Saimm 201504 apr

VOLUME 115 NO. 4 APRIL 2015

Many SolutionsOne SourceFLSmidth offers you a complete line of equipment and services with proven reliability

and enhancing performance from the leading brand names of ABON, Buffalo, CEntry,

Conveyor Engineering, Dawson Metallurgical Laboratories, Decanter Machine, Dorr-

Oliver, EIMCO, ESSA, FFE, Fuller-Traylor, KOCH, Knelson, Krebs, Ludowici, Maag Gear,

Mayer, Meshcape, MIE, Möller, MVT, PERI, Phillips Kiln Services, Pneumapress, RAHCO,

Raptor, Roymec, Shriver, Summit Valley, Technequip and WEMCO.

FLSmidth is your One Source because:

We own our technology

We deliver One Source mineral solutions

World renowned OEM equipment & brands

We have operating references across Africa

We understand Africa’s challenges

We understand our customer’s operating needs

We underwrite project funding

For more information contact us

www.flsmidth.com

�

ii APRIL 2015 The Journal of The Southern African Institute of Mining and Metallurgy

OFFICE BEARERS AND COUNCIL FOR THE2014/2015 SESSION

Honorary PresidentMike TekePresident, Chamber of Mines of South Africa

Honorary Vice-PresidentsNgoako RamatlhodiMinister of Mineral Resources, South AfricaRob DaviesMinister of Trade and Industry, South AfricaNaledi PandoMinister of Science and Technology, South Africa

PresidentJ.L. Porter

President ElectR.T. Jones

Vice-PresidentsC. MusingwiniS. Ndlovu

Immediate Past PresidentM. Dworzanowski

Honorary TreasurerC. Musingwini

Ordinary Members on Council

V.G. Duke T. PegramM.F. Handley S. RupprechtA.S. Macfarlane N. SearleM. Motuku A.G. SmithM. Mthenjane M.H. SolomonD.D. Munro D. TudorG. Njowa D.J. van Niekerk

Past Presidents Serving on CouncilN.A. Barcza J.C. NgomaR.D. Beck S.J. Ramokgopa J.A. Cruise M.H. RogersJ.R. Dixon G.L. SmithF.M.G. Egerton J.N. van der MerweG.V.R. Landman W.H. van NiekerkR.P. Mohring

Branch ChairmenDRC S. MalebaJohannesburg I. AshmoleNamibia N. NamatePretoria N. NaudeWestern Cape C. DorflingZambia H. ZimbaZimbabwe E. MatindeZululand C. Mienie

Corresponding Members of CouncilAustralia: I.J. Corrans, R.J. Dippenaar, A. Croll,

C. Workman-DaviesAustria: H. WagnerBotswana: S.D. WilliamsUnited Kingdom: J.J.L. Cilliers, N.A. BarczaUSA: J-M.M. Rendu, P.C. Pistorius

The Southern African Institute of Mining and Metallurgy

PAST PRESIDENTS

*Deceased

* W. Bettel (1894–1895)* A.F. Crosse (1895–1896)* W.R. Feldtmann (1896–1897)* C. Butters (1897–1898)* J. Loevy (1898–1899)* J.R. Williams (1899–1903)* S.H. Pearce (1903–1904)* W.A. Caldecott (1904–1905)* W. Cullen (1905–1906)* E.H. Johnson (1906–1907)* J. Yates (1907–1908)* R.G. Bevington (1908–1909)* A. McA. Johnston (1909–1910)* J. Moir (1910–1911)* C.B. Saner (1911–1912)* W.R. Dowling (1912–1913)* A. Richardson (1913–1914)* G.H. Stanley (1914–1915)* J.E. Thomas (1915–1916)* J.A. Wilkinson (1916–1917)* G. Hildick-Smith (1917–1918)* H.S. Meyer (1918–1919)* J. Gray (1919–1920)* J. Chilton (1920–1921)* F. Wartenweiler (1921–1922)* G.A. Watermeyer (1922–1923)* F.W. Watson (1923–1924)* C.J. Gray (1924–1925)* H.A. White (1925–1926)* H.R. Adam (1926–1927)* Sir Robert Kotze (1927–1928)* J.A. Woodburn (1928–1929)* H. Pirow (1929–1930)* J. Henderson (1930–1931)* A. King (1931–1932)* V. Nimmo-Dewar (1932–1933)* P.N. Lategan (1933–1934)* E.C. Ranson (1934–1935)* R.A. Flugge-De-Smidt

