J.S. Atr. Inst. Min. Metal/" vot 91,no. 7. Jut 1991. pp. 221.231.

Surface effects of total coal-seam extractionby underground mining methods*

by H. WAGNERt and E.H.R. SCHOMANNt

SYNOPSISThe paper gives a detailed overview of the history of subsidence engineering on South African collieries.

Surface displacements resulting from total-seam extraction are discussed, and the important role of localgeology is highlighted. It is shown that the differences in subsidence behaviour between South Africanand European collieries can be attributed to the presence of competent strata in the overburden of SouthAfrican coal seams. Criteria are given for the prediction of the most important subsidence parameters, anda new model for the simulation of surface subsidence above total-extraction panels is introduced.

The final section of the paper deals with the damage to surface structures caused by total-seam extraction.A brief summary of the most important undermining experiments on South African collieries is given, andit is concluded that local experiences of damage compare with those in Europe.

SAMEVATTINGDie referaat gee 'n uitvoerige oorsig oor die geskiedenis van insakkingsingenieurswese by Suid-Afrikaanse

steenkoolmyne. Oppervlakverskuiwings as gevolg van die totale uitmyning van 'n laag word bespreek endie belangrike rot van plaaslike geologie word beklemtoon. Daar word getoon daat die verskille tussen dieinsinkingsgedrag van Suid-Afrikaanse en Europese steenkoolmyne toegeskryf kan word aan dieteenwoordigheid van sterk lae in die deklaag van Suid-Afrikaanse steenkoollae. Daar word maatstawweaangegee vir die voorspelling van die belangrikste insinkingsparameters en 'n nuwe model word voorgestelvir die simulasie van oppervlakinsinking bo totaal uitgemynde panele.

Die laaste deer van die referaat handel oor die skade aan bogrondse strukture as gevolg van die totaleuitmyning van steenkoollae. Daar word 'n kort opsomming van die belangrikste ondermyningseksperimenteby Suid-Afrikaanse steenkoolmyne gegee en die gevolgtrekking word gemaak dat die plaaslike ondervindingvan skade met die in Europa vergelyk kan word.

INTRODUCTIONIt is now generally accepted that South Africa's coal

reserves, although enormous by any standards, are limitedand that every effort must be made to extract coal seamsas completely as is practicable. Apart from strategic andreserve reasons, there are economic and operationalreasons that favour total-seam extraction. The trendtowards more complete extraction of the country's coalreserves is clearly evident from the changes in miningmethods that have taken place over the past twenty years.Much of the growth of the South African coal-miningindustry can be attributed to the introduction of total ex-traction methods, either in the form of opencast surfacemining or in the form of pillar extraction and longwallmining. There is every reason to assume that this trendwill not only continue, but, in fact, intensify.

A specific feature of total-extraction underground coal-mining methods is that, when the coal seam is extracted,the support to the overlying rock strata is removed andthe strata are allowed to subside. The mining-inducedground movements not only change the surface topo-graphy but introduce strains in the sub-surface. There cantherefore be no question that total seam extraction can

*Paper presented at the School on the Total Utilization of CoalResources, which was held by The South African Institute of Min-ing and Metallurgy at Witbank in October/November 1989.

t Chamber of Mines of South Africa, P .0. Box 61809, Marshalltown,2107 Transvaal.

@ The South African Institute of Mining and Metallurgy, 1989. SAISSN 0038-223X/3.00+0.00.

have a significant influence on the land surface and itsuse.

The purpose of this paper is to briefly review the stateof the art of subsidence engineering in South Africa andto identify potential problem areas.

SUBSIDENCE ENGINEERING IN SOUTH AFRICANCOLLIERIES

Subsidence engineering can be defined as the predic-tion of mining-induced surface movements and the as-sessment of how these movements affect the surface andsurface structures. The discipline of subsidence engineer-ing was developed in the coal-mining districts of GreatBritain and Europe during the second half of the last cen-tury. Methods for the prediction of subsidence weredeveloped 1-3 so that estimates could be made of thedegree and extent of ground settlement and damage tostructures that were likely to occur as a result of exten-sive coal-mining operations in built-up areas. One of themain motivations for the development of subsidence en-gineering in Great Britain and in Europe was a need forthe realistic and meaningful allocation of compensationcosts for subsidence damage to individual mines. Thisneed was particularly great where mining took place indensely populated areas and where neighbouring minesextracted coal seams.

Subsidence engineering in Great Britain and Europedeveloped into a highly specialized and recognized dis-cipline, and is seen by many as the forerunner of modernstrata control. Today it has reached a very advanced stage

JOURNAL OF THE SOUTH AFRI(;AN INSTmJTE OF MINING AND METALLURGY JULY 1991 221

and is an integral part of coal-mine planning. Majorcities, roads, waterways, and harbours have been under-mined at greater depth with no danger to the populationand with minimal damage to the surface structures.

Subsidence engineering in South Africa is a relativelynew discipline. The reasons for this are straightforward.Until fairly recently, bord-and-pillar mining was thealmost exclusive method of coal extraction from under-ground. Provided the pillars are designed to adequatesafety factors, the surface movements associated with thismethod of mining are insignificant and generally muchless than the seasonal movements due to the swelling ofsoils4,s. A second, and equally important, reason for thelate emergence of subsidence engineering in South Africais the fact that most coal-mining operations took placein the less populated areas of South Africa and, in thecase of the Natal, in mountainous regions. For thesereasons very few subsidence measurements were taken.Heslop6 and Steart 7, reporting subsidence observationsabove pillar-extraction panels (stooping panels) in Natal,noticed that local coal-mine subsidence did not conformto the European and British subsidence models. In par-ticular, the maximum subsidence above the centre of thepillar-extraction panels was less than that observed over-seas, and the limit angle was steeper.

Systematic subsidence observations started with the in-troduction of longwall mining in South Africa in the1960s. While much of the earlier work was concerned withthe behaviour of dolerite sills above total-extractionpanels, Salamon et al.8, Galvin9, and, more recently,SchiimannlO-1S, Wagner and Schiimanns, Jack et al. 16,JackJ7, and Jack and Schiimannl8 were concerned withthe effects of subsidence on the surface and on surfacestructures. Among others, the effects of total seam ex-traction on farm dwellings, different types of roads, rail-way lines, and power lines were studied. Most of thepresent knowledge on mining subsidence is confined tosingle-seam extraction. Only at Durban Nagivation Col-lieries in NataI19,2o, and at Sigma Collieries in the VaalBasin21-23,have subsidence observations been carried outin connection with single- and double-seam extraction.

