state of the art - laubscher

7/27/2019 State of the Art - Laubscher

1/21

55

Cave Mining--The State of the Art

D.H. Laubscher*

INTRODUCTION

The expression "cave mining" will be used here to refer to all

mining operations in which the ore body caves naturally afterundercutting and the caved material is recovered through

drawpoints. The term encompasses block caving, panel caving,inclined-drawpoint caving, and front caving. Caving is the

lowest-cost method of underground mining, provided that

drawpoint size and handling facilities are tailored to suit thecaved material and that the extraction horizon can be

maintained for the life of the draw. 1

The daily production from cave-mining operations throughout

the world is approximately 370 kt/d. Table 55.1shows

production broken down by layout.

By way of comparison, the South African gold mines produce

350 kt/d.

Today, several open-pit mines currently producing in excess of50 kt/d are examining the feasibility of implementing low-cost,

large-scale underground mining methods. Several undergroundcave mines that produce high tonnages are planning to

implement dropdowns of 200 m or more. This will result in aconsiderable change in their mining environments. These

changes will necessitate more detailed mine planning, ratherthan the simple projection of current mining methods to greater


2/21

depths.

As more attention is directed to mining large, competent orebodies with low-cost underground methods, it is necessary todefine the role of cave mining. In the past, caving has beenconsidered for rock masses that cave and fragment readily. The

ability to better assess the cavability and fragmentation of ore

bodies, the availability of robust load-haul-dump (LHD)machines, an understanding of the draw control process,

suitable equipment for secondary drilling and blasting, andreliable cost data have shown that competent ore bodies with

coarse fragmentationl can be mined at a much lower cost using

caving rather than with drill-and-blast methods. However, once

a cave layout has been developed, there is little scope forchange.

Aspects that have to be addressed are cavability, fragmentation,

draw patterns for different types of ore, drawpoint or drawzonespacing, layout design, undercutting sequence, and support

design.

Table 55.2shows that there are significant anomalies in thequoted performance of different cave operations.

It is common to find that old established mines that havedeveloped standards during the course of successful mining in

the upper levels of an ore body are resistant to change and do

not adjust to the ground control problems that occur as miningproceeds to greater depths or as rock types change. Mines that

have experienced continuous problems are more amenable toadopting new techniques to cope with a changing mining

situation. Detailed knowledge of local and regional structural

geology, the use of an accepted rock mass classification to

characterize the rock mass, and knowledge of regional andinduced stress environments are prerequisites for good mine

planning. It is encouraging to note that these aspects are

receiving more and more attention.


3/21

The Laubscher rock mass classification system provides both in

situ rock mass ratings (IRMR) and rock mass strength. Such a

classification is necessary for design purposes. The IRMR defines

the geological environment, and the adjusted or mining rockmass ratings (MRMR) consider the effects of the mining

operation on the rock mass. Figure 55.1is a flow sheet of theMRMR procedure with recent modifications. The reader is

encouraged to read the paper "The IRMR/MRMR Rock MassClassification System for Jointed Rock Masses" by Laubscher andJakubee, which is included in this volume. The ratings describe

in detail cavability, subsidence angles, failure zones,fragmentation, undercut-face shape, cave-front orientation,

undercutting sequence, overall mining sequence, and supportdesign.

FACTORS AFFECTING CAVING OPERATIONS


4/21

The 25 parameters that should be considered beforeimplementing any cave mining operation are set out in Table

55.3. The parameters in capital letters are a function of theparameters that follow in the same box. Many parameters areuniquely defined by the ore body and the mining system and are

not discussed further. The parameters considered later arecommon to all cave-mining systems and need to be addressed if

any form of cave mining is contemplated.

CavabilityMonitoring a large number of caving operations has shown that

two types of caving can occur--stress and subsidence caving.

However, it is better to use the terms "vertical extension" to

mean upward propagation of the cave and "lateral extension" tomean the propagation of the cave as a caved block is expanded.

Vertical extension caving occurs in virgin cave blocks when the

stresses in the cave back exceed the strength of the rock mass.Caving can stop when a stable arch develops in the cave back.

The undercut must be increased in size or the boundariesweakened to induce further caving. High horizontal stresses

acting on steep dipping joints increase the MRMR. This was the

situation in block 16 at the Shabanie Mine. A stable back was

formed when the undercut had a hydraulic radius of 28 and theMRMR was 64. When block 7 (adjacent to block 16) caved, the

horizontal confining stress was removed, which resulted in a

reduction in the MRMR to 56 in block 16. At this point, caving

occurred.

Lateral extension caving occurs when lateral restraints on theblock being caved are removed by mining adjacent to the block.This often results in a large stress difference, leading to rapid

propagation of the cave and limited bulking.