(1935–1936)* T.K. Prentice (1936–1937)* R.S.G. Stokes (1937–1938)* P.E. Hall (1938–1939)* E.H.A. Joseph (1939–1940)* J.H. Dobson (1940–1941)* Theo Meyer (1941–1942)* John V. Muller (1942–1943)* C. Biccard Jeppe (1943–1944)* P.J. Louis Bok (1944–1945)* J.T. McIntyre (1945–1946)* M. Falcon (1946–1947)* A. Clemens (1947–1948)* F.G. Hill (1948–1949)* O.A.E. Jackson (1949–1950)* W.E. Gooday (1950–1951)* C.J. Irving (1951–1952)* D.D. Stitt (1952–1953)* M.C.G. Meyer (1953–1954)* L.A. Bushell (1954–1955)

* H. Britten (1955–1956)* Wm. Bleloch (1956–1957)* H. Simon (1957–1958)* M. Barcza (1958–1959)* R.J. Adamson (1959–1960)* W.S. Findlay (1960–1961)

D.G. Maxwell (1961–1962)* J. de V. Lambrechts (1962–1963)* J.F. Reid (1963–1964)* D.M. Jamieson (1964–1965)* H.E. Cross (1965–1966)* D. Gordon Jones (1966–1967)* P. Lambooy (1967–1968)* R.C.J. Goode (1968–1969)* J.K.E. Douglas (1969–1970)* V.C. Robinson (1970–1971)* D.D. Howat (1971–1972)

J.P. Hugo (1972–1973)* P.W.J. van Rensburg (1973–1974)* R.P. Plewman (1974–1975)

R.E. Robinson (1975–1976)* M.D.G. Salamon (1976–1977)* P.A. Von Wielligh (1977–1978)* M.G. Atmore (1978–1979)* D.A. Viljoen (1979–1980)* P.R. Jochens (1980–1981)

G.Y. Nisbet (1981–1982)A.N. Brown (1982–1983)

* R.P. King (1983–1984)J.D. Austin (1984–1985)H.E. James (1985–1986)H. Wagner (1986–1987)

* B.C. Alberts (1987–1988)C.E. Fivaz (1988–1989)O.K.H. Steffen (1989–1990)

* H.G. Mosenthal (1990–1991)R.D. Beck (1991–1992)J.P. Hoffman (1992–1993)

* H. Scott-Russell (1993–1994)J.A. Cruise (1994–1995)D.A.J. Ross-Watt (1995–1996)N.A. Barcza (1996–1997)R.P. Mohring (1997–1998)J.R. Dixon (1998–1999)M.H. Rogers (1999–2000)L.A. Cramer (2000–2001)

* A.A.B. Douglas (2001–2002)S.J. Ramokgopa (2002-2003)T.R. Stacey (2003–2004)F.M.G. Egerton (2004–2005)W.H. van Niekerk (2005–2006)R.P.H. Willis (2006–2007)R.G.B. Pickering (2007–2008)A.M. Garbers-Craig (2008–2009)J.C. Ngoma (2009–2010)G.V.R. Landman (2010–2011)J.N. van der Merwe (2011–2012)G.L. Smith (2012–2013)M. Dworzanowski (2013–2014)

Honorary Legal AdvisersVan Hulsteyns Attorneys

AuditorsMessrs R.H. Kitching

Secretaries

The Southern African Institute of Mining and MetallurgyFifth Floor, Chamber of Mines Building5 Hollard Street, Johannesburg 2001P.O. Box 61127, Marshalltown 2107Telephone (011) 834-1273/7Fax (011) 838-5923 or (011) 833-8156E-mail: [email protected]

ContentsJournal Commentby H.R. Phillips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivPresident’s Corner by J.L. Porter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

Special ArticlesSouth African National Committee on Tunelling Young Members Group –SANCOT - YMGby L. Nene. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viHandover of model stope to Wits School of Mining Engineeringby S. Braham. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii

New head of Wits mining school announcedby S. Braham. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ixWits-SRK link boosts rock engineering skillsby S. Braham. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x

Re-aligning the cutting sequence with general support work and drafting a support sequence at Simunye Shaftby K. Lombard. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

Mining through areas affected by abnormal stress conditions at Syferfontein Collieryby C. Legote. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265

A critical evaluation of the water reticulation system at Vlaklaagte Shaft, Goedehoop Collieryby R. Lombard. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271

Optimization of shuttle car utilization at an underground coal mineby P.R. Segopolo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285

Explosives utilization at a Witwatersrand gold mine by M. Gaula. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297

Critical investigation into the problems surrounding pillar holing operationsby J.P. Labuschagne, H. Yilmaz, and L. Mpolokeng. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307

LHD optimization at an underground chromite mineby W. Mbhalati . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313

The viability of using the Witwatersrand gold mine tailings for brickmakingby M. Malatse and S. Ndlovu. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321

Evaluation of some optimum moisture and binder conditions for coal fines briquettingby P. Venter and N. Naude . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329

Air drying of fine coal in a fluidized bed by M. Le Roux, Q.P. Campbell, M.J. van Rensburg, E.S. Peters, and C. Stiglingh. . . . . . . . . . . . 335

International Advisory Board

R. Dimitrakopoulos, McGill University, CanadaD. Dreisinger, University of British Columbia, CanadaE. Esterhuizen, NIOSH Research Organization, USAH. Mitri, McGill University, CanadaM.J. Nicol, Murdoch University, AustraliaH. Potgieter, Manchester Metropolitan University, United KingdomE. Topal, Curtin University, Australia

The Journal of The Southern African Institute of Mining and Metallurgy APRIL 2015


▲iii

Editorial BoardR.D. BeckJ. Beukes

P. den HoedM. Dworzanowski

M.F. HandleyR.T. Jones

W.C. JoughinJ.A. LuckmannC. MusingwiniR.E. Robinson

T.R. StaceyR.J. Stewart

Editorial ConsultantD. Tudor

Typeset and Published byThe Southern African Instituteof Mining and MetallurgyP.O. Box 61127Marshalltown 2107Telephone (011) 834-1273/7Fax (011) 838-5923E-mail: [email protected]

Printed by Camera Press, Johannesburg

AdvertisingRepresentativeBarbara SpenceAvenue AdvertisingTelephone (011) 463-7940E-mail: [email protected] SecretariatThe Southern AfricanInstitute of Mining andMetallurgyISSN 2225-6253 (print)ISSN 2411-9717 (online)

THE INSTITUTE, AS A BODY, ISNOT RESPONSIBLE FOR THESTATEMENTS AND OPINIONSADVANCED IN ANY OF ITSPUBLICATIONS.Copyright© 1978 by The Southern AfricanInstitute of Mining and Metallurgy. Allrights reserved. Multiple copying of thecontents of this publication or partsthereof without permission is in breach ofcopyright, but permission is hereby givenfor the copying of titles and abstracts ofpapers and names of authors. Permissionto copy illustrations and short extractsfrom the text of individual contributions isusually given upon written application tothe Institute, provided that the source (andwhere appropriate, the copyright) isacknowledged. Apart from any fair dealingfor the purposes of review or criticismunder The Copyright Act no. 98, 1978,Section 12, of the Republic of SouthAfrica, a single copy of an article may besupplied by a library for the purposes ofresearch or private study. No part of thispublication may be reproduced, stored ina retrieval system, or transmitted in anyform or by any means without the priorpermission of the publishers. Multiplecopying of the contents of the publicationwithout permission is always illegal.

U.S. Copyright Law applicable to users Inthe U.S.A.The appearance of the statement ofcopyright at the bottom of the first page ofan article appearing in this journalindicates that the copyright holderconsents to the making of copies of thearticle for personal or internal use. Thisconsent is given on condition that thecopier pays the stated fee for each copy ofa paper beyond that permitted by Section107 or 108 of the U.S. Copyright Law. Thefee is to be paid through the CopyrightClearance Center, Inc., Operations Center,P.O. Box 765, Schenectady, New York12301, U.S.A. This consent does notextend to other kinds of copying, such ascopying for general distribution, foradvertising or promotional purposes, forcreating new collective works, or forresale.