The results of all the subsidence work done to date in-dicate that there are significant differences between thesubsidence behaviour of South African coal measures andthat of European coal measures. These differences canbe attributed to the generally shallower coal seams, andparticularly to the more competent rock strata, found inSouth African coalfields.

As a result of these differences the models for subsi-dence prediction developed for British and Europeancoal-mining conditions have been found unsuitable forlocal conditions. Recently, Schiimann and Hamenstaedt24developed a composite subsidence-prediction model thatseems to be well-suited to local conditions.

From the brief history of subsidence work on SouthAfrican coalfields, it is evident that much progress wasmade in recent years in the study of coal-mine subsidence,its effect on surface structures, and its prediction. Themost important findings of this work are summarizedbelow.

EFFECTS OF MINING ON THE SURFACE AND SURFACESTRUCTURES

The effects of mining on the surface and on surface

222

structures can be divided into two separate but equallyimportant components:

(i) the prediction of mining-induced surface displace-ments, and

(ii) the response of the surface and surface structuresto these displacements.

The sub-division of the problem into two componentsis of great practical importance since it makes it possibleto take full advantage of experiences gained elsewhere.While mining~induced surface displacements are greatlydependent on the local situation (Le. the local geologicalconditions, the depth of mining, the thickness and angleof dip of the extracted coal seam, the type of supportused, and the length and width of the extraction panel),it should be noted that the response of the surface andsurface structures to mining-induced surface displace-ments depends largely on the type of structure and thenature of the land surface. Since similar structuresrespond to given surface displacements in a similar man-ner, it is possible to transfer experiences gained in othermining areas, and even in other countries, to the assess-ment of local subsidence problems. This aspect is of thegreatest practical significance since it means that use canbe made of the wealth of experience that has been gainedoverseas. Accordingly, it is proposed to deal with thetopic of subsidence engineering under the above twoheadings.

Mining-induced Ground DisplacementsDuring an underground excavation, displacements are

induced in the rock mass, and their magnitude dependson the excavation geometry, the stress field in which theopening is excavated, and the physical properties of therock mass. Often the displacements induced by miningexcavations are not noticed on the surface, but there aremany situations when surface displacements as a resultof mining are noticeable or even damaging.

Ground displacements consist of absolute movementssuch as vertical subsidence, horizontal displacements, anddifferential movements. The last named are relative dis-placements of various points of the surface in respect toone another. In the case of relative horizontal displace-ments, adjacent surface points can either come closertogether, resulting in a compressive strain, or can movefurther apart, resulting in a tensile strain. The horizon-tal strains in different directions are frequently of differ-ent magnitude or even different signs, leading todistortion of the surface. Similarly, the differential ver-tical movements of adjacent surface points will result ina change in the slope of the surface. Since a change inslope is particularly critical for tall buildings, the termtilt is frequently used in subsidence engineering.

Experience has shown that, in most instances, abso-lute ground displacements do not cause damage to sur-face structures. However, vertical ground movements,even if they are uniform over considerable distances, cancause inconvenience or give rise to the formation of pans.Relative ground displacements, on the other hand, canseriously affect surface structures, resulting in tensioncracks or the buckling of surface structures, which is de-pendent on the magnitude of the horizontal strains andwhether these are tensile or compressive.

JULY1991 JOURNAL OF 11-IE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY

Uniaxialcompressive Flexural

strength strength BulkingRock type MPa MPa factor, ki

Fine 70-120 5Sandstone Medium 50-70 3 1,3-1,5

Coarse 30-50 2Shale 40-80 2 1,1-1,2Dolerite 250-390 10 ?Coal 15-40 1,5 1,3

In subsidence engineering, it has become accepted prac-tice to refer to absolute and relative surface displacementsas follows:

Absolute ground displacements

Vz Vertical subsidence (m)Vz maxMaximum vertical subsidence (m)

(Vz maxusually occurs above the centre of amined-out area)

VxVy Horizontal displacement (m)

Relative ground displacements

ex,Ey Horizontal strain in the x and y directions respec-tive (mm/m)Ground curvature (Um)Change in slope (mm/m).

."T

Fig. 1 shows a vertical section through the centre lineof a very wide total-extraction panel. The face of the ex-traction panel advances to the left. The various elementsof mining-induced ground movements are shown inschematic form. The position and magnitude of thehorizontal and vertical ground movements, and thehorizontal strain and tilt, depend, among other things,on the local geological conditions, the ratio between theextracted mining height, h, and the depth below surfaceof the seam, H, and the ratio between the width of thetotal-extraction panel, W, and its depth below surface, H.

Strain "'."Vx Horizontal

',.disPlacement

+ X

Subsidence

//

//

I

+emax

. (compression)

I "I ,I ......

xE ,]

I

:Tmax

I

Tilt

i

H

F

\ I /\ : /\ I /\ I /

Vzmax

legend: A Trough limitB Point of maximum tensile strainC Subsidence completeD Transition point TP

E Point of maximum compressionF Face position

'Y Influence angle or angle of draw{3 Angle of break

'"Tangent angle

Fig. 1-Schematlc model of the various elements of mining.Induced ground movements

Effect of GeologyAmong the most important factors influencing mining-

induced surface displacements are the strength proper-ties of the various rock strata that overlie the coal seams.Since the formation of coal seams is a sedimentaryprocess, the thickness of individual rock strata tends tobe fairly uniform over considerable distances. However,the strength of individual strata can vary greatly depend-ing on the rock type and the thickness of the strata. Thedominant rock formations in South African coalfields aresandstones, ranging in grain size from more than 1mm(coarse-grained) to a few micrometres (fine-grained), andshales. The shale bands often contain carbonaceousmaterial and are usually thinly bedded. From the pointof view of strata control, relatively massive, fine-grainedsandstone layers can bridge spans of many tens of metreswithout failing, whereas the self-supporting spans of thin-ly laminated shales are generally much less than 10m inlength.

Apart from the differences in strength properties, thereare other significant differences between sandstone andshale formations. When sandstone layers fail, they tendto break up into comparatively large blocks with manyvoids between the blocks. Sandstone formations have,therefore, good bulking properties, which is an impor-tant factor as far as surface subsidence is concerned. Incontrast, shale formations tend to fail in thin slabs, result-ing in tightly packed caved material with poor bulkingproperties25.

