The Measurement and Control of Dilution in an Underground Coal Operation

M Nopp1
ABSTRACT
Dilution in underground coal operations can be underestimated by up to five times during feasibility studies, with disastrous economic consequences. Every one per cent of stone waste introduced can reduce an in situ specific energy of 25 MJ/kg by some 0.25 MJ/kg and increases the abrasiveness of the run-of-mine coal. Where a wash plant is not in place, penalties imposed for not meeting contractual specifications alone can amount to some $A 1 million per year. This paper presents a method for monitoring, reporting and ultimately controlling dilution in an underground coal mining operation and presents the dilution trends from a continuous miner/longwall mine over an eight year period. The overall dilution level for the mine can be back-calculated by comparing the run-of-mine coal quality to the predicted in situ, undiluted quality. However, this will not identify the actual contaminant types, areas of the mine responsible for the dilution and the reasons for the dilution. This additional information is essential for controlling dilution to acceptable levels. Quality and horizon control using existing sampling information is generally only adequate for longer-term control purposes, and is not suitable for monitoring mining-induced dilution. The cost of close-spaced surface drilling to collect the information necessary for short-term control is prohibitive, particularly for deeper mines. However, much of this data can be collected through underground channel sampling and horizon control measurements taken during continuous miner development and longwall extraction. The mine runs the risk of not meeting contractual quality specifications and of increased equipment wear if this additional information is not available and the sources of dilution cannot be controlled. The information can also be used to plan longwall extraction heights on a short-term scale with resulting improvements in productivity.

analyses. To allow the information to be computer modelled, additional information such as seam thickness, sample co-ordinates, and the elevation of the seam (or horizon) roof or floor, may be required. Channel sample spacing is partially controlled by the layout of the mine sections, and spacings of between 50 m and 200 m may be typical for various mines. Geostatistical techniques can be applied to select optimal sample configurations within these constraints (Nopp, 1994). Grade control practices using existing sampling data are generally quite adequate, except for monitoring mining-induced dilutions. In the case of a longwall operation mining the full-seam thickness, every one per cent of stone introduced into the ROM coal reduces an in situ SE of 24.70 MJ/kg by some 0.25 MJ/kg. While the overall dilution for the mine may be estimated by comparing run-of-mine (ROM) coal quality and in situ, undiluted quality predictions, the actual dilution types, dilution areas, and reasons for the dilution will remain unclear from such estimations. Personal experience has shown that this dilution can be readily measured and/or estimated in the underground situation, even for longwall panels where the mined-out area is rendered inaccessible in the goaf. The monitoring system involved is outlined below, together with some actual results for a longwall colliery.

DILUTION CONTROL
Waste dilution is unavoidable in a longwall mine extracting the full seam thickness. The dilution from continuous miner (CM) development may be planned in order to create operating head room. This paper is concerned with the unplanned inclusion of dilution in the ROM production. The sources of dilution may be divided into the following three main classes:

INTRODUCTION
Grade control, as practiced on any mine, is concerned with predicting routine quality and tonnage values for short- to medium-term mining operations, and the reconciliation of these predictions against actual results. For a longwall mine supplying steam coal, such routine parameters may include the proximate analyses, specific energy (SE), abrasiveness and sulphur, as well as float and sink analyses if the coal is beneficiated. In addition, grade control is concerned with identifying the mineable horizon (or horizons) in a selective-mining operation, together with monitoring how well these horizons have been extracted or adhered to. Close-spaced surface drilling can be costly for any mining venture, particularly for longwall mines operating in deeper seams. For this reason, much of the data for short-term grade control requirements are acquired during actual bord-and-pillar development. The underground data are derived from so-called channel samples. These samples are cut or chipped from the coal seam, perpendicular to the mining horizon, with dimensions suitable for obtaining sufficient representative sample for analyses. For a 2 m thick mining horizon, a channel sample 300 mm wide and 50 mm deep over the thickness of the horizon would provide some 4.5 kg of sample, sufficient for most of the necessary
1. MAusIMM, Snowden Mining Industry Consultants Pty Limited, PO Box 2207, Brisbane Qld 4001. E-mail: mnoppe@snowden.com.au

Primary dilution
Cutting of the stone floor or roof (accidental or planned) by the longwall shearer or CM machines.

