure 72 The lateral thickness variation of coal Zone 11 over the Study area ...................... 124

Figure 73 The standard deviation of the GDCDD for the zones at a Rd of 1.60 and

2.0 based on sample and zone information for Geelbekpan..................................... 128

Figure 74 The standard deviation of the GDCDD for the zones at a Rd of 1.60 and

2.0 based on sample and zone information for Groenfontein .................................. 129

Figure 75 The standard deviation of the GDCDD for the zones at a Rd of 1.60 and

2.0 based on sample and zone information for Gannavlakte ................................... 130

Figure 76 The standard deviation of the GDCDD for the zones at a Rd of 1.60 and

2.0 based on sample and zone information for Ringbult .......................................... 131

Figure 77 The standard deviation of the GDCDD for the zones at a Rd of 1.60 and

2.0 based on sample and zone information for all farms .......................................... 132

Figure 78 Graph of the averaged and combined standard deviation at a Rd of 1.6 and 2.0, for all the zones. (Trend lines are dashed) .................................................. 133

13

LIST OF TABLES

Table 1 SACS (1980) subdivision as used at the Groenfontein Area (highlighted in red).............................................................................................................................. 52

Table 2 Typical geophysical las data supplied by Weatherford (units below in the table) ........................................................................................................................... 70

Table 3 Reaction of geophysical probe over the contact between coal and mudrock for Sample 32A .......................................................................................................... 76

Table 4 Distribution diagram of densities in borehole G250038 – for specific samples. ...................................................................................................................... 81

Table 5 Summary of Sample 01A data. .................................................................................... 95

Table 6 Standard deviation calculated at two Rd’s from the GCCD data for Zone 11 to Zone 01. Zone 05-11 forms the Volksrust Formation (highlighted in brown) whist the un-highlighted values of Zones 01-04 form part of the

Vryheid Formation. .................................................................................................. 126

Table 7 Standard deviation calculated at two Rd’s from the GCCD data for Zone 01 to Zone 11 as deduced from combined individual sample information.

Zone 05-11 forms the Volksrust Formation (highlighted in brown) whist the un-highlighted values of Zones 01-04 form part of the Vryheid

Formation. ................................................................................................................ 127

14

ACT

CIM

CMMI

CRIRSCO

ECE

GDCDD

IMMM

JORC

JSE

MSD

SAIMM

SAMREC

SAMVAL

SANS

SME

TISD

UN-ECE

UNFC

ABBREVIATIONS

Advanced Coal Technologies – A company doing coal analysis in Pretoria

SAMREC Code

Canadian Institute of Mining, Metallurgy

Council of Mining and Metallurgical Institutions

Committee for Mineral Reserves International Reporting

Standards

Economic Commission for Europe

Geophysically Derived (Relative density) Cumulative

Distribution Line Diagram

Institute of Materials, Minerals, and Mining

Joint Ore Reserves Committee

Johannesburg Stock Exchange

Multiple seam deposit type coal

South African Institute of Mining and Metallurgy

South African Mineral Resources Committee

The South African Code for Reporting of Mineral

Resources and Mineral Reserves

The South African Mineral Asset Valuation Committee

South African National Standard

Society for Mining, Metallurgy, and Exploration

Thick interbedded seam deposit type

United Nations Economic Commission for Europe

United Nations Framework Classification for Fossil

Energy and Mineral Reserves and Resources

15

DEFINITIONS

ADEN (in las file)

CADE (in las file)

Calliper

CODE (in las file)

Competent Person

Compensated density

A las file is data readings as received from the geophysical probe usually in text format. ADEN is also known as Vectar processed density readings.

Calliper width from density tool.

A device for measuring the diameter of the internal wall of a borehole or tubing using an arm.

Compensated density.

A person who has a minimum of five years’ experience relevant to the style of Mineralisation and type of deposit or class of deposit under consideration and to the activity he or she is undertaking and is registered with one or more recognized professional organisations.

A density log that has been corrected for the effect of mud and mudcake by using two or more detectors at different spacing’s from the source.

The shorter the spacing, the shallower the depth of investigation and the larger the effect of the mudcake. Thus, a short-spaced detector, which is very sensitive to the mudcake, can be used to correct a long-spaced detector, which is only slightly sensitive to it.

In a typical two-detector compensation scheme, the density measured by the detector furthest spaced from the receiving part of the probe and this value is corrected by a factor, which is a deduced function from the difference between near- and long-spaced densities. The correction is found to depend on the difference between formation and mudcake density multiplied by mudcake thickness.

Although there are three unknowns, simple functions are reliable for moderate corrections.

Experimental results are often presented in the form of a spine and ribs plot. There are other schemes using, for example, more detectors.

16

DENB (in las file)

DENL (in las file)

DEPT (in las file)

DEPO (in las file)

Factoring

Gamma ray (in las file)

GRDE (in las file)

HQ/TNW

Degree of Compensation

Propagation logs rely on measuring the difference in properties of a wave at two receivers. The borehole influences this difference if the tool is tilted or if there is a cavity opposite one of the receivers. The effect can be compensated for by using two transmitters that radiate sequentially in opposite directions. In ideal conditions, the effect of a tilt or a cavity is exactly opposite for the two transmitters, so that an average gives the correct result. Borehole compensation is different from borehole correction.

Short-spaced density readings. Also referred to as near-spaced density readings.

Long-spaced density readings.

Depth in metre.

Density porosity readings.

Some mathematical correction applied to make anomalous data acceptable.

The principle of an activation log, which is a log of elemental concentrations derived from the characteristic energy levels of gamma rays emitted by a nucleus that has been activated by neutron bombardment.

Gamma ray readings from density tool.

Specific diameter or size of diamond core drilling.

17

Las file

Study area

A las file is data readings as received from the geophysical probe, usually in text format.

Geographical area on which the work for this thesis was done. The area consists of nine farms in the Steenbokpan district.

Proximate data

The proximate analysis of coal was developed as a simple means of determining the distribution of products obtained when the coal sample is heated under specified conditions. Proximate analysis separates the products into four groups:

(1) moisture,

(2) volatile matter, consisting of gases and vapours driven off during pyrolysis,

(3) fixed carbon, the non-volatile fraction of coal, and

(4) ash, the inorganic residue remaining after combustion. Proximate analysis is the most often used analysis for characterizing coals in connection with their utilisation.

Samples

The coal formations display such a well-developed repetition of coal-mudstone assemblages that it can be divided into seven sedimentation cycles or coal zones. Smaller sub-cycles or samples are contained within these coal zones.

Vectar (Vertical Enhancement by Method by which a digital filter is used to smear

Combination and Transformation of

Associated Responses) the short-spaced log to match the resolution of the long-spaced density (and also compensated) log.

Zones

The coal formations display such a well-developed repetition of coal-mudstone assemblages that it can be divided into seven sedimentation cycles or coal zones.

18

1 INTRODUCTION

The SAMREC Code (the Code), launched in March 2000, governs all forms of public disclosure relating to exploration results, Mineral Resources and Mineral Reserves by Mineral companies in South Africa. The SAMREC Code (2007) defines a ‘Mineral resource’ is a concentration or occurrence of material of economic interest in or on the earth’s crust in such form, quality and quantity that there are reasonable and realistic prospect for eventual economic extraction. The location, quantity, grade, continuity and other geological characteristics of a Mineral resource are known, or estimated from specific geological evidence, sampling and knowledge interpreted from an appropriately constrained and portrayed geological model. Mineral Resources are subdivided, and must be so reported, in order of increasing confidence in respect of geoscientific evidence, into Inferred, Indicated or Measured categories. Portions of a deposit that do not have reasonable and realistic prospects for eventual economic extraction should not be classified as a Mineral resource.

The Code also provides for a two-way relationship between Mineral Resources and Mineral

Reserves. This accounts for uncertainties associated with any of the modifying factors considered when converting Mineral Resources to Mineral Reserves which may result in there being a lower level of confidence in the Mineral Reserves than in the corresponding Mineral

Resources. Allocation into the appropriate reserve or resource category must be made by a

Competent Person. General concepts for reporting coal Resources and coal Reserves were established with the publication of the SAMREC Code, prepared by the South African Mineral

Committee under the auspices of the SAIMM. The Code provides the framework and standards for public reporting for the Johannesburg Stock Exchange (JSE).

This section of the listing rules (Section 12) of the JSE, sets out the criteria for the listing of, and the additional disclosure requirements for, Mineral Companies, and in certain circumstances, substantial Mineral assets of non-Mineral Companies.

19

The Code has been adopted by the relevant National Reporting Organisations (NROs) including SAIMM and SAMREC member organisations and is incorporated in the JSE listing requirements.

In South Africa, the Code sets out a required minimum standard for the public reporting of exploration results, Mineral Resources and Mineral Reserves. References in the Code to public report or public reporting pertain to those reports detailing exploration results, Mineral

Resources and Mineral Reserves and prepared as information for investors or potential investors and their advisers. The Code takes into account issues of a global nature while addressing certain circumstances unique to South Africa. The following principles govern the application of the Code:

• Materiality: A public report contains all the relevant information that investors and their professional advisors would reasonably require, and expect to find, for the purpose of making a reasoned and balanced judgement regarding the exploration results,

Mineral Resources and Mineral Reserves being reported on.

• Transparency: The reader of a public report must be provided with sufficient information, the presentation of which is clear and unambiguous, to understand the report and not be misled.

• Competency: The public report is based on work that is the responsibility of suitably qualified and experienced persons who are subject to an enforceable professional code of ethics. Public reports dealing with Exploration Results, Mineral Resources and

Mineral Reserves must use only the terms Proved or Probable Mineral Reserves,

Measured Mineral Resource, Indicated and Inferred Mineral Resources and Exploration

Results. Thus only Resources/Reserves whose presence was proven beyond a certain

level of certainty are reported. Figure 1 indicates the relationships between the various

categories of reporting of Resources and Reserves as used in the SAMREC Code

(2009)

20

Figure 1 The relationships between the various categories of reporting of Resources and

Reserves according to the SAMREC Code

The SAMREC Guidelines stipulate that:

• Evaluation techniques and key assumptions must be disclosed as per Clause 27,

SAMREC code (2009);

• metallurgical recovery factors must be included in public reports, SAMREC code

(2007); and

• that if any of the data is materially adjusted or modified for the purpose of making the estimate, SAMREC code, (2000), this fact must be clearly disclosed in the public report”.

21

During the compilation of the SAMREC Code, it was soon apparent, SANS standard, (2004) that additional guidelines and parameters were required to standardise the reporting of coal

Resources and coal Reserves for both Securities Exchange requirements and for the South

African National Coal resource and Coal reserve Inventory (the National Coal Inventory). The standard was prepared in conjunction with the Code by the SAMREC Coal Commodity

Specific Sub-committee. The mechanistic application of the SANS standard by Competent

Persons has led to instances where Resources were downgraded, based on the standard recommended in the SANS standard and not on real, practical issues. There are dispute mechanisms available, but these are not preferred as an amicable solution should be arrived at before going into dispute. The SANS standard makes specific mention of two basic coal deposit types, which are representative of South African coal deposits, i.e.; multiple seam and thick

interbedded seam deposit types (Figure 2),.

Figure 2 Multiple seam deposit type Coal resource (<50%ash) and thick interbedded seam deposit type on right, SANS (2004)

22

• Multiple seam deposit type coal (MSD) is characterised by a discrete number of coal

seams (see Figure 2), typically between 0,5 m and 7,0 m in thickness, separated by

inter-burden units whose thickness generally significantly exceeds the thickness of the individual coal seams (typically Witbank and Highveld coal types). The typical associated drilling densities, required in this deposit type, which is subject to the opinion of the Competent Person, is depicted in Figure 3.

Figure 3 Drilling densities required for a multiple seam deposit

• Thick interbedded seam deposit type (TISD), characterized by a succession of multiple, thin interbedded coal and non-coal layers with a total thickness of typically between 40

m and 70 m (Figure 2) (typically encountered in the Ellisras Basin). The typical

associated drilling density, subject to the expert opinion of the Competent Person is

depicted in Figure 3 and Figure 4.

23

Figure 4 Drilling densities required for a thick interbedded seam deposit

One can note from Figure 4 that the Ellisras Basin deposits have twice the recommended

distance between the boreholes for the Inferred Mineral Resource and Indicated Mineral

Resource categories because the inherent thickness and lateral continuity of the coal seams in the area makes correlation possible. The exception is the 350 m distance between data points recommended for both deposit types in the measured Mineral resource categories. The reason this was done at the time the standard was written, was because insufficient enough data was available at the time on the Ellisras Basin which could shed light on the recommended drilling densities for measured Mineral resource purposes, Dingemans, D, (2012, personal comment).

During the course of the author’s work, a due diligence study was carried out by external auditors on the status of resources necessary as feedstock for the planned Medupi power station. The auditors of that study made the following important conclusion regarding the status of the resources namely, that the resources, that were expected to be at a measured state, were classified as indicated based solely on the fact that the drilling density did not meet the

350 m or 8 boreholes per 100 hectare specified in the SANS standard. The distances mentioned

24

in the SANS standard were not utilised as indicative but as prescriptive. Opinions gathered from other consulting companies confirmed that while indications in the SANS standard could be interpreted as prescriptive, they may be used as additional information by the Competent

Person:

• The mechanistic application of the SANS standard by Competent Persons has led to disputes in the classification of resources at some mines;

• the cost of analysis of one borehole in the Ellisras Basin, based on determining the metallurgical and heating properties of the coal, is in the region of R1,5 million per borehole. This excludes costs for testing the coal for gasification suitability; and

• the graphical depiction, of the position of these eight boreholes, is not given in the

SANS standard and is open to interpretation. There is general consensus that these eight

boreholes must be evenly distributed. Some indication is given in Figure 5.

Figure 5 Distribution of boreholes to get to eight boreholes per 100 hectare

Borehole logging or well logging, commencing in 1928, makes use of an instrument to describe a specific physical feature of the intersected lithological units present in a drilled borehole. Logging may be done during or after the drilling process is completed. The logging

25

procedure consists of lowering a logging tool, attached to a cable, and a measuring system into the borehole, and measuring one or several parameters while the probe is hoisted up the borehole at a pre-determined speed. The logging tool is designed to measure radioactivity, electrical or a variety of other properties of the surrounding rocks or fluid present in the borehole. In this thesis, the main emphasis is on the collection of density measurements. The data is captured in real time through the wire line cable and supplied to the client in paper or electronic format. In the 1960’s, it was discovered that coal can be delineated accurately by high resolution density, neutron and gamma ray logs.

“Slimline logging in coal has been in use since 1960 and the use thereof has spread over the whole world. It is the industry standard”, Firth (1999). The method for correlating well logs was patented in America on the 13 December 2007 - US 7,295,926 B2. The inventor was

Benjamin Peter Jeffryes, of Riston (UK) and the assignee, Schlumberger Technology

Corporation, Ridgefield, CT (US), US Patent (2007).

All boreholes in this study were geophysically logged and the various coal seams correlated over the entire Study area. When the need arose to have the relative density information of the various coal seams independently verified, an innovative method was devised to determine the relative densities of the various coal samples. The geophysically derived relative density values were compared to the analytically determined relative densities and a correlation of more than

95% was proved. This method was so successful that it was used as basis for checking for weighing errors made by the technical assistants during the course of their work.

It was also found that a specifically derived relative density value, the Vectar processed density, was the best when compared with the analytically determined relative densities. Using the geophysical data to “fingerprint” each sample became standard practice at this time, eventually resulting in a method whereby the Vectar processed density graphs of the various coal samples could be correlated and compared over the entire Study area.

26

The aim of this thesis includes proving that other drilling data (a combination of percussion drilling and geophysical logging), distinct from cored drilling data alone, can provide data points for proving the lateral continuity of the ore body, both in terms of structure and quality, and may be used by a Competent Person to estimate the tonnage, grade and yield with a higher level of confidence.

2 THE AIM OF THIS THESIS

The aim of this thesis includes proving that, other drilling data (a combination of percussion drilling and geophysical logging), distinct from cored drilling data alone, can provide data points for proving the lateral continuity of the ore body, both in terms of structure and quality, that may be used by a Competent Person to estimate the tonnage, grade and yield of the coal deposit with a higher level of confidence.

The SANS working group formulated the SANS10320 standard that deals specifically with coal resource classification. The SANS 10320 standard states that, for the purpose of classifying thick, inter-bedded seam deposit Resources, such as those found in the Ellisras

Basin, in the Mineral resource category, there has to be quality data points at a recommended spacing of 350 m apart, which is similar in practice to the other coal basins in South Africa.

All the distance requirements, for classifying the resources in the Inferred or Indicated Mineral

Resource category, were double for the Ellisras Basin than those of other areas in South Africa.

The latter would usually be classified as multiple seam deposits (such as found in the Witbank area).

27

When officially reporting resources, it was found that some Competent Persons viewed the recommended spacing requirement of 350 m, not as a guideline, but as a ”law”. The primary aim of this thesis is determining whether the devised method of reworking the geophysical data

(GDCDD–method) could give a Competent Person enough confidence in the validity of specific data, to influence their opinion; resulting in the possible reclassification of the resources to a higher category. For instance, from Inferred to the Indicated category or from

Indicated to the Measured Mineral Resource category.

During the course of the project the most applicable geophysical density measurement was identified which correlated the best with the relative densities as measured with the

Archimedes’ method. These density measurements as extremely important as they are used in the modelling process to determine the in-situ tonnes.

The method devised was used to prove that data from percussion drilling could also serve as a dependable point of observation to classify resources.

• The first step in this process was determining how accurately the geophysical data predicted the relative density of the coal seams on the Study area;

• a method (GDCDD-method) was then devised using the geophysical log to ‘fingerprint’ the various coal seams and specific parts (samples) of the coal seams;

• the devised method was used to illustrate the correlation between the samples over a widely spread geographic area and was used to demonstrate that the ‘fingerprint’ for the percussion boreholes is identical to the cored boreholes;

• the devised method was used to illustrate the correlation between the zones over a wide area and to demonstrate that the ‘fingerprint’ for the percussion boreholes is identical to the cored boreholes; and

• the suitability of the proposed method on the classification of the coal resources, was discussed in table format.

28

3 LITERATURE STUDY

The literature study was done upon four aspects of the study:

• The requirements of the SAMREC code;

• applicability and accuracy of geophysical density logging of coal;

• distance based drilling patterns; and

• the responsibilities of the Competent Person.

3.1 SAMREC Code

Structural changes in the mining sector during the past decade, mainly as a consequence of globalisation, have had a profound impact on the financing of new mining and exploration projects, Frick (2002). The present structure relies heavily on obtaining financing on capital markets in the developed world. Most of the capital used for direct foreign investment in the mining sector in Africa is raised on a small number of stock exchanges in Europe, North

America, Australia and South Africa. Over time, all these institutions developed codes of corporate governance requiring transparent financial public disclosure.

By the early 1990’s, two international groups, the Council of Mining and Metallurgical

Institutions (CMMI) and United Nations Economic Commission for Europe (UN-ECE),were working independently towards the development of international definitions for Mineral resource and Mineral Reserve classification. A sub-committee of the CMMI, Committee for

Mineral Reserves International Reporting Standards (CRIRSCO) was made up of representatives from Australia, the Canadian Institute of Mining, Metallurgy (CIM), the South

African Institute of Mining and Metallurgy (SAIMM), the Institute of Materials, Minerals, and

Mining (IMMM), and the Society for Mining, Metallurgy, and Exploration (SME). These groups met for the first time in 1994 during the 15 th

CMMI Congress in Sun City, South

Africa.

29

This CRIRSCO sub-committee also stated that since 1992, however, the UN-ECE had been developing its own set of definitions (the UNFC). The UNFC enabled the comparison of different national Mineral resource and Mineral Reserve classification systems, particularly for those countries in transition to market economies. These two groups soon realised that their efforts would be more fruitful if the results were merged. In 1998 and 1999, the UN-ECE and

CRIRSCO met in Geneva where the UN-ECE agreed to adopt, with minor modifications, the

CRIRSCO definitions into the UNFC for Mineral Resources and Mineral Reserves that were common to both systems. The UN-ECE suggested that CRIRSCO definitions be reduced into shorter sentences to facilitate translation and to give the definitions true international status. .

The UN-ECE has published and approved the United Nations Framework Classification for

Fossil Energy and Mineral Reserves and Resources (UNFC-2009) in 2009 in cooperation with

CRIRSCO and the Society for Petroleum Engineers (SPE), the World Petroleum Council

(WPC), the American Association of Petroleum Geologists (AAPG) and the Society of

Petroleum Evaluation Engineers (SPEE) United Nations, (2010). This code represents an umbrella classification of a universal character which reconciles, unifies and expands, at a high-level, the classification definitions of the CRIRSCO template for Minerals and coal and the Petroleum Resource Management System (PRMS) for oil and gas. The UNFC-2009 framework remains outside the immediate scope of this study, which is linked to the South

African coal industry and therefore to the national South African code”.

The coordination of Mineral resource and Mineral Reserve definitions by UN-ECE and

CRIRSCO gave consistency in the accompanying Guidelines to the definitions. The effect of this was that the reporting codes of individual countries were then revised along similar lines.

The similarity of the reporting codes of CRIRSCO member countries has reached a point where there is now development of:

30

• An international definition for the Competent Person;

• a list of principles which would provide minimum requirements for professional institutions overseeing the Competent Person(s); and

• International Reporting Code and Guidelines.

To this end, the following draft documents have been prepared by CRIRSCO for submission to its member countries for review:

• International Guidelines for Reporting Mineral Resources and Mineral Reserves;

• International Definition of the Competent Person; and

• International Rules of Conduct for the Competent Person.

The SAMREC Code, launched in March 2000, governs all forms of public disclosure relating to exploration results, Mineral Resources and Mineral Reserves by mineral companies in South

Africa. According to Mullins (2008), “the JSE has also seen huge growth in market capitalisation and one-third of the JSE is still resource dominated, which is comparable to other stock exchanges, such as those in Australia and Canada. The SAMREC and SAMVAL Codes must be viewed in the context of the industry which they were designed to support. Confidence in our resources, reserves and the valuation of these, underpin this trillion dollar industry.”

Mineral resource confidence classification should take into account practical considerations such as drilling, sampling and assay integrity, borehole spacing, geological control and continuity, grade continuity, estimation method and block size, potential mining method and reporting period, Snowden (2001).

Lundell (2006), states that during the late 1990’s, unfortunate developments within the mining industry made clear the need to develop uniform Canadian standards to govern how issuers disclose scientific and technical information about mineral projects to the public. The response

31

was National Instrument 43-101, a rule issued by the Canadian Securities Administrators which aimed to restore public confidence in mining related stocks by enhancing the accuracy and integrity of public disclosure in the mining sector. The Instrument governs the public disclosure of scientific and technical information by publicly traded mining companies, covers oral statements as well as written documents and websites, and requires that all disclosure be based on advice by Qualified Person who, in some cases, must be independent of the mining company and the property. A Qualified Person as defined in the Instrument is an individual who:

• Is an engineer or geoscientist with at least five years of experience in mineral exploration, mine development or operation or mineral project assessment, or any combination of these;

• has experience relevant to the subject matter of the mineral project and the technical report; and

• has been a member in good standing of a recognised professional association.

Vaughan and Felderhof (2002), mentioned that the choice of the appropriate category of mineral reserve is determined primarily by the relevant level of confidence of the mineral resource.

Figure 1

(

on p

. 21)

shows the direct relationship between Indicated Mineral Resources and

Probable Mineral Reserves and between Measured Mineral Resources and Proved Mineral

Reserves used in the SAMREC code. The code also provides for a two-way relationship between Measured Mineral Resources and Probable Mineral Reserves. This accounts for any of the uncertainties associated with any of the modifying factors considered when converting

Mineral Resources to Mineral Reserves which may result in there being a lower level of confidence in the Mineral Reserves than in the corresponding Mineral Resources. An Indicated

Mineral Resource could never be converted to a Proved Mineral Reserve. Allocation into the appropriate category must be made by a Competent Person.

32

Gluskoter (2000), mentions that the National Coal resource Assessment of the United States of

America uses the following categories when classifying coal resources:

• Measured Mineral Resource: coal within a radius of 400 m of a control point where the thickness of the coal has been measured;

• Indicated Resources: coal within a radius of 400 m to 1 200 m of a control point;

• Inferred Resources: within a radius of 1 200 m to 5 km of a control point; and

• Hypothetical Resources: coal beyond a radius of 5 km from a control point.

The JORC code (2004) mentions the following:

• An Inferred Mineral Resource is that part of a mineral resource for which tonnage, grade and mineral content can be estimated with a low level of confidence;

• Indicated Mineral Resource is that part of a mineral resource for which tonnage, densities, shape, physical characteristics, grade and mineral content can be estimated with a reasonable level of confidence; and

• Measured Mineral Resource is that part of a mineral resource for which tonnage, densities, shape, physical characteristics, grade and mineral content can be estimated with a high level of confidence.

What is important to note is that the JORC Code does not make any mention of actual distances from control points.

3.2 Applicability and accuracy of geophysical density logging of coal

Geophysical logging has long been an invaluable aid to correlating lithologies world-wide and provide economical information Wood et al (1983). Ryder (2002) states that the use of geophysical logging as a tool in sequence stratigraphy is “seriously under-developed”. The use of gamma rays (to determine specific lithological boundaries), spontaneous potential logs

33

(which measures changes in electrical potential between an electrode on the sonde and one at the surface) and resistivity logs (which measures the resistivity of fluids in the surrounding rock to an applied electrical current) is commonplace Ryder (2002). The use of density measurements as a means of correlating stratigraphy is not mentioned anywhere in press. An internet search and direct approaches to various organizations and individuals has confirmed this state of affairs.