A special feature of many South African coalfields isthe presence of igneous intrusions that intersect the coalseams as near-vertical dykes or that form massive sillsabove or below the coal seams. These dolerite dykes andsills have exceptional strength properties when un-weathered, but can be very weak in their decomposedstate close to the surface. Dolerite sills of 30 or 40m inthickness can span several hundred metres without fail-ing and, by doing so, prevent surface subsidence.However, the longterm stability of undermined sills canbe critical, and the possibility of surface movements tak-ing place many years after the completion of mining can-not be ruled out.

Table I summarizes the more important strength andbulking properties of the major rock formations foundin South African coalfields.

TABLEITYPICAL STRENGTH AND BULKING PROPERTIES OF ROCK

FORMATIONS IN SOUTH AFRICAN COALFIELDS

An estimate of the self-supporting spans of differenttypes of rock strata can be obtained from equation (1):

(Ux)max = p.g'L2(1)

2t

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METAu.uAGY JULY1991 223

where p = density of the rock strata (kg/m3)g =gravitational acceleration (9,81 m/s2)L = unsupported span (m)t = thickness of strata (m).

Based on the values of flexural strength given in TableI, the relationship between strata thickness and self-supporting spans for the different rock strata found inSouth African coalfields is shown in Fig. 2. The simplemodel, equation (1), holds remarkably well, in a com-parison of the calculated spans with observed spans atwhich the strata failed in total-extraction panels9,lO,2S.

From the significant differences in the strength proper-ties of various rock strata and their influence on the self-supporting spans of rock beams, it is clear that the localgeology plays an important role in the subsidence process.Careful consideration has therefore to be given to thespecific geological features of the coal-mining district dur-ing the prediction of the surface subsidence caused bytotal coal-seam extraction.

The role of dolerite sills on surface subsidence abovetotal-extraction areas deserves particular attention. Fieldobservations and theoretical studies have shown thatstrong dolerite sills behave like elastic plates close to the~oint of failures. Based on these findings and the analy-SISof numerous total-extraction panels beneath massivedolerite sills, Galvin9 developed a semi-empirical modelfor the prediction of the mining span, Serih required toinduce failure of the sill:

t2Serit= 1165 td-935

~d + 2tp tan (f3-90),"""

(2)

where td = thickness of dolerite sills (m)Dd =depth of base of dolerite sill below surface

(m)tp = thickness of parting between sill and coal

seam (m)(3 = caving angle of parting measured from the

coal seam ahead of the working face e).10080 FLEXURAL STRENGTH:6 DOLERITE :10MPa

4 SANDSTONE:5SHALE: 2COAL: 1.5

20

~ 10E I!~

6

~ 4wz

~ 2I....

1~ 8~ .g: .6If'I

.4

.2

0.11 2 I. 6 81:1 20 l,() 60 100 200 1.00 1000

SELF-SUPPORTING SPAN (m)

Fig. 2- Typical self.supportlng spans for the rock formationsfound In South African coalfields

224 JULY 1981 JOURNAL OF THE SOUTH AFRICAN INSTI11JTE OF MINING AND METALLURGY

Surface subsidence above total-extraction panels whosewidth is smaller than the critical span given by equation(2) is governed by the elastic deflection of the doleritesill. A systematic analysis of subsidence above longwallpanels of subcritical dimensions in the top seam at Dur-ban Navigation Collieries (DNC) by Schiimann2o hasshown that the maximum subsidence in the centre of atotal-extraction panel is given by

Vz max = 0,04 (S/Dd)2-0,06(m), (3)

where S is the width of the total-extraction panel.

In cases where S exceeds the critical width Serit it wasI: 12 '

,

lound that the vertical and horizontal surface displace-ments in the region of the first break of the dolerite sillwere larger than anywhere else"'inthe subsidence trough.Furthermore, the presence of dolerite sills resulted in sub-sidence troughs that were narrower and shallower13 thantroughs calculated by use of the subsidence model deve-loped by the National Coal Board (NCB) in Great Bri-tain. The effect of dolerite sills on the maximum amountof surface subsidence above total-extraction panels isclearly evident from Fig. 3, which compares 14local sub-sidence data with the NCB subsidence model. To accountfor the important influence of the width of extractionpanels, the subsidence data were plotted against the ratioof panel width to mining depth (WIll). Two importantobservations can be made. First, in the presence ofdolerite sills, the panel widths have to be much greaterthan on British coalfields for significant subsidence to oc-cur. Second, the maximum amount of surface subsidenceis only about two-thirds of that observed on British coal-fields. The reasons for these differences are related en-tirely to differences in the geological conditions.

~ 1.0 VzmaxE -~ hg 0.8

w~ 0.6w0Ui 0.4m::J(/)

0.2x

i DJ0

H =200m.(sin le seam I

H = 100m~C<a 0--'-

%5" ~,..-'--,/' /00.6'

/ ./ K""'""

./ ..-o,~( ~e , ./' ~

r ,;w e e, /,,/, ~e

,,/0 ,/. not f.il.d/./..-'

0:2 0'.4 0'.6 0'.8 1:0 1:2 1'.4 1'.6 1.8 2.0 2.2Panel Width / Depth. W I H ,

~gend. 1%Dolerite.Sondstonel- 0124. 331: 8136.501: 6134.371; 0105.541: LONGWALLs

010. 831:~(35.37L RIB PILLAREXTRACTIONs

..

Fig. 3-Maxlmum subsidence versus ratio of panel width todepth for 8lgma, Brandsprult, and Coalbrook Collieries

The presence of massive, strong sandstone beds in thestrata overlying coal seams has a similar effect on sur-face subsidence to that of massive dolerite sills. However,in quantitative terms the differences are less pronouncedbecause of the lower strength of the sandstone beds.

Mention must be made of the effects of vertical or near-vertical dolerite dykes on the process of subsidence. Thecontact of a dolerite dyke with the surrounding strataoften constitutes a zone of weakness. Consequently, thestrata, as a result of total coal-seam extraction, will tendto move along the contact area. The probability for thisis greater in cases where the dyke runs parallel or at very

acute angles to the panel abutment. Localized departuresfrom otherwise normal subsidence behaviour can be ex-pected in these instances, and have to be taken into con-sideration in the evaluation of the effects of total-seamextraction on surface structures. Fortunately, manydolerite dykes tend to weather when exposed to water andthe atmosphere, and lose much of their strength close tothe surface. As a result, the changes in surface subsidencedue to the presence of these dykes. are not as abrupt aswith strong, unweathered dykes.