Secondary dilution
Slabbing or break-up of the roof or floor during mining and tramming, and the subsequent loading of this material together with the coal (rather than being stowed in back areas).

Tertiary dilution
Waste material loaded with the coal during section-cleaning operations. Incorporation of waste material from the longwall gate roads (CM generated material) with the coal mined from the retreating longwall face. Dilution is not necessarily a problem, providing its amount and affect on ROM coal have been correctly predicted during feasibility study and mine planning. The mine, and the mines clients, would then be prepared for the diluted ROM coal, with supply contracts, and plant and machinery designed accordingly. However, experience shows that dilution from longwall and CM operations in many coalfields has been under-estimated by as

5th International Mining Geology Conference

Bendigo, Vic, 17 - 19 November 2003

243

M NOPP

much as five times. Actual monthly dilution levels have been known to run at over ten per cent, with levels more commonly in the order of four to six per cent (Barker, 1986; Chapman and Falcon, 1991; Nopp, 1992; Hookham, 2003). These under-estimations have occurred despite past experience and the application of sophisticated rock testing. Dilution is generally only considered a problem when the levels are sufficiently high to impact on a mines ability to meet its coal supply contracts. There are three main ways in which such problem contamination levels may be addressed, namely: 1. Install a coal treatment plant to separate the more-dense contaminants from the coal Depending on the volume of ROM coal to be treated, this could add some 20 per cent to the estimated capital cost of a new mine (Fourie, 1985). Washing may also result in significant losses of coal. For example, most of the ten per cent dilution may be removed together with ten to 20 per cent of the ROM coal. 2. Re-negotiate coal supply contracts to accommodate the lower-quality coal This may translate to finding a new client, particularly if the mines product is radically different to the initial estimates. Since this is not possible for a power station-tied mine, the options are essentially re-negotiate or reduce. 3. Reduce the amount of dilution at its source This involves firstly identifying the sources of dilution. This must be done with regard to the types and proportions of contaminants (roof, floor, waste, etc) and the distribution of dilution areas. Only once this has been achieved, can the mine consider ways of reducing the dilution at its source. Depending on the origin of the bulk of the dilution, methods of tackling the problem may include: education (awareness) of production personnel; redefining cutting and/or loading procedures; minor or major design changes to cutting and/or loading equipment. The remainder of this discussion is concerned with the control of dilution at its source. The method to be described is based on a system implemented by the author over a period of three years, and the results are typical of those actually achieved. Certain facts and figures have been modified to protect the confidentiality of the data. The discussion is concerned more with the acquisition and manipulation (interpretation) of data, rather than with details of how the results may actually be used to control dilution levels. The methods used to control dilution may differ from mine to mine, depending on the nature of the dilution. Detailed report formats are also not presented, as these too depend on the requirements of individual mines.

The measuring pattern must also take account of the CM operating procedure. In this example, the full width of a roadway or split (say 6 m to 7 m) cannot be cut during a single advance, and is therefore cut during two passes. This may result in a step cut into the roof and/or floor, particularly in dipping seams (Figure 1).

FIG 1 - Cross-section of a roadway mined by a continuous miner machine, through a gently-dipping coal seam.

DILUTION MEASUREMENTS
Some form of measuring and monitoring system is required to identify and quantify the sources and levels of dilution. This should include a procedure for checking (reconciling) such measurements or estimates. The following example is directly applicable to a singe-seam mining operation, where the total seam thickness is being extracted.