According to Binzhong and Esterle (2008), coal responds well to most geophysical logging methods because its properties contrast well with those of other lithological units commonly intersected in coal-bearing sequences. Coal has in general a low density, a lower seismic velocity, lower magnetic susceptibility, a higher electrical resistivity, and lower radioactivity compared to surrounding rocks. Density logs are routinely and easily acquired during the exploration process, but they are influenced inter alia by the decay of the source, mud caking, borehole cavities, and the presence of water, logging speed, and good pre-shift calibration.

The density tool logging technique is based on subjecting the surrounding formation to a bombardment of medium energy (0,2-2,0MeV) collimated gamma rays and measurement of their attenuation between the energy source and the energy detector, Ryder (2002). The attenuation (or Compton scattering) is a function of the number of electrons that the formation contains – its electron density (electrons/cm3) – which is in turn very closely related to its common density; the gamma log does not record the electron density, it measures the attenuation, resulting from the electron density, Ryder (2002).

Binzhong and Esterle (2008) also described how density logs were derived from the interaction of gamma rays from a gamma ray source (usually Caesium (Cs) 137) through Compton scattering within the formation rock surrounding the borehole. A density log is a record of the electron density (which is the number of electrons per cm

3

) of rocks adjacent to a drill hole.

The log measures the induced gamma rays emitted by the rocks after bombarding them with a gamma ray source encased in a probe and lowered into the drill hole. The more dense the

34

rocks, the more gamma rays they absorb. Ryder (2002) discussed how accurate densities could

be measured with geophysics (Figure 6).

Figure 6 Statistics for coal density measurement by various means illustrating that accurate densities can be measured with geophysics. *Density = induced density, Ryder (2002).

One can deduce from the above data that the geophysical data corresponds well to the actual relative density values. The geophysical density data, on which this thesis is based, was obtained by use of a dual density sonde supplied by an internationally known company,

Weatherford. This tool has a diameter of 76 mm (3 inches) and collects the following information:

• short and long-spaced density;

• compensated density;

• degree of compensation;

• calliper and borehole volume; and

• gamma ray - naturally occurring gamma radiation which originates from the radioactive isotope of potassium (K

4

0), and from isotopes in the decay chains of uranium 238 and thorium 232. Among clastic rocks, these tend to have low abundance in sandstones and coals, but generally a high abundance of in clay Minerals.

35

Figure 7 Dual density sonde (from Weatherford promotional material) showing various operating parameters and physical measurements

36

This sonde is a dual detector density tool in which the near and far measurements of density are characterised and calibrated independently, and then combined in model-based algorithms which correct for borehole size, fluid density variations and mudcake effects. This results in a compensated density field log which can be further corrected using Vectar processing of the compensated density log. This produces a log with the spatial resolution of a short-spaced log.

In the coal mining industry, these tools are used for:

• Coal identification;

• coal quality;

• porosity/Lithology;

• enhanced resolution logging; and

• gas detection.

3.2.1 Vectar Processing

Firth (1999), mentions that high resolution devices suffer one major drawback: borehole effects become more pronounced as vertical resolution increases. The result is that rough boreholes will adversely affect results, and thus detectors with longer spacing’s (and hence lower vertical resolution, but reduced borehole effects are normally preferred.

There is a way, however, to partially circumvent this dilemma. From any tool that has at least two measurements, it is possible to extract the high resolution log and impose it on the poorer resolution (generally compensated) log. This is Vectar (ADEN) processing. It’s most notable application is in density logging where it produces the high resolution compensated curve. The key to Vectar processing is in the digital filter used to smear the short -spacing log to match the resolution of the long (and the compensated) log. Subtracting the unfiltered short spaced log from the smeared original, leaves bed boundary information while eliminating stick-offs due to borehole effects. Adding the boundary information to the compensated curve (weighted by a

37

function of the degree of compensation) gives an unconditional improvement to the resolution of the compensated log. To get the best results from density logs, the good vertical resolution of the short-spaced log has to be combined with the good quantitative measurement of the compensated log. This is done by the mechanism known as Vectar processing.

According to Samworth (2010), Vectar processing can be considered in two ways:

• Imposition of the short-spacing log resolution upon the compensated, or long spaced log; or

• continuous dynamic calibration of the short-spacing log by the compensated or long spaced log.

Vectar processing, which uses compensated density as a base, produces a log that is preferable to using either the short spaced or long spaced log.

An example of the effect of this is shown in Figure 8

38

Figure 8 Example of a geophysical density log. The graph on left is natural gamma and calliper log and the one on the right is the various density logs

In Figure 8 the following can be observed when comparing the Vectar log with the near-

spaced density log and compensated density log:

• The Vectar log overlays the compensated log in value; and

• the effects of the cavities on the near-spaced log are suppressed.

39

3.3 Distance only based drilling patterns

The distance between the data points can be related to the sample, drill-hole distance spacing

and/or to the variography range calculated by means of geostatistics (Figure 9). One of the

disadvantages of this method is that the resources are classified in a range of concentric or rectangular circles for which the area needs to be calculated. Generating these shapes depends

on the particular modelling or drafting package. Minex calculates rectangles (as in Figure 10)

and ArcGIS, circles as in Figure 11.

Figure 9 Map illustrating a resource classification scheme using only distance criteria. The solid stars represent the location of drill holes. The filled circles represent the area covered by the search parameters surrounding a hole. The filled circles may represent the portion of a deposit, potentially classified as Indicated Resources, Pincock Perspectives (2009)

40

Figure 10 Rectangles drawn around data points (boreholes) as modelled in the geological modelling package - Minex

41

Figure 11 Circles drawn around data points (boreholes) at the study area, with ArcGIS, an

ESRI product

42

3.3.1 Resource classification based on drillhole density

The basis for classifying the resources is based on how many boreholes occur in a specific area, and the presumption is that the denser the pattern, the more reliable the information, (as in

Figure 12).

Figure 12 Resource classification based on drilling density, Pincock Perspectives (2009)

Figure 12 illustrates the common method for classifying resources in the Ellisras Basin.

43

3.4 Rules of conduct for a Competent Person

When preparing reports, Competent Persons need to be confident they can face their peers and demonstrate competence in the commodity, type of deposit, and situation under consideration.

Where doubt exists, they should seek advice from appropriately experienced colleagues or decline to act as a Competent Person in a particular situation. There are no specific rules for the resolution of disputes between a company and a consultant, between a company and another company, or between a company, individual or stock exchange. The parties can go to court, or seek advice from consultants /experts, or the matter may go to arbitration with some professional body (like the South African Council for Natural Scientific Professions

(SACNASP)), or referred to a panel of experts - as in the case of public reports published by the JSE. Any person (or company) being under suspicion of misconduct, negligence or involved in a dispute should have, and has, the right to defend themselves with the means considered most appropriate, Camisani (2010, personal comment). According to Felderhof and

Vaughn (2002), resources and reserves public reporting regulations are becoming more stringent in many mining countries, requiring the particular expertise and experience of professional resources and reserves auditors.

The Code recognises that estimation of Mineral Resources and Mineral Reserves:

• Is a team effort compiled under the direction of a Competent Person;

• each Competent Person should accept responsibility for his or her own contribution where, there is a clear division of responsibilities within a team;

• the Competent Person is responsible for the report as a whole report; and

• the Competent Person should ensure that the work of the other contributors is acceptable.

François-Bongarçon (1998) stated that “The most striking fact for a consultant doing regular detailed due-diligence studies are how differently people, companies and financing institutions react to having their work, or the object of their interest, audited”. He also stated that there is a trend towards the rule of thumb that reserves be known on an annual basis within ± 15 relative percent at 90 percent confidence. This would mean that, on average, for one year in 20, the

44

tonnage, grade or contained metal will be less than 85 percent of the estimate, which could be considered a reasonable business risk.

The orebody is the only real asset in a mining project and the only one that will generate revenue. If it should be misrepresented, it would be a major concern to bankers, Benning

(2000). To gain comfort, potential lenders will want to see that extensive exploration programmes have been undertaken and that the results vetted by appropriate independent parties, and finally, that they have been audited by credible consulting engineers.

There are several ways in which risk, in respect of confidence, can be quantified and methods such as geostatistics have been exhaustively described by many authors in a great many articles. One of the newer techniques is conditional simulation, a method which allows one to evaluate practically any variable that can impact on a reserve. Some forms of risk are impossible to quantify, such as geological interpretation and the competence of the team doing the field work.

3.5 Comment on the literature study

An internet search, as well as email contact with universities and geophysical organisations, revealed no information on the subject of using and applying relative density measurements for

‘fingerprinting’ coal seams. The use of geophysics for density measurement is proven technology and very accurate. The classification of coal resources and the data on which this is based on, is subject to the scrutiny of a Competent Person and his/her professional opinion as regards to the quality of data on which the assumptions are based.

45

4 THE STUDY AREA

The Study area is situated in the southwestern part of the Limpopo Province of the Republic of

South Africa (Figure 13) and is geologically located in the Ellisras Basin (cf Waterberg

Coalfield). The exploration project was based on collaboration between two large role players in the South African coal mining industry, namely, Sasol and Exxaro. They collaborated for the purpose of identifying whether the coals in the Ellisras Basin was suitable for gasification purposes in the Sasol process, and that sufficient reserves existed on the farms on which the two companies have coal exploration rights. The Study area is situated within the boundaries of the Lephalale Magisterial District, in close proximity to the village of Steenbokpan. The

Permo-Carboniferous Ellisras Basin, forming part of the Karoo Supergroup, and effectively thus, the Ellisras Basin, extends approximately 90 km east-west and approximately 40 km north-south near the northwestern border of the South Africa, and also extends into Botswana, where it passes into the much larger Kalahari (Karoo) Basin.

Study area

Figure 13 Location of the Study area after Sambo ( 2008)

The Study area (Figure 14)

consists of nine farms and extends roughly 20 km from north to south and 10 km east to west (20 000 hectares).

46

27°19’ E

North

NORTH

CENTRAL

SOUTH

10 km

Figure 14 Location of the nine farms and geographic areas in the Study area

20 km

23°32’ S

47

4.1 Regional structural setting

The Ellisras Basin forms a small part of the Kalahari Basin, which is a widespread basin

occurring in South Africa, Botswana and Namibia (Figure 15).

Figure 15 The Ellisras Basin and other Karoo basins in southern Africa, Mtimkulu (2009)

The structural history of this area has been well documented by Mtimkulu (2009) and the following is an abridged version of that work. The Ellisras Basin is fault bounded along its

southern and northern margins and is a graben type deposit (Figure 16). In this area, the basin

is influenced by three major fault coal zones, the Eenzaamheid Fault delineating its southern limit, the Zoetfontein Fault near its northern limit, and the Daarby Fault, with a down-throw of approximately 350 m towards the northeast. The Daarby fault divides the coalfield into a deep lying north-eastern portion and a shallow south-western portion.

48

Sedimentological facies changes also influence the continuity of the coal zones, with deterioration in coal development. Mtimkulu, (2009) also describes how the depositional patterns in the Ellisras Basin were influenced by mobile geological structures such as lineaments or faults which were continuously active synsedimentary geological structures.

27°10’ E

23°40’ S

Figure 16 Sketch map of the Ellisras Basin showing major structural elements

The Ellisras, Mopane and Mmamabula (see Figure 15) coalfields formed within a larger

intracratonic trough (Soutpansberg) and were subjected to tectonic/structural reactivation during Permian to early Triassic times, Arnot and Williams (2007). Several structural lineaments in the basement rocks to the Ellisras Basin were reactivated over time, probably due to the intrusion of the Bushveld Igneous Complex (ca. 2056 Ma), Buick et al (2001). Tectonic reactivation, caused in part by sediment loading, took place during the Karoo era and this controlled locally the sedimentation taking place within the basin. Energy levels influenced the types of sediments and this played a role in the formation of specific lithotypes. Coal formation took place during deposition of the Ecca Group (±280 Ma) and these are the Grootegeluk

(Volksrust) and Swartrant Formations (Figure 17). Coal from the Grootegeluk Formation is

49

considered to be autochthonous, while the coal from the Swartrant Formation is allochthonous,

C. Dreyer, (2004, personal comment).

Figure 17 Coal-bearing formations within the stratigraphy of the Ellisras basin-fill, Bordy et al (2010). Coal zones 11-4 are of the thick interbedded coal type with Coal zones 3-1 being multiple seam deposits, according to SANS (2004)

4.2 Local structural setting

At the Study area, the northern boundary of the coal deposit (Figure 18) is formed by faulting

that has a displacement of about 50 m towards the south, while the most southerly margin is delineated by a fault that has a throw of about 150 m to the north. Numerous smaller faults criss-cross the orebody; and most major faults strike roughly east-west and the minor faults strike north-east. There is a major northeast-southwest trending faulted coal zone in the southern part of the area.

50

LEGEND

Faults

Throw

10 km

Figure 18 Local structural geology and borehole positions at the Study Area

North

4.3 Stratigraphy

The Ellisras Basin forms part of the Karoo Supergroup. Subdivision of the Karoo Supergroup as encountered in the Ellisras Basin is based mainly on lithological boundaries. All the

51

“classic” units of the Karoo succession as found in the Main Karoo Basin Sequence are present

in the basin area and hence, the same nomenclature is applied (Table 1). Local stratigraphic

terminology for this basin (Figure 17) also shown, for comparison.. Only the Beaufort and

Ecca Group rocks (highlighted in red) are found in the Study area

Table 1 SACS (1980) subdivision as used at the Groenfontein Area (highlighted in red).

GROUP FORMATION

(SACS – 1980)

FORMATION

(Cilliers 1951)

Bordy et al

Subdivision

2010

Representative Rock

Type

Average

Thickness

Drakensberg

Basalt

Clarens

Sandstone

Drakensberg

Cave

Sandstone

Clarens

Sandstone

Lava, purplish to red, amygdaloidal

95 m

Sandstone, fine grained, white to yellow-brown to reddish

80 m

STORMBERG

Elliot Red Beds

Lisbon

Formation

Molteno

BEAUFORT Beaufort

Molteno

Beaufort

Greenwich

Formation

Eendragtpan

Formation chocolate brown, clayey

Sandstone, white, medium to coarse grained, scattered pebbles

Mudstone, purple and greenish grey, alternating at top, light

90 m grey at base

Volksrust

Mudstone

Upper Ecca

Grootegeluk

Formation

Intercalated mudstone and bright coal

60 m

ECCA

DWYKA

Vryheid

Pietermaritzburg

Mudstone

Lower Ecca

Dwyka

Middle Ecca

Dwyka

Goedgedacht

Formation

Sandstone and grit, intercalated

55 m carbonaceous mudstone, siltstone, few thick coal seams, mainly dull

Mudstone and sandstone, grit in lower portions

150 m

Swartrant

Formation

Wellington and

Waterkloof

Formations

Tillite 3 m

52

4.4 Coal types and distribution

4.4.1.1 Grootegeluk Formation

The upper part of the coal deposit in the Study area, the Grootegeluk Formation (Figure 19),

comprises intercalated mudstone and bright coal layers (Figure 20).

Figure 19 The Grootegeluk Formation coals at Groenfontein, adapted from Bordy et al (2010)

This formation has an average thickness of 72m ± 5m ( Figure 21). It displays such a well-

developed repetition of coal-mudstone assemblages that it can be divided into seven

sedimentation cycles (Figure 27) or coal zones. Smaller sub-cycles (samples) are contained

within these coal zones; these were sampled individually during exploration of the deposit.

53

Figure 20 Photograph of typical cores from Grootegeluk Formation coals. The yellow pencil marks are lithology boundaries with associated measured depths below surface. White pencil writing denotes sample numbers and the amount of plastic bags needed to place the material in weighing purposes

Figure 21 Thickness (general) statistics of the Grootegeluk Formation at the Study area

54

The Grootegeluk Formation (Figure 19) coal zones typically start with bright coal at the base,

with the ratio of coal-mudstone decreasing from the base of each coal zone upwards (Figure

22).

Mudrock

Sample 07A

Coal

Sample 08A

Sample 09A

Figure 22 Typical upwards coarsening cycles on which sample boundaries are based. CO9 is coal Zone 09 and the other units are depth in meters

The Grootegeluk Formation mudstone shows an increase in carbon content with depth and range from a massive bluish-grey mudstone at the top to carbonaceous mudstone towards the

55

Basal coal zone. Although the thickness and coal quality of the Grootegeluk Formation is reasonably constant across the coalfield, a large variation in the yield of semi-soft coking coal occurs vertically in the coal succession, but the upper coals have a higher coking coal yield.

4.4.1.2 Swartrant Formation

The Swartrant Formation (

± 55m thick) forms the lower part of the coal deposit. It consists of carbonaceous mudstone and sandstone with inter-bedded dull coal seams varying in thickness

from 1.5 m to 17 m (Figure 23). The Swartrant Formation displays the characteristics of a

Multiple Seam Deposit Type, according to the guidelines of the SANS 10320 standard.

Figure 23 Thickness statistics (MINEX) of the thickness of the coal zones in the Swartrant

Formation

There are three coal seams or coal zones in the Swartrant Formation, consisting predominantly

of (Figure 25 and Figure 26) dull coal, with some bright coal and interbedded sandstones

developed at the base of coal zones 1, 2, 3 (Figure 27). Due to lateral facies changes and

changes in the depositional environment, these coal zones are characterized by a large variation

in thickness and quality (Figure 24).

56

Zone 03

Zone 02 Zone 01

Density readings from geophysical logging

Figure 24 Variation in lithology of the Swartrant Formation coals seams. Red lines denote zone boundaries and the illegible writing on the left side of the lithology log and to the right of the geophysical log is the depth in one meter intervals

57

Figure 25 Swartrant coal seams with laminated siltstone

Figure 26 Swartrant coal seams, showing dull coal interbedded with sandstone (coarse, white rock) and mudrock (dark grey laminated rocks). The yellow pencil marks are lithology boundaries with associated measured depths below surface. White pencil writing denotes sample numbers

In the Waterberg Coalfield (Ellisras Basin) there is a convention on how to sub-divide coals

into coal zones (Figure 28). This was done to find a logical scheme for creating benches in the

mining operations as well as for aiding the mining of coal, as a specific unit then has certain characteristics. The numbering convention of the various coal zones as applied in the Ellisras

basin sequence is shown in Figure 29.with the Study area on left and Grootegeluk mine on

right.

58

60%

samples Nrs 24A-F

2%

Sample 32A

Figure 27 Sub-dividing the various coals into coal zones and samples in the Ellisras Basin. Study area on left and Grootegeluk mine on right

59

As an example of how the coal zones can be mined is shown below (Figure 28

).

The

Grootegeluk mine’s mining benches are based on the coal zones and all the coal seams are mined except for the lowest coal seam in the succession (coal Zone 01 or sample 32A), as it is below a 10 m thick sandstone and is not currently economically viable.

Figure 28 zones and mining benches as defined at Grootegeluk mine

60

4.5 Sub-outcrop maps of the various coal zones

Figure 29 shows the position of the various areas identified for the purpose of illustrating the

sub-outcrop positions of the various coal zones at the Study area.

27°19’ E

A

Northern Area

Central Area

23°32’ S

Southern Area

A’

Figure 29 Geographical areas identified in the Study area

10 kilometer

61

The fault in the northern part of the Grootwater farms can be clearly seen (Figure 30) where

the coal seams dip towards the South and West. The following sub-outcrop and section maps show the occurrence of the various upper coal zones in the Groenfontein area.

WELGELEGEN

GROOTWATER

Swartrant Fm

GEELBEKPAN

Grootegeluk Fm

5 KILOMETER

Figure 30 Northern Area coal zones.

The top part is a plan showing the sub-outcrop positions of the coal zones and the bottom parts illustrates sections through the stratigraphy at various positions to illustrate faults and the dip of the rocks. Vertical exaggeration is ten times

62

N

In the central area, the extensive faulting in the eastern part is clearly visible (Figure 31). On

Vlakfontein, the Swartrant Formation coals sub-outcrop against the overlying aeolian sand cover.

VLAKFONTEIN

GROENFONTEIN

TAMBOOTIEVLEI

Grootegeluk Fm

DUIKERFONTEIN

Swartrant Fm

5 KILOMETER

5 KILOMETER

Figure 31 Central Area coal zones. The top part is a plan showing the sub-outcrop positions of the coal zones and the bottom parts illustrates sections through the stratigraphy at various positions to illustrate faults and the dip of the rocks. Vertical exaggeration is ten times

63

In the southern area (Figure 32), the effect of the eastern part is clearly seen. It is interesting

to note that on Vlakfontein, the Swartrant Formation coals occur in sub-outcrop west and is shallow dipping along a north-south profile.

GANNAVLAKT

Swartrant Fm

DUIKERFONTE

Grootegeluk Fm

5 KILOMETER

RINGBULT

Figure 32 Southern Area top coal zones.

The top part is a plan showing the sub-outcrop positions of the coal zones and the bottom parts illustrates sections through the stratigraphy at various positions to illustrate faults and the dip of the rocks. Vertical exaggeration is ten times

In the southern area (Figure 32) the effect of the Eenzaamheid fault is clearly seen. It is also

interesting to note that the rocks dip towards the west and are shallow dipping along a north-

south profile. A complete profile along A-A’ is shown in Figure 33

64

1 5 0 m e t e r

Grootegeluk Fm

Swartrant Fm

2 0 k m

F i g u r e 3 3 N o r t h - s o u t h s e c t i o n t h r o u g h a l l f a r m s a l o n g l i n e A - A ’ i n F i g u r e 2 9 . S u r f a c e t o p o g r a p h y i s n o t s h o w n

1 4 0 m

6 5

5 TECHNICAL INFORMATION RELATED TO THE PREPARATION

OF THIS THESIS

5.1 Prospecting methods used at Groenfontein

The prospecting method used at Groenfontein comprised the drilling of cored exploration boreholes on a random spacing of ± 1 000 m x 1 000 m and infill percussion drilling, with the exception of Groenfontein where detailed cored drilling was done to delineate the deposit geometry in detail (as a bulk sample was mined from this area for gasification tests). All the boreholes lithologies were described in detail and the boreholes were geophysically logged.

The coal samples recovered from the cored boreholes were analysed by Advanced Coal

Technology (ACT) laboratory in Pretoria.

5.2 Drilling techniques, planning, control and material recovery

5.2.1 Cored drilling

Cored boreholes were drilled at the Study area by drilling the initial part of the borehole

(piloting) through the loose overburden using tricone or similar methods. Initially (pre-

2006), HQ/TNW boreholes, with percussion piloting through the overburden, were drilled at

Groenfontein. Thereafter, all boreholes were drilled to large-diameter PQ (123 mm) coring to retrieve sufficient representative sample material to cater for the large suite of analyses to be performed. All holes were geophysically logged by Weatherford within a few hours or a day from being completed. All boreholes drilled after 2006, was done by one contractor, Diabor, and all percussion drilling by Ellisras Boorwerke, which has been drilling in the area for more than 20 years.

5.2.2 Percussion drilling

Percussion infill drilling (usually 115 mm in diametre) is geologically described within a few days of being completed and geophysical logging finished immediately after completion of

66

the borehole. Percussion drilling is mostly used for structural delineation and determining the position of the coal seams. The geophysical logging is done for verification purposes and determining the position of the various samples in each borehole.

Cored boreholes were drilled at the Study area by drilling the initial part of the borehole

(piloting) through the loose overburden using tricone or similar methods. According to contract, cored boreholes were only accepted for payment if:

• The core is clean;

• the depth of a borehole has been verified by a geologist; and

• core recovery for the coal seams has been confirmed to be 95% or higher.

Random depth checks are conducted during the drilling of each borehole and the geophysical logging is also used to verify reported drilling depths. Statistical analyses indicate that the reported drilling depths are within acceptable level of accuracy.

5.3 Geophysical density logging performed at the Study area

All boreholes were geophysically logged in the Study area. These included 164 cored and

261 percussion boreholes. Each of the cored boreholes had approximately 40 samples, consisting of about 4 bags each, which is a total of 25 760 bags that had to be weighed in water and in air, a total of 51 520 measurements. Five percent of these had to be reweighed making a total of 2576 samples; a grand total thus of 54 096 measurements. At an average of

150 m per borehole there were 24 600 m of geophysical logging that were completed and with the 261 percussion holes a further 39 150 m of logging were added, making the total logged length to 63 750m. Each centimeter of the total logged length has its own record making up for a dataset extending close to 6.4 million lines of information. This data was used in the compilation of this thesis.

67

Weatherford, the geophysical wireline logging company, used a DD 2 sonde which has a 7

cm vertical resolution. Figure 34 and Figure 35 illustrates the standard setup of equipment in

the field.

Figure 34 Standard setup of geophysical logging equipment in the field

68

Figure 35 Diagram of setup of geophysical logging equipment in the field

The data received by the geologist in the field are usually supplied as a text file (Table 2) and

typically contains tabulations of relevant readings of gamma, density types, porosity, depth and calliper width, which the geologist must interpret.