From this discussion of the effects of local geology onsurface subsidence above total-extraction panels, it is clearthat this aspect requires careful attention. An integral partof subsidence engineering is therefore the evaluation oflocal geological conditions and their assessment from asubsidence point of view. Where total-extraction coalmining has already taken place, this assessment is rela-tively simple and can often be confined to a comparisonof borehole logs from panels that had been extractedpreviously with panels that have been planned and aredue for total extraction. The main purpose of this com-parison is to ensure that there are no sudden or signifi-cant changes in the composition of the overburden stratanor in the thickness of the overlying dolerite sills.

The assessment of the effects of local strata conditionson surface subsidence is more difficult for areas wherethere is no local experience oftotal extraction. However,over the years much experience has been gained and muchinformation collected concerning total coal-seam extrac-tion under a variety of geological conditions. This makesit possible for the information gained in other areas to

Vl!lU!ENICINC. SI>.SOLBURC SOlJTI{ RAND

be used for the purpose of subsidence prediction in coal-mining areas where no total-extraction experience exists.A critical examination of Fig. 3 shows a remarkable con-sistency in subsidence behaviour as a result of strata com-position. It should be noted that longwall panels withsimilar strata composition, as indicated by the propor-tion of dolerite and sandstone in the overburden strata,show similar subsidence values when allowance is madefor the important influence of the ratio of panel widthto mining depth.

Fig. 4 gives an overview of the typical stratigraphy ofthe major South African coalfields. Detailed subsidenceinformation is available for the mines marked with anasterisk.

Effect of Depth of Mining on Surface SubsidenceThe depth of mining affects surface subsidence in sever-

al ways.Firstly, as total-extraction mining starts, the mining

span increases until a point is reached where the self-supporting span of the immediate roof beam is exceededand the beam fails. With further increases in mining span,this process continues, and more and more roof strataparticipate in the caving process. An important aspect ofthe caving process is that the caved material bulks andgradually fills the void created by the seam extraction.Once this point has been reached, the caved material startsexerting some resistance to further strata movement. Theresistance is initially low, but increases rapidly as thecaved material becomes more and more compressed bythe weight of the overlying strata. Ultimately a point of

OOAIBROOK SPRlNCFlElD

HlCllVElD

NEW DENMARK BIlANDSPRUIT VRYHEIDSlCMA CORNELIA

,', ~E~ij

RI

~:

jj

:~;.~~

-:100.00

-3\0.00

E.TVL

MAjUIIA

KUPRlVER UfRECIIT & VRYHEID

DNC ufRECIIT

ij -

~~

.'

~!

.!

.;:.::

i~:.

~j::::::::::

at

-

mSOIU>

~ SHALE/SILTSTONB

0 SANDSl'ONE

I2J DOLEIIITE

. COAL

Fig. 4- Typical stratlgraphy at collieries conducting total extraction

JULY1981 225JOURNAL OF THE SOUTH AFRICAN INST11UTE OF MINING AND METAWJRGY

equilibrium is reached where the resistance of the cavedbut compressed roof strata equals the weight of the over-burden strata. Obviously, the greater the depth of min-ing, the more compressed will the caved material becomeand the greater will be the surface subsidence.

The above process has been described in detail in Sala-mon's model of strata reconsolidation26. According toSalamon, the initial height of caving, he, is given by

Pe ki"f'hhe = - . [exp -1], , (4)

ki"Y Pe(ki -1)

where Pe = a factor describing the compressibility ofthe caved material,

ki = bulking factor,

'Y = specific weight of the caved material, andh = height of the extracted seam.

For cases where the depth of mining, H, exceeds thecaving height, he, the normalized surface subsidence,Vz, is given byh

Vz= 1-(ki-l)'Pe

h ki'Y. hki'Y'H+Pe. loge

k p'(5)

i'Y(H - he) + e

Wagner and Schiimann5 have shown that this modeldescribes surface subsidence above wide total-extractionpanels. As far as the parameters ki and Pe are concerned,the following typical values have been found for SouthAfrican coalfields:

Pekshaleksandstone

3-5 MPa1,21,4-1,5.

Secondly, the ratio of the extracted seam height, h, tothe depth of mining, H, is critical as far as the mode ofsurface subsidence is concerned. Salamon and Oravecz27have defined three distinctly different modes of subsi-dence depending on the value of the ratio, hi H, namely:

(1) discontinuous subsidence (hIH>O,I),(2) subsidence with surface cracks (0,1> hlH> 0,01),

and(3) smooth subsidence (hlH < 0,01).

Since most South African coal seams have a thicknessof more than 2 m and are situated at depths of less than300 m, it follows that subsidence with surface cracks isthe dominant mode of subsidence in South African coal-fields. In the case of thick, shallow coal seams, discon-tinuous subsidence with major surface cracks occurs;also, at instances pronounced surface steps occur abovethe edge of extraction panels.

This mode of surface subsidence is likely to becomemore important for the total extraction of more than onecoal seam. In this instance, the combined extractionheight of all the extracted seams is the critical parameter.Smooth surface subsidence will gain in importance as thedepth of coal mining increases. First examples of thismode of surface subsidence were seen at Durban Navi-gation Colliery and at New Denmark Colliery near Stan-derton.

From the point of view of surface damage, the effectsof total-seam extraction become less and less significant

as the ratio hlH decreases. In deep South African goldmines, this ratio is less than 0,001, and damage to sur-face structures as a result of subsidence is virtuallyunknown.

Thirdly, the width of total-extraction panels, W, rela-tive to the depth of mining, H, is a critical parameter fromthe point of view of surface subsidence. When the widthof total extraction panels is small compared with theirdepth, WIH<O,I, then most of the effects of total ex-traction are confined to the immediate vicinity of the ex-traction panel. However, as WIH increases, more andmore of the strata overlying total-extraction panels comeinto tension (Fig. 5).

1.0 z/H

0,8COMPRESSIVE REGION

0.6 - - -'<1-- - (Jz/q=O~- " -/ ./ (BUDAVARI.19831

/ Ip109;H/{=O.66)

//'

/:

0,4

0,2TENSILE REGION

00 2 3 W/H

Fig. 5-lncrease In height of the tensile zone above the panelcentre as a function of the wldth.to-depth ratio of shallow panels

Since the strength of rock in tension is very much low-er than that in compression, strata caving is facilitated.The results of theoretical studies reported by Salamon28are in support of subsidence observations. In subsidenceengineering, it is well known that the lateral dimensionsof total-extractfon panels have to meet a certain criticalvalue to allow for the full development of surface subsi-dence in the centre of the panels. Panel dimensionssmaller than the critical dimensions are referred to as sub-critical, while panels that exceed the critical dimensionsare referred to as super-critical. Fig. 3 provides some use-ful information on the relationship between WIH andthe maximum surface subsidence in the centre of total-extraction panels. The value of WIH at which the verti-cal subsidence reaches its maximum value correspondsto the critical value, We. From Fig. 3 it can be seen thatWeis about 1,4 H for British coal mines, while the valuefor South African collieries ranges from 1,7 H to 3 H,depending on the composition of the superincumbentstrata. As pointed out earlier, the presence of very mas-sive dolerite sills could result in even greater critical panelwidths.