Board-and-pillar sections
The amount of stone or floor removed during CM mining can be directly measured in the mined-out areas. However, detailed measurements, of say 1 m intervals, are not possible from a logistical or financial point of view. Instead, a relatively wide-spaced measuring pattern must be designed which is capable of providing a reasonable estimate of the dilution for a given area.

CM and shuttle-car (loader) tramming may also result in the break-up of the floor behind the actual advancing face (potential secondary or tertiary dilution). This is common in the central, travelling way of roadways and splits, particularly at intersections, and must be accounted for by the measuring patterns. Measurements are taken perpendicular to the mined horizon (or seam section). The mined height, seam thickness and amount of floor and/or roof material cut or removed are measured at the side of roadways and splits (against solid coal pillars and barriers). Measurements in the centre of roadways are only of the mined height, together with a record of whether roof and/or floor material have been removed. The seam thickness for these mid-road measurements may then be estimated as the average of the two nearest side-road thickness measurements. The amount of floor and/or roof removed at each mid-road point is then calculated as the measured mined height less the estimated seam thickness. The spacing and pattern of the measurements required is determined largely by empirical methods (trial and error), aided by classical statistical analyses of the results. Geostatistical continuity analysis (semivariogram construction and modelling) may also be applied to determine optimal sample spacings (Nopp, 1994). A typical measuring pattern is illustrated in Figure 2, together with an example of a report and calculation form for CM-section monitoring. Primary dilution is estimated directly from the side-road measurements (Figure 2). This is possible because the area immediately against the solid-coal pillar or barrier is generally not affected by secondary floor break-up. Secondary dilution, derived predominantly from the central portion of the roadways, is readily calculated from the mid-road and intersection measurements (Figure 2). The relative proportions of primary and secondary dilution can be estimated providing the correct ratio of mid-road to side road measurements is used. A ratio of about 1:2, and the pattern shown in Figure 2, proved adequate for the given example. Spacing between measurements may be in the order of 10 m to 20 m. A typical three-road CM sections monthly advance of between 200 m and 300 m may then require some 100 to 200 measurements. These measurements may be recorded by one observer in a matter of a few hours, at a rate of some 30 measurements per hour. Height measuring may be facilitated by using readily-available electronic measurement tools.

244

Bendigo, Vic, 17 - 19 November 2003

5th International Mining Geology Conference

THE MEASUREMENT AND CONTROL OF DILUTION IN AN UNDERGROUND COAL OPERATION

SECTION: WEST DATE: 15 NOV

12 1 1 2 13 2 16 20 12 10 11 9 17 22 13 11 14 12 15 13 21 23 16 14 SPLITS 3 19 10 8 25 9 7 19 3 4 14 4 18

SPLIT No.: 12 - 15 NAME: G. DAY .

15 5 26 5 6

8 24

17

15

18

SIDE-ROAD MEASUREMENTS No. 1 2 3 ... 26 SUM / AVG. STD. DEV. CALCULATIONS AVG. SEAM THICKNESS Mined height 2000 2050 1950 ... 2150 2010 67 Floor/roof cut 0 F30 0 ... F100 60 25

MID-ROAD MEASUREMENTS No. Mined height 2050 2070 2000 ... ... 2210 2060 73 Floor/roof cuts F F+ R F ... F

= avg. Mined height avg roof/floor cut

= 2010 = 1950 - 60 mm (RANGE = 1800 to 2250 )

AVG. MINED HEIGHT

= ( SUMside + SUMmid ) / (No side + No.mid )

= (52260 = 2031 mm + 39140 ) / ( 26 +
19

) to 2300 )

(RANGE = 1800

PRIMARY DILUTION

= (avg.roof / floor cut) x (Rdstone) x 100 (avg.seam thickness) x (Rdcoal) =( ( 60 1950 ) x (2.30 ) x 100 ) x (1.50 )

= 2.83 % (DILUTION RANGE = 0 mm to 150 mm)

TOTAL DILUTION

= (avg.mined ht. avg.seam thick.) x (Rdstone) x 100 avg.seam thickness x (Rdcoal) = ( 2031 ( = 6.37 % - 1950 ) x (2.30 ) x 100 ) x (1.50 ) 1950

FIG 2 - Dilution measurement pattern and example report-from for a continuous minor board-and-pillar section.