69

Table 2 Typical geophysical las data supplied by Weatherford (units below in the table)

Depth GRDE CODE CADE

160.210

88.732

2.385

120.933

DENL

2.353

160.200

80.265

2.389

121.055

2.357

160.190

79.588

2.392

121.182

2.362

160.180

106.342

2.395

121.215

2.366

160.170

100.585

2.401

121.244

2.372

160.160

103.294

2.403

121.371

2.375

160.150

103.633

2.402

121.498

2.377

160.140

102.278

2.402

121.557

2.379

160.130

81.620

2.399

121.760

2.379

160.120

80.604

160.110

76.201

160.100

78.571

2.397

2.401

2.409

121.887

121.986

122.027

2.380

2.385

2.392

160.090

77.217

160.080

79.926

160.070

78.571

160.060

75.185

2.411

122.099

2.396

2.421

122.127

2.405

2.426

122.254

2.411

2.423

122.301

2.411

160.050

70.105

160.040

70.443

160.030

78.910

160.020

80.604

160.010

87.377

160.000

87.038

159.990

81.281

159.980

71.798

159.970

72.814

159.960

72.814

2.423

2.426

2.426

2.436

2.430

2.418

2.404

122.346

122.376

122.574

123.018

123.009

123.290

123.494

2.413

2.417

2.419

2.428

122.599

2.422

2.439

122.799

2.431

2.435

122.868

2.430

2.432

2.428

2.419

2.408

159.950

82.297

2.396

123.888

2.401

159.940

102.278

2.382

124.200

2.388

159.930

110.745

2.367

124.460

2.375

159.920

116.164

2.355

124.632

2.362

159.910

125.308

2.338

124.795

2.345

2.386

2.368

2.374

2.409

2.434

2.422

2.421

2.419

2.459

2.439

2.506

2.492

2.528

2.534

2.479

2.278

2.317

2.370

2.333

2.429

2.311

2.326

2.321

2.407

2.338

DENB

2.270

2.295

2.337

2.275

2.338

2.317

13.593

15.262

12.949

13.499

12.625

13.055

11.895

13.969

13.226

14.374

16.780

16.164

16.460

17.573

18.907

DEPO

15.774

16.328

15.891

15.031

15.252

14.791

14.444

15.439

15.296

15.151

15.639

15.115

14.249

12.898

14.620

12.500

2.419

2.365

2.403

2.426

2.458

2.435

2.452

2.413

2.466

2.432

2.465

2.474

2.513

2.513

2.452

2.362

2.380

2.428

2.388

2.470

2.354

2.375

2.384

2.436

2.396

ADEN

2.370

2.380

2.424

2.370

2.422

2.401

70

6 ACCURACY OF THE GEOPHYSICAL DATA IN PREDICTING

THE RELATIVE DENSITY OF THE COAL SEAMS ON THE

STUDY AREA

The following method was used:

• The lithological unit’s intersected and other geological features are described from the cored or percussion cutting;

• the geological log as captured by the geologist of that borehole is depicted in SABLE

(database program) next to the geophysical log received (Figure 36);

• the geophysical log versus the geological log is visually interpreted by the geologist, marked on the paper log and then the various samples marked out on the core in the field; and

• the samples are collected and weighed and the relative density of the coal samples determined.

Short-spaced density

Vectar processed density

Long-spaced density

Figure 36 Plot of lithology and relative densities determined by geophysical logging. Three curves are long-spaced density (off-set for clarity), short-spaced density (serrated line) and

Vectar processed density (smoothed line)

71

During 2009 and 2010, Sasol excavated a bulk sample at Groenfontein to determine if the coal could be gasified and used in the Sasol

TM

synfuels process. Figure 37 illustrates the

correlation between the geophysical logging of a borehole drilled in the pit and the geology as found in the bulk sample site at Groenfontein.

Figure 37 Photo of the local geology and a superimposed geophysical log of a borehole in a bulk sample pit illustrating the correlation between the geophysical logging and the lithotypes

72

6.1 Statistical evaluation of the measured relative densities of cored material

The relative density of a sample is extremely important as it is used to determine the tonnage of that sample in the geological model. Relative density measurements are carried out per sample by weighing each sample in air and in water respectively, and dividing the weight obtained in air by the corresponding sample’s weight obtained in water. The scale used for these measurements is calibrated using a control sample which has a known relative density.

An important requirement for SAMREC compliance is obtaining independent verification of the relative densities used for modelling purposes. This could have been done using traditional methods at the laboratory, or, as in the case of this study,, using the results from the geophysical logging done by Weatherford.

6.2 Reliability of measurements of sample masses

The reliability of weight measurements is checked in three ways, which indicate whether or not a weighing error has been made.

• The sample’s relative density (Rd) measurements are compared with the core recovery recorded for each sample;

• the weight of each sample (in air) is measured again by ACT upon receipt of the sample at the laboratory in Pretoria. ACT’s sample weight measurements are then compared with the original sample weights on the list which accompanied the samples to the laboratory, and should a significant discrepancy exist, the geology personnel at the field office is informed immediately. In general, only minimal weight measurement discrepancies were detected between the samples weights performed by the project team, and those performed by ACT; and

• all relative densities (Rd’s) measured are compared to the geophysically derived values generated by Weatherford. The Rd’s that are not within 95% of the theoretical values calculated by geophysical means are re-measured in the field. The overall

correlation between the two values for Vectar processed density is 95% (Figure 38),

73

for compensated density it is 80.9% (Figure 39), for long spaced density 66.2%

(Figure 40) and for short spaced density 81.3% (Figure 41).

Figure 38 Regression plot of geophysically derived Vectar processed densities versus weighed densities

Figure 39 Regression plot of geophysically derived compensated densities versus weighed densities

74

Figure 40 Regression plot of geophysically derived long-spaced densities versus weighed densities

Figure 41 Regression plot of geophysically derived short-spaced density versus weighed density

It was found that the density determined empirically in the field was the most accurately mimicked by the densities determined by means of the Vectar processed density reported by

Weatherford.

75

7 STATISTICAL DEPICTION OF THE GEOPHYSICAL DATA AND

ASSOCIATED CALCULATIONS

7.1 Using the geophysical log to ‘fingerprint’ the various coal zones and specific parts (samples) of the coal seams.

A method (Geophysically Derived Cumulative Distribution Density - GDCDD-method) was devised by the author to determine how geophysical logs between two adjacent boreholes could be correlated. As an opening remark it must be noted that the depiction of the density per individual sample could be inaccurate due to the fact that, as the geophysical probe approaches a specific sample, it is at the same time detecting the influence of the surrounding

lithological units as is illustrated below (Table 3).

Table 3 Reaction of geophysical probe over the contact between coal and mudrock for

Sample 32A

Dept h

Rd

76

The following must be taken into account when looking at the figure above:

• Coal (at a depth of 197,03 m) as a rock type has a Rd of 1.65;

• mudrock (at 196,92 m) as a rock type has a Rd of 2.56;

• as the probe moves from mudrock (at 196,92 m) towards coal (at 197,03 m) it gives readings between 2.56 and 1.65 depending on how close it is to the coal. When the probe reaches coal it reads 1.65;

• the inverse is true when the probe moves from coal to mudrock;

• this factor is common to all geophysical logs; and

• the actual Rd reading should be 2.3, 2.3 and 2.29 as it approaches the contact, and then 1.65, 1.65 etc. as the lithology changes to coal. A reading such as 1.80, is a calculated average between mudrock and coal as the probe moves from one rock type

to the other (Table 3 and Figure 42).

Figure 42 Detail illustration of gradual increase in Rd near the mudstone/coal contact as the probe moves upwards in the borehole. “Deduced” data is created in the process

77

When the geologist receives the geophysical log and starts correlating the various layers, he

does so based on the visual characteristics of the geophysical graph. In the case of Figure 43

below it can be seen that the geophysical log, correlate with the lithology in the boreholes as well as between boreholes themselves.

Figure 43 Correlation between boreholes using the geophysical profile

To what extent this visual, alignment takes place, is difficult to quantify. It is still unclear if one can prove, through some statistical process, that the two graphs are similar. Is it possible to know that the ‘fingerprint’ of the one graph is the same as the ‘fingerprint’ of the other graph, thereby proving that the rock types contained in that specific fingerprint are the same in terms of the lithology - a critical fact in coal geology and beneficiation?

A method of ‘fingerprinting’ the various coal samples was devised where it was decided to calculate the distribution of relative densities in the chosen geological intersection. As

78

mentioned before in this chapter, an Excel spread sheet (Table 3) was used to give a value of

1 to each Rd, depending on in which class it reported to. From (Table 3) it can be noted that:

• The use of 0.05 relative density intervals (ie, 1.35, 1.40, 1.45 etc.) was used as the basis for distinction up to a Rd of 2.30. From a Rd of 2.3 to 3.0 no washability data were done at the laboratory and only a sink value was done at a Rd of 3.0;

• if a specific Rd value, at a specific depth, classified into a certain interval, then a value of 1 was inserted at that specific depth interval; for instance;

• at a depth of 196.92 m, the Rd reported was 2.56 and this was registered as a 1 in the column where the class interval is between 2.30 and 3.00;

• at a depth of 196.99 m, the Rd reported was 2.03 Rd. This triggered a 1 value in the in the column where the class interval is between 2.00 and 2.05; and

• the 1 value for the various class intervals was then summed to determine the

distribution for the various classes of Rd’s (Table 4).

79

Figure 44 illustrates how a specific sample interval would be classified using the count of

Rd’s measured per Rd interval.

Figure 44 Detail illustration of how an identified sample would be classified using the various Rd intervals and counting the number of times a specific Rd value would fall in a specific class interval. The data of the graph is reported per centimeter and everything to the left of the vertical R 1,35 line would report to the 1,35 interval on the distribution diagram and so on

80

In the instance of Sample 1A in , this sample has a thickness of 4.23 m, the average Rd for the intersection was 2,26 and there were 288 Rd’s in the intersection classed between 2.3 and

3.00. It had nine readings classing between 2.25 and 2.3, and four readings between 2.20 and

2.25.

Table 4 Distribution diagram of densities in borehole G250038 – for specific samples.

Variations in the shape and position of individual geophysical (Figure 43) graphs indicate the

following:

• Faulting influences the relationship between the amount of coal and the amount of mudrock present in the specific intersection;

• if there are lateral facies variations, the shape of the graph, between boreholes also changes; and

• local post-depositional conditions, such as weathering and erosion, also change the shape of the density graph.

81

The cumulative data derived, per sample, for all the boreholes in Figure 45 were used to

portray geophysically derived cumulative distribution diagram (GDCDD) of the cumulative distribution of the relative density of the samples. As an example, a GDCDD was drawn for

all the samples intersected in borehole G250038 (Figure 45).

Figure 45 Geophysically derived cumulative distribution (GDCDD) line diagram for various samples in borehole G2500038 on Groenfontein. The horizontal axis depicts the relative density and the vertical axis the cumulative value of the density data per sample

82

The method devised was used to compare the behaviour of a specific lithostratigraphic unit in

adjacent boreholes (Figure 46).

Figure 46 Geophysically derived cumulative distribution line diagram of Sample 01B in 18 boreholes . One line represents a cumulative distribution graph for a sample of a specific borehole

In Figure 46, the same sample (Sample 01B) is shown for 18 different boreholes . A different

line colour is used in each instance. It can be seen that the derived curves have a similar pattern but are displaced vertically in relation to each other. This illustrates that there is a relatively good correlation between the same lithostratigraphic units between adjacent boreholes in which the stratigraphic unit occurs.

It was found that the method could also be used to check the accuracy of the raw data which the geologist uses to define the sub-units. The following example illustrates this. The Las file in the Las database gave the following information for the curve of Sample 08A for G250019

(Figure 47). From the figure it can be seen that the sample of G250019 reveals an anomalous

trend.

83

G250019

Sample 08A

Figure 48).

Figure 47 Anomalous curve for G250019 Sample 08A

The first part of the investigation into why this borehole appeared anomalous, consisted of interrogating the proximate analysis data. The proximate analysis data showed that G250019

had a very low yield (

G250019

Sample 08A

84

Figure 48 Proximate information for the yield at various fractions for borehole G250019

Sample 08A.

The printed geology log from the database (Figure 49) revealed that Sample 08A consists of

90% mudstone (with a high Rd), and a small layer of coal at the bottom of the intersection.

This distribution of the lithotypes makes it impossible for this Sample to have such high

GDCDD graph while consisting of only 10% coal. Upon further investigation, it was found the Las data file in the database was incorrect.

Figure 49 lithological/geophysical log for G250019 sample 08A

This method was easily applied to other anomalous GDCDD values and depending upon the outcome of such investigations, the database corrected accordingly. The database could thus be cleared of deduced data.

85

8 DISCUSSION ON THE USE OF STATISTICALLY DERIVED

GDCDD WHEN EVALUATING SAMPLES OVER THE STUDY

AREA

The GDCDD method was tested in boreholes on four farms in the Study area along an

irregularly-curved section line A-A’ (Figure 50). The farms were Geelbekpan – representing

the northern part of the Study area (blue circle), Groenfontein representing the central Study area (green circle), and Gannavlakte/Ringbult – representing the southern part of the Study

area (red circle in Figure 50). The test consisted of proving whether the sample lithology,

determined in boreholes with continuous coring, could be reliably followed in the adjoining percussion boreholes along the section line to the detail required for mining purposes. If successful, this would confirm the practical usefulness of the method and result in considerable saving to the coal mining companies without compromising the results of the mining operations.

86

A

NORTH

CENTRAL

A’

A’

SOUTH

Figure 50 Four farms (Geelbekpan, Groenfontein, Gannavlakte and Ringbult) chosen to determine whether the method could be applied to more than one adjacent area

Figure 51 shows a profile of geophysically logged boreholes along the A-A’ line of Figure

50, which illustrates the lateral continuity of the coal zones.

87

GEELBEKPAN GROOTWATER GROENFONTEIN GANNAVLAKTE RINGBULT

F i g u r e 5 1 G e o p h y s i c a l p r o f i l e a l o n g A - A ’ i l l u s t r a t i n g v a r y i n g g e o l o g y o v e r t h e S t u d y a r e a . C o l o u r e d l a b e l s d e n o t e t h e b o t t o m o f c o a l z o n e s .

R e d l i n e s d e n o t e m a j o r s t r a t i g r a p h i c b o u n d a r i e s

8 8

8.1 Method for discussing the GDCDD of samples

The method explained in Chapter 7 was applied in the three non-immediately adjacent areas of the Study area (north, central and south). The parameters used in the description of the curves and significant results of the GDCDD graphs were condensed under a series of headings in relevant tables accompanying each graph. These headings with their meanings are listed below.

8.1.1 Graphs

Four graphs illustrating GDCDDs from the central area top left, the northern area, bottom left and the southern area, top right as well as a combined graph of all three areas, bottom right

(Figure 53), are given and used as a base for a tabulated discussion of the selected samples.

8.1.2 Tables

A table consisting of thirteen points of discussion for each sample is given. The points of discussion in the tables are summarised below.

8.1.2.1 Angle of repose of graphs

The angle of a curve in a graph is the angle between the X-axis and the best-fitted straight line of the relevant curve. In the GDCDD graphs the wording “angle of repose” commonly refers to the percent value (read in the Y-axis) that the GDCDD curve assumes at a certain relative density (read in the X-axis). There are two significant angles of repose, one between

Rd 1.35 (where the curve intersects the X-axis) and Rd 2.25 (portion one), which can be fitted with a straight line and a second one from Rd 2.25 to Rd 3.0 which could be fitted with another straight line at a different slope than the first one. The point of the curve at Rd 2.25 represents generally the inflexion point of the GDCDD curve for each sample. The second portion of the curve is sometimes very steep, and in any event, much steeper than the first portion and depends on the analytical yield of the sample. These Rd values are based on the

89

standard used for the reporting of proximate data on the Study. No analysis is done on the fractions between 2.25 and 3.0.

The best indicator of angle of repose is the angle of the curve in portion one. A low angle of repose equates up to 40% increase over the first part of the curve (portion one, up to Rd

2.35). A medium angle of repose equates up to 80% increase over the same portion, while a

high angle of repose would amount to an 80% increase (Figure 52).

8.1.2.2 Start point of graph

All the graphs start at a Rd of 1.35.

8.1.2.3 Spread

The spread indicates how spread the lines are from each other at a certain Rd. A Rd value of

2.0 was chosen as reference point as it is usually the cut point at which a typical power station product coal would be produced from the Ellisras Basin coals. In this thesis, a narrow spread signifies that the envelope of the GDCDD lines is within 20% at a Rd of 2.0.

Narrowly spread data would indicate that the lithology is similar, while widely spread data would indicate larger variation in lithology.

8.1.2.4 Yield at a Rd 2.0

The analytical (proximate analysis) yield at a Rd of 2.0 is given. The Rd value of 2.0 was chosen as reference point as it is usually the cut point at which a typical power station product coal would be produced from the Ellisras Basin coals, as stated above.

8.1.2.5 Remarks – Yield at a Rd 2.0

General remarks regarding the yield at a Rd of 2.0 are discussed in the table.

90

8.1.2.6 Raw Rd Sample

The raw Rd of the sample will be indicated in the table. This Rd is determined by weighing the sample in water and in air in the field and was discussed previously.

8.1.2.7 Thickness (in meter)

Thickness of the sample is reported in the table.

8.1.2.8 Remarks regarding raw Rd and thickness

Remarks regarding the data of the thickness and raw Rd are included in the table. There is an inverse correlation between the raw Rd and yields as a low raw Rd is a reflection of the high amount of low Rd material (coal) in the sample, and this would equate to a higher yield.

8.1.2.9 Standard deviation at a Rd of 1.60 and 2.0

The table will contain a tabulation of the standard deviations calculated at two Rd points namely 1.60 and 2.0. These two Rd points were chosen on the following basis;

• 1.60 = the cut point on the washtable, at which nearly all the vitrinite would be removed from the coal and 2.00 = the cut point at which powerstation coal can be produced from the Grootegeluk Formation coals.

• In order to quantify the standard deviation of the data, two reference points were chosen per graph and the standard deviation of the curve values at those points were calculated and compared. The two points of interest are at a Rd of 1.60 and at 2.0

(Figure 52).

91

Figure 52 Standard deviation calculations at specific Rd's (1,60 and 2,0) - green blocks

8.1.2.10 Comments on standard deviation

General remarks regarding the standard deviation are dealt with. A higher standard deviation signifies a wider spread of the GDCDD envelope, as explained above. If the volume of mudrock increases in the stratigraphic unit in a specific borehole, as compared to neighbouring boreholes, then it has the effect of increasing the variance at the Rd 2.0 value.

8.1.2.11 Correlation accuracy by geologist

Comments, with a certain degree of subjectivity, concerning the complexity or otherwise at which a geologist could correlate the coal seams in different boreholes by means of geological and geophysical data, are included.

92

8.1.2.12 Modelling accuracy

Comments, with a certain degree of subjectivity on the ease and accuracy at which a geologist could set up and maintain a geological model of the samples or zones, are included in the table.

8.1.2.13 Use as SANS 10320 data correlation point

In South Africa, the SAMREC code is the standard for reporting resources and reserves. For coal the reporting is based on the SANS 10320 standard. Therefore, a subjective comment on the adherence to the SANS 10320 standard, is included in the table. Although the comments are subjective, this does not imply non-adherence to the SANS 10320 standard. A Competent

Person would necessarily be the final judge of any remark made in a report on resources and reserves.

8.2 Specific example from the study area

8.2.1 Sample 01A

In this chapter, Sample 01A is dealt with in detail with its relevant tabulation of comments as an example of the work undertaken to prove the usefulness of the method in interpreting the continuity of a coal seams (samples) or coal seam mining units (zones) in percussion holes.

All the samples shown in the stratigraphic column (Figure 19) have already been dealt with

in this thesis. Discussing the rest of the samples would merely be a repetitive exercise; all graphs of these samples and their relevant tabulations are shown in the Addendum on the attached CD.

93

F i g u r e 5 3 C o m b i n e d G D C D D o f S a m p l e 0 1 A ( S p r e a d a t a n d R d o f 2 . 0 i n d i c a t e d w i t h l i g h t b l u e d o t t e d l i n e )

9 4

T a b l e 5 S u m m a r y o f S a m p l e 0 1 A d a t a .

BASIC INFORMATION

ABOUT SAMPLE 01A

Angle of repose of graphs

SUMMARY

S a m p l e 0 1 A

GEELBEKPAN GROENFONTEIN RINGBULT

L o w L o w L o w

GANNAVLAKTE

L o w

I n t h e c a s e o f S a m p l e 0 1 A ( F i g u r e 5 3 ) t h e l o w a n g l e o f r e p o s e i s d u e t o t h e f a c t t h a t t h e r e a r e n o t m u c h l o w

d e n s i t y m a t e r i a l ( c o a l ) i n t h i s S a m p l e ( F i g u r e 5 4 ) , w h i c h w o u l d e q u a t e t o a l o w y i e l d .

Start point of graphs at

Rd 1.35

D a t a v a l u e s s t a r t a t t h e s a m e p o i n t .

Spread

I n t h e s o u t h , t h e s p r e a d i s ± 3 4 % ( b e t w e e n 1 8 % a n d 5 2 % ) i l l u s t r a t i n g t h a t t h e l i t h o l o g y i s m o r e v a r i a b l e i n t h i s a r e a .

Yield at Rd 2.0 –

Sample 1A

S a m p l e 0 1 A 1 1 . 2 7 1 3 . 8 2 2 6 . 7 1 3 1 . 8 3

Remarks – Yield at Rd

T h e l o w y i e l d c a n b e s e e n f o r G r o e n f o n t e i n a n d G e e l b e k p a n b u t t o w a r d s t h e s o u t h - R i n g b u l t a n d G a n n a v l a k t e –

t h e y i e l d s a r e h i g h e r a s s h o w n a b o v e . F i g u r e 5 8 ( p a g e 1 0 2 ) a l s o i l l u s t r a t e s

t h a t u p t o R d 1 . 7 0 t h e h i g h e r v a l u e s i n

2.0

t h e G D C D D g r a p h c o r r e s p o n d t o h i g h e r y i e l d v a l u e s i n t h e y i e l d g r a p h .

Raw Rd Sample 01B

Thickness (m)

0 , 2 1 v a r i a n c e 2 . 2 3 2 . 2 7 2 . 0 2 2 . 0 9

0 . 9 3 m v a r i a n c e 3 . 8 2 2 . 8 9 3 . 3 3 3 . 2 9

Remarks on raw Rd and

T h e r e i s a n i n v e r s e c o r r e l a t i o n b e t w e e n t h e r a w R d a n d y i e l d s o f S a m p l e 0 1 A . T h e t h i c k n e s s i s r e a s o n a b l y

9 5

thickness

c o n s t a n t ( F i g u r e 5 5 , p a g e 9 9 ) , b u t f o r t w o v e r y d e e p f a u l t e d b o r e h o l e s i n t h e c e n t r a l p a r t o f t h e S t u d y a r e a . O n

G e e l b e k p a n S a m p l e 0 1 A i s t h i c k e r d u e t o t h e f a c t t h a t a n e x t r a t h i n c o a l s e a m i s d e v e l o p e d a b o v e S a m p l e 0 1 A .

Standard deviation at a

Rd of 1.60 and 2.0

T h e s e e x t r a c o a l s a m p l e s o c c u r r e g u l a r l y a t t h e G r o o t e g e l u k m i n e a r e a .

Relative density 1.6 2.0 1.6 2.0

S a m p l e 0 1 A 3 . 1 3 . 6 4 . 5 5 . 1

1.6 2.0 1.6 2.0

2 8 . 3 2 8 . 4 7 . 4 6 . 0

Comments on standard

T h e s t a n d a r d d e v i a t i o n (σ ) f o r R i n g b u l t f a r m i s a b o u t ei g h t ti m e s h i g h e r th a n f o r t h e ot h e r f a r m s in t h e S t u d y a r e a .

deviation

T h e r e a s o n f o r t h i s i s t h e l o c a l w e a t h e r i n g o n t h i s f a r m ( F i g u r e 5 9 ) .

T h e d e g r e e t o w h i c h t h e g e o l o g i s t c a n c o r r e l a t e S a m p l e 0 1 A i s g o o d a n d t h i s s a m p l e i s e a s y t o c o r r e l a t e ( F i g u r e

5 4 ) , b u t R i n g b u l t / G a n n a v l a k t e i s m o r e d i f f i c u l t d u e t o t h e f a c t t h a t t h e g e o l o g y v a r i e s m o r e o v e r t h i s a r e a .

Correlation accuracy by

P r o b l e m s w i t h c o r r e l a t i o n , l i k e f a u l t i n g a n d c o r r e l a t i o n e r r o r s m a d e b y t h e g e o l o g i s t , w o u l d b e p r o m i n e n t i n t h e

geologist

G D C D D c u r v e s . I t c a n b e s e e n f r o m F i g u r e 5 4 t h a t a l t h o u g h t h e t h i c k n e s s o f t h e i n d i v i d u a l c o a l l a y e r s i n S a m p l e

0 1 A v a r i e s a b i t , t h e f a c t o f t h e m a t t e r i s t h a t t h e r e a r e t w o c o a l s e a m s i n t h e s a m p l e a n d t h i s c a n b e f o l l o w e d o v e r t h e w h o l e S t u d y a r e a . A t G e e l b e k p a n , l o c a l l y , a t h i r d c o a l s e a m i s d e v e l o p e d i n S a m p l e 0 1 A .