Relationship between Vertical Subsidence and OtherSubsidence Parameters

So far, the discussion of subsidence phenomena hasbeen largely concerned with the maximum surface subsi-dence that can be expected as a result of total coal-seamextraction. The reason for this is that vertical surface sub-sidence is the best understood and most readily measuredsubsidence parameter. However, from the viewpoints ofsubsidence and, in particular, surface damage, some ofthe other subsidence parameters are more important.

226 JULY 1991 JOURNAL OF ll-tE SOUTH AFRICAN INSTITUTE OF MINING AND METALWAGY

-0,2 0

0,1

0,2

VZ/h0,3

0,4

h H c a b

Colliery m m t d t d t d

NCB 0,43 6,1 7,6Brandspruit 2,9 130 0,43 0,43 65 57 11.8 10,2Coalbrook 2,7 160 0,49 0,49 12 22,S 7,3 5,6Sigma 2,6 110 0,46 0,46 7,9 25,4 5,7 3,8

Maximum Horizontal DisplacementsAs shown in Fig. 1, the maximum horizontal displace-

ment, Vxmu.,occurs above the mined-out area some dis-tance from the edge of the extraction panel. Localexperience indicates that horizontal displacements due tothe total extraction of the first seam are somewhat smallerthan those found over areas of multi-seam total extrac-tion. The relative magnitude of horizontal surface dis-placement compared with the maximum verticaldisplacement, Vz max,is

Vx mu ... 0,2 Vz max.

Maximum Horizontal StrainGround strains above total-extraction panels on South

African coal mines show large variations due to theblocky nature of the caved overburden strata and the ab-sence of a thick soil cover.

Based on numerous field observations, Wagner andSchiimanns established a relationship between the ratioof the extracted mining height, h, and the depth of min-ing, H, and the maximum surface strain, Emax:

Emu = 0,43 (h/H).The corresponding relationship for British coal-mining

areas is

Et max = 0,65 (h/H) and

Ec max = 0,51 (h/H).

When the average maximum subsidence of 55 per centof the extracted mining height replaces h, the strain fac-tor is 0,8 for critical panels, which is larger than the aver-age strain factor of 0,6 used by the NCB3, but it is similarto the theoretical value of 0,75 derived by Salamon28.

Maximum TiltThe position of the point of maximum tilt, T, in the

subsidence trough coincides with the position of maxi-mum horizontal displacement. Wagner and Schiimannsfound that the average value of maximum tilt, T can beexpressed as a function of (h/H):

T = 1,5 (h/H).

The local tilt factor of 1,5 compares with a theoreticalvalue of 3,02 derived by Salamon28 and a value of 2,75found in Great Britain. When h is replaced by the aver-age maximum subsidence of 55 per cent of the extractedmining height, a tilt factor of 2,73 is obtained, whichcompares well with the theoretical value and practical ex-periences in Great Britain.

Subsidence ProfilesSo far, only the maximum values of the various subsi-

dence parameters have been considered. However, it iswell known that surface subsidence varies across andalong total-extraction panels. Numerous observationshave shown that subsidence profiles above total-extraction panels in South Africa are remarkably similar(Fig. 6), and can be expressed by the following formuladescribing a half-profiles:

CVz/h =

1 +ae-bx', (6)

Parameter c describes the maximum subsidence thatcan be expected, a is a measure of the distance from theabutment at which significant subsidence can be expect-ed, and b is a measure of the steepness of the subsidencecurve. Table 11 summarizes the values of the variousparameters for a number of South African collieries andthe standardized British subsidence profile. The latter hasbeen adjusted to allow for the smaller magnitude of ver-tical subsidence found in South Africa.

X/H

0,2 0,4 0,8 1,00,6

- BrandspruitColliery(L/W)

--- Coalbrook Colliery (L/W)

- ,- Greenside Colliery (P/Exl)

~~,

, Kriel Colliery (P/Ext)

~ ,'-."," """"'-, '-'-

" -'-'.....- - -"':.-=0,5

Fig. 6-Dynamlc subsidence profiles measured In the centreline of total-extraction panels on different collieries

Modelling of Surface SubsidenceUp to now the discussion has been concerned with

idealized situations and the prediction of subsidenceparameters using simple semi-empirical relationships. Un-til recently this was the only practical means of predict-ing surface subsidence above total-extraction panels.Attempts to develop numerical models for subsidence en-gineeringmet with limited success. In particular, it wasfound that the presence of competent beds in the over-burden strata could not be handled by subsidence modelsdeveloped overseas. This difficulty was overcome recently

TABLE 11PARAMETER VALUES FOR TRANSVERSE(t) AND DYNAMIC (d)

SUBSIDENCE CURVES

JOURNAL OF THE SOUTH AFRICAN 1NS111UTE OF MINING AND METAUUAGY JULY1991 227

by Schiimann and Hamenstadt24, who adopted a two-stepapproach to simulate surface subsidence above local total-extraction panels. The two-step approach consists, first,of modelling the effects of strong, competent beds, whichtend to overhang the total-extraction area, therebypreventing or minimizing strata caving along the edgesof the total-extraction panel. The second step consists ofmodelling subsidence in the inner part of the extractionpanel, that is the central area, which is no longer affect-ed by the overhanging competent beds. Finally, the ef-fects of the two simulations are added, resulting in acomposite subsidence trough. The principles of the two-step approach are depicted in Fig. 7, which clearly showsthe edge zone and the central zone.

A total of seventeen case studies have been conductedto evaluate the applicability of this simulation methodto real situations. The results obtained so far have beenmost encouraging.

C,: ~

0'I<>-RI--1

'I'I

IIII ,

I I

1- d;

I 'IIIIIIII

t~tll ~:/;'500/0

ROOF CONVERGENCE

--- 0

CAVITIES-----

&. BULKINGVOIDSC, C2 E

Fig. 7-Slmulatlon of two.step subsidence due to Incompleteclosure In the edge zone and larger closure In the central zone

50%

Summary of Mining-induced Surface DisplacementsThe results of field observations, theoretical studies,

and numerical simulations of surface subsidence due tototal coal-seam extraction can be summarized as follows.