5th International Mining Geology Conference

Bendigo, Vic, 17 - 19 November 2003

245

M NOPP

The large amount of regularly-spaced data generated from the measurements lends itself to classical statistical analysis and for storage and manipulation in spreadsheets. The various thickness and height values are typically normally-distributed and arithmetic averages therefore provide adequate estimates of the means of these values. Calculating the average seam thickness from the mean dilution and mean mined height (as shown in Figure 2), produces essentially identical results to those obtained if each measurement were first reduced to its individual components (namely mined height, seam thickness and amount of dilution). The observer could include additional features of geological interest on the section plan (Figure 2), such as floor and roof types, faults, joints, scours (washouts) and dykes. Some of these features are needed to properly interpret the dilution results, and the extra time spent gathering such data is worth the effort. The CM sections can be measured at any given time interval, depending on their individual rates-of-advance, and the manpower available. Fortnightly measurements may be reasonable from a logistical viewpoint, and would also provide dilution trends during the month, thus reducing the amount of measuring required at month-end. Production personnel, say the miners in each section, are encouraged to measure the primary contamination during each shift. A brief report of these results should be submitted to the geology section. The report format would be similar to that shown in Figure 2, except that measurements would only be along the road or split sides and would be closely spaced, at say 2 m intervals. The calculations could also be omitted, but a section for comments on general geological conditions could be included.

A simple and effective, although somewhat subjective, method for predicting seam thickness and dilution in a longwall panel is as follows:

On an overlay of the relevant section plans, plot the

measured seam thickness for individual points, as measured during dilution monitoring of the CM sections surrounding the longwall panel. longwall panels, making use of mapped trends for features such as floor and roof rolls (scours), and faults.

Contour the seam thickness data, extrapolating across the Superimpose the grid of mined heights over the seam

thickness contour plan and calculate the amount of stone cut at each grid-point (mined height less seam thickness). Reported comments of roof and floor cut (or loaded) may be used to update the thickness contour plan.

The CM-derived seam thickness measurements lend themselves to the application of computer-based estimations (and contouring) for within-panel seam thickness. In practice though, the extrapolation range of 200 m or more across the longwall panel may not produce realistic results and hand-contouring may prove more suitable.

General
Monthly dilution values are based on large amounts of data and generally return reasonable (and reconcilable) results. However, even on a daily or shift-level basis, the measurements and estimates can identify the source of a significant dilution problem. For example, if a routine ROM sample quality indicates an increase in dilution, the daily or shift reports from the production sections should include sufficient information to allow the origin of the problem to be identified. The success, or otherwise, of this system obviously depends on reliable measuring and reporting by production personnel. Knowledge and confidence of within-panel seam thickness and other geological features (faults, joints, floor rolls, etc), are important for longwall horizon control. This knowledge allows ideal mining heights across the face to be determined in advance, while reducing the risk of encountering unexpected geological difficulties.