Modelling accuracy

G o o d c o r r e l a t i o n , e a s y t o m o d e l . N o f a c t o r i n g n e c e s s a r y t o c o m p e n s a t e f o r m o d e l l i n g a c c u r a c y a t G r o e n f o n t e i n a n d G e e l b e k p a n . S o m e f a c t o r i n g m a y b e n e c e s s a r y a t R i n g b u l t / G a n n a v l a k t e d u e t o t h e w e a t h e r i n g o f S a m p l e 0 1 a i n t h i s a r e a .

T h e u s e o f t h e G D C D D p r o v e s t h a t t h e m e t h o d c a n b e u s e d t o p r o v e t h e c o n t i n u i t y o f S a m p l e 0 1 A i n t h e a r e a w i t h

Use as SANS 10320 data

a h i g h m e a s u r e o f a c c u r a c y i n t h e n o r t h a n d a g o o d m e a s u r e o f a c c u r a c y i n t h e s o u t h . U s i n g t h e G D C D D m e t h o d

correlation point

i t i s p o s s i b l e t o c o r r e l a t e b o r e h o l e s i n a s e c t i o n a n d s t r u c t u r a l a n o m a l i e s h a s a d e f i n i t e e f f e c t o n t h e s h a p e o f t h e c u r v e s i n t h e G D C D D d i a g r a m s . B y c o m p a r i n g t h e s h a p e o f t h e v a r i o u s G D C D D d i a g r a m s o f a c h o s e n s a m p l e o r

9 6

z o n e i t i s p o s s i b l e t o s e e t h e e f f e c t o f f a u l t i n g a n d / o r w e a t h e r i n g o n a n y s p e c i f i c b o r e h o l e . T h e l a t e r a l v a r i a t i o n o f

S a m p l e 0 1 A o n R i n g b u l t i l l u s t r a t e s t h i s f a c t c l e a r l y . T h e g r a p h s f o r S a m p l e 0 1 A i s c l o s e l y s p a c e d n o t s h o w i n g l a r g e v a r i a t i o n b u t f o r t h e n o r t h . T h e y i e l d s f o r t h e s o u t h a r e h i g h e r t h a n t h o s e o f t h e n o r t h e r n a r e a . T h e t h i c k n e s s v a r i e s b e t w e e n 2 , 8 9 m a n d 3 , 8 2 m . G e e l b e k p a n h a s t h e t h i c k e s t S a m p l e 0 1 A o f t h e f a r m s a s t h e r e i s a n e x t r a t h i n c o a l s e a m d e v e l o p e d i n S a m p l e 0 1 A . I t w a s a l s o p o s s i b l e t o n o t e h o w t h e s t r a t i g r a p h y c h a n g e s l a t e r a l l y f r o m f a r m t o f a r m o r a r e a t o a r e a . F o r S a m p l e 0 1 A t h e σ v a l u e s a r e ei g h t ti m e s h i g h e r i n t h e s o u t h t h a n t h a t f o r t h e ot h e r f a r m s . T h i s i s l a r g e l y d u e t o w e a t h e r i n g i n t h e s o u t h . T h e r e i s a g o o d c o r r e l a t i o n b e t w e e n y i e l d s a n d t h e G D C D D c u r v e s . T h i s f a c t s h o u l d b e e a s i l y u n d e r s t o o d a s t h e b a s e s o f t h e G D C D D i s c o u n t i n g t h e v a r i o u s a m o u n t s o f d e n s i t i e s w i t h i n a c e r t a i n i n t e r v a l a n d t h e m o r e d e n s e m a t e r i a l f o u n d i n t h e i n t e r v a l t h e l o w e r t h e y i e l d . F o r m o d e l l i n g p u r p o s e s , S a m p l e 0 1 A s h o u l d b e e a s y t o m o d e l . N o f a c t o r i n g n e c e s s a r y t o c o m p e n s a t e f o r m o d e l l i n g a c c u r a c y a t G r o e n f o n t e i n a n d G e e l b e k p a n . S o m e f a c t o r i n g m a y b e n e c e s s a r y a t R i n g b u l t / G a n n a v l a k t e . T h e u s e o f t h e G D C D D p r o v e s t h a t t h e m e t h o d c a n b e u s e d t o p r o v e t h e c o n t i n u i t y o f S a m p l e 0 1 A i n t h e a r e a w i t h a h i g h m e a s u r e o f a c c u r a c y i n t h e n o r t h a n d a g o o d m e a s u r e o f a c c u r a c y i n t h e s o u t h .

9 7

Mudstone

Coal

Density log

S A M P L E 0 1 A

G E E L B E K P A N

S A M P L E 0 1 B

20 km

G R O E N F O N T E I N G A N N A V L A K T E

F i g u r e 5 4 G r a p h i c a l r e p r e s e n t a t i o n o f S a m p l e 0 1 A a n d S a m p l e 0 1 B o n t h e S t u d y a r e a i l l u s t r a t i n g t h e v a r y i n g l i t h o l o g y o f t h i s s a m p l e o v e r t h e

S t u d y a r e a

9 8

Figure 54 graphically illustrates the lithology as described by the geologist and the associated

geophysical log as captured by Weatherford, that illustrates the varying lithology. In Figure 55 the

thickness of Sample 01A over the Study area indicates that the thickest development of Sample 01A is in the south (blue area).

Welgelegen

Geelbekpan

Groenfontein metres

Local modelling inaccuracies due to lack of boreholes.

Ringbult

Gannavlakte

Figure 55 Thickness of Sample 01A over the Study area showing thickest development of Sample

01A in the in the south (blue area)

99

8.3 Relationship between GDCDD and yield calculated from proximate data and the relationship between cored and percussion boreholes’ GDCDD curves

A specific section line (B-B’) was chosen on the Study area to illustrate the concept of how the

GDCDD curves compare with the actual proximate data for chosen samples (Figure 56). In this

exercise both cored and percussion boreholes were selected along a section line. Furthermore,

GDCDD curves were plotted for the two drilling methods (in the top part of Figure 58) in the same

section line to find out what type of variation would occur in the same lithological unit when using a cored or a percussion borehole.

B

B’

Figure 56 Plan of section line (B - B’) on Groenfontein

100

Figure 57 Profile along section B - B’ on Groenfontein, Different color represents different Samples

Figure 57 illustrates the stratigraphic position of the various samples in the boreholes. Each colour

represents a sample. The sample cross-section was chosen to test whether percussion boreholes could be used as a data point as defined in SANS 10320. This was done for all the samples found in the

Study area but only the following samples (01A, 01B, 02A, 03A, 04A, 05A, 7A and 11A) are reported here for the sake of simplicity and clarity.

101

It was found that the two drilling types yielded the same curves, in terms of position, angle of repose, start point etc. and whatever yield curves the cored boreholes has, the percussion drilling would

replicate them (Figure 58). Figure 59 illustrates how local geological conditions influence the

spatial distribution of coal seams over the whole Study area where G250736 is in the north and the other boreholes occur in the south.

Figure 58 Sample 01A, GDCDD (on Groenfontein) at the top and yield curves along a selected

profile at the bottom. Yellow vertical lines are discussed in point 8.4 on p. 104

102

F i g u r e 5 9 L o c a l l i t h o l o g i c a l v a r i a t i o n , b a s e d o n g e o p h y s i c a l l o g g i n g , f r o m n o r t h t o s o u t h i n S a m p l e 0 1 A , e n d i n g o n R i n g b u l t ( o n t h e r i g h t o f t h e f i g u r e ) . C o l o r e d l a b e l s d e n o t e t h e b o t t o m o f i n d i v i d u a l c o a l z o n e s . R e d l i n e i s t h e c o r r e l a t i o n o f Z o n e 1 1

1 0 3

This figure shows how Sample 01A is influenced, primarily, by weathering due to the fact that it is the top coal seam in the area and that the rocks are not totally horizontal but influence by structural conditions in the Ellisras Basin.

8.4 Comparison between analytically derived yield and the GDCDD curves

It was previously mentioned in this thesis that the GDCDD method has a ‘built in’ error because the geophysical tool interprets relative densities as it moves, for instance, from mudrock to coal. These densities could not be an actual true reflection of the lithotype encountered, for instance Rd 2.7 for mudrock or Rd 1.6 for coal, but an interpreted Rd depending on the distance to the next lithotype encountered in the borehole, or if the probe is situated inside the lithotype then the actual relative densities is returned.

The method of comparing the CDCDD curves and actual proximate data yield (Figure 58)

has illustrated that there was a specific correlation between the two values and that the

GDCDD curves resemble the yield curves. The yellow vertical lines in Figure 58 are used to

illustrate this point. In the lower part of the figure, the actual yield for Sample 01A, shows that at a Rd of 1.70 the yield is between 9,9 and 13,7, while in the upper figure (GDCDD curves) the values for all the boreholes are between 8 and 20. At a Rd of 2.0 the actual yield is between 13.3 and 16.3 while on the GDCDD curves the values are between 15 and 35 – nearly double those of the yield curves. In short the red line for borehole G250011 is higher up on the graph for both the yield graph and GDCDD.

After seeing these trends in the graphs a combined graph was then prepared to investigate this

relationship in more detail (Figure 60). In this example the correlation between the two sets

of data from the farm Welgelegen is shown and this figure illustrates the effect of the ‘built in deduced’ data (the deduced Rd values reported as the probe moves from coal to mudrock).

How this comes about was discussed in detail in Chapter 7.1 (on p. 76) above.

104

The same borehole’s

(W228027) data for the analytical yield (darker brown) and the

GDCDD data (lighter brown).

Geophysical data

Analytical data

Figure 60 Combined actual yield and geophysically derived data for Welgelegen showing divergence starting at about an Rd of 1.75. Darker lines are usually actual analytical densities and lighter colored lines the GDCDD curves. Horizontal axis is relative density and vertical axis is percentage

It can be seen in the graph (Figure 60) that up to a Rd of 1.75 the curve envelope for the

analytical yield and GDCDD envelope have a similar position on the graph. Above an Rd of

1.75 the two datasets diverge. The GDCDD shows an increase, at a higher trend, than the analytical yields. This divergence is caused by the derived “deduced” data.

The conclusion that can be drawn from Figure 60 is that the GDCDD curves mimic the

actual analytical yield up to an Rd of 1.75.

105

8.5 Anonymous check of information from neighbours

During the latter stages of this thesis the author was sent anonymous data from four boreholes by a friendly neighbour to determine whether the GDCDD method can be used on this data and expose any correlation problems.

Although the data differed from the Weatherford data in that the readings are taken 10 cm

apart, it still clearly shows local variation in geology for Sample 32A (Figure 61) and

Sample 2A (Figure 62).

Figure 61 Anonymous data for Sample 32A illustrating faulting in the borehole with the green curve. Jumps in lines are a function of the data being collected in 10 cm intervals

106

Figure 62 Anonymous data for Sample 02A illustrating effect of weathering in the area

The anomalous curves were queried at the source of the data. The reply, in all the cases where anomalous graphs were seen, was that those specific samples were influenced by local anomalous geology. In the case of Sample 32A – faulting (basically no coal left in the intersection) and in the case of Sample 02A – weathering (coal material weathered away and only higher density mudrock left behind).

107

8.6 Summary on the use of statistically derived GDCDD curves when evaluating samples over the Study area

When considering all the data discussed in this chapter on the application of the GDCDD method to the classification of samples in the Ellisras Basin coals, the following conclusions and observations can be made:

Angle of repose of graphs

• The angle of repose in the GDCDD was found to be similar to and indicative of the yield up to a Rd of 1.75, when ‘deduced’ readings start to play a role and the curves then deviate from the wash curves derived from the proximate analysis;

• the angle of repose of the GDCDD curves can be subdivided into high, medium or low and the amount to which this happens is a function of how much low density material there is in the sample. The more low density material (coaly material), the higher the angle of repose;

• the GDCDD for samples 28A to 32A has a very high angle of repose due to the inherent nature of these coals – having a low ash content and associated high yield; and

• the amount of interbedded waste material, like sandstone or mudstone, influences the shape of the curves.

Start point of graph at Rd 1.35

• The graphs do not have the same point of origin; and

• the more low density material, the higher up on the y-axis the curve starts which is indicative of the amount of coal found in the sample.

108

Spread

• The GDCDD method allows the user to define spread in terms of percentage;

• the percentage spread for samples in individual geographical areas, depositional environments or simply various farms can be compared and quantified. The spread of the curves indicates how varying the depositional environments were; and

• the amount of mudrocks in the sample influences the spread for a specific sample and if the geologist does not take care when defining from-to depths of samples in boreholes accurately, it shows up as anomalous on the GDCDD curves for that specific sample.

Yield at 2.0

• The data for the analytically derived yield curves (proximate data) are mimicked by the values derived in the GDCDD curves;

• currently the yield at a specific ash product cannot be deduced for the GDCDD curves as a method for deducing the ash content from the density measurements has not yet been invented;

• the variance in yield is clearly shown by the GDCDD method;

• there are variances in yield due to the lithological changes in the area;

• when comparing the GDCDD curves for cored boreholes to those of the percussion boreholes it is clearly seen that there are no differences in the two drilling methods:

• the GDCDD method of evaluating the lithology is independent of the drilling method;

• the method allows a geologist to use the GDCDD curve as a method of comparing unknown boreholes with known boreholes and makes deductions about the lithological similarities; and

• when drilling at a certain grid spacing, it is possible to fill some of the drill points with percussion drilling and use that information, reworked by means of the GDCDD method, to come to a conclusion regarding the lateral continuity, thickness and to a certain extent yield, and form an opinion regarding the SAMREC status of the resources.

109

Raw Rd

• It has been shown that the geophysically derived (Vectar) density compares to within

95% and more of those measured with conventional methods;

• the geophysical data can give a geologist accurate predictions of the relative densities of any rock types or intersections chosen in the boreholes; and

• the geophysical measurements of the relative density can be used as an independent verification of the field Rd’s.

Thickness

• The thickness of individual samples can vary significantly over the Study area;

• there are instances where the thickness varies very little over the Study area; and

• the thickness can vary due to the depositional environment or it can be caused by faulting and weathering.

Standard deviation

• By using the σ at two Rd points (1.60 and 2.0) it is possible to compare the σ between boreholes and to make some deduction regarding the accuracy of the correlation between boreholes;

• some samples, depending on the variations in lithology, has a small σ value and if there is a large variation in lithology a large σ results; and

• the Grootegeluk Formation samples – samples 01A to 22F has an average σ value of

10 (Figure 63) and the Swartrant Formation -

Sample 28A to 32A has an average σ

value of 14 (Figure 63). The reason for this is the fact that there are interbedded

sandstones in these samples that complicate correlation.

110

Correlation accuracy by geologist

• By using the GDCDD curves it is possible to clearly highlight anomalous geology like faulting and weathering;

• sample boundary, correlation errors is shown up in the GDCDD curves;

• deduced raw data captured in the database is highlighted in the GDCDD method;

• combining the standard deviation and GDCDD curves helps the geologist to decide on how accurately samples can be correlated; and

• it is shown that anonymous data can be correlated using the GDCDD method.

Modelling accuracy

• Factoring could be necessary to compensate for modelling accuracy in instances where there is a lot of local variation in lithology;

• care should be taken when doing correlation in some parts of the Study area (such as

Gannavlakte and Ringbult);

• the inherent good lateral correlation would mean that modelling can be done to a high degree of accuracy; and

• In terms of modelling it was shown there is good lateral correlation between the samples on the various farms, which should make it easy to model the samples in the

Study area.

111

Use as SANS 10320 data correlation point

• The GDCDD method can be used to trace the lateral continuity of samples with a high degree of accuracy;

• percussion drilling has no detrimental effect on the shape of the GDCDD curve;

• there is no difference in the shape of the curves produced from percussion and cored boreholes. The two drilling methods are equally reliable with regard to their GDCDD curves;

• it is possible to make use of infill percussion drilling and use the GDCDD method to compare the results with known cored drilling boreholes; and

• there is a direct correlation between GDCDD curves and analytical yield curves below an Rd of 1.75.

112

GROOTEGELUK

10

FORMATION

SWARTRANTFORMATION

14

F i g u r e 6 3 S t a n d a r d d e v i a t i o n l i n e g r a p h f o r t h e c o m b i n e d a v e r a g e s t a n d a r d d e v i a t i o n o f R d 1 . 6 a n d 2 . 0 , f o r a l l t h e s a m p l e s i n t h e E l l i s r a s b a s i n f o u n d i n t h e S t u d y a r e a ( p e r f a r m ) a n d t h e a s s o c i a t e d t r e n d l i n e s

1 1 3

9 DISCUSSION ON THE USE OF STATISTICALLY DERIVED

GDCDD WHEN EVALUATING COAL ZONES

At the only productive mine in the area, Grootegeluk, the coal is not mined on a per sample

basis but per coal zone and these coal zones are divided into benches (Figure 64). The Study

area subdivision for the zones is shown in Figure 65. The GDCDD method, used for

evaluating the various coal samples in the Ellisras Basin area, was applied to the coal zones in the Study area to determine whether the method is applicable to these zones and could assist a Competent Person in classifying the resources.

Figure 64 Grootegeluk subdivision of coal zones into mining benches, Dreyer (1994)

114

Figure 65 Study area subdivision of coal samples into coal zones

The Groenfontein and Welgelegen areas (Figure 14) were used as reference to evaluate the

coal zones using the GDCDD method.

115

9.1 Methodology used

In this investigation the depths of the top and bottom contacts of the coal zones were identified and used as basis for the GDCDD evaluation. The geophysical data were evaluated

using the GDCDD method (Figure 66) and the curves were drawn. The values along the x-

axis at two chosen cut-points, Rd 1.60 and 2.00, were tabled. These data were then processed using the general statistics function in Excel functions to determine the following:

• general statistics;

• the 2x standard deviation (2 σ) at the chosen Rd cut-points; and

• whether the standard deviation falls within the maximum and minimum values of the data range.

Data areas used in the calculation of statistics

histogram (Figure 67

and others) in thesis

G250726

G250005

G250025

Figure 66 Trend of the GDCDD curves along profile A-A’ for Zone 11 in the boreholes, on

Groenfontein (borehole number listed below the graph). Each curve represents the GDCDD in one borehole. The average trend curve is also indicated (thick red line) for all the curves combined. The lowest (borehole G250005), highest (G250725) and mean (G250025) curves are indicated on the graph

116

9.2 Coal Zone 11

Coal Zone 11, the uppermost coal zone in the stratigraphy, consists of Samples 01A and 01B.

It has two bright coal seams of varying thickness and interbedded dark grey mudstones. On

Geelbekpan there is, locally, a third coal seam developed. The general statistics of Zone 11 at

the two chosen Rd’s is shown in (

Figure 67

).

Figure 67 General statistics at a Rd of 2 and 1.6 of the GDCDD’s of coal Zone 11 at

Groenfontein, p lus and minus two standard deviations (2σ) in yellow blocks

117

The statistics were derived using the add-in function in Excel, giving a range of parameters in

the reporting, including standard deviation (σ). In the yellow blocks on the graph (Figure

67),

data for the 2σ (standard deviation =σ) is given. Comparing this value to the minimum and maximum values for all the data for Zone 11, it indicates that for an Rd of 2 the minimum value is within the mean -

2σ and the maximum value is outside the mean + 2σ.

This would suggest that the Rd 1.60 values fall within the mean +

2σ, which is a narrow spread, and the Rd 2.0 values fall inside the mean +

2σ, which is a narrow spread as well.

This equates to Zone 11not varying significantly over Groenfontein.

There is very little lithological variation in coal Zone 11 across Groenfontein in the GDCDD

(Figure 68). However, due to a high incidence of faulting significantly influencing the dip

and depth of the coal seams,

For illustrative purposes (Figure 69) the GDCDD curves for the same zone was plotted for

an adjacent farm Welgelegen to prove the point that local geological conditions play a role in how the curves for the various boreholes present on the graph. In the case of Welgelegen there are boreholes of which the top part of Zone 11 has been weathered, and this shows up clearly on the GDCDD curves of the specific boreholes (W228015, W228713 and W228715)

(Figure 69).

Figure 71 illustrates the section across some of the boreholes on Welgelegen to show how

local weathering played a role in removing the top part of Zone 11 , so influencing the

GDCDD of these boreholes.

118

G 2 5 0 7 2 6 G 2 5 0 0 0 5 G 2 5 0 0 2 5

F i g u r e 6 8 C o a l Z o n e 1 1 g e o l o g i c a l / g e o p h y s i c a l p r o f i l e a c r o s s G r o e n f o n t e i n . N o t e t h a t t h e l i t h o l o g y i n G 2 5 0 7 2 6 w a s i n a d e q u a t e l y l o g g e d , b u t t h e g e o p h y s i c a l l o g h a s h i g h l i g h t e d t h i s f a c t

1 1 9

W 2 2 8 0 1 5

W 2 2 8 7 1 5

W 2 2 8 7 1 3

W 2 2 8 7 1 7

F i g u r e 6 9 C o m b i n e d G D C D D o f C o a l Z o n e 1 1 o n W e l g e l e g e n i l l u s t r a t i n g t h e e f f e c t o f e r o s i o n o n c o a l Z o n e 1 1

1 2 0

Figure 70 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of coal Zone 11 at Welgelegen.

Plus and minus two standard deviations in yellow blocks

In Figure 70, the data in the yellow blocks, for the 2

σ indicates, that for a Rd of 2.0, the minimum value is within the Mean -

2σ and the maximum value is also within the Mean +

2σ. This would suggest a narrow spread of the curves for both limits. The same is true for the

RD 1.6 values. This is due to erosion and proves that more of the lower part of coal Zone 11

is preserved over Welgelegen. The lower curves, (Figure 69), are widely spread vertically

while the upper curves (those above the mean) are spaced more widely apart indicating thicker and/or more numerous mudstone layers being present there.

121

Strata eroded away

Strata eroded away

W 2 2 8 0 1 5 W 2 2 8 7 1 3 W 2 2 8 7 1 5 W 2 2 8 7 1 7

3 km

F i g u r e 7 1 C o a l Z o n e 1 1 p r o f i l e , i l l u s t r a t i n g w h y l o w , h i g h a n d a v e r a g e v a l u e s o c c u r i n t h e G D C D D d i a g r a m o n W e l g e l e g e n a s a r e s u l t o f l o c a l g e o l o g i c a l c o n d i t i o n s

1 2 2

In Figure 71, borehole W228717 shows the normal development of Sample 01A, but in the

other boreholes, the complete sequence of Sample 01A, is not present as the top coal seam has been eroded away. This has clearly affected the behaviour of the GDCDD curves of coal

Zone 11 on Welgelegen (Figure 69). All the scattered graphs at the top of Figure 69 have

been affected by erosion.

Based on the statistical method applied in Figure 69 and Figure 70 two tables of information

were produced that give the σ values at a Rd of 1.60 and 2.0 for the eleven zones identified

on the Study area. Table 6 was set up using the x-axis value of the GDCDD curves, while

Table 7 shows the

averaged σ values for the zones made up from the x-values of the individual samples that make up the various zones thus combining the individual sample

values of each zone into a single value. The values in Table 7 reflect the variability of the

individual samples making up the zones while the values in Table 6 reflect less variability, as

they are based on total zone information.

The values of these two tables are illustrated figuratively for each farm in Figure 73 to

Figure 78. All

these figures illustrate that the σ values for the zones are lower than the σ values calculated for the zones based on the sample information. The lower variability of the

Volksrust Formation (Zones 11 to 5) is clearly illustrated in all these figures. The average σ value for the Volksrust Formation is 8.0 while the σ value for the Vryheid Formation is 12.1

(Table 6)

The average σ value for the zones, based on zone information is 9.2 while the average σ value for the zones, based on sample information is 13.1.These data confirm the fact that the individual samples are, clearly, more variable than the individual zones.

Zone 11 is the thickest developed over Geelbekpan where a third coal seam is developed in

Sample 01A (Figure 72). This figure shows the thicker Zone 11 in a blue-green colour on the

eastern side of Geelbekpan. The red parts in the figure is where Zone 11 has either been

123

faulted out of the stratigraphy (and is not exposed in the area) or as in the case of the south – been eroded due to the dip of the coal seams.

Welgelegen

Geelbekpan metres

Groenfontein

Ringbult

Gannavlakte

Figure 72 The lateral thickness variation of coal Zone 11 over the Study area

124

9.3 The importance of the standard deviation when evaluating zones:

• Geelbekpan (Figure 73): The average

σ value on this farm for zones 11 to 4 is between 4 and 7.1, (blue bars) and for zones 3 to 1 it is between 9.5 to 19.2.This would concur with the local geology where there is numerous thin sandstone layers within these coal zones. This would also concur with the SANS 10320 requirement that the multiple seam coal deposit needs to be drilled to a spacing that is closer than those of the thick interbedded coal seam deposit like the Volksrust Formation coals.

What this figure also illustrates is that using samples to calculate zones directly returns a higher average

σ value as the sample σ values is more variable than zones.

The average

σ for the zones based on the value for the samples (between 3.8 and

23.2).is nearly double those of the zones (between 4.0 and 19.2) (bars versus the lines).