(1) Surface subsidence above local total-extractionpanels differs significantly from that observed inEuropean coal-mining districts. The main reasonsfor these differences is the presence of competentdolerite sills and massive sandstone layers in mostlocal coal-mining areas. As a result of these com-petent strata, the magnitude of surface subsidencetends to be much less than that observed in Europe,and the extent of significant surface displacementsis confined to the mined-out area.

(2) Surface displacements follow a well-defined pat-tern and can be predicted with a fair degree of relia-bility and accuracy. Semi-empirical relationshipshave been derived for all the important subsidenceparameters.

(3) The modelling of surface subsidence by advancednumerical techniques has been achieved successful-ly for a number of situations.

(4) There is a wealth of subsidence information cover-ing more than two decades of total-extractionmining in South Africa.

REsPONSE OF SURFACE STRUcruRES TO TOTAL-SEAM EXTRACTION

In recent years considerable local knowledge has beengained concerning the behaviour of a variety of surfacestructures that were undermined by either longwall facesor pillar-extraction panels. Most of the investigationswere sponsored by the Strata Control Advisory Commit-tee of the Coal Mining Research Controlling Committee(CMRCC).

In view of the importance of these investigations forthe future of total-extraction coal mining in South Afri-ca, some of the more significant investigations aredescribed briefly.

Railway LinesAt Greenside Collieries, a railway siding was under-

mined by a pillar-extraction panel situated at a depth of55 m. The surface subsidence resulting from the extrac-tion of the 2,0 m thick coal seam was 0,9 m. Throughoutthe six-week period of undermining, rail transport wasnot interrupted. Ballast was introduced for concurrentlifting of the railway tracks to maintain acceptable gra-dients and to compensate for any lateral tilt. This experi-ment showed that railway lines can be undermined byrelatively shallow total-extraction panels provided ade-quate precautions are taken and close contact is main-tained between the mining personnel and the operatorsof the railway line.

RoadsAt Secunda Collieries, a provincial road has been un-

dermined several times by longwall panels situated atdepths ranging from 120 to 130 m. Surface subsidenceresulting from the extraction of the 3 m thick coal seamranged from 1,3 to 1,5 m. Tensile cracks and bucklingof the upper layer of the road structure were observedin regions of tensile and compressive strains. Somedamage was done to the sub-base of the road. The over-all damage was within acceptable limits, and the roadswere repaired. In order to minimize compressive damage,vertical slots were cut at regular intervals into the roadsurface at two of the roads in the Secunda area. The 0,5 mdeep slots, which were cut prior to undermining, virtual-ly eliminated compressive damage and significantlyreduced the costs of road repair.

At New Denmark Colliery, a provincial road was un-dermined by several212m wide longwall panels situatedat a depth of about 200 m. The effects on the road ofthe extraction of the 2 m thick coal seam were minimal.The maximum subsidence in this instance did not exceed0,4m.

At Matla Collieries, several district gravel roads wereundermined by longwall faces situated at a depth of about60 m. The extracted seam thickness in this instance was1,8 m. The roads were kept fully operational throughoutthe period of undermining, and the maximum subsidenceto which they were subjected was 1,0 m.

Numerous farm roads have been undermined by long-wall and pillar-extraction panels at Usutu, Sigma, Coal-brook, and Durban Navigation Collieries without anysubsidence damage and without being rendered unser-viceable.

228 JULY 1991 JOURNAL OF THE SOUTH AFRICAN 1NS11TUTEOF MINING AND METAllURGY

0.010IW. KARAGANtIt<)<>

D

c..J (U)Z <> 0

~0~()(IJ(/)

0Z

~OC) 0 0,010 0.020

GROUND SLOPE. (TI LT). T .

Fig. 8-C1..... ofdamage In term. of ground strain end .Iope.

Transmission TowersFour experimental 400 kV transmission towers were un-

dermined by a 212 m wide longwall panel at Secunda Col-lieries. The surface subsidence resulting from theextraction ofthe 2,9m thick coal seam situated at a depthof 121m was 1,3 m. A maximum tilt of 35 mm/m andmaximum strains of 8 mm/m in tension and 7 mm/m incompression were observed at the position of the tow-ers. There was no damage to the towers, and only oneof sixteen footings showed hairline cracks in the concretecollarll.

BuildingsThe most comprehensive subsidence investigation in

South Africa was conducted at Brandspruit Colliery inSecunda, where a whole farm complex was underminedby a 212 m wide longwall panel situated at a depth of130m. The buildings that were undermined included abrick farm house, a sandstone building, two long steel-framed brick structures, and a number of small structuressuch as a circular water reservoir. The conditions of thevarious structures were carefully recorded prior to the ex-traction of the 3 m thick coal seam, and the structureswere monitored on a virtually continuous basis through-out the undermining process. Damage to the structureswas very limited-a maximum subsidence of 1,3 m. Sub-sequent to undermining, most of th~ buildings were re-stored at moderate cost to a condition equal to or betterthan before total-seam extraction took place, and theynow form the basis of the Secunda Mines Centre forGround Conservation.

Recently, a club complex consisting of two buildingsand a swimming pool was undermined by a 212 m widelongwall face at New Denmark Colliery, near Stander-ton. The maximum subsidence resulting from the extrac-tion of the 2 m thick coal seam at a depth of 185m was0,36 m. The induced tilt in the vicinity of the club housewas 1mm/m, and the tensile strain was 0,2 mm/m. TheDivision of Building Technology of the CSIR found thatthe crack pattern in the club house had changed little,except for a small extension of a few existing cracks, anda slight widening of some of these cracks or a slight com-pression of others. The overall rating of damage to thebuilding before and after undermining remained un-changed. Furthermore, it was concluded that the damagecould be regarded as negligible, particularly when com-pared with. that due to heaving clay. The swimming pool,consisting of a glass-fibre shell set in sand, remaitied un-damaged after approximately 30mm of subsidence18.

. SUMMARY OF OBSERVATIONSThe most important finding of all the undermining

studies is that local experiences of the damage to surfacestructures due to mining-induced surface displacementsis similar to that in other mining countries. Consequent-ly, overseas experience can be applied with a fair degreeof confidence in South Africa.

The findings of numerous undermining experiments inSouth Africa can be summarized as follows.

(1) Damage to surface structures is closely related tothe magnitude of the mining-induced surfacestrains and tilts.