Longwall sections
It is impractical, if not impossible, to routinely obtain direct measurements of the roof or floor cut across a longwall face. The coal face is generally unstable and not protected by the support shields, while mining would have to stop, and large volumes of coal be removed to expose the coal-floor contact. In addition, the mined-out area is inaccessible in the goaf. The dilution estimate is therefore less direct than for the CM sections. One parameter that can be measured routinely across the face is the mined height. This may be measured between chock shields (hydraulic support shields), at essentially any spacing or time interval. For example, the mined height may be measured at every tenth shield (15 m for 1.5 m wide chocks) at a set time during each shift, or after a specific face advance (retreat) of say every 15 m. The production personnel should be involved in these measurements. A series of such measurements would result in a grid of some 170 mined height values per 200 m advance (for a 200 m wide panel). As was the case for the mid-road measurements in the CM sections, the prediction of dilution requires a value for seam thickness at each grid-point. Since the seam thickness cannot be measured directly, it must be estimated from the available data. The data available for predicting seam thicknesses within a longwall panel include the detailed dilution measurements along the CM-mined gate-road developments, and occasional in-panel borehole intersections. Thickness predictions may also be checked against comments recorded when measuring the mined heights. If a mined height exceeds a predicted seam thickness at a given point, then the difference is inferred as the amount of stone (roof and/or floor) cut or removed. Production personnel may record whether roof and/or floor have indeed been cut (or broken) and loaded at each of the mined height measurement points. A typical recording sheet for a longwall section is illustrated in Figure 3. Obviously, independent checks of these measurements may occasionally be required to ensure the correct recording and reporting of the facts.

Reporting results (combined dilution report)

Individual sections dilution results should be recorded in such a way that a cumulative result is available at any time of the production month. The cumulative dilution estimate for each section may then be combined (tonnage weighted) to produce a dilution summary for the whole mine (Figure 4). Additional columns may be included in such a report to compare production personnel dilution measurements (primary dilution) to those determined by the geology section. A column of target dilution levels for each section may also be inserted. The total measured dilution for the whole mine may be reconciled by comparison with the dilution calculated from the ROM coal quality and the estimated in situ, undiluted coal quality. This comparison monitors the reliability of the measurements, providing the various measurement, analytical and estimation errors are known. Confidence limits can also be estimated and reported.

CASE STUDY Background

The mine where this dilution program was introduced was a 200 m deep, longwall operation producing more than six million tonnes per year from a single 2 m thick seam.

246

Bendigo, Vic, 17 - 19 November 2003

5th International Mining Geology Conference

THE MEASUREMENT AND CONTROL OF DILUTION IN AN UNDERGROUND COAL OPERATION

SECTION: LW1 DATE: 28 Nov

No. CHOCK No.

SPLIT(CHAINAGE): 1250m SHIFT: Morning (10 am)

MINED HEIGHT mm 2250 16 31 46 61 76 91 106 121 2130 2080 2030 2330 2180 2220 2250 2300 SEAM THICKNESS mm (completed by geologist.)

NAME: L. WALL
FLOOR/ROOF CUT OR LOADED
FLOOR CUT BY C.M. GRADING DOWN FROM C.M. COAL FLOOR COAL FLOOR - ROOF FALL COAL FLOOR FLOOR FLOOR

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

MAIN-GATE

TAIL-GATE

130

1950

FLOOR CUT BY C.M.

GENERAL CONDITIONS
ROOF FALL - 5m x 1m x 0.01m ( CHOCK 60-64) -LOADED INTO GOAF.

FIG 3 - Example of a dilution measurement report for a longwall section.

DILUTION SUMMARY REPORT MONTH END - NOV

SECTION TONNAGE MINED HEIGHT MEASURED FLOOR CUT PRIMARY mm TOTAL mm

DATE COMPILED

29-Nov

MEASURED CONTAMINATION PRIMARY % TOTAL %

DATE OF LAST TARGET MEASUREMENT DILUTION TOTAL %

28 NOV '91 TONS X 1000

mm

CM's AA BB CC DD CM TOTAL AVG. LW's LW1 LW2 TOTAL LWS TOTAL MINE 130 120 250 355 2200 2010 2109 2119 51 67 59 62 3.6 5.3 4.4 4.7 5.0 28-Nov 28-Nov 4.5 5.0 4.7 5.0 20 30 30 25 105 2010 1880 2270 2410 2142 55 37 50 21 40 96 75 82 28 70 4.3 3.1 3.5 1.3 3.0 7.7 6.4 5.7 1.8 5.4 16-Nov 17-Nov 26-Nov 27-Nov 6.4 7.1 6.0 2.5 5.6

DILUTION % CALCULATED FROM SPECIFIC ENERGY

FIG 4 - Example of a dilution summary report combining results for the sections of a mine.