• Groenfontein (Figure 74): The

σ value on this farm for all the zones is between 4.1 and 15.5 with Zone 4 equals to 15.5 and Zone 3, 13. The reason for this is interbedded sandstones for Zone 3 and the ‘separated’ coal seam at the bottom of Zone 4. In the case of this farm it should be possible to use the borehole spacing of the thick interbedded coal seam deposit type to prove the classification of the resources on this farm. The average

σ for the samples is nearly double those of the zones between 6.9 and 32.5.

• Gannavlakte (Figure 75): The

σ value for this farm for zones, in general, is higher than that in the north (average = ±10) with values between 4.8 and 16.2. The samples has nearly double the σ value (average = ±15) than those of the zones. These two points proves that the lithology in the south is more variable than the north.

• Ringbult (Figure 76): The

σ value for this farm, in general, is higher than that in the north (average = ±8) with values between 4.7 and 16.6. The samples also have nearly double the σ value (average = ±15) than that of the zones. These two points prove that the lithology in the south is more variable than the north and this is mostly due to the fact that the seams dip towards the north and over nearly half the total farm area the seams is subjected to weathering as they would sub-outcrop.

125

T a b l e 6 S t a n d a r d d e v i a t i o n c a l c u l a t e d a t t w o R d ’ s f r o m t h e G C C D d a t a f o r Z o n e 1 1 t o Z o n e 0 1 . Z o n e 0 5 - 1 1 f o r m s t h e V o l k s r u s t F o r m a t i o n

( h i g h l i g h t e d i n b r o w n ) w h i s t t h e u n - h i g h l i g h t e d v a l u e s o f Z o n e s 0 1 - 0 4 f o r m p a r t o f t h e V r y h e i d F o r m a t i o n .

Standard deviation of zones as calculated from averaged Zone information

FARM

06

05

04

03

02

01

Zones

11

10

09

08

07

GEELBEKPAN

1.6

5 . 4

5 . 1

5 . 1

3 . 9

4 . 0

3 . 5

1 . 9

2 . 5

1 7 . 4

1 3 . 4

2.0

7 . 3

2 . 9

6 . 3

6 . 3

5 . 4

1 0 . 8

9 . 1

9 . 3

2 0 . 8

5 . 6

Avg

6 . 3

4 . 0

5 . 7

5 . 1

4 . 7

7 . 1

5 . 5

5 . 9

1 9 . 1

9 . 5

GROENFONTEIN

1.6

3 . 9

7 . 0

5 . 8

8 . 3

5 . 3

4 . 0

5 . 0

1 4 . 4

1 5 . 7

9 . 7

2.0

4 . 4

4 . 5

9 . 0

1 0 . 6

6 . 4

1 0 . 9

7 . 6

1 6 . 6

1 0 . 4

4 . 3

Avg

4 . 1

5 . 8

7 . 4

9 . 5

5 . 8

7 . 4

6 . 3

1 5 . 5

1 3 . 0

7 . 0

GANNAVLAKTE

1.6

8 . 9

7 . 2

8 . 0

7 . 6

8 . 1

3 . 4

3 . 8

4 . 1

1 9 . 2

1 7 . 0

2.0

9 . 3

2 . 4

9 . 2

6 . 4

8 . 0

1 7 . 6

1 6 . 9

1 7 . 7

8 . 4

5 . 1

Avg

9 . 1

4 . 8

8 . 6

7 . 0

8 . 0

1 0 . 5

1 0 . 4

1 0 . 9

1 3 . 8

1 1 . 0

1.6

1 6 . 5

1 0 . 9

7 . 6

1 0 . 9

9 . 1

4 . 7

2 . 3

4 . 2

1 2 . 7

1 0 . 5

RINGBULT

2.0

1 6 . 8

6 . 2

8 . 1

1 4 . 3

7 . 2

1 6 . 0

7 . 2

1 6 . 1

6 . 7

6 . 5

Avg

1 6 . 6

8 . 6

7 . 9

1 2 . 6

8 . 2

1 0 . 3

4 . 7

1 0 . 2

9 . 7

8 . 5

1.6

8 . 7

7 . 6

6 . 6

7 . 7

6 . 6

3 . 9

3 . 2

6 . 3

1 6 . 3

1 2 . 6

TOTAL

2.0

9 . 4

4 . 0

8 . 2

9 . 4

6 . 7

1 3 . 8

1 0 . 2

1 4 . 9

1 1 . 6

5 . 4

Avg

9 . 1

5 . 8

7 . 4

8 . 5

6 . 7

8 . 9

6 . 7

1 0 . 6

1 3 . 9

9 . 0

2 9 . 3 9 . 1 1 9 . 2 1 2 . 2 6 . 7 9 . 4 1 6 . 0 1 6 . 3 1 6 . 2 1 4 . 9 1 5 . 0 1 4 . 9 1 8 . 1 1 1 . 8 1 4 . 9

Grand Total 7.8 7.9 8.4 7.7 7.8 8.3 8.7 10.0 10.0 8.8 10.2 10.2 8.3 8.9 9.2

1 2 6

Zones

11

10

09

08

07

06

05

04

03

02

01

Grand Total

T a b l e 7 S t a n d a r d d e v i a t i o n c a l c u l a t e d a t t w o R d ’ s f r o m t h e G C C D d a t a f o r Z o n e 0 1 t o Z o n e 1 1 a s d e d u c e d f r o m c o m b i n e d i n d i v i d u a l s a m p l e i n f o r m a t i o n . Z o n e 0 5 - 1 1 f o r m s t h e V o l k s r u s t F o r m a t i o n ( h i g h l i g h t e d i n b r o w n ) w h i s t t h e u n - h i g h l i g h t e d v a l u e s o f Z o n e s 0 1 - 0 4 f o r m p a r t o f t h e

V r y h e i d F o r m a t i o n .

3.2

21.5

17.9

26.3

10.7

3.5

9.6

7.9

10.1

9.2

5.6

3.4

1.6

Standard deviation of zones as calculated from averaged individual Sample information

FARM

GEELBEKPAN

2.0 Avg

GROENFONTEIN

1.6 2.0 Avg

GANNAVLAKTE

1.6 2.0 Avg 1.6

RINGBULT

2.0 Avg 1.6

TOTAL

2.0 Avg

4.0

7.4

3.8

12.1 10.0

6.8

8.5 10.4 6.5

7.4

7.0

9.3

6.9

8.4

8.3

7.8 5.9 6.9 16.0 16.0 16.0 8.5

14.4 7.2 10.8 7.9 6.2 7.1 10.6

10.2 14.4 12.3 13.1 16.3 14.7 9.7

8.2

6.8

8.4

8.7

13.0 11.4

11.2 9.5 13.3 16.3 13.5 14.1 16.7 14.7 16.0 18.3 16.5 13.4 15.6 13.5

11.0 10.1 7.2 9.6 8.4 14.9 10.9 12.9 15.3 13.6 14.4 11.6 11.3 11.5

14.8 10.2

12.9 8.2

7.5

8.3

14.9

10.9

11.2

9.6

7.4

7.6

25.4

28.1

16.4

17.9

8.0

7.5

25.8

23.8

16.9

15.7

7.1

6.7

20.2

18.9

13.7

12.8

14.7 8.9 11.9 12.4 12.1 9.9 25.8 17.8 12.9 25.0 19.0 9.5

24.8 23.2 19.0 16.9 17.9 23.1 5.7 14.4 15.3 4.1 9.7 19.7

19.5

12.9

14.5

16.3

10.4 14.2 20.2 12.9 16.6 25.5 7.3 16.4 19.1 6.7 12.9 20.7 9.3 15.0

8.9 17.6 35.4 29.6 32.5 13.5 8.5 11.0 13.5 8.3 10.9 22.2 13.8 18.0

12.0 11.3 13.4 13.3 13.2 13.5 14.2 13.8 13.1 14.9 14.0 12.7 13.6 13.1

1 2 7

GROOTEGELUK FORMATION

Based on the total Zone illustrating less deviation

Based on compositing Samples in the Zone illustrating higher deviation

SWARTRANTFORMATION

F i g u r e 7 3 T h e s t a n d a r d d e v i a t i o n o f t h e G D C D D f o r t h e z o n e s a t a R d o f 1 . 6 0 a n d 2 . 0 b a s e d o n s a m p l e a n d z o n e i n f o r m a t i o n f o r G e e l b e k p a n

1 2 8

GROOTEGELUK FORMATION

Based on the total Zone illustrating less deviation

Based on compositing Samples in the Zone illustrating higher deviation

SWARTRANTFORMATION

F i g u r e 7 4 T h e s t a n d a r d d e v i a t i o n o f t h e G D C D D f o r t h e z o n e s a t a R d o f 1 . 6 0 a n d 2 . 0 b a s e d o n s a m p l e a n d z o n e i n f o r m a t i o n f o r G r o e n f o n t e i n

1 2 9

GROOTEGELUK FORMATION SWARTRANTFORMATION

F i g u r e 7 5 T h e s t a n d a r d d e v i a t i o n o f t h e G D C D D f o r t h e z o n e s a t a R d o f 1 . 6 0 a n d 2 . 0 b a s e d o n s a m p l e a n d z o n e i n f o r m a t i o n f o r G a n n a v l a k t e

1 3 0

GROOTEGELUK FORMATION SWARTRANTFORMATION

F i g u r e 7 6 T h e s t a n d a r d d e v i a t i o n o f t h e G D C D D f o r t h e z o n e s a t a R d o f 1 . 6 0 a n d 2 . 0 b a s e d o n s a m p l e a n d z o n e i n f o r m a t i o n f o r R i n g b u l t

1 3 1

GROOTEGELUK FORMATION SWARTRANTFORMATION

F i g u r e 7 7 T h e s t a n d a r d d e v i a t i o n o f t h e G D C D D f o r t h e z o n e s a t a R d o f 1 . 6 0 a n d 2 . 0 b a s e d o n s a m p l e a n d z o n e i n f o r m a t i o n f o r a l l f a r m s

1 3 2

GROOTEGELUK FORMATION SWARTRANTFORMATION

F i g u r e 7 8 G r a p h o f t h e a v e r a g e d a n d c o m b i n e d s t a n d a r d d e v i a t i o n a t a R d o f 1 . 6 a n d 2 . 0 , f o r a l l t h e z o n e s . ( T r e n d l i n e s a r e d a s h e d )

1 3 3

9.4 Discussion on the use of GDCCD when evaluating the coal zones of the

Study Area.

By combining the standard deviation values at a Rd of 1.60 and 2.0 and averaging the two

values, the graph in Figure 78 can be drawn. The graph illustrates the following points

previously made:

• The large variation in σ values for Zone 11 is due to the fact that weathering in the

south plays a role in altering the nature of Zone 11 (Figure 68);

• the uncertainty about the ‘separated’ bottom coal layer developed in Zone 4

(displayed in the Addendum) influenc es the σ values of Zone 4 on Groenfontein in as much as the average σ values for the preceding zones hovers around 6 and then it increases to 13 at Zone 4.

• Interbedded sandstone plays a role in elevating the σ values for zones 1 to 3 on all the farms;

• the σ value increase from Zone 11 to Zone 1 for three of the four areas investigated

(Geelbekpan, Groenfontein and Gannavlakte), but shows a decrease for Ringbult

(Table 6);

• on Geelbekpan (G226) the σ value increases three-fold (3 to 12), on Groenfontein

(G250), doubles (5-10), on Ringbult (R303) it shows a slight a decrease (10-9) and doubles on Gannavlakte (G299) (6-12);

• the variability of Zone 01 is seen on all the farms (Figure 78). Interbedded sandstone

plays a role in elevating the σ values for zones 1 to 3 on all the farms;

• the Vryheid Formation (Zones 11 to 05) is less variable than the Volksrust Formation

(Zones 04 to 01) as can be seen in Table 6 and Table 7;

• the σ value for the zones are generally less variable than the σ value for the zones

made up of σ values combined from individual samples making up the zones (Figure

73 to Figure 77);

134

• Figure 69 illustrates that there are six boreholes on Welgelegen that are influenced by

anomalous local conditions, but the other 35 boreholes are closely spaced around the average.

• The GDCDD curves can be used to determine whether the geology in the percussion holes are the same as those of the cored boreholes.

• The GDCDD curves clearly indicated anomalous geology of the zone in specific

boreholes (W228015, W227715, W228713 in Figure 69).

• The Competent Person can make up his/her mind about variability of the geology on this farm by using the GDCDD curves of all the boreholes (cored and percussion).

• Figure 66 illustrates that there are possibly two slightly anomalous intersections of

Zone 10 (G250726 and G250005) on Groenfontein out of 30 boreholes. Zone 10 is thus very regular over the farm and should not present any great challenges when interpreting the lateral extent of Zone 10 on Groenfontein. Percussion boreholes confirm the geology of Zone 10 in collaboration with cored boreholes.

• The GDCDD curves for Zone 09 on Groenfontein (displayed in the Addendum on

CD) shows that there are two boreholes with anomalous GDCDD curves (G250721 and G250726) when compared to the average curve of G250020. When observing the section drawn through the mentioned boreholes it can be conclusively seen that

G250721 was subjected to faulting that removed the top part of Zone 09 and that

G250726 is anomalously thin with the lower portion on Zone 09 only developed to half of the “normal” Zone 09 illustrated in G250020.

There are two anomalous boreholes for Zone 09 on Groenfontein and they have been clearly identified by the anomalous GDCDD curves they present (displayed in the

Addendum on CD). The other 29 boreholes have closely grouped GDCDD curves which illustrates that the lithologies of Zone 09 is very regular over the whole farm.

Percussion boreholes confirm the geology of Zone 09 in collaboration with cored boreholes.

135

10 CONCLUSION

The first geophysical tools were placed in a borehole in 1927 and the use thereof in the next

50 years was based, primarily, in the oil industry. In the 1960’s major technological advances included transistors and integrated circuits while the 1970’s brought computers to process the data quickly and efficiently.

The use of down-hole logging is standard practice by mining companies in the area and the use of geophysics is absolutely essential in some large coal basin deposits. Without using geophysics as correlation aid, it would often be very difficult for even the most competent of geologists to define the exact roof and floor of the various coal Samples and/or Zones. With slimline geophysical logging, the various coal Samples and Zones can be correlated to a high degree and discrepancies easily identified. In the remote chance that the drilling company tries to hide core losses in the coal, the slimline logging would show exactly where the floor and roof contacts are and the discrepancy can then be taken up with the drilling company.

The relative densities of the various coal samples were determined by means of the

Archimedes method. The relative density of the same samples were predicted by means of the geophysically determined relative density and it was found that an accuracy of more than

95% was achieved when predicting the relative density by means of the Vectar processed derived densities..

136

The graphical representation of the relative density of the rocks, by means of a line graphs showing many different traces, does not adequately indicate lithology variation. A method was then devised by the author that enables the scientist to portray the graphical data of the geophysically derived density log, in a derived cumulative distribution diagram line diagram

(GDCDD) curves.

• It was possible to use the method to check on the accuracy of the database and errors made during the capture of geophysical data.

• It was also possible to use the method to check on the accuracy of the database and errors made during the capture of geological data.

• It was found that the method enables the scientist to produce similar curves for specific Samples or Zones chosen in the stratigraphy.

• The line curves (which equate to an easily understood depiction) of any Sample or

Zone can be correlated accurately over the area of the investigation (200 km

2

).

• Any geological dissimilarity, in any chosen Sample or Zone, produces anomalies in the shape or position of the GDCDD curve of that anomalous Sample or Zone.

• The lateral correlation between lithologies can be described accurately and this would be beneficial to a Competent Person and demonstrates that this method is valuable in classifying coal resources in large coal occurrences.

• It would enable the scientist to accurately portray any chosen coal intersection’s relative density distribution when measured by geophysical means and compare that to those measured in adjacent boreholes.

The GDCDD method was further refined to calculate the standard deviation at the two chosen relative density points of 1.6 and 2.0 and the derived values can be used as basis for evaluating how the various Samples and Zones vary laterally.

137

Geophysical logging is thus an essential aid when evaluating where the various Samples or

Zones occur in the stratigraphy. Although pictures say a thousand words, some mathematical method had to be found to persuade a Competent Person that the curves tell a convincing scientific story as the following points illustrate.

• The use of geophysics, and in particular the slimline down-hole logging used in the

Study area, has proved that the relative densities derived by the Vectar processed method, can be used to control and improve the field measurement of the relative densities. Accuracies of more than 95% are the norm.

• By evaluating the density readings of a specific sample or zone boundary in the stratigraphy by means of a cumulative distribution line diagram, it can be shown that any local variation in the seam structure is immediately noticeable when combining the curves of the area/farm as it modifies either the origin of the curve on the Y-axis, its inclination and/or shape.

• Weathering and/or faulting changes the local composition and the associated relative density of the lithologies in that specific sample or zone and this has a clear impact on the GDCCD curves. Zone 11, the top coal zone in the stratigraphy, is sometimes influenced by weathering and even the loss of 30 cm of this zone, significantly modifies the shape and position of the GDCDD curve of that zone. This proves the high sensitivity of the GDCDD method.

By combining cored and percussion drilling for a specific area or farm it was illustrated that there is a clear, close relationship between the shapes of the GDCDD curves for that area/farm independent of the drilling method used.

138

With regards to the spacing of boreholes, the SAMREC Code (and thus the SANS 10320 standard) prescribes a specific drilling density based on the thickness of the coal seams.

Barring special conditions such as structural complexity, the spacing of boreholes is prescribed as 350 m between boreholes to conform to Measured Mineral Resource status for thick interbedded coal seam deposits such as those in the Ellisras Basin.

• The SANS committee made the classification of measured Mineral resource the same for both the multiple coal seam deposits (Witbank coal basin types in South Africa) and the thick interbedded coal seam type deposits (Ellisras basin types in South

Africa).

• For Indicated and Inferred Resource classification, in thick interbedded deposits, wider borehole spacing is allowed which could be in the order of three times that allowed for the multiple coal seam deposit resources.

• The reason for this decision was based on conservatism due to a lack of quantifiable data at that time. In the meanwhile, new data emerged as other role players became involved in exploration activities in the Ellisras basin and this new data should be recognized.

• Structurally complex areas can be evaluated in detail by means of infill percussion drilling and the lithology accurately compared to adjacent cored boreholes.

• Evaluating the additional percussion boreholes in combination with original boreholes by using the GDCDD method leads to quantifiable information which is as accurate in interpreting the lithology as would have been the case with additional core drilling.

The advantage of using percussion drilling is that it is considerably cheaper than core drilling.

• In the latest update of the SANS 10320 standard, in its final draft stage, the 350 m borehole spacing requirement has been altered to a recommended 500 m.

139

The issue of drilling density is a critical issue all over the world where the classification of resources needs to be done. By applying the GDCDD method, in coal deposits, to both percussion and cored boreholes in a specific project, a Competent Person is able to reliably decide to which category the Resources can be assigned.

140

11 WORKS CITED

Arnot, D. and Williams, C. (2007). CIC Energy Corp.: Mmamabula Project. south-eastern

Botswana. National Instrument 43-1-1 technical Report, Project No. J912. Snowden

Consultants. 287 pp.

Binzhong, Z. and Esterle, J. (2008). Toward improved coal density estimation from geophysical logs. Exploration Geophysics, 2008, 39, CSIRO Publishing. p124–132.

Benning, I. (2000). Bankers’ perspective of mining project finance. The Journal of The South

African Institute of Mining and Metallurgy. May/June 2000.p145-152.

Bordy, E. M., Knoll, F. and Bumby, A. (2010). New data on the palaeontology and sedimentology of the Lower Jurassic Lisbon Formation (Karoo Supergroup), Ellisras Basin,

South Africa. – N. Jb. Geol. Paläont. Abh., 258: 145–155; Stuttgart.

Buick, I.S., Lana. C and Gregory. C. (2011). A LA-ICP-MS and SHRIMP U/PB age constraint on the timing of REE Mineralisation associated with Bushveld granites. South

African Journal of Geology (March 2011), 114(1):1-14.

Dreyer, J.C. (1994). Total Utilization of the Coal resources: The Grootegeluk Experience.

XVth CMMI Congress. Johannesburg, SAIMM. P 153-164.

Firth, D. (1999). Log Analysis for Mining Applications. Reeves Information Manual. Reeves

Wireline Services, Editor Peter Elkington, Contributor Reeves Wireline Services, Publisher

Reeves Wireline Services, 156pp

141

François-Bongarçon, D.M. (1998). Due-Diligence Studies And Modern Trends In Mining.

Vice President Geostatistics and Sampling. With contributions from S.D. Long, Principal

Geochemist and Dr H.M. Parker. Mineral Resources Development, Inc. San Mateo,

California, USA. 15 pp.

Frick, C. (2002). Direct Foreign Investment and the Environment: African Mining Sector.

OECD Global Forum on International Investment Conference on Foreign Direct Investment and the Environment. Lessons to be learned from the Mining Sector 7 - 8 February 2002.

29pp.

JORC CODE. (2004). The 2004 Australasian Code For Reporting Exploration Results,

Mineral Resources And Ore Reserves (The JORC Code). 32pp.

Gluskoter, H. (2000). Coal resource assessment at the US Geological Survey and the United

Nations Framework Classification for reserves/resources. Committee on Sustainable Energy

Tenth Session, 31 October-2 November 2000. 18 pp.

Lundell, LLP. (2006). National Instrument 43-101: Amendments to Canadian Rules

Concerning Mineral Project Disclosure and Technical Reports. 2pp

Mtimkulu, MN. (2009). A provisional basinal study of the Waterberg-Karoo, South Africa,

MSc dissertation, University of Pretoria, Pretoria, viewed 2012/01/04. http://upetd.up.ac.za/thesis/available/etd-08172010-191251/. 134 pp.

Mullins, M. (2008). The SAMREC 2007 and SAMVAL 2008 Codes. The South African

Mineral Reporting Codes. Breakfast launch of SAMREC 2007 and SAMVAL 2008. Hilton

Hotel, 10 April 2008. 4pp.

142

Pincock Perspectives. (2009). Mineral resource Classification Schemes: Some Thoughts.

Issue 99, November 2009. 4pp.

Ryder, M. (2002). The Geological Interpretation of Well Logs. Published by Rider-French

Consulting Ltd., P.O. Box 1, Sutherland. lV283XL Scotland. 2nd Edition. Revised, 2002.

290pp.

Sambo, M.V. (2008). Characterizing the Hanging Wall to Zone 1, Vryheid Formation,

Grootegeluk Coal Mine-Investigating its potential as aggregate stone. Unpublished Bachelor of Science Honours Degree in Geology Report, School of Geosciences, University of the

Witwatersrand. Johannesburg, 125 pp.

SAMREC CODE (2000 edition). (2000). The South African Code For The Reporting Of

Exploration Results, Mineral Resources And Mineral Reserves. Prepared By the South

African Mineral resource Committee (SAMREC) Working Group under the Joint Auspices of the Southern African Institute of Mining and Metallurgy. 39pp.

SAMREC CODE (2007 edition). (2007). The South African Code For The Reporting Of

Exploration Results, Mineral Resources And Mineral Reserves. Prepared by the South

African Mineral resource Committee (SAMREC) Working Group under the Joint Auspices of the Southern African Institute of Mining and Metallurgy and the Geological Society of

South Africa. 29 pp.

SAMREC CODE (2007 edition amended in 2009). (2009). The South African Code For The

Reporting Of Exploration Results, Mineral Resources And Mineral Reserves. Prepared by the

South African Mineral resource Committee (SAMREC) Working Group under the Joint

143

Auspices of the Southern African Institute of Mining and Metallurgy and the Geological

Society of South Africa. 61 pp.

SAMVAL CODE. (2008). The South African Code For The Reporting Of Mineral Asset

Valuation as amended July 2009. Prepared by The South African Mineral Asset Valuation

(SAMVAL) Working Group. www.samcode.co.za. 17pp.

Samworth, J.R. (2010). The Use Of The Vectar Density And Compensated Density Instead

Of The Long And Short Spaced Density Log In Slimline. 9pp.

SANS 10320:2004, Edition 1. (2004). South African National Standard. South African Guide

To The Systematic Evaluation Of Coal Resources And Coal Reserves. 142pp.

Snowden, D.V. (2001). Practical Interpretation of Mineral Resource and Ore Reserve

Classification Guidelines, in Mineral Resource and Ore Reserve Estimation – The AusIMM

Guide to Good Practice (Ed: A C Edwards), The Australasian Institute of Mining and

Metallurgy: Melbourne. p643–652

US Patent. (2007). Method for Correlating Well Logs, U.S. Patent Documents. Patent No.:

US 7,295,926 B2. 14pp.

United Nations. (2010). United Nations Framework Classification for Fossil Energy and

Mineral Reserves and Resources 2009. (UNFC-2009). The ECE Energy Series No. 39.United

Nations Publication. 20 pp.

144

Vaughan, W.S. and Felderhof, S. (2002). International Mineral Resource and Mineral

Reserve Classification and Reporting Systems. Paper Prepared for and Presented at: The 48th

Annual Rocky Mountain Mineral Law Institute, Lake Tahoe, Nevada, July 24-26. 31 pp.

Wood, G. H., Jr., Kehn, T. M., Carter, Devereux, M. and Culbertson, W. C. (1983), Coal

Resource Classification System of the U.S. Geological Survey. GEOLOGICAL SURVEY

CIRCULAR 891. 65pp.