(2) No single incidence of catastrophic and unexpect-

ed structural failure was observed.(3) In most instances, structural damage was well

within the range of repairs.(4) Damage to buildings was generally less than that

observed overseas owing to the very shallow foun-dations of local structures (no basements).

CLASSIFICATION OF DAMAGE

Based on observations of local and overseas damage,various categories of damage can be established. Exam-ples of damage criteria for buildings are given in Figs.8 and 9 and in Table Ill. An important aspect is the in-fluence of the length of a building on the value of sur-face strain that leads to a certain degree of structuraldamage. Fig. 8 shows that small buildings can toleratevery much greater surface strains without being damagedthan very long buildings.

CONCLUSIONSBased on more than twenty years of detailed studies

into the effects of total coal-seam extraction on the sur-face and surface structures, the following conclusionshave been reached.

(1) The magnitude and extent of the surface displace-ments resulting from total coal-seam extraction canbe predicted with a fair degree of accuracy for mostgeological conditions.

(2) The magnitude of the vertical subsidence abovetotal-extraction panels in South Africa is general-ly smaller than observed on European coal mines.This is due to the presence of competent beds inthe overburden strata.

c..J

Z«er~(/)

0Z::J

~OC) 0 10 20 30 1.0 50m

LENGTH OF STRUCTURE.t .

Fig. 9-Clesse. of demege In tenn. ofground strain end length ofstructure.

JOURNAL OF THE SOUTH AFRICAN INSTITUTEOF MINING AND METALWRGY JULY1991 229

Poland USSRGreat Britain (S. Knothe, 1952; Donetsk district

Class of damage (ref. 3, p. 49) Rimant, 1958) (VNIMI 1958)

I Very slight Haircracks in plaster, isolated slight Haircracks, no damagefracture in building, not visible outside

~1=30mm/60m T<2,5 mm/m T< 4 mm/mE=0,5mm/m 1/<20'JO-6/m p >20 km

E < 1,5 mm/m E < 2 mm/m

11 Slight Several slight fractures inside, doors Damage easy to repair Limit for five-storey buildingsand windows stick

~1= 30-60 mm/60 m T< 5 mm/m T<4,5 mm/mE=O,5-1 mm/m 1/<50'JO-6/m p > 18 km

E <3 mm/m E <2,5 mm/m

III Appreciable Slight fracture outside or one main frac- Appreciable damage, build- Three- and four-storeyture, doors and windows stick ing still useable buildings

~1= 60-120 mm/60 m T<JO mm/m T< 5mm/mE= 1-2 mm/m 1/<90.JO-6/m p > 12 km

E < 6 mm/m E < 3,5 mm/m

IV Severe Severe open cracks, floors sloping, walls Severe damage, anchoring Two-storey buildingsleaning, anchoring required required

~I=0,12-0,2 m/60 m T<15mm/m T<8 mm/mE = 2-3 mm/m 1/<120'JO-6/m p >5,5 km

E < 9 mm/m E <6 mm/m

V Very severe Part or complete rebuilding required Destruction of building One-storey onlyT<JO mm/m

~1>0,2 m/60 m p>3kmE>3 mm/m 1/>160'JO-6/m E < 7,5 mm/m

TABLE IIICATEGORIES AND LIMITS OF DAMAGE

Note: The radius of curvature p =!'1/

(3) Most of the mining-induced surface displacements,particularly damage to structures, is localized andconfined to mined-out areas.

(4) Surface subsidence due to total-seam extraction isa gradual and controlled process, and results in thegradual development of damage.

(5) In the majority of cases, more than 90 per cent ofthe surface displacement occurs within a period ofless than three months of undermining.

(6) Residual surface movements are generally verysmall and do not give rise to damage.

(7) Most surface structures can be maintained in anoperational condition during undermining provid-ed proper precautions are taken.

(8) The cost of subsidence monitoring is small com-pared with the longterm benefits that can be gainedfrom the improved local knowledge.

(9) Most of the subsidence investigations have beenconfined to evaluations of the effects of single-seam extraction on the surface. Much work needsto be done to improve the knowledge on the ef-fects of multi-seam extraction on the surface.

OUTLOOKThe coal locked up in underground coal seams, the land

surface, and the ground water are important naturalresources that are essential for the future developmentof the country. Optimal utilization of these resources istherefore of national importance.

In an effort to improve.the utilization of the country's

limited coal reserves, the coal industry is moving towardsthe wider application of total-extraction mining methodsin the form of opencast mining and underground long-wall mining and pillar extraction. At present, about onehalf of the country's annual coal production is mined bytotal-extraction methods-a proportion that is likely toincrease in the future.

Questions related to the effects of total coal-seam ex-traction on the land surface and its use will therefore beraised more frequently. As far as the agricultural sectoris concerned, it is hoped that the work of the Van Nie-kerk Commission will provide the basis for a harmoni-ous co-existence between that sector and mining, whichare both vitally important pillars of South Africa's econ-omy. As for the interaction between coal mining andother users of the surface, it is hoped that the rigid atti-tude that exists in some quarters will give way to a moreflexible approach based on sound engineering and eco-nomics principles. Only on this basis will it be possibleto lift unnecessary restrictions or to avoid unnecessaryduplication of costly surface infrastructures.

A mechanism will have to be found to pool allknowledge on the effects of total coal-seam extractionon the environment, and to promote improved commu-nication between all parties who have a vested interestin this important area.

ACKNOWLEDGEMENTSMuch of the work described in this paper is the result

of a long-range research programme into subsidence

230 JULY 1991 JOURNAL OF THE SOUTH AFRICAN IN5TI1UTE OF MINING AND METALLURGY

above coal-mine workings that is being sponsored by theCoal Mining Research Controlling Committee (CMRCC)and conducted under the auspices of its Strata ControlAdvisory Committee. The Coal Mining Division of theChamber of Mines' Research Organization (COMRO)has carried out most of the subsidence researchprogramme under contract to the CMRCC.

The coal-mining industry, particularly Sasol Corpora-tion, the Trans Natal Collieries Ltd, Amcoal, Iscor, andGold Fields, have supported these research efforts overa long period of time. In recent years, the Road Researchand Building Research Divisions of the CSIR have madevaluable contributions to the evaluation of subsidence-related damage to roads and buildings. The collabora-tion of Eskom in the work on power-line pylons is greatlyappreciated.

Finally, the substantial progress of subsidence engineer-ing in South Africa was made possible only through theenthusiastic support given to the research teams by thevarious mines that participated in the investigations, andthrough the understanding and continuing supportprovided by the various inspectorates under the Govern-ment Mining Engineer.