5th International Mining Geology Conference

Bendigo, Vic, 17 - 19 November 2003

247

M NOPP

The mine had completed some two and a half years of CM development, before commissioning its first longwall. The average dilution estimated from SE during this period was 3.9 per cent, with 95 per cent of the monthly dilution levels below 5.7 per cent (Figure 5), compared to an expected average dilution of two per cent.
2 longwall Control system introduced 1 longwall
st nd

DILUTION LEVEL %

6 4 2 0
1 2 3 5 6 4 PRODUCTION YEAR 7 8

60

40 20 0

% LONGWALL TONNAGE

80

monthly dilution for the first three years of the systems operation was 5.7 per cent, with a trend towards four per cent dilution over the last five months. This represents a reduction in dilution of up to 50 per cent over the expected level if no action had been taken. The monthly average for the dilution measurements are compared to those calculated from SE in Figure 6. The average monthly difference between these two values is 0.88 per cent, with a trend for the measured dilution to be below the calculated values by some one per cent. Since the calculated values are considered actual levels, a measured value below the calculated value implies additional ROM dilution to that measured, which may well be expected from secondary or tertiary sources. The interpretation of these differences also depends on the reliability of the SE estimates and ROM sampling, as well as the reliability of the dilution measurements. In general, the combined monthly values for measured dilution have an associated error of about five per cent, while the error for poorly-calculated estimates may be in the order of ten per cent (at 90 per cent confidence).
10.0 9.0

DILUTION LEVEL %

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 5 6 7 8

FIG 5 - Graph of monthly dilution levels (as calculated from SE), illustrating the effect of longwalling, and the impact of the dilution control system (two month moving average values).

The high dilution levels were investigated and assigned primarily to weaker-than-anticipated floor conditions. Some measurements were taken, but these did not form part of a routine control system, possibly because ROM coal qualities still met the coal supply agreement stipulations and the area had better than average thickness and qualities. It was also hoped that dilution would improve, or at least remain static, with the introduction of longwalling. The first longwall panel operated for a year and during that time, the average monthly dilution increased to as high as nine per cent. The increase in dilution during this period appears to have been directly related with the increasing longwall tonnage (Figure 5). When a second longwall was brought into production, the dilution again increased to over eight per cent and this prompted the start of the dilution monitoring program. The average monthly dilution during the seven months prior to introducing the program was 7.7 per cent, with a high one month of 9.0 per cent (Figure 5). The effect of virtually doubling the dilution, from the initial monthly average of 3.9 per cent to 7.7 per cent, was to place the ROM coal outside of contractual specifications for more than only SE. Together with penalties for ash content and abrasiveness, the cost of the penalties was in the order of $A 1 million per year. The dilution measuring system, similar to that outlined above, was introduced in the fifth year of operation. The number of sections and rates-of-advance on the mine meant that additional manpower was required to run the program. The system operated with the full-time services of a geological observer (no formal geological training), and a portion of the time of a geological assistant (with formal geological training). The mine geologist maintained overall responsibility for the program and final reporting and follow-up of results.

PRODUCTION YEAR
CALC. FROM SPECIFIC ENERGY MEASURED

FIG 6 - Measured versus calculated monthly average dilution levels.