145

ADDENDUM 1

Samples 01B and 02A

Figure 1 Combined GDCDD of Sample 01B (Spread at a Rd of 2.0 indicated by light blue dotted line)

Figure 2 Combined GDCDD of Sample 02A (Spread at a Rd of 2.0, indicated by light blue dotted line)

SAMPLE 02A

SAMPLE 04A

SAMPLE 05A

SAMPLE 03A

SAMPLE 06A

GEELBEKPAN GROENFONTEIN GANNAVLAKTE

Figure 3 Graphical representation of Sample 02A to Sample 06A making up coal Zone 10 on the Study area

Table 1 Summary of Sample 01B and 02A data.

BASIC INFORMATION

ABOUT SAMPLE 01B

AND 02A.

SUMMARY

Angle of repose of graphs

Sample 01B Medium Medium

Sample 02A Medium Medium

The angle of repose is indicative of the high yield of these specific samples.

Medium to high

Medium

Medium to high

Medium

Start point of graphs at Rd

Data values do not have the same point of origin for sample 01B (Figure 78) but for sample 02 they have the same

start point (Figure 79). Gannavlakte and Ringbult start at a higher point of origin (more values at the lower Rd

1.35

range.

Spread

Groenfontein has a slightly wider spread of 25% for Sample 01B, but a narrow spread for Sample 02A. Geelbekpan has a narrow spread for both samples and Ringbult/Gannavlakte the widest spread (up to 40%) illustrating the varying conditions towards the south and the regularity in the north.

Yield at Rd 2.0 – Sample

1B

Yield at Rd 2.0 – Sample

2A

11.6% variance

8.3% variance

GEELBEKPAN

68.9

39.6

GROENFONTEIN

69.2

42.2

RINGBULT

80.7

40.8

GANNAVLAKTE

75.0

33.9

Remarks – yield at 2.0

For Sample 01B, the data shows that the yields for Ringbult and Gannavlakte are higher than the other sample areas.

For Groenfontein, the higher values in the GDCDD graph corresponds to a higher yield value in the yield data.

Raw Rd Sample 01B

Raw Rd Sample 02A

Remarks on raw Rd

Thickness Sample 01B

Thickness Sample 02A

There is a 12% variation over the whole area. In the case of Sample 02A, there is an 8,3% variation over the whole area. The GDCDD curves for Sample 02A along the Groenfontein profile shows very narrow correlation between

the known yield boreholes and the unknown boreholes (Figure 83 and Figure 84).

0.09 variance

0.09 variance

0.77 meter variance

1.64

1.92

1.44

1.64

1.91

The south has the lowest raw Rd for Sample 01B, and Groenfontein has the lowest raw Rd for Sample 02A and

1.38

1.55

1.92

2.15

1.59

2.00

1.92

Remarks on thickness

0.32 meter variance 2.80 2.91 2.87 2.59

The thickness of Sample 01B varies in thickness (Figure 58), but Sample 02A has less variation of thickness

(Figure 80, Figure 81, Figure 82 and Figure 86) and does not occur everywhere, due to geological variations,

such as faulting.

Standard deviation at a Rd

Relative density 1.6

Sample 01B 4.0

2.0

4.3

1.6

9.1

2.0

8.9

1.6

3.7

2.0

3.7

1.6

8.2

2.0

5.8

of 1.60 and 2.0

13.

Sample 02A

6.2 5.9 7.3 6.4 5.9 2 11.4 7.1

Comments on standard

It can be seen that the σ for Groenfontein is higher than for the other areas in the Study area for Sample 01B.

deviation

Ringbult and Gannavlakte have higher values for Sample 02A.

Correlation accuracy by

Good, easy to correlate. Problems with correlation, like faulting and correlation errors, will be prominent.

geologist

Modelling accuracy

Good correlation, easy to model. No factoring necessary to compensate for modelling accuracy at Geelbekpan.

Groenfontein has to be investigated in more detail due to some erosion of this sample in the southern part of the

sampled area. (Figure 31). The south shows that care should be taken when doing correlation and factoring will probably be necessary.

The use of the GCCD proves that the method can be used to prove the continuity of Sample 01B with a good

measure of accuracy. Figure 85 illustrates that the coal seam next to the green 11 label in the figure has a varying

thickness and this results in the variance illustrated in the GDCDD (Figure 78). Ringbult has some variance, but

this can be geologically declared (Figure 67). The spacing in the GDCDD graphs for Sample 01B is very narrow for Geelbekpan, wider for Groenfontein and wide for the graphs in the south (Ringbult and Gannavlakte). Sample

01B is easy to correlate over the Study area, but it is somewhat variable in the south. There is an 11.6% variance in yield over the Study area and a thickness variation of 0,77 m. For modelling purposes, there is good lateral

Use as SANS 10320 data

correlation.

correlation point

No factoring necessary to compensate for modelling accuracy at Geelbekpan. Groenfontein has to be investigated in more detail due to some erosion of this sample in the southern part of this farm. The south shows that care must be taken when doing correlation and factoring will probably be necessary. There is a direct correlation between

GDCDD’s and yield curves. In terms of modelling, there is good lateral correlation, which should make modelling straightforward.. The south shows that care should be taken when doing correlation and factoring will probably be necessary.

The use of the GCCD proves that the method can be used to prove the continuity of these samples with a good measure of accuracy, as a SANS data point. Ringbult has some variance but these can be geologically declared.

Groenfontein

Welgelegen

Gannavlakte

Figure 4 Thickness of Sample 01B over the Study area

Geelbekpan meters

Ringbult

Groenfontein

Welgelegen

Gannavlakte

Figure 5 Thickness of Sample 02A over the Study area

Geelbekpan meters

Ringbult

Figure 6 Groenfontein Sample 01B, GDCDD at the top and yield curves along a selected profile. (Illustrated in Figure 55)

Figure 7 Groenfontein Sample 02A GDCDD at the top and yield curves along a selected profile. (Illustrated in Figure 55)

Figure 8 Occurrence of Sample 01B on Groenfontein, illustrating variances in Sample 01B. Green labels denote bottom of coal zones

Figure 9 Occurrence of Sample 02A on Ringbult, illustrating variances in Sample 02A. The yellow labels denote bottom of coal zones (From geophysical data)

Samples 03A, 04A and 05A

Figure 87 to Figure 89

Figure 10 Combined GDCDD of Sample 03A (Spread at a Rd of 2.0, indicated by light blue dotted line)

Figure 11 Combined GDCDD of Sample 04A (Spread at a Rd of 2.0, indicated by light blue dotted line)

Figure 12 Combined GDCDD of Sample 05A (Spread at a Rd of 2.0, indicated by light blue dotted line)

SAMPLE 02A

SAMPLE 04A

SAMPLE 05A

SAMPLE 03A

SAMPLE 06A

GEELBEKPAN GROENFONTEIN

Figure 13 Graphical representation of Sample 02A to Sample 06A on the Study area

GANNAVLAKTE

Table 2 Summary of Sample 03A, 04A and 05A data.

BASIC INFORMATION

ABOUT SAMPLE 03A, 04A

AND 05A

SUMMARY GEELBEKPAN GROENFONTEIN RINGBULT GANNAVLAKTE

Sample 03A

Sample 04A

High

High

High

High

High and medium

High and medium

High and medium

High and medium

Angle of repose of graphs

Sample 05A High High High High

High angles of repose for the various areas and associated higher yields. The amount of interbedded

inorganic material in the south (in certain areas) influences the angle of repose of some graphs (Figure 87

to Figure 89).

Start point of graphs at Rd 1.35

Data values do not start at the same point of origin, illustrating the variable amount of low density material in the various areas.

Spread

Yield at Rd 2.0 -Sample 03A

Yield at Rd 2.0 - Sample 04A

Yield at Rd 2.0 - Sample 05A

For Sample 03A (Figure 87), the spread is around 30% for all the areas; for the south, the yields vary

considerably. For Sample 04A (Figure 88), the spread for the south is very wide (55%) due to mudstones

on top and at the bottom of this sample (Figure 90). Sample 05A is narrowly spread for Groenfontein and

Geelbekpan (± 10%), but only slightly wider in the south due to one borehole (Figure 89).

14.9% variance

18.8% variance

5.1% variance

57.4

73.2

86.3

55.7

68.9

82.8

53.1

64.8

83.7

68.0

54.4

81.2

Remarks- Yield

Raw Rd Sample 03A

Raw Rd Sample 04A

Raw Rd Sample 05A

Remarks on raw Rd

Thickness Sample 03A

Thickness Sample 04A

Thickness Sample 05A

Remarks on thickness

The lowest yield is found in sample 03A. Sample 04A has the highest variance in yield and Sample 05A the lowest. The highest yield is found in Sample 05A.

0.45 variance

0.12 variance

0.08 variance

1.72

1.64

1.55

1.95

1.66

1.56

1.82

1.67

1.59

1.50

1.76

1.63

Highest variance is found in Sample 03A, due to the lateral facies changes and the lowest variance in

Sample 05A, with the least lateral facies variance. The lower yield are reflected in the higher raw Rd value for Sample 03A.

0.18 meter variance

0.36 meter variance

1.79

2.07

1.78

1.69

1.75

2.02

1.61

1.71

0.93 meter variance 4.29 4.73 4.03 3.80

Sample 04A varies between 1 m and 3 m (Figure 93). Good correlation of thickness for all three samples.

The south is thinner than the north for Sample 03A (Figure 92) and Sample 05A (Figure 94). Sample

05A is quite thick in the Groenfontein area.

Standard deviation at a Rd of

1.60 and 2.0

Relative density 1.6

Sample 03A 6.6

Sample 04A 4.4

Sample 05A 10.2

2.0

11.4

3.6

4.3

1.6

5.9

9.3

8.2

2.0

7.6 11.7 8.0 10.4

4.5

1.2

1.6

1.5

8.9

Comments on standard

The standard deviation varies between 0.0 and 21.6, averaging around 8.0%.

deviation

2.0

0.0

6.5

1.6

21.6

20.1

2.0

7.1

17.2

4.2

Good, easy to correlate, except for the south. Problems with correlation, such as faulting and correlation

geologist

Modelling accuracy

(Figure 90 and Figure 91

). The σ for Gannavlakte (Sample 04A) is high and care should thus be taken

when correlating.

Good correlation, easy to model, except for the south. Here some factoring would be necessary to compensate for modelling accuracy.

The use of the GCCD proves that the method can be used to prove the continuity of Sample 03A and

Sample 04A in the north with a good measure of accuracy; except for the south where care has to be taken when correlating the geology. The reason is that Sample 03A and Sample 04A have to a large extent, been weathered away (Figure 32). There is a narrow correlation between the various graphs for Groenfontein and Geelbekpan, except for the south which is slightly wider spaced. The start-point of the graphs for the south in particular, shows a variable amount of low density material in the samples, with Sample 04A

Use as SANS 10320 data

being especially variable. The raw Rd for Sample 03A is also quite variable (0.45), but the thickness is remarkable even (0.28 m over the Study area). Sample 05A varies by 0.93 m, but this Sample is quite

correlation point

thick (+3.80 m). The standard deviation for Sample 04A is the highest (17.2) because of the inclusion of high density material in this sample in the south. Of the three samples, Sample 05A has the lowest variance in yield. The lower yields for the south resulted because the south has a higher raw Rd, in particular, Sample 04 for Gannavlakte. When comparing the yield with the GDCDD curves with the analytical yield along a specific profile they show that high values for the GDCDD curves corresponds

with a higher value in the analytical yield (for sample 03A-Figure 95, for sample 04A-Figure 96 and for

sample 05A-Figure 97).There is good correlation of thickness for all three samples. The south is thinner

than the north for Sample 03A and Sample 05A. The thickness of Sample 03A is remarkably even over the whole area, having only a 0.18 cm variance.

The lateral lithofacies correlation over the whole area of these three samples is constant, and it is relatively easy to correlate, except for the south. Problems with correlation, such as faulting and correlation errors, will be prominent. The variation in yields in the south can be ascribed to the geological variations. The comparison between the yield and the GDCDD shows a clear relationship between the derived curves and

the yield where the curves with the higher value correlate well to the higher yield curves (Figure 95 to

Figure 97

).The σ for Gannavlakte Sample 04A is high and care should be taken when correlating. In

terms of modelling accuracy, the use of the GCCD demonstrates that the method can be used to prove the continuity of Sample 03A and Sample 04A in the north with a good measure of accuracy; except for the south where care has to be taken when correlating the geology. The reason being that Sample 03A and

Sample 04A have been weathered to a large extent.

R30302

G299007

Figure 14 The lateral lithological variation in Sample 05A in the south. The geophysical graphs look dissimilar and mudstone lenses are not laterally consistent

Groenfontein

Welgelegen

Gannavlakte

Figure 15 Thickness of Sample 03A over the Study area

Geelbekpan meters

Ringbult

Groenfontein

Welgelegen

Gannavlakte

Figure 16 Thickness of Sample 04A over the Study area

Geelbekpan meters

Ringbult

Welgelegen

Groenfontein

Gannavlakte

Figure 17 Thickness of Sample 05A over the Study area

Geelbekpan meters

Ringbult

Table 3 Standard deviation calculated at two RDs from the GCCD data for Sample 03A to

Sample 05A.

GEELBEKPAN GROENFONTEIN

FARM

GANNAVLAKTE RINGBULT TOTAL

Sample 1.6 2.0 Avg 1.6 2.0 Avg 1.6 2.0 Avg 1.6 2.0 Avg 1.6 2.0 Avg

03A

04A

6.6

4.4

11.4

3.6

9.0

4.0

5.9

9.3

7.6

4.5

6.8

6.9

10.4

21.6

7.1

17.2

8.8

19.4

11.7

1.5

8.0

0.0

9.8

0.8

8.6

9.2

8.6

6.3

8.6

7.8

05A

10.2 4.3 7.3 8.2 1.2 4.7 20.1 4.2 12.2 8.9 6.5 7.7 11.9 4.1 8.0

Figure 18 Groenfontein Sample 03A GDCDD at the top and yield curves along a selected profile (illustrated in Figure 55)

Figure 19 Groenfontein Sample 04A GDCDD at the top and yield curves along a selected profile (illustrated in Figure 55)

Figure 20 Groenfontein Sample 05A GDCDD at the top and yield curves along a selected profile (illustrated in Figure 55)

Sample 06A and Sample 07A

Figure 21 Combined GDCDD of Sample 06A (Spread at a Rd of 2.0, indicated by light blue dotted line)

28

Figure 22 Combined GDCDD of Sample 07A (Spread at a Rd of 2.0, indicated by light blue dotted line)

29

SAMPLE 08A

SAMPLE 09A

SAMPLE 07A

SAMPLE 08A

SAMPLE 09A

GEELBEKPAN GROENFONTEIN GANNAVLAKTE

Figure 23 Graphical representation of Sample 07A to Sample 09A. Sample 08A not present in this specific Groenfontein borehole

30

Table 4 Summary of Sample 06A and Sample 07A data.

BASIC

INFORMATION

ABOUT SAMPLE 06A

AND SAMPLE 07A.

SUMMARY GEELBEKPAN GROENFONTEIN RINGBULT GANNAVLAKTE

Angle of repose of graphs

Sample 06A

Sample 07A

High

Medium

High

Medium

High

Medium

High

Medium and high

Start point of graphs at

Data values do not start at the same point. Gannavlakte and Ringbult start higher (more values at the lower Rd

Rd 1.35

range.

For Sample 06A, the south has the narrowest spread of 10% (Figure 98). For Sample 07A, the south has a wide

Spread

The angle of repose in all the areas varies. The Ringbult/Gannavlakte graph has a high angle of repose and a kink at

a Rd of 1.45 (Figure 99). spread of 55% (Figure 99). These two samples, in general, have a wider spread than the previous samples,

illustrating varying lithology over the Study area.

Yield at Rd 2.0-Sample

06A

Yield at Rd 2.0 - Sample

07A

Remarks - Yield

25.9% variance

13.5% variance

58.4

23.0

57.9

30.4

83.8

20.6

86.7

16.9

The graph (Figure 98) shows that the GDCDD curve data for the south reaches about 80% at a Rd of 1.60, and this

31

Raw Rd-Sample 06A

Raw Rd-Sample 07A

Remarks on raw Rd

is 30 percent higher than the north. This is supported by the proximate data (yields at a Rd of 2.0) for these areas in the table. The low angle of repose for Sample 07A in the Ringbult/Gannavlakte GDCDD is reflected in the low yield of 16.9%.

0.22 variance

0.28 variance

1.72

2.05

1.68

1.91

1.51

2.08

1.50

2.19

The raw Rd values are echoed in the yield values. The south has a higher Rd value for Sample 07A and corresponding lower yield at a Rd of 2.0.

Thickness-Sample 06A

Thickness-Sample 07A

Remarks on thickness

0.70 meter variance

1.52 meter variance

1.20

1.42

1.90

1.42

1.28

1.86

1.89

2.94

The prominent lateral variation in Sample 06A thickness and geology over the whole area is shown in Figure 101.

The south is nearly double the thickness of the north for Sample 07A and there is quite a big lateral variation (Figure

102). Figure 103 indicates the thickest Sample 06A on Groenfontein (Figure 104) indicates that Sample 07A is best

developed on Gannavlakte.

Relative density 1.6 2.0 1.6 2.0 1.6 2.0 1.6 2.0

Standard deviation at a

Sample 06A 20.3 11.6 21.2 12.7 11.7 3.5 8.3 0.3

Rd of 1.60 and 2.0

Sample 07A 10.5 16.7 9.2 15.0 8.2 14.5 3.4 13.6

Comments on standard deviation

T

he σ value for samples 06 is high in the north and central areas. The table also shows that at a Rd of 1.60, the σ value is smaller than for the

σ value at a Rd of 2.0. There is a gradual decrease in σ towards Ringbult for both samples. The average σ vale for Sample 06A = 11.2, for Sample 07A = 11.4 and for Sample 08A = 10.5.

Correlation accuracy by Sample 6A would be more difficult to correlate in the north as it is just over one metre thick. Local variation and any

32

geologist

Modelling accuracy

degree of faulting would make it difficult to identify. Correlation would then be based on the samples surrounding

Sample 6A and not so much Sample 6A in itself.

Sample 07A would be easy to identify in the north, but care must be taken as the sandstone in Sample 07A varies in thickness in the south. Identification is relatively easy on Ringbult due to the increased thickness of the sandstone.

Sample 06A would be more difficult to correlate and care should be taken when doing correlation. Factoring will probably be necessary to compensate for modelling accuracy at Geelbekpan and Groenfontein.

The use of the GCCD proves that the method can be used to prove the continuity of Sample 06A and Sample 07A with some measure of accuracy. The three samples are variable over the thesis area and in places either faulted away or poorly developed. For Sample 06A, the variance in thickness is about 60% of the total thickness, while the thickness variation for Sample 07A is greater than the thickness of this sample in the north. Sample 06A has higher yields in the south and Sample 7A has higher yields in the north. There are prominent lateral variation in Sample

06A thickness and geology over the whole area. The south is nearly double the thickness of the north for Sample

Use as SANS 10320 data

07A and there is quite a big lateral variation. The thickest occurrence of Sample 06A occurs on Groenfontein and

correlation point

the thickest Sample 07A is found on Gannavlakte.

Sample 6A would be more difficult to correlate in the north as it is just over one metre thick. Local variation and any degree of faulting would make it difficult to identify. Correlation would then be based on the samples surrounding

Sample 6A and not so much Sample 6A itself. Sample 07A would be easy to identify in the north, but care must be taken as the sandstone in Sample 07A varies in thickness in the south. Identification is relatively easy on the farm

Ringbult due to the increased thickness of the sandstone. When modelling, Sample 06A would be more difficult to

33

correlate and care should be taken when doing correlation.

Factoring will probably be necessary to compensate for modelling accuracy at Geelbekpan and Groenfontein.

The use of the GCCD demonstrates that the method can be used to prove the continuity of Sample 06A and 07A with a good measure of accuracy.

34

Table 5 Standard deviation calculated at two Rd’s from the GCCD data for samples 06A and Sample 07A on the various farms.

Sample

06A

07A

GEELBEKPAN

1.6

20.3

10.5

2.0

11.6

16.7

Avg

16.0

13.6

VRYHEID FORMATION

FARMS

GROENFONTEIN

1.6 2.0

21.2 12.7

9.2 15.0

Avg

17.0

12.1

GANNAVLAKTE

1.6

8.3

3.4

2.0

0.3

13.6

Avg

4.3

8.5

1.6

11.7

8.2

RINGBULT

2.0

3.5

Avg

7.6

14.5 11.4

1.6

15.4

7.8

TOTAL

2.0

7.0

Avg

11.2

14.9 11.4

35

GEELBEKPAN

G226014

GROENFONTEIN

G250006

RINGBULT

R303020

GANNAVLAKTE

G299007

Figure 24 Lateral variation of Sample 06A over the Study area. Blue line denotes possible correlation points, but this is uncertain

36

RINGBULT

R303020

RINGBULT

R303010

Figure 25 Lateral variation of Sample 07A in the south

37

GANNAVLAKTE

G299011

GANNAVLAKTE

G299007

Welgelegen

Groenfontein

Gannavlakte

Geelbekpan

Figure 26 The lateral variation of Sample 06A over the Study area

38 meters

Ringbult

Welgelegen

Groenfontein

Gannavlakte

Geelbekpan

Ringbult

Figure 27 The lateral thickness variation of Sample 07A over the Study area meters

39

Samples 08A, to 14A

Figure 28 Combined GDCDD of Sample 08A (Spread at a Rd of 2.0, indicated by light blue dotted line)

40

Figure 29 Combined GDCDD of Sample 09A (Spread at a Rd of 2.0, indicated by light blue dotted line)

41

GEELBEKPAN has no Sample 10A coals

Figure 30 Combined GDCDD of Sample 10A (Spread at a Rd of 2.0, indicated by light blue dotted line)

42

Figure 31 Combined GDCDD of Sample 11A (Spread at a Rd of 2.0, indicated by light blue dotted line)

43

Figure 32 Combined GDCDD of Sample 12A (Spread at a Rd of 2.0, indicated by light blue dotted line)

44

Figure 33 Combined GDCDD of Sample 13A (Spread at a Rd of 2.0, indicated by light blue dotted line)

45

Figure 34 Combined GDCDD of Sample 14A (Spread at a Rd of 2.0, indicated by light blue dotted line)

46

SAMPLE 10A

SAMPLE 13A

SAMPLE 14A

SAMPLE 14B

SAMPLE 11A

GEELBEKPAN GROENFONTEIN

Figure 35 Graphical representation of Sample 10A to Sample 14B

47

SAMPLE 12A

GANNAVLAKTE

Table 6 Summary of Sample 08A, 09A, 10A, 11A, 12A, 13A and 14A data.

BASIC INFORMATION

ABOUT SAMPLE 08A, 09A,

10A, 11A, 12A, 13A AND 14A

DATA

SUMMARY GEELBEKPAN GROENFONTEIN RINGBULT GANNAVLAKTE

Sample 08A

Sample 09A

Low

Medium to high

Low

High

Low to high

Medium to high

Low to high

Medium to high

Angle of repose of graphs

Sample 10A

Sample 11A

N/A

Low

Sample 12A

Sample 13A

High

High

Sample 14A Med to high

Low

Low

High

Med to high

High

Low to high

Low to high

Low to high

Low to high

Med to high

Low to high

Low to high

Low to high

Low to high

Med to high

The graphs in this selection all exhibit a lower angle of repose with a variable rise (Figure 105, Figure 106,

Figure 107, Figure 108, Figure 109, Figure 110 and Figure 111). Ringbult/Gannavlakte illustrates the

variability of the geology in the south. This low angle is reflected in the yields which is quite low, except

for two samples, 09A (Figure 106) and 14A, (Figure 111) which have yields higher than 50%. The yield

for Sample 10A (Figure 107) is the lowest of this group and this Sample is not developed on Geelbekpan

(Figure 112).

48

Start point of graphs at Rd

Data values for all the graphs (Figure 105 to Figure 111) start at the same point, except for

1.35

Ringbult/Gannavlakte in the south, Sample 11A (Figure 108) and Sample 12A (Figure 109).

Spread

For Sample 08A, the graphs are narrowly spread, except for Ringbult/Gannavlakte which has a 40% spread.

For Sample 09A, Geelbekpan and the south, has about a 40% spread. Sample 10A does not occur on

Geelbekpan, but the south has a spread of about 70%. Sample 11A has an 80% spread in the south, 30% on

Groenfontein and 20% on Geelbekpan. Sample 12A has a 30% spread for Geelbekpan and Groenfontein and 80% for the south. Sample 13A has a 50% spread, and the south a 70% spread. For Sample 14A, all the graphs show a wide spread between 45% and 75%. All these graphs are indicative of the variability of the

geology in the south (Figure 112).

Yield at Rd 2.0 -Sample 08A

Yield at Rd 2.0 - Sample 09A

Yield at Rd 2.0 - Sample 10A

Yield at Rd 2.0 -Sample 11A

Yield at Rd 2.0 - Sample 12A

Yield at Rd 2.0 - Sample 13A

Yield at Rd 2.0 -Sample 14A

Remarks- Yield

Raw Rd Sample 08A

4.1% variance

16.2% variance

8.4% variance

7.1% variance

9.7% variance

12.8

66.4

Na

13.4

37.1

14.8

63.6

9.4

20.4

39.5

15.8

53.9

6.6

19.3

37.3

16.9

50.2

15.0

15.8

29.8

8.0% variance 35.4 39.0 31.0 37.0

37.2% variance 74.3 37.1 63.8 58.5

The highest yields are found in Sample 09A and Sample 14A, while the lowest yield is found in Sample

10A. The largest variance in yield is found in Sample 14A and the smallest variation in Sample 08A. The variance in Sample 14A is due to geological variances in this sample.