REFERENCES

1. WHITIAKER, B.N., and REDDISH, D.J. Subsidence-occurrence,prediction and control. Developments in geotechnical engineering,

56. Amsterdam, Elsevier Science Publishers, 1989.

2. KRATZSCH, H. Mining subsidence engineering. Berlin, SpringerVerlag, 1983.

3. NATIONAL COAL BOARD, U.K. Subsidence engineer's handbook.2nd ed. 1975.

4. ORAVECZ, K.!. Loading of coal pillars in bord and pillar work-ings. Johannesburg, Chamber of Mines of South Africa, Research

Report 40/73, 1973.

5. WAGNER, H., and SCHUMANN, E.H.R. The effect of total coal

seam extraction on the surface and surface structures. Colloquiumon Recent Mining and Metallurgical Developments in the EasternTransvaal, Witbank, South African Institute of Mining and Metal-lurgy, act. 1985.

6. HESLOP, W.T. Coal pillar extraction from two seams and sur-

face effects. J. Chem. Metall. Min. Soc. S.Afr., act. 1921.7. STEART, F.A. Mining subsidence with particular reference to the

coal fields of South Africa. Coal Base Miner. S.Afr., vol. 3, Dec.1955, pp. 42-46; Jan. 1956, pp. 30-38; Dec. 1956, pp. 28-34.

8. SALAMON, M.D.G., ORAVECZ, K.I., and HARDMAN, D.R. Rockmechanics problems associated with longwall trials in South Afri-

ca. Proceedings Fifth International Strata Control Conference, Lon-don, 1972.

9. GALVIN, J .M. The development of the theory of dolerite sill be-haviour. Proceedings Fifth Congress of International Society of

Rock Mechanics, Sec D. Melbourne, Australian GeomechanicsSociety, 1983.

10. SCHUMANN, E.H.R. Controlled subsidence of the Blackhill-Greenside railway line, Part!. Johannesburg, Chamber of Minesof South Africa, Research Report 8/83, 1983.

11. SCHUMANN, E.H.R. Surface subsidence due to the extraction ofmoderately thick coal seams at shallow depth. Proceedings 3rd In-

ternational Conference Ground Movements and Structures, Cardiff,

Jul. 1984.

12. SCHUMANN, E.H.R. The control of surface subsidence bywidth/depth ratio and chain pillar size in the presence of compe-

tent coal measures: 7th International Conference on Ground Con-

trol in Mining, Morgantown (USA), 1988.

13. SCHUMANN,E.H.R. Surface subsidence above longwall extrac-tion panels at South African collieries. SANGORM Symposium,Swaziland, Nov. 1988.

14. SCHUMANN,E.H.R. Subsidence and prevention of damage fromfirst seam shallow extraction of coal under competent strata. In-ternational Mining and Metallurgy Conference on Mineral Extrac-tion, Utilisation and the Surface Environment, Harrogate (UK),Dec. 1988.

15. SCHUMANN,E.H.R. Subsidence above total extraction panels atSouth African collieries. Proceedings 8th International Strata Con-trol Conference, Diisseldorf (West Germany), May 1989.

16. JACK, B.W., STEIJN, J.J., and GAY, N.C. The effect of subsi-dence as a result of shallow mining operations on surface struc-ture. A quantitative case study. SANGORM Symposium,Johannesburg, 1984.

17. JACK,B.W. Case studies of the effects of surface subsidence ongravel and bituminous roads. SANGORM Symposium, Johannes-burg, 1986.

18. JACK, B.W., and SCHUMANN,E.H.R. Verifying prediction ofsurface subsidence and its effects on surface structures. SANGORMSymposium, Witbank, 1989.

19. SCHUMANN,E.H.R. Surface subsidence over longwalls at Dur-ban Navigation Colliery. Johannesburg, Chamber of Mines ofSouth Africa, Research Report 19/79, 1979.

20. SCHUMANN,E.H.R. The influence of a dolerite sill on the de-velopment of surface subsidence at Durban Navigation Colliery.Ibid., Research Report UR 169, Jul. 1987.

21. SCHUMANN,E.H.R. Surface subsidence due to multiseam totalextraction at Sigma Colliery. Ibid., Research Report 41/87, Sep.1987.

22. VANDERMERWE,J.N. Design methods to arrive at the optimalplacing and mining of inter panel pillars to alleviate their effectson surface. SANGORM Symposium, 1986. pp. 133-144.

23. VAN DERMERWE, J.N. Analysis of surface subsidence over alongwall panel of 150m below surface. Ibid., pp. 145-150.

24. SCHUMANN,E.H.R., and HAMENSTAEDT,R. The method of sub-sidence prediction by means of influence functions, including resultsfrom case studies. SANGORM Symposium, Witbank, 1989.

25. WAGNER,H., and STEIJN, J.J. Effect of local strata conditionson the support requirements on longwall faces. Proceedings TenthCongress, International Society Rock Mechanics, Montreux, 1979.pp. 557-564.

26. SALAMON,M.D.G. Reconsolidation of caved areas. Johannes-burg, Chamber of Mines of South Africa, Report 58/66, 1966.

27. SALAMON,M.D.G., and ORAVECZ,K.!. Rock mechanics in coalmining. Johannesburg, Chamber of Mines of South Africa, P.R.D.Series no. 198, 1976.

28. SALAMON,M.D.G. Rock mechanics in underground excavations.Chap. 8, Surface subsidence. Proceedings 3rd Congress, Interna-tional Society Rock Mechanics, Denver (USA), Sep. 1974.

BIBLIOGRAPHY

GALVIN, J.M., STEIJN, J.J., and WAGNER, H. Rock mechanics of

total extraction. Paper D-Increased underground extraction of coal.Symposium, South African Institute of Mining and Metallurgy, 1981.

LEAR, C.D., and SCHUMANN, E.H.R. The influence of dolerite sills

on total extraction underground workings and surface subsidence. SAN-GORM Symposium, Witbank, 1989.

SALAMON, M.D.G. Linear models for predicting surface subsidence.

Proceedings 5th Congress, International Society for Rock Mechanics,

Sec E. Melbourne (Australia), 1983. pp. 107-114..

ScHDMANN, E.H.R., and HARDMAN, D.R. Subsidence caused by coal

seam extraction beneath a competent overburden. Fifth Australia-New

Zealand Conference on Geomechanics, Sydney (Australia), Aug. 1988.

JOURNAL OF lHE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY JULY1991 231