Figure 55.2, based on a worldwide experience base, illustrates

caving and stable situations in terms of the hydraulic radius

(area divided by perimeter) for a range of MRMR values. An

additional curve has been added to account for the stability thatoccurs with equidimensional shapes.

All rock masses will cave. The manner of their caving and thefragmentation need to be predicted if cave mining is to be

implemented successfully. The rate of caving can be slowed by


5/21

control of the draw since the cave can propagate only if there isspace into which the rock can move. The rate of caving can be

increased by a more rapid advance of the undercut, butproblems can arise if an air gap forms over a large area. In thissituation, the intersection of major structures, heavy blasting,

and the influx of water can result in damaging airblasts. Rapid,uncontrolled caving can result in an early influx of waste.

In conventional layouts, the rate of undercutting (RU) should be

controlled so that the rate of caving (RC) is faster than the rateof damage (RD) due to abutment stresses. Thus, RC > RU > RD.

However, in areas of high stress, the rate of caving must be

controlled to maintain an acceptable level/amount of seismicactivity in the cave back. Otherwise, rock bursts can occur insuitably stressed areas (pillars and rock contacts). As advance

undercutting will be used in these situations, damage to the

undercut and production levels will not be a problem.

The stresses in the cave back can be modified to some extent bythe shape of the cave front. Numerical modeling can be a useful

tool for helping an engineer determine the stress patternsassociated with possible mining sequences. An undercut face

that is concave towards the caved area provides better controlof major structures. In ore bodies having a range of MRMRs, the

onset of continuous caving will be in the lower-rated zones if

they are continuous in plan and section. This effect is illustratedin Figure 55.3B, where the class 5 and 4B zones are shown to becontinuous. In Figure 55.3A, the pods of class 2 rock are

sufficiently large to influence caving, and cavability should bebased on the rating of these pods. Good geotechnicalinformation, as well as information from monitoring the rate of

caving and rock mass damage, is needed to fine-tune this

relationship.

Particle Size Distribution

In caving operations, the degree of fragmentation has a bearing

on

Drawpoint spacing

Dilution entry into the draw column

Draw control

Draw productivity


6/21

Secondary blasting and breaking costs

Secondary blasting damage

The input data needed to calculate the primary fragmentationand the factors that determine the secondary fragmentation as afunction of the caving operation are shown in Figure 55.4.

Primary fragmentation can be defined as the size distribution of

the particles that separate from the cave back and enter thedraw column. The primary fragmentation generated by

subsidence caving is generally more coarse than that resultingfrom stress caving. The reasons for this are (1) the more rapid

propagation of the cave, (2) the rock mass disintegrates

primarily along favorably oriented joint sets, and (3) there is

little shearing of intact rock. The orientation of the cave front orback with respect to the joint sets and the direction of principal

stress can have a significant effect on the primary

fragmentation.

Secondary fragmentation is the reduction in the size of the

original particles as they move through the draw column. Theprocesses to which particles are subjected determine the size

distribution which reports to the drawpoint. A strong, well-

jointed material can result in a stable particle shape at a low

draw height. Figure 55.5shows the decrease in particle size fordifferent draw heights and coarse (less jointed) to fine (well

jointed) rock masses. A range of RMRs will result in a wider

range of particle sizes than that produced by rock with a single

rating. This is due to the fact that the fine material tends tocushion larger blocks and prevents further attrition of these

blocks. This difference is illustrated in Figure 55.3B, in which

class 5 and class 4 material is shown to cushion the largerprimary fragments from class 3 material. A slow rate of draw

results in a higher probability of time-dependent failure as thecaving stresses have more time to work on particles in the draw

column.


7/21


8/21


9/21

Fragmentation is the major factor determining drawpoint

productivity. Experience has shown that 2 m3 is the largestblock that can be moved by a 6-yd3 LHD and still allow an

acceptable rate of production to be maintained. In Figure 55.6,the productivity of a layout using 3.5-, 6-, and 8-yd3 LHDs

loading to a grizzly is related to the percentage of fragments

larger than 2 m3. The use of secondary explosives is based on

the amount of oversized rock that cannot be handled by a 6-yd3LHD.

A computer simulation program has been developed forcalculating primary and secondary fragmentation. The results

obtained from this program are confirmed by underground

observations.

Drawzone SpacingDrawpoint spacings for grizzly and slusher layouts reflect the

spacing of the drawzones. However, in the case of LHD layoutswith a nominal drawpoint spacing of 15 m, drawzone spacing

can vary from 18 to 24 m across the major apex (pillar),

depending on the length of the drawbell. This situation occurs


10/21

when the length of the drawpoint crosscut is increased to ensurethat an LHD is straight before it loads. In this case, the major

consideration (optimum ore recovery) is compromised byincorrect use of equipment or a desire to achieve ideal loadingconditions. The drawzone spacings for 30 m center-to-center

production drift spacings are shown in Figure 55.7, with thecritical distance being across the major apex. It can be seen that

the length of the drawbell can be increased from 8 to 10 m andthe production drift spacing to 32 m without affecting the major

apex spacing. There still will be interaction in the drawbell.