0.11 variance 2.24 2.13 2.21 2.22

49

Raw Rd Sample 09A

Raw Rd Sample 10A

0.15 variance

0.10 variance

0.10 variance

1.69

Na

2.17

1.69

2.31

2.16

1.76

2.26

2.12

1.84

2.21

2.22

Raw Rd Sample 11A

Raw Rd Sample 12A

Raw Rd Sample 13A

Raw Rd Sample 14A

0.11 variance

0.11 variance

1.92

1.90

1.90

1.88

1.96

1.98

2.01

1.99

Remarks on raw Rd

0.27 variance 1.66 1.93 1.72 1.77

The highest Rd is found on the farm Groenfontein, Sample 10A, and the lowest Rd on Geelbekpan, Sample

14A.

1.69 meter variance 1.84 2.03 3.53 3.53

Thickness Sample 08A

Thickness Sample 09A

Thickness Sample 10A

Thickness Sample 11A

Thickness Sample 12A

Thickness Sample 13A

0.78 meter variance

1.39 meter variance

0.94 meter variance

1.63 meter variance

0.81 meter variance

0.55 meter variance

2.12

Na

2.40

1.51

1.78

1.38

2.82

1.95

2.18

1.62

1.75

1.42

2.04

1.98

2.77

3.14

2.50

1.22

2.47

3.34

3.12

2.22

2.56

1.77

Thickness Sample 14A

Remarks on thickness

Standard deviation at a Rd of

The lateral variation in thickness of Sample 14A is shown in Figure 112. The reason why Sample 19A does

not occur in some areas on Geelbekpan, is due to a small coal seam that is not developed below Sample

09A; this is illustrated in Figure 100. The 1.63 metre variance in thickness for Sample 12A is more than the

thickness of this sample on Geelbekpan.

Relative density 1.6

50

2.0 1.6 2.0 1.6 2.0 1.6 2.0

1.60 and 2.0 deviation

Correlation accuracy by geologist

Modelling accuracy

Sample 08A 4.9

Sample 09A 8.4

Sample 10A

8.3

11.4

3.9

9.1

5.9

5.4

7.5

14.5

12.3 19.3 9.7

18.9 15.1 17.6

5.3

13.4

16.0

13.2

Sample 11A 3.7 5.0

Sample 12A 10.0 10.1

14.5

13.9

30.1 12.0 20.1 8.6

15.2 17.0 22.0 23.4

25.8

26.1

Sample 13A 7.2 16.1 13.8 15.0 13.9 17.0 10.5 10.4

Sample 14A 18.7 23.2 19.1 21.1 15.5 19.7 13.4 19.1

The σ values for samples 08A to 11A are quite variable. For Sample 12A, there is a gradual increase in σ for Sample 08A = 9.7, for Sample 11A = 15.0, Sample 12A = 17.2, Sample 13A = 12.9 Sample 14A = 18.7.

The combined average is 13.2 (Figure 115).

Good, easy to correlate, but for the south. Problems with correlation, such as faulting and correlation errors, would be prominent on the GDCCD plot. The best correlation for all the graphs in this series is found on the

farm Groenfontein(Sample 09A)(Figure 106). The variation in yields in the south can be ascribed to the

local geological variation (Figure 100).

Good correlation, easy to model, but for the south. Here some factoring would be necessary to compensate

for modelling accuracy, in particular, Sample 14A (Figure 113). There is a general increase in thickness of

the samples towards the south, but the yields are generally lower. The thicknesses of interbedded mudstones is generally greater towards the south.

51

Sample

08A

09A

10A

11A

12A

13A

14A

The use of the GDCCD proves that this method can be used to prove the continuity of Sample 08A to

Use as SANS 10320 data

Sample 14A in the north with a good measure of accuracy; except for the south where care has to be taken when correlating the geology. Sample 14A has a large variance in yield and care must be taken when using

correlation point

data from this sample for correlation purposes and sample 10A, which is a this sample is not laterally

continuous over the Study area (Figure 114).

Table 7 Standard deviation calculated at two RDs from the GCCD data for samples 08A to Sample 14A.

VRYHEID FORMATION

FARM

GEELBEKPAN

1.6

4.9

8.4

3.7

10.0

7.2

18.7

2.0

8.3

11.4

5.0

10.1

16.1

23.2

Avg

GROENFONTEIN

1.6 2.0 Avg

6.4

9.4

4.2

3.9

9.1

5.4

7.5

3.9

7.8

5.9 14.5 9.5

14.5 30.1 21.6

8.2

9.5

13.9

13.8

15.2

15.0

12.2

12.7

23.4

10.5

20.4 19.1 21.1 18.2 13.4

GANNAVLAKTE

1.6

9.7

17.6

5.3

8.6

2.0 Avg 1.6

RINGBULT

2.0 Avg 1.6

13.4 11.0 12.3 19.3 15.7 7.7

16.0 16.1 18.9 15.1 14.7 13.5

13.2 9.2

25.8 17.0 12.0 20.1 15.0

5.6

9.7

26.1

10.4

19.1

23.3

9.1

15.8

17.0

13.9

15.5

22.0

17.0

19.7

19.1

15.1

16.9

16.1

11.4

16.7

TOTAL

2.0 Avg

11.6 9.7

12.5 13.0

13.8 9.7

20.3 15.0

18.4 17.2

14.6 13.0

20.8 18.7

52

G226011 G226006

G250012 G250006

Figure 36 Lateral variation of Sample 14A over the Study area

G299007

53

G299011 R303010 R303019

G250006

G250012

Figure 37 The small coal seam in Sample 10A on Groenfontein is not present on Geelbekpan

54

G226011

13.2

Figure 38 Standard deviation graph at Rd 1.6 and Rd 2.0, combined and averaged, for Sample 08A to Sample 14A

55

Samples 14B to 23D

Figure 39 Combined GDCDD of Sample 14B (Spread at a Rd of 2.0, indicated by light blue dotted line)

56

Figure 40 Combined GDCDD of Sample 15A (Spread at a Rd of 2.0, indicated by light blue dotted line)

57

Figure 41 Combined GDCDD of Sample 15B (Spread at a Rd of 2.0, indicated by light blue dotted line)

58

Figure 42 Combined GDCDD of Sample 16A (Spread at a Rd of 2.0, indicated by light blue dotted line)

59

Figure 43 Combined GDCDD of Sample 17A (Spread at a Rd of 2.0, indicated by light blue dotted line)

60

Figure 44 Combined GDCDD of Sample 18A (Spread at a Rd of 2.0, indicated by light blue dotted line)

61

SAMPLE 17A

SAMPLE 15A

SAMPLE 15B

SAMPLE 16A

SAMPLE 18A

GEELBEKPAN GROENFONTEIN

Figure 45 Graphical representation of Sample 15A to Sample 18A

62

GANNAVLAKTE

Figure 46 Combined GDCDD of Sample 19A (Spread at a Rd of 2.0, indicated by light blue dotted line)

63

Figure 47 Combined GDCDD of Sample 20A (Spread at a Rd of 2.0, indicated by light blue dotted line)

64

G250723\21A

G250729\21A

G250737\21A

G250746\21A

G250752\21A

G250724\21A

G250730\21A

G250738\21A

G250748\21A

G250753\21A

GROENFONTEIN Sample 21A

G250725\21A

G250731\21A

G250742\21A

G250748\21A

G250037\21A

G250726\21A

G250731\21A

G250743\21A

G250749\21A

G250038\21A

G250727\21A

G250735\21A

G250744\21A

G250750\21A

G250 AVERAGE 21A

G250728\21A

G250736\21A

G250745\21A

G250751\21A

100

80

60

40

20

0

Figure 48 Combined GDCDD of Sample 21A (Spread at a Rd of 2.0, indicated by light blue dotted line)

65

SAMPLE 19A

SAMPLE 20A

SAMPLE 21A

GEELBEKPAN GROENFONTEIN

Figure 49 Graphical representation of Sample 19A to Sample 21A

66

GANNAVLAKTE

Figure 50 Combined GDCDD of sample 22A (Spread at a Rd of 2.0, indicated by light blue dotted line)

67

Figure 51 Combined GDCDD of Sample 22B (Spread at a Rd of 2.0, indicated by light blue dotted line)

68

Figure 52 Combined GDCDD of Sample 22C (Spread at a Rd of 2.0, indicated by light blue dotted line)

69

Figure 53 Combined GDCDD of Sample 22D (Spread at a Rd of 2.0, indicated by light blue dotted line)

70

Figure 54 Combined GDCDD of Sample 22E (Spread at a Rd of 2.0, indicated by light blue dotted line)

71

Figure 55 Combined GDCDD of Sample 22F (Spread at a Rd of 2.0, indicated by light blue dotted line)

72

SAMPLE 22A

SAMPLE 22B

SAMPLE 22C

SAMPLE 22D

SAMPLE 22E

SAMPLE 22F

GEELBEKPAN

Figure 56 Graphical representation of Sample 22A to Sample 22F

GROENFONTEIN

73

GANNAVLAKTE

Figure 57 Combined GDCDD of Sample 23A (Spread at a Rd of 2.0, indicated by light blue dotted line)

74

Figure 58 Combined GDCDD of Sample 23B (Spread at a Rd of 2.0, indicated by light blue dotted line)

75

Figure 59 Combined GDCDD of Sample 23C (Spread at a Rd of 2.0, indicated by light blue dotted line)

76

SAMPLE 23A

SAMPLE 23B

GEELBEKPAN GROENFONTEIN

Figure 60 Graphical representation of Sample 23A to Sample 23D

77

SAMPLE 23C

SAMPLE 23D

GANNAVLAKTE

Table 8 Summary of Sample 14B to Sample 23C data.

BASIC INFORMATION

ABOUT SAMPLE 14C tot 23C SUMMARY

DATA

Angle of repose of graphs

Sample 14B

Sample 15A

Sample 15B

GEELBEKPAN GROENFONTEIN RINGBULT

High

High

High

Sample 16A Low to med

Sample 17A Med to high

Sample 18A Med to high

Sample 19A Low to med

Sample 20A Med to high

Sample 21A

Sample 22A

Sample 22B

Low to high

Low to high

Low to high

Sample 22C Med to high

Sample 22D Med to high

Sample 22E

Sample 22F

High

High

High

High

High

Med

Med

Med to high

Low to high

Med to high

Med to high

Med to high

High

High

High

High

High

High

High

High

High

Med to high

High

Low to high

Med to high

Low to high

Low to med

Low to high

Med to high

Med to high

Low to high

Low to high

GANNAVLAKTE

High

High

High

High

Med to high

High

Low to high

Med to high

Low to high

Low to med

Low to high

Med to high

Med to high

Low to high

Low to high

78

Spread

Sample 23A High

Sample 23B High

Sample 23C Low to high

High

High

Med to high

Med to high

Low to high

Low to high

Med to high

Low to high

Low to high

The angle of repose varies over these samples, and it is clear from the graphs, that the lithology in the

south is more variable than in the north (Figure 116, Figure 117, Figure 118,Figure 119, Figure 120,

Figure 121, Figure 123, Figure 124, Figure 125, Figure 127, Figure 128, Figure 129, Figure 130,

Figure 131, Figure 132, Figure 134, Figure 135 and Figure 136).

Start point of graphs at Rd 1.35

Data values start at the same point, except for Ringbult/Gannavlakte in the south: Sample 17A (Figure

120) and Sample 18A (Figure 121).

For Sample 14B and Sample 15A, the spread for Groenfontein and Geelbekpan is very narrow (<10%), while in the south it is ±25%. Sample 15B and Sample 16A have a spread of 25% for the north and

±40% for the south. Sample 17A has a spread of 35% on Groenfontein, 60% on Geelbekpan and 75% in the south. Sample 18A has the narrowest spread in the south and Geelbekpan has the widest distribution.

Sample 19 has a 40% spread in the north and 93% in the south. Sample 20A has a 50% spread for the north and 90% for the south. Sample 21A has a 50% spread for Groenfontein, 65% for Geelbekpan and

93% for the south. Sample 22A has a spread of 25% for Groenfontein, 55% for Geelbekpan and the south. Sample 22B has a spread of 50% for Groenfontein, 65% for Geelbekpan and 85% for the south.

Sample 22C has a spread of 40% for Groenfontein, 50% for Geelbekpan and 80% for the south.

Sample 22D has a 20% spread for Groenfontein, 45% for Geelbekpan and 78% for the south. Sample 22E

79

Yield at Rd 2.0 -Sample 014B

Yield at Rd 2.0 - Sample 15A

Yield at Rd 2.0 - Sample 15B

Yield at Rd 2.0 -Sample 16A

Yield at Rd 2.0 - Sample 17A

Yield at Rd 2.0 - Sample 18A

Yield at Rd 2.0 -Sample 19A

Yield at Rd 2.0 -Sample 20A

Yield at Rd 2.0 - Sample 21A

Yield at Rd 2.0 - Sample 22A

Yield at Rd 2.0 -Sample 22B

Yield at Rd 2.0 - Sample 22C

Yield at Rd 2.0 - Sample 22D

has a spread of 25% for the north and 100% for the south. Sample 22F has a spread of 38% for

Groenfontein, 25% for Geelbekpan and 85% for the south.

Sample 23A has a very narrow spread for Groenfontein, a narrow spread for Geelbekpan and a 68% spread for the south. Sample 23B has as spread of about 35% for the north and 80% for the south. Sample

23C has a spread of 45% for Groenfontein, 70% for Geelbekpan and 75% for the south. The data as a whole indicates that the lithology in the south is more variable than on the northern farms.

24.8 % variance

8.2 % variance

8.6 % variance

26.7 % variance

89.9

61.9

49.4

20.9

85.0

58.0

42.9

26.1

82.0

64.9

51.4

40.7

65.1

56.7

51.5

47.6

34.1 % variance

27.1 % variance

6.3 % variance

11.2 % variance

48.5

34.5

16.2

23.2

33.6

34.2

16.8

27.7

53.3

53.9

14.3

16.5

67.7

61.3

10.5

22.0

22.2 % variance

10.4 % variance

18.2 % variance

14.0 % variance

21.9 % variance

37.3

17.4

22.2

28.3

32.1

80

36.7

19.8

35.2

28.1

47.5

20.1

9.4

22.0

38.9

25.6

15.1

9.8

25.2

42.1

36.6

Yield at Rd 2.0 -Sample 22E

Yield at Rd 2.0 -Sample 22F

Yield at Rd 2.0 - Sample 23A

Yield at Rd 2.0 - Sample 23B

Yield at Rd 2.0 -Sample 23C

Remarks- Yield

Raw Rd Sample 14B

Raw Rd Sample 15A

Raw Rd Sample 15B

Raw Rd Sample 16A

Raw Rd Sample 17A

Raw Rd Sample 18A

Raw Rd Sample 19A

Raw Rd Sample 20A

Raw Rd Sample 21A

Raw Rd Sample 22A

Raw Rd Sample 22B

Raw Rd Sample 22C

17.6 % variance

10.9 % variance

21.5 % variance

34.9 % variance

18.5 % variance

46.8

67.5

80.4

72.9

52.9

58.5

69.6

85.6

68.4

34.4

42.8

58.7

64.1

56.5

39.8

40.9

62.1

64.6

38.0

36.4

The largest variation in yield is found in Sample 17A and Sample 23B, and the highest yield in Sample

14B on Geelbekpan and Sample 23A on Groenfontein.

0.12 variance

0.03 variance

1.62

1.73

1.62

1.72

1.65

1.72

1.74

1.75

0.06 variance

0.28 variance

1.80

2.12

1.86

2.02

1.80

1.87

1.82

1.84

0.21 variance

0.14 variance

0.16 variance

0.15 variance

0.40 variance

0.17 variance

0.21 variance

0.13 variance

1.83

1.91

2.12

2.04

1.78

2.08

2.07

2.02

1.93

1.92

2.10

2.00

1.89

2.08

1.90

1.97

1.80

1.80

2.20

2.09

2.09

2.21

2.11

1.96

1.72

1.78

2.26

2.13

2.18

2.25

2.10

1.89

81

Raw Rd Sample 22D

Raw Rd Sample 22E

Raw Rd Sample 22F

Raw Rd Sample 23A

Raw Rd Sample 23B

Raw Rd Sample 23C

Remarks on raw Rd

Thickness Sample 14B

Thickness Sample 15A

Thickness Sample 15B

Thickness Sample 16A

Thickness Sample 17A

Thickness Sample 18A

Thickness Sample 19A

Thickness Sample 20A

Thickness Sample 21A

Thickness Sample 22A

Thickness Sample 22B

Thickness Sample 22C

0.21 variance

0.16 variance

0.15 variance

0.20 variance

0.24 variance

2.01

1.95

1.84

1.78

1.87

1.89

1.87

1.81

1.70

1.82

2.10

2.03

1.96

1.89

1.92

1.97

2.03

1.93

1.90

2.06

0.16 variance 1.95 2.01 2.05 2.11

The largest variance in yield occurs in Sample 21A and the smallest variance in raw Rd in Sample 15A.

0.7 meter variance 1.14 1.53 1.71 1.90

0.49 meter variance

0.23 meter variance

2.87

3.12

3.06

2.92

2.67

2.94

2.57

2.89

0.59 meter variance

0.90 meter variance

0.88 meter variance

1.07 meter variance

0.23 meter variance

0.77 meter variance

0.95 meter variance

0.69 meter variance

0.32 meter variance

2.54

1.95

2.18

2.69

1.67

1.53

3.86

2.90

2.32

2.13

2.21

2.47

1.80

1.61

1.60

4.38

2.47

2.25

1.95

2.85

2.54

2.47

1.84

2.22

4.07

3.16

2.55

2.03

2.21

1.66

2.87

1.69

1.45

3.43

2.77

2.23

82

Thickness Sample 22D

Thickness Sample 22E

Thickness Sample 22F

Thickness Sample 23A

Thickness Sample 23B

Thickness Sample 23C

0.79 meter variance

0.35 meter variance

0.55 meter variance

0.56 meter variance

1.03 meter variance

0.55 meter variance

2.62

3.06

2.10

1.26

1.31

1.49

2.85

2.74

1.95

1.18

1.06

2.04

2.51

2.86

2.50

1.74

1.37

1.97

2.06

3.09

2.24

1.63

2.09

1.83

Remarks on thickness

Standard deviation at a Rd of

1.60 and 2.0

Sample 14B has a fairly constant thickness (Figure 112).The thickness of sample 15A, 15B and 17A is

fairly constant (Figure 122), while the thickness of sample 16A and 18A becomes thinner towards

Gannavlakte (Figure 122). The largest variance in thickness occurs in Sample 19A and the smallest

variance in Sample 15B and Sample 20A. To have only a maximum variance in thickness of 1.07 m over

a few hundred square km is remarkable (Figure 126). Figure 139, which shows the thickness contours of

Sample 15B, shows an equal thickness over the area. The purple circle shows modelling inaccuracies which are caused by a lack of borehole information. The thickness of sample 22A to 22F id constant

over the whole area (Figure 133). Sample 23D only occurs in the south ().Figure 137

Relative density 1.6

Sample 14B 10.7

Sample 15A 4.2

Sample 15B 8.6

2.0

1.3

2.4

5.9

1.6

12.7

6.6

4.7

2.0

2.0

3.9

8.3

1.6

21.8 12.7 23.7

8.1 4.4

12.5

2.0

7.5

1.6

12.7

12.8

2.0

7.1

7.6

9.6

83

Sample 16A 8.4 12.2

Sample 17A 15.7 15.9

Sample 18A 9.1 18.7

Sample 19A 3.4

Sample 20A 7.9

11.0

14.0

Sample 21A 5.4

Sample 22A 2.4

Sample 22B 8.6

Sample 22C 5.2

Sample 22D 0.9

19.3

12.8

18.9

18.1

13.0

9.3

8.8

6.6

5.7

9.9

6.9

3.4

12.8

7.4

5.7

13.2 12.5 17.6 10.0

12.4 27.4 16.5 18.8

10.4 15.8 22.2 20.3

13.6

16.8

10.5

14.7

5.4 22.8

14.3 10.5 24.0

8.0 30.5

7.4 15.9

7.2 20.2

1.2

9.0

12.1

1.6

3.5

15.5 13.7 25.1 20.4

7.7 3.3 20.8 14.8

Sample 22E 0.8

Sample 22F 2.8

Sample 23A 4.1

8.5

6.4

6.3

8.1

12.6

17.2

6.2

10.9

2.2

8.2 29.4

5.4 31.4

9.1 22.7

3.8

1.8

6.4

39.3

34.1

27.7

Sample23B 2.4 12.5 9.8 17.4 9.8 31.0 7.5 31.4

Comments on standard deviation

Sample 23C 10.7 1.3 12.7 2.0 21.8 12.7 23.7 7.1

The largest σ value, of 39.3, can be found at a Rd of 2.0 in Sample 22E on Gannavlakte, and the smallest

σ value at a Rd of 1.60 for Sample 22E on Geelbekpan The σ value in the south are generally higher than in the central and northern areas. In general, Geelbekpan has the lowest σ value of all the samples and

Ringbult and Gannavlakte the highest σ value. When combining the σ values of 1.60 an d 2.0 it can be

clearly seen that the south has higher value than the north for all the samples (Figure 138). On this graph

84

36.4

12.9

29.0

22.8

30.4

9.6

16.3

11.3

13.3

26.5

(Figure 138)

it is illustrated that basically all the samples in the north has a σ value of less than 10 and the south and average of about 12.5.

Correlation accuracy by

Correlation is good and easy to do, but care will have to be taken in the southern part of the Study area.

geologist

Modelling accuracy

Care must be taken in modelling these samples, as there is some variation on a per sample basis.

Sample 14B to Sample 15B are very closely spaced and very similar for the north and only slightly wider for the south. From Sample 16A to Sample 23B, the graphs illustrate a wider association of the graphs and the south is quite wide to haphazard in places, based on the local geology. Sample 23C is influenced by the mudrock parting below the coal sample; where the parting was greater than 2 m, the coal seams below were added to Sample Zone 04 in which Sample 23 occurs. The angle of repose of the graphs is influenced by the amount of low density material (coal) present in the sample. Sample 23B and Sample

Use as SANS 10320 data

17A have the highest variance in yield and Sample 14B and Sample 23A some of the highest yields.

Sample 22A has the lowest yields. Sample 21A has the highest variance in raw Rd because it consists

correlation point

virtually of only of mudrock material in places. Sample 15A has the lowest variance with a Rd of only

0.03.

In term of thickness, Sample 15B has the least variance in thickness and Sample 19A the highest variance. Sample 22A is the thickest individual sample and Sample 23A on Groenfontein the thinnest sample. It should be a simple matter for a geologist to correlate these samples over the Study area, but when modelling, care must be taken with some of the samples. In general, the lithology variation in the south is higher than in the north.

85

FARM

Sample

14B

15A

15B

16A

22D

22E

22F

23A

23B

23C

17A

18A

19A

20A

21A

22A

22B

22C

Table 9 Standard deviation calculated at two RDs from the GCCD data for Sample 14B to Sample 23C.

GEELBEKPAN

14.0

19.3

12.8

18.9

18.1

13.0

8.5

12.2

15.9

18.7

11.0

2.0

1.3

2.4

5.9

6.4

6.3

12.5

25.4

2.8

4.1

2.4

3.0

7.9

5.4

2.4

8.6

5.2

0.9

0.8

1.6

10.7

4.2

8.6

8.4

15.7

9.1

3.4

10.9

12.4

7.6

13.7

11.6

7.0

4.7

Avg

6.0

3.3

7.3

10.3

15.8

13.9

7.2

4.6

5.2

7.4

14.2

GROENFONTEIN

VRYHEID FORMATION

GANNAVLAKTE

14.3

16.8

10.5

14.7

15.5

7.7

6.2

13.2

12.4

10.4

13.6

2.0

2.0

3.9

8.3

10.9

2.2

17.4

17.5

9.9

6.9

3.4

12.8

7.4

5.7

8.1

9.3

8.8

6.6

5.7

1.6

12.7

6.6

4.7

12.6

17.2

9.8

8.7

Avg

7.3

5.3

6.5

11.2

10.6

8.5

9.7

12.1

11.8

7.0

13.8

11.5

6.7

7.1

11.7

9.7

13.6

13.1

9.0

12.1

1.6

3.5

20.4

14.8

3.8

1.6

23.7

12.7

12.8

10.0

18.8

20.3

1.2

1.8

6.4

7.5

11.1

26.5

36.4

12.9

29.0

22.8

30.4

39.3

9.6

16.3

11.3

13.3

2.0

7.1

7.6

9.6

34.1

27.7

31.4

26.2

17.7

24.3

7.2

16.3

21.6

22.6

21.5

Avg

15.4

10.2

11.2

9.8

17.6

15.8

7.3

17.9

17.1

19.4

18.6

10.5

8.0

7.4

7.2

13.7

3.3

8.2

1.6

21.8

8.1

12.5

12.5

27.4

15.8

5.4

5.4

9.1

9.8

18.5

RINGBULT

24.0

30.5

15.9

20.2

25.1

20.8

29.4

2.0

12.7

4.4

7.5

17.6

16.5

22.2

22.8

31.4

22.7

31.0

27.0

17.3

19.3

11.6

13.7

19.4

12.0

18.8

Avg

17.3

6.2

10.0

15.0

21.9

19.0

14.1

18.4

15.9

20.4

22.8

86

TOTAL

19.7

25.8

13.0

20.7

20.4

18.0

20.8

13.1

15.3

15.7

15.2

2.0

5.8

4.6

7.8

20.7

14.7

23.1

24.0

9.3

8.1

3.7

8.0

11.7

6.2

5.2

1.6

17.2

7.9

9.7

10.0

17.7

12.9

3.9

5.6

9.2

7.4

10.3

14.5

16.9

8.3

14.4

16.0

12.1

13.0

Avg

11.5

6.2

8.7

11.6

16.5

14.3

9.6

13.2

12.0

15.2

17.2

10.0

Figure 61 Standard deviation graph at of Rd 1.6 and 2.0, combined and averaged, for Sample 14C to Sample23C

87

Welgelegen

Geelbekpan

Groenfontein meters

Gannavlakte

Ringbult

Figure 62 The lateral thickness variation of Sample 15B over the Study area. Purple circle delineates area of modelling inaccuracies due to lack of information

88

Samples 28A to 32B

Figure 63 Combined GDCDD of Sample 28A (Spread at a Rd of 2.0, indicated by light blue dotted line)

89

Figure 64 Combined GDCDD of Sample 29A (Spread at a Rd of 2.0, indicated by light blue dotted line)

90

Figure 65 Combined GDCDD of Sample 30A (Spread at a Rd of 2.0, indicated by light blue dotted line)

91

Figure 66 Combined GDCDD of Sample 30B (Spread at a Rd of 2.0, indicated by light blue dotted line)

92

Figure 67 Combined GDCDD of Sample 31A (Spread at a Rd of 2.0, indicated by light blue dotted line)

93

Figure 68 Combined GDCDD of Sample 31B (Spread at a Rd of 2.0, indicated by light blue dotted line)

94

Figure 69 Combined GDCDD of Sample 32A (Spread at a Rd of 2.0, indicated by light blue dotted line)

95

Table 10 Summary of Sample 28A to Sample 32A data.

BASIC INFORMATION

ABOUT SAMPLE 28A tot

SAMPLE 32A DATA

SUMMARY GEELBEKPAN GROENFONTEIN RINGBULT GANNAVLAKTE

Sample 28A High

Sample 29A Med to high

Sample 30A Med to high

Sample 30B

Sample 31A

High

High

High

High

High

High

High

High

High

High

High

High

High

High

High

High

High

Angle of repose of graphs

Sample 31B High High High High

Sample 32A High High Med to high Med to high

The nature of the coals in these samples has changed to a mat coal, and these coals are low ash, high yield coals

which is reflected in the GDCDD curves (Figure 140, Figure 141, Figure 142, Figure 143, Figure 144,

Figure 145 and Figure 146). The lower curves in Sample 32A, for the south, are those reflected by the

inclusion of interbedded sandstones in the samples.

Start point of graphs at Rd

The graphs do not have a common starting point at the origin of the graph but it varies quite a lot and this is

1.35

because of the varying quality of the coal in the samples

Spread

Sample 28A and Sample 29A have a narrow spread for Groenfontein and the south, but the presence of interbedded sandstones makes the spread at Geelbekpan wider. Sample 30A has similar spreads for Geelbekpan

96

and the south of 35% - 45%. samples 30B, 31A and 31B have a narrow spread for all samples, while Sample

32A has a spread of 30% plus for Geelbekpan and the south. This narrow spread is indicative of the type of coals found in these samples.

Yield at Rd 2.0 -Sample

28A

Yield at Rd 2.0 - Sample

29A

Yield at Rd 2.0 - Sample

30A

Yield at Rd 2.0 -Sample

30B

Yield at Rd 2.0 - Sample

19.7% variance

21.6% variance

5.2% variance

9.0% variance

71.2

66.7

78.5

86.1

85.7

82.3

82.6

86.3

89.6

88.2

79.4

77.3

31A

Yield at Rd 2.0 - Sample

31B

Yield at Rd 2.0 -Sample

32A

Remarks- Yield

Raw Rd Sample 28A

Raw Rd Sample 29A

8.3% variance

17.0% variance

9.0% variance

89.2

90.9

80.4

87.2

89.4

87.1

89.7

78.4

89.4

The largest variation in yield is found in Sample 29A and the highest yield in Sample 31B.

0.17 variance 1.76 1.66 1.60

0.18 variance 1.77 1.70 1.59

97

90.9

79.4

83.7

84.0

81.4

73.9

85.1

1.59

1.67

Raw Rd Sample 30A

Raw Rd Sample 30B

Raw Rd Sample 31A

Raw Rd Sample 31B

Raw Rd Sample 32A

Remarks on raw Rd

Thickness Sample 28A

Thickness Sample 29A

Thickness Sample 30A

Thickness Sample 30B

Thickness Sample 31A

Thickness Sample 31B

Thickness Sample 32A

Remarks on thickness

-Standard deviation at a

Rd of 1.60 and 2.0

0.07 variance

0.12 variance

0.16 variance

0.20 variance

0.23 variance

1.70

1.61

1.54

1.67

1.72

1.67

1.61

1.53

1.53

1.60

1.63

1.70

1.63

1.68

1.49

The largest variance in yield occurs in Sample 32A and the smallest variance in raw Rd in Sample 30A.

0.17 meter variance

0.57 meter variance

1.49

1.65

1.50

1.51

1.36

1.38

1.33

1.08

0.15 meter variance

0.11 meter variance

1.33

1.31

1.29

1.36

1.20

1.25

1.35

1.36

1.64

1.73

1.69

1.73

1.58

0.18 meter variance

0.13 meter variance

0.51 meter variance

1.34

1.31

1.30

1.36

1.32

1.53

1.15

1.29

1.53

1.18

1.19

1.81

The largest variance in thickness occurs in Sample 29A and the smallest variance in Sample 31B. Sample 32A has a scattered occurrence, in particularly on Groenfontein, but does not occur over a large area. Geelbekpan’s

Sample 32A also has a variable thickness as depicted in Figure 150.

The σ value increases to the south on the right of the figure when interbedded sandstone begins to play a role in the subdivision of the samples.

Relative density 1.6 2.0

Sample 28A 24.4 26.1

98

1.6

20.7

2.0 1.6 2.0

18.8 17.9 3.1

1.6

25.9

2.0

2.1

Sample 29A 18.7 23.5

Sample 30A 15.8 17.5

Sample 30B 18.0 2.4

Sample 31A 22.6 19.7

Sample 31B 15.3 2.0

17.3

23.3

17.9

20.3

19.3

15.0 12.7 5.2

13.4 23.2 9.9

6.1 21.8 1.9

17.4 19.7 6.7

14.8 11.7 8.2

20.3

21.4

25.2

27.0

28.5

9.4

8.5

4.3

6.2

10.3

Sample 32A 26.3 8.9 35.4 29.6 13.5 8.3 13.5 8.5

The σ value in the north and central areas are generally higher than in the south. The lowest σ per sample is

Comments on standard

found in Sample 30B. The bar chart in Figure 147 shows that Geelbekpan and Groenfontein, in general, have

deviation

a higher σ value than the other two sampled areas and that the Volksrust Formation has a higher σ value that that of the Vryheid Formation.

Correlation accuracy by

Care will have to be taken when correlating Sample 28A to Sample 32B due to the presence of interbedded

geologist

Modelling accuracy

sandstones in certain areas (Figure 148 and Figure 149).

Care must be taken when modelling these samples, as there is some variation on a per sample basis.

Use as SANS 10320 data

Care must be taken when correlating these samples, as there is some variation on a per sample basis.

correlation point

99

Sample

28A

29A

30A

30B

31A

31B

32A

Avg

Table 11 Standard deviation calculated at two RDs from the GCCD data for Sample 28A to Sample 32A.

GEELBEKPAN

1.6

24.4

18.7

15.8

18.0

22.6

15.3

26.3

20.2

2.0

26.1

23.5

Avg

25.3

21.1

VOLKSRUST FORMATION

GROENFONTEIN

Farm

GANNAVLAKTE

1.6

20.7

17.3

17.5 16.6 23.3

2.4 10.2 17.9

19.7 21.2 20.3

2.0

8.9

14.3

8.7

16.6

17.1

19.3

35.4

22.0

2.0

18.8

15.0

Avg

19.7

16.2

13.4 18.3

6.1 12.0

17.4 18.9

14.8 17.1

29.6

16.4

32.5

19.2

1.6

25.9

20.3

21.4

25.2

27.0

28.5 10.3

13.5

23.1

2.0

2.1

9.4

8.5

4.3

6.2

8.5

7.0

Avg

14.0

14.9

11.0

15.1

1.6

17.9

12.7

13.5

17.2

RINGBULT

2.0

3.1

5.2

8.3

6.2

Avg

10.5

8.9

10.9

11.7

1.6

22.2

17.2

22.2

20.6

TOTAL

2.0

12.5

13.3

13.8

11.0

Avg

17.4

15.3

14.9 23.2 9.9 16.5 20.9 12.3 16.6

14.8 21.8 1.9 11.9 20.7 3.7 12.2

16.6 19.7 6.7 13.2 22.4 12.5 17.4

19.4 11.7 8.2 9.9 18.7 8.8 13.8

18.0

15.8

100

15.8

Figure 70 Standard deviation bar graph at a Rd of 1.6 and 2.0, combined and averaged, for Sample 28A to Sample 32B

101

Figure 71 Profile on Geelbekpan for Sample 32A, illustrating variability. Boreholes in no specific order

Figure 72 Profile on Groenfontein for Sample 32A, illustrating variability. Boreholes in no specific order

102

Groenfontein

Welgelegen

Gannavlakte

Geelbekpan meters

Ringbult

Figure 73 The lateral thickness variation of Sample 32A over the Study area. The purple oval is a function of the modelling package and does not denote anything in particular

103

Coal Zone 10

G250020

Figure 74 Combined GDCDD of Zone 10

104

G250010

G250727

The high values (upper line in the graph) in Figure 151 are represented by G250020, in

which the upper coal zones have been eroded away (Figure 153).

Figure 75 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of Zone 11 at Groenfontein. Plus and minus two standard deviations in yellow blocks

In Figure 152

, the data in the yellow blocks, for 2σ, indicates that for a Rd of 2, the

minimum value is outside 2σ and the maximum value is outside the 2σ. This would suggest that the lower value fall outside 2σ, which is a wide spacing, but that the maximum value fall outside the 2σ which also illustrates a wide distribution. The same is true for the RD 1.6 value. This is a result of erosion and that more of the lower part of coal Zone 11 is preserved

over Welgelegen. The central curves (Figure 151) are closely spaced and the upper and

lower curves are spaced more widely apart. There is some variability in the lithotypes.

105

Eroded away

G250020

2 km

G250010 G250727

Figure 76 Zone 10 geological profile below illustrating why low, high and average values occur in the GDCDD diagram

106

Groenfontein

Welgelegen

Geelbekpan

Gannavlakte

Figure 77 The lateral thickness variation of Zone 10 over the Study area

Zone 10 is the thickest on Groenfontein and Welgelegen (Figure 154).

metres

Ringbult

107

Coal Zone 09

G250721

Figure 78 Combined GDCDD of Zone 09

108

G250726

G250020

The GDCDD of coal Zone 09 is illustrated in Figure 155. This figure illustrates that two

boreholes (G250721 and G250726) are far removed from the other graphs. In the case of

G250721, the top of coal Zone 09 has been weathered away (Figure 157). Using only the

lower part of the coal zone leads to an erroneous interpretation of coal Zone 09 as an entity.

G250020 has thick sandstone developed in Sample 09A and the total thickness of coal Zone

09 for this borehole is 5,5 m compared to a normal 3,5 m for the coal zone (as represented by

borehole G250726) (Figure 157).

In Figure 156

, the data in the yellow blocks, for 2σ, indicates that for a Rd of 2, the

minimum value is inside 2σ and the maximum value is outside 2σ. This would suggest that the lower values fall inside

2σ, which is a narrow spacing, but that the maximum values fall outside 2σ which also illustrates a wide distribution. For a Rd of 1.6, both the minimum and maximum value are inside 2σ. This would suggest that there is a narrow spacing between the

individual line graphs. Figure 156 illustrates that the few outlying data points does not

influence the statistical result too much at a Rd 1.6, but at a Rd of 2 it has an impact. The lithotypes are reasonably constant throughout coal Zone 09.

109

Figure 79 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of Zone 09 at Groenfontein. Plus and minus two standard deviations in yellow blocks

110

Removed by faulting

G250721 G250020

2 km

Figure 80 Zone 09 geological profile illustrates why low, high and average values occurs in the GDCDD diagram

G250726

111

Groenfontein

Welgelegen

Geelbekpan metres

Gannavlakte

Figure 81 The lateral thickness variation of Zone 09 over the Study area

Ringbult

Zone 09 is the thickest developed over Gannavlakte and Ringbult (Figure 158).

112

Coal Zone 08

G250031

Figure 82 Combined GDCDD of Zone 08

113

G250727

G250007

Figure 83 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of Zone 08 at Groenfontein. Plus and minus two standard deviations in yellow blocks

In Figure 160

, the data in the yellow blocks, for 2σ, indicates that for a Rd of 2, the

minimum and maximum value are inside 2σ. This would suggest a narrow spacing for the σ value. For the Rd 1.6 values, the minimum value is outside 2σ and the maximum value is just inside 2σ. This would suggest that the lower values have a wider spacing, but that the maximum values fall inside 2σ, which also illustrates a narrow distribution. All this evidence

suggests that there is a wider spacing between the individual line graphs. Figure 160

illustrates that the one or two outliers do not influence the statistical result too much at a Rd of 1.6, but at a Rd of 2.0, it has an impact. The lithotypes in this Zone vary to some extent.

114

G250031

G250727

Figure 84 Zone 08 geological profile illustrates why low, high and average values occur in the GDCDD diagram)

115

G250007

G250727 has a thick coal zone, Zone 08, with thick sandstone near the top (Figure 161)

leading to low GDCDD values (Figure 159)

Welgelegen

Geelbekpan

Groenfontein metres

Ringbult

Figure 85 The lateral thickness variation of Zone 08 over the Study area

Gannavlakte

Zone 08 is the thickest developed over Gannavlakte (Figure 162).

116

Coal Zone 07

G250018

Figure 86 Combined GDCDD of Zone 07

117

G250024

G250016

Figure 87 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of Zone 07 at Groenfontein. Plus and minus two standard deviations in yellow blocks

In Figure 164

, the data in the yellow blocks, for 2σ, indicates that for a Rd of 2, the

minimum and maximum value is inside 2σ. This suggests a narrow spacing for the σ value.

For the Rd 1.6 value, the min imum and maximum value is inside 2σ. All the evidence

suggests that there is a narrow spacing between the individual line graphs. Figure 163

illustrates that the line graphs have a narrow spacing, meaning that the lithotypes are similar.

118

G250018

G250024

Figure 88 Zone 07 geological profile illustrating why low, high and average values occur in the GDCDD diagram

G250016

119

The lower part of G250018 is not well developed (Figure 165) which leads to higher values

in the GDCDD for Zone 07 (Figure 163).

Zone 07 is the best developed over the north (Figure 166) while the area in the purple circle

is a modelling inaccuracy due to the absence of boreholes.

Welgelegen

Geelbekpan

Groenfontein metres

Ringbult

Figure 89 The lateral thickness variation of Zone 07 over the Study area

Gannavlakte

120

Coal Zone 06

Figure 90 Combined GDCDD of Zone 06

G250021

121

G250031

G250023

In Figure 168

, the data in the yellow blocks, for 2σ, indicates that for a Rd of 2, the

minimum value falls inside the 2σ and the maximum value is just outside 2σ. This suggests a narrow spacing for the σ value.

Figure 91 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of coal Zone 06 at Groenfontein.

Plus and minus two standard deviations in yellow blocks

For the Rd 1.6 value, the minimum and maximum value is just outside 2σ. This suggests that the values have a wider distribution. All the evidence suggests that there is a wider spacing

between the individual line graphs. Figure 167 illustrates that the line graphs have a wider

spacing. This suggests varying geological lithotypes in this Zone, which is supported by the

profile in Figure 169.

122

G250021

G250031

Figure 92 Zone 06 geological profile illustrating why low, high and average values occur in the GDCDD diagram

G250023

123

Figure 170 shows that G250021 has no thin mudrock near the top of coal Zone 06 as in the

other two boreholes. G250031 has low GDCDD values (Figure 167) as it has a very thick

mudrock developed at the top of coal Zone 06 (Figure 169).

Zone 06 is the thickest developed on Geelbekpan (Figure 170).

Welgelegen

Geelbekpan

Groenfontein metres

Ringbult

Figure 93 The lateral thickness variation of Zone 06 over the Study area

124

Gannavlakte

Coal Zone 05

G250021

Figure 94 Combined GDCDD of Zone 05

125

G250031

G250023

In Figure 172

, the data in the yellow blocks, for 2σ, indicate that for a Rd of 2, the minimum value falls inside the 2σ and maximum value is outside 2σ.

Figure 95 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of Zone 05 at Groenfontein. Plus and minus two standard deviations in yellow blocks

This suggests a narrow spacing for the σ value. For the Rd 1.6 value, the minimum value is inside 2σ, while the maximum value is outside 2σ. All the evidence suggests that there is a

wider spacing between the individual line graphs. Figure 171 illustrates that the line graphs

have a slightly wider spacing. This suggests slightly varying geological lithotypes in this

Zone, which is supported by the profile in Figure 173.

126

G250008

G250005

Figure 96 Zone 05 geological profile below illustrating why low, high and average values occurs in the GDCDD diagram

G250746

127

There is little variation in the geology of Zone 05 (Figure 171 and Figure 173) except for

thickness.

Zone 05 is the thickest developed over the south and on Geelbekpan Figure 174).

Welgelegen

Geelbekpan

Groenfontein metres

Ringbult

Figure 97 The lateral thickness variation of Zone 05 over the Study area

Gannavlakte

128

Coal Zone 04

G250009

G250036

G250728

Figure 98 Combined GDCDD of Zone 04. Red oval denotes second population which is influenced by sandstones included in the zone

129

In Figure 176

, the data in the yellow blocks, for 2σ, indicate that for a Rd of 2, the minimum

and maximum value fall inside 2σ as the data is very widely spaced and the σ value is three times greater tha n most of the previous σ values.

Figure 99 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of Zone 04 at Groenfontein. Plus and minus two standard deviations in yellow blocks

This would suggest that a wide spacing exists for these values. For the Rd of 1.6 value, the minimum value is inside 2σ, while the maximum value is outside 2σ. The evidence suggests

that there is a wider spacing between the individual line graphs. Figure 175 illustrates that

the line graphs have a wider spacing. This suggests varying geological lithotypes in this zone,

which is supported by the profile in Figure 177.

130

G250009

G250036

Figure 100 Zone 04 lithological profile illustrates why low, high and average values occurs in the GDCDD diagram

131

G250728

Zone 04 has layers of coal of varying thickness at the bottom of the zone. How thick the interbedded mud/sandstone was, determined whether the separated lower coal layer was included in Zone 04 or whether it was disregarded and made part of Zone 04A. As Zone

04A’s distribution is random and thus of no economic interest, it was not included in this study. Zone 04 consists of Sample 24A to Sample 24D which occurs randomly throughout

the thesis area. In Figure 178 the circle denotes a second population in the GDCDD of Zone

04. This population is based on the amount of interbedded waste in Zone 04. G250036 in

Figure 177 illustrates this clearly.

There is another identified coal zone in the stratigraphy that occurs between coal Zone 04 and coal Zone 03, namely coal Zone 04A. This zone consists of dull coals, like Zone 04 and the lower coal zones, but because the coal are so irregularly distributed, they are not discussed in

this thesis. This zone is illustrated in Figure 179 but not discussed in any detail.

132

Groenfontein Boreholes

Problem area

Ringbult Boreholes

Figure 101 ‘Separated’ coal at the bottom of Zone 04 (above thick red line). The area in which the geologist will have difficulty in determining the floor of Zone 04, is shown in the red circle. Horizontal lines on graph are at one metre intervals

133

Figure 102 Irregular occurrence of the coals of Sample 24A to Sample 24D in coal Zone 04A, which is not discussed in this thesis due to their irregular nature

134

Zone 04 is the thickest developed in the south (Figure 180).

Welgelegen

Groenfontein

Geelbekpan metres

Ringbult

Figure 103 The lateral thickness variation of Zone 04 over the Study area

Gannavlakte

135

Coal Zone 03

G250021

Figure 104 Combined GDCDD of Zone 03

G250716

136

G250012

In Figure 182

, the data in the yellow blocks, for 2σ, indicates that for a Rd of 2, both the

minimum and maximum values fall inside the 2σ as that the data is very widely spaced and the σ value is two times greater than most of the previous σ values.

Figure 105 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of Zone 03 at Groenfontein .

Plus and minus two standard deviations in yellow blocks

For the Rd 1.6 value, the minimum value is inside 2σ, while maximum value is outside 2σ.

The evidence suggests that there is a wider spacing between the individual line graphs. Figure

181 illustrates that the line graphs have a wider spacing. This suggests varying geological

lithotypes in this zone, which is supported by the profile in Figure 183.

137

G250021

G250716

G250012

Figure 106 Zone 03 lithological profile, illustrating why low, high and average values occurs in the GDCDD diagram

138

Zone 03 is the thickest developed over the north (Grootwater) (

Figure 184

) due to the

presence of interbedded sandstones.

Welgelegen

Geelbekpan

Groenfontein metres

Gannavlakte

Zone 03 is the thickest developed over the north (Grootwater) (

Figure 184

) due to the

presence of interbedded sandstones.

139

Coal Zone 02

G250729

Figure 108 Combined GDCDD of Zone 02

140

G250027

G250005

In Figure 186

, the data in the yellow blocks, for 2σ, indicate that for a Rd of 2 and Rd 1.6,

both the minimum and maximum value fall inside 2σ as the data is widely spaced.

Figure 109 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of Zone 02 at Groenfontein. Plus and minus two standard deviations in yellow blocks

The evidence suggests that there is a wider spacing between the individual line graphs. Figure

185 illustrates that the line graphs have a wider spacing. This suggests some variance in

geological lithotypes in this zone, which is supported by the profile in Figure 187. Figure

188 illustrates that the line graphs have a wider spacing.

141

G250027

G250729

Figure 110 Zone 02 lithological profile illustrates why low, high and average values occur in the GDCDD diagram

142

G250005

W228015

W228716

W228027

Figure 111 Combined GDCDD of Zone 02 on Welgelegen, illustrating the presence of interbedded sandstones

W228713

143

The higher

σ value illustrated in Figure 185 are due to the amount of inorganic material

found in the coals of Zone 02 (Figure 187). On Welgelegen, north-east of Groenfontein, the

variance in Zone 2 is even more prominent.

In Figure 189

, the data in the yellow blocks, for 2σ, indicates that for a Rd of 2, the

minimum values fall outside 2σ and maximum values fall inside 2σ, as the data is very widely spaced and the σ value is 28.7, which is very high. For the Rd 1.6 value, the minimum and maximum values are inside 2σ.

Figure 112 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of Zone 02 at Welgelegen. Plus and minus two standard deviations in yellow blocks

This suggests varying geological lithotypes in this zone, which is supported by the profile in

Figure 190.

144

W228015 W228716 W228027

W228713

Figure 113 Zone 02 lithological profile illustrates why low, high and average values occur in the GDCDD diagram on Welgelegen

145

Zone 02 is the thickest developed over Welgelegen (

Figure 191

), due to the presence of

interbedded sandstones.

Welgelegen

Geelbekpan

Groenfontein metres

Ringbult

Figure 114 The lateral thickness variation of Zone 02 over the Study area

Gannavlakte

146

Coal Zone 01

G250021

Figure 115 Combined GDCDD of Zone 01

G250017

G250020

147

Coal Zone 01 is not developed over the whole Study area and on a large part of Groenfontein , not developed. Due to its thickness of, mostly, less than 2 meters it consists only of one sample (32A) and is not influenced by interbedded sandstones in the way Zone 02 is influenced. In only limited instances has a geologist included a small coal seam above Zone 01 that is separated by a thin sandy

layer from Zone 01 (

Figure 194

) and the influence of this is shown upon the GDCDD of the

borehole in

Figure 192

(borehole G250017).

In

Figure 193

, the data in the yellow blocks, for

2σ, indicates that for a Rd of 2 and Rd 1.6, both the minimum and maximum values fall inside 2σ as the data is widely spaced.

Figure 116 Statistics at a Rd of 2 and 1.6 of the GDCDD’s of Zone 01 at Groenfontein. Plus and minus two standard deviations in yellow blocks

148

The suggests that there is a wider spacing between the individual line graphs.

Figure

illustrates that the line graphs have a narrower spacing. This suggests some variance in

geological lithotypes in this zone, which is supported by profile in

Figure 194

.

149

G250017 G250021

Figure 117 Zone 01 lithological profile illustrates why low, high and average values occur in the GDCDD diagram

G250020

150

Zone 01 is the not developed over most of the Groenfontein area (

Figure 195

).

Welgelegen

Geelbekpan

Groenfontein

Ringbult metres

Figure 118 The lateral thickness variation of Zone 01 over the Study area

151

Gannavlakte