STUDENT EDITION


s part of its commitment to supporting youngprofessionals entering the mining andmetallurgical industries, the SAIMM holds a

Student Colloquium every year. The ten papers in thisedition of the Journal are based on presentations madeat that event by students and recent graduates inmining engineering, metallurgy, and mineralsprocessing. The opportunity to present their final yearresearch projects is restricted to those students fromeach institution achieving the best results for thissubject, and the papers presented here have beenfurther selected following scrutiny by a panel of seniorprofessionals, acting as judges at the Colloquium.

This year there is an emphasis on coal, with four ofthe six mining papers based on vacation work at coalmines, while two of the four mineral processing papersdeal with the utilization of coal fines. However, theresearch topics reflected in these papers are of lessconcern than the overall quality. Four and five yearsago the number of learners entering the degreeprogrammes represented by these papers showed amarked increase, which means that competition tohave a paper published in this edition of the Journal isintense. Even if a particular topic is only of marginalinterest to the reader, I would recommend that youread the abstract and glance through the paper toappreciate the high quality of the work of these youngprofessionals.

Of course there is another and more disturbing sideto the rising number of graduates. This comes at a timeof great difficulty for the global mining industry, andnowhere more so than in Southern Africa. Afterdecades of undersupply, we are now seeing graduatesstruggling to find employment. Even more surprising isthat some of the best graduates, having receivedmining industry bursaries throughout theirundergraduate studies, are being cut loose ongraduation to find a job anywhere they can. Theeducation they have received and the skills they havelearnt generally equip them not only for their narrowspecialization, but for a place in the workforce, oftenfar from the mining industry. Once gone they areunlikely to return.

This raises the questions of whether the SouthAfrican mining industry has a future, what sort offuture will it be, and what sort of professionals will itrequire in order to be successful. This once greatindustry is at a low ebb as it struggles to come to termswith the forces imposed on it by the past decade or so,not least of which is the present slump in commodity

prices. This was an industry, particularly the goldsector, that relied on brawn and where physical effortdelivered the product. To remain competitive in the21st century, mining and mineral processing hasalready changed significantly and will continue to doso. To achieve further gains in competitiveness, tocontinue the path to zero harm and to achieve theoverall goals of sustainability, it will be brains and notbrawn that will make the difference. It is all too easy toregard young graduates, at the start of their workinglives, as a cost rather than an asset. While it is wellunderstood that many graduates are unsuitable for acareer within the narrow confines of ‘production’, thesuccess of minerals industry companies is increasinglydependent on the ‘service departments’, wheretechnical skills are in short supply.

Instead of employing young graduates only inproduction posts, while constantly bemoaning the lowpass rates in industry certificates of competence, itcould prove hugely beneficial for industry to broadenits vision of how to employ mining engineers andmineral processing graduates. Mine planning, rockengineering, mine ventilation, research anddevelopment, plant optimization, project management,and mechanization are some of the specialities thathave a huge impact on the future of a company. Whileuniversities are seen by industry as a source ofgraduates with a broad-based education in theirdiscipline, they also have a significant role to play indeveloping specialists through postgraduatequalifications. With a limited market for these coursessome rationalization between universities is required,and this is where institutions such as the SAIMM andthe other professional associations can help facilitatediscussions.

We all know that mining is a cyclic business andalso that ‘the darkest hour is just before dawn’. Thehigh standard of papers in this edition of the Journal,coupled with the fact that these students have appliedtheir minds to solving practical problems, gives a clearindication of the raw talent available to our industry.Our industry must now have the foresight to developthis talent in order to ensure we are well positioned totake advantage of the next upturn in the commoditiescycle.