Sand model tests have shown that there is a relationshipbetween the spacing of drawpoints and the interaction ofdrawzones. Widely spaced drawpoints develop isolated

drawzones with diameters defined by the fragmentation. When

drawpoints are spaced at 1.5 times the diameter of the isolateddrawzone (IDZ), interaction occurs. Interaction also improves as

drawpoint spacing is decreased, as shown in Figure 55.8. Theflow lines and the stresses that develop around a drawzone are

shown in Figure 55.9. The sand model results have been

confirmed by observation of the fine material extracted duringcave mining and by the behavior of material in bins. The

question is whether this theory, which is based on isolateddrawzones, can be wholly applied to coarse material, where

arching (spans) of 20 m have been observed. The collapse of

large arches will affect the large area overlying the drawpoint,as shown in Figure 55.10. The formation and failure of arches

lead to wide drawzones in coarse material, so that drawzone

spacing can be increased to the spacings shown in Figure 55.10.


11/21

The frictional properties of the caved material must also be

recognized. Low-friction material can flow greater distances

when under high overburden load, and this can mean widerdrawzone spacings. It is therefore logical to expect that when a


12/21

line of drawbells is drawn this creates a large low-density zone.Similarly, when lines of drawbells are drawn on alternate shifts

good interaction will result. The draw program at the HendersonMine broadly follows this pattern.

There is a need to continue with three-dimensional model tests

to establish some poorly defined principles, such as the

interaction across major apexes when the spacing of groups ofinteractive drawzones is increased. Numerical modeling could

possibly provide the solutions for the draw behavior of coarselyfragmented material.

Draw Control

Draw control requirements are shown in Figure 55.11and Table55.4. The grade and fragmentation in the dilution zone must beknown if sound draw control is to be practiced. Figure 55.12

shows the value distribution for columns A and B. Both columns

have the same average grade of 1.4%, but the valuedistribution is different. The high grade at the top of column B

means that a larger tonnage of waste can be tolerated beforethe shutoff grade is reached.


13/21

In LHD layouts, a major factor in poor draw control is thedrawing of fine material at the expense of coarse material. Strict

draw control discipline is required so that the coarse ore isdrilled and blasted at the end of the shift in which it reported inthe drawpoint.

It has been established that the draw will angle towards less

dense areas. This principle can be used to move the materialoverlying the major apex by creating zones of varying density

through the differential draw of lines of drawbells.

Dilution

The percentage of dilution is defined as the percentage of the

ore column that has been drawn before the waste materialappears in the drawpoint. It is a function of the amount ofmixing that occurs in the draw columns. Mixing is a function of--

Ore draw height

Range in fragmentation of both ore and waste

Drawzone spacing

Range in tonnages drawn from drawpoints

The range in particle size distribution and the minimumdrawzone spacing across the major apex will give the height of

the interaction zone (HIZ). This is illustrated in Figure 55.13.There is a volume increase as the cave propagates so that a

certain amount of material must be drawn before the cavereaches the dilution zone. The volume increase, or swell factors,

is based on the fragmentation and is applied to column height.Typical swell factors for fine, medium, and coarse fragmentation

are, respectively, 1.16, 1.12, and 1.08.


14/21

A draw control factor is based on the variation in tonnages from

working drawpoints (Figure 55.14). If production data are notavailable, the draw control engineer must predict a likely draw

pattern. A formula based on the above factors has beendeveloped to determine the dilution entry percentage.

Dilution entry = (A - B)/A #8734;C #8734;100,

where A = draw-column height H swell factor

B = height of interactionand C = draw control factor.

The graph for dilution entry was originally drawn as a straightline, but underground observations show that, where the

dilution occurs early, the rate of influx follows a curved line with

a long ore "tail" (Figure 55.15). Figure 55.16shows that dilutionentry is also affected by the attitude of the drawzone, which canangle towards higher overburden loads.


15/21

Layouts

Eight different horizontal LHD layouts and two inclined drawpoint

LHD layouts are used at various operating mines. An example ofan inclined LHD layout is shown in Figure 55.17. The El Teniente

layout is shown in Figure 55.18and the Henderson Mine layout

in Figure 55.19. Numerical modeling of different layouts showedthat the Teniente layout was the strongest and had severalpractical advantages. For example, the drawpoint and drawbell

are on a straight line, which results in better drawpoint support

and flow of ore. If there is brow wear, the LHD can back into theopposite drawpoint. The only disadvantage is that the layout

does not have the flexibility to accommodate the use of electric

LHDs such as does the herringbone.