BENEFITS
The production from the initial longwall sections appears to have been largely responsible for the significantly increased dilution levels. The understanding of the dilution from the measuring system were used to improve longwall horizon control and planning, and this alone probably resulted in almost halving the longwall-sourced dilution. The control system also identified floor-stone material as responsible for some 90 per cent of total dilution, with 40 per cent to 60 per cent of this resulting from primary stone cutting. Production personnel awareness, combined with simple modifications or adjustment to cutting procedures and CM machines, were largely responsible for improvements in CM-related dilution levels. The overall better control and lower levels of dilution resulted in improved short-term planning, ROM qualities and fewer penalty payments, directly saving the mine up to $A 1 million per year. The detailed measurements generated by the program also provided a vast amount of additional geological information to improve ROM thickness and quality estimates and mine planning capabilities. For example, the seam floor could be divided into zones displaying particular behaviour with respect to both CM and longwall mining. Applying this knowledge to borehole core interpretations, allowed diluted, ROM coal qualities to be predicted with greater accuracy. The improved

RESULTS
The average monthly dilution level during the first year of the dilution systems operation was 5.8 per cent, with only one month exceeding 6.5 per cent. This was attained despite a continued increase in longwall tonnage (Figure 5). The average

248

Bendigo, Vic, 17 - 19 November 2003

5th International Mining Geology Conference

THE MEASUREMENT AND CONTROL OF DILUTION IN AN UNDERGROUND COAL OPERATION

understanding of dilution sources also allowed the problem of predicting non-linear parameters, such as abrasive index, to be addressed (Nopp, 1990; Chapman and Falcon, 1991).

REFERENCES
Barker, O B, 1986. An Overview of the Coal Reserve Position and Certain Technical and Geological Constraints to Coalfield Development in South Africa, South African Mining World, 5(9; 10):94-107; 57-71. Chapman, G and Falcon, R M S, 1991. The Abrasive Characteristics of the Gus Seam at Majuba Colliery (Pty) Limited, in Abstracts: Conference on South Africas Coal Resources: Their Origin, Characterization and Utilization, Nov 1991 (Geological Society of South Africa, Eastern Transvaal Branch: Witbank). Fourie, D B, 1985. An Investigation into the Coal Quality Requirements for Matla Coal Ltd, B Eng (Mining), thesis (unpublished), Pretoria University. Hand, P E, 1991. The Rationalization of Mine and Preparation Plant Output to Enhance the Exploitation of Coal Reserves in South Africa, Journal of the South African Institute of Mining and Metallurgy, 91(7):249-252. Hookham, M, 2003. NDC Smartens Up, Australias Longwalls, June, pp 8-10. Nopp, M A, 1990. Prediction of Abrasive Indices for Contaminated Coal Samples, paper presented at Workshop: Abrasiveness of Coal, Eskom, Megawatt Park, Johannesburg, November. Nopp, M A, 1992. Geological Controls for Coal Exploration and Mining, MSc Dissertation (unpublished), Rhodes University, Grahamstown. Nopp, M A, 1994. Practical Geostatistics for On-site Analysis A Coal Example, in Proceedings of the Mining Geostatistics Conference, pp 168-181 (Geostatistical Association of South Africa, Kruger National Park, South Africa).

CONCLUDING REMARKS
Geological control encompasses a wide range of functions in various fields, including mineral economics, mineral evaluation, exploration, mining and management. This control begins well-before the first borehole is drilled on a prospect or target, and continues through to the end of a mining-ventures life. The form these geological controls take tends to vary depending on the context in which they are applied. The emphasis during the early stages of planning and exploration is generally on making the right decisions, or doing the right job. This emphasis shifts as one moves into the realm of mining, and control during these phases is often directed at ensuring the job is done right or correctly. This is generally achieved by way of a variety of grade control systems. The success of many decisions or production-type controls in mining and exploration depend ultimately on the reliability of geological predictions and interpretations. Geologists must therefore ensure they provide the best-possible estimates and interpretations at all levels of the management-explorationmining continuum. The coal-related example presented here illustrates the application of some statistical techniques towards improving geological estimation and reporting capabilities. The dilution control system provides a simple, but effective method for monitoring and potentially reducing dilution on a longwall operation.

5th International Mining Geology Conference

Bendigo, Vic, 17 - 19 November 2003

249