H.R. Phillips

Journal Comment

�

iv APRIL 2015 The Journal of The Southern African Institute of Mining and Metallurgy

Schoo

lofM

iningEn

gine

eringScho

olof

MiningEn

gine

eringSc

hoo

lofMiningEn

ginee

ring

Scho

olof

MiningEn

gine

eringScho

oloff

MiningEngineeringSchoolofMiningEngineerin

gSchoolofMiinin

g Engine

ering

SchooloffMinnin

g Eng

ineeri

ngSch

oolof M

ining Eng

ineerin

g School

of Minin

g Enginee

ringSch

ooolof M

ining Engi

neering

SizeIt matters! Our high-level academic

staff offers specialist higher degrees

and produces the largest number of

graduates in the English-speaking

world.

SubstanceOur accredited world-class programmes in uence outh

frica frica and the orld. he chool is student-centred

and academicall relevant with de ned elds of mining. e

bring the environment to our students.

StyleOur multi-disciplinary and specialist postgraduate

programme leads to sustained research output,

fascilitating regular step changes in industry.

School of Mining EngineeringUniversity of Witwatersrand, Johannesburgniversity of Witwatersrand, JohaUniveversrsityty of WiWitwawatersrsraand,d, Johannesburgrg

�

vi APRIL 2015 The Journal of The Southern African Institute of Mining and Metallurgy

South African National Committee on TunellingYoung Members Group – ‘SANCOT – YMG’

Young professionals and the youth at large can make valuable contributions in the civil and mining industries. Like manyother organizations within different industries, SANCOT has the responsibility of ensuring that there is effective youth

involvement in all professional activities within the industry. This way, the youth are also able to make a meaningfulcontribution towards their professional and technical development.

At a meeting held on 14 January 2015 at AECOM offices in Centurion, the South African National Committee onTunnelling approved the formation of a Young Members Group (’SANCOT –YMG’). Mr Lucky Nene was nominated andaccepted as the chairman of this Young Members Group.

SANCOT–YMG has adopted its mandate from its mother bodies, SANCOT and the SAIMM, and is working very closelywith the International Tunnelling Association (ITA) and the youth body of ITA, which is ITA-YM. This is to ensure allianceand compliance on various aspects that affect the young professionals. The mandate as adopted from the ITA-YM isstructured as follows:

a) To provide a technical networking platform within the ITA for young professionals and studentsb) To bridge the gap between generations and to network across all experience levels in the industryc) To create awareness of the tunnelling and underground space industry to new generationsd) To provide young professionals and students with a voice in the ITA, including the Working Groupse) To look after the next generation of tunnelling professionals and to pass on the aims and ideals of the ITA.Through general interactions via other professional platforms, young professionals have shown interest in this youth

structure and a desire to take part. It is therefore envisaged that all interested companies would encourage their youngprofessionals, within both mining and civil engineering, to have representatives within SANCOT-YMG. This participation andinvolvement is encouraged to extend beyond place(s) of work and will also include those young professionals that are atacademic institutions.

The focus areas for SANCOT-YMG would be to mirror the mother body activities and objectives in a way that ensuresfun, enthusiasm, and the ongoing participation of young professionals in all aspects of the mother body and industry atlarge. These focus areas are as adopted form the ITA-YM mandates and include the following:

� Arranging events for international networking, and exchange of experience and technology between youngprofessionals and students

� Inspiring the young generation to join and actively participate in ITA� Encouraging member nations to establish domestic YM groups for each individual ITA member nation.To date, SANCOT-YMG has embarked on a number of activities, including researching other existing professional youth

organizations and groups in order to understand how they are structured, what their current involvement is, and whereSANCOT-YMG can participate in the promotion of young professionals’ interests. To date these include ITA–YM,SAIMM–YPC, and CESA-YPF.

As a way forward, SANCOT–YMG intends to embark on the following activities:a) Requesting assistance from the mother body in the formulation of the working committee/councilb) Continue engaging with various young professionals’ organizations and other related stakeholders in an attempt to

strengthen relationships and pursue youth interestsc) Start implementation of the ITA-YM mandates in association with the SAIMM mandatesd) Continue to participate in the activities of the ITA-YM, SAIMM-YPC, CESA–YPF, and other youth groups both locally

and internationally.All interested young professionals and those who would like to participate in general and offer assistance in the

sustenance of this your professional entity are invited to contact the SANCOT–YMG chairman directly [email protected] or via Raymond van der Berg on [email protected]

L. Nene