16/21

A factor that needs to be resolved is the correct shape of the

major apex. It is thought that a shaped pillar will assist in therecovery of fine ore. Also, in ore bodies characterized by coarse

fragmentation, there is less chance that stacking will occur(Figure 55.20). The main area of brow wear is immediately

above the drawpoint. If the vertical height of the pillar above thebrow is small, failure of the top section will reduce the strength

of the lower section and result in aggravated brow wear.


17/21

More thought must be given to the design of LHD layouts to

provide a maximum amount of maneuver room within aminimum size of drift opening. Thus larger machines can be

used within the optimum drawzone spacings. Another aspectthat needs attention is LHD design, that is, to reduce the length

and increase their width. While the use of large machines mightbe an attraction, it is recommended that caution be exercised

and that a decision on machine size be based on a correctassessment of the required drawzone spacing in terms of

fragmentation. The loss of revenue that can result from dilution

far exceeds the lower operating costs associated with largermachines.

Undercutting

Undercutting is one of the most important aspects of cave

mining since not only is a complete undercut necessary toinduce caving, but the undercut method can reduce thedamaging effects of induced stresses.

The normal undercutting sequence is to develop the drawbell

and then to break the undercut into the drawbell, as shown inFigure 55.18. In environments of high stress, the pillars andbrows are damaged by the advancing abutment stresses. The

Henderson Mine technique of developing the drawbell with long

holes from the undercut level reduces the time interval andextent of damage associated with postundercutting. To preserve

stability, the Henderson Mine has also found it necessary todelay the development of the drawbell drift until the bell must

be blasted (Figure 55.19).

The damage caused to pillars around drifts and drawbells byabutment stresses is significant and is the major factor in brow

wear and excavation collapse. Rock bursts also occur in theseareas. The solution is to complete the undercut before

development of the drawpoints and drawbells. The "advanced


18/21

undercut" technique is shown in Figure 55.21.

In the past, it was considered that the height of the undercut

had a significant influence on caving and, possibly, the flow ofore. The asbestos mines in Zimbabwe had undercuts of 30 m

with no resulting improvement in caving or fragmentation. Thelong time involved in completing the undercut often led to

ground control problems. Good results are obtained withundercuts of minimum height, provided that complete

undercutting is achieved. Where gravity is needed for the flow ofblasted undercut ore, the undercut height needs to be only half

the width of the major apex. This results in an angle of repose of45#176; and allows the ore to flow freely.

Support RequirementsIn areas of high stress, weak rock will deform plastically andstrong rock will exhibit brittle, often violent, failure. If there is a

large difference between the RMR and MRMR values, yielding

support systems are required. This is explained in Figure 55.22.


19/21

Prestressed cables have little application in underground

situations unless it is to stabilize fractured rock in a low-stressenvironment. The need to constrain the rock laterally and for

lining surfaces such as concrete cannot be too highlyemphasized. Support techniques are illustrated in Table 55.5.

The use of cable bolts in brows is common practice, but oftenthese bolts are not installed in a pattern that takes into account

joint spacing and orientation. Cable bolts should not be installed

in highly jointed ground, as the bolts do not apply lateralrestraint to the brow and serve only hold blocks in place.


20/21

CONCLUSIONS

Cavability can be assessed provided accurate geotechnical dataare available and geological variations are recognized. The

MRMR system provides the necessary data for an empiricaldefinition of the undercut dimension in terms of the hydraulic

radius.

Numerical modeling can assist an engineer in understanding anddefining the stress environment.

Fragmentation is a major factor in assessing the feasibility of

cave mining in large, competent ore bodies. Programs are beingdeveloped for predicting fragmentation, and even the less-

sophisticated programs provide good design data. The economicviability of caving in competent ore bodies is determined by LHD

productivity and the cost of breaking large fragments.

Drawpoint and drawzone spacings for coarser material need tobe examined in terms of recovery and improved mining

environments. Spacings must not be increased to loweroperating costs at the expense of ore recovery.

The interactive theory of draw and the diameter of an isolated

drawzone can be used in the design of drawzone spacings.

Complications occur when drawzone spacings are designed onthe basis of primary fragmentation and the secondary

fragmentation is significantly different.

ACKNOWLEDGMENTS

This paper presents an update of the technology of cave mining.

It is not possible to quote references since the bulk of the datasupporting the contents of this paper have not been published,and the basic concepts are known to mining engineers.

However, it is appropriate to acknowledge the contributions

from discussions with the following people in Canada, Chile,South Africa, and Zimbabwe: R. Alvarez, P.J. Bartlett, N.J.W.Bell, T. Carew, A.R. Guest, C. Page, D. Stacey, and A. Susaeta.

The simulation program for the calculation of primary and

secondary fragmentation was written by G.S. Esterhuizen at


21/21

Pretoria University.

1. Bushman's River Mouth, South Africa.

state of the art - laubscher

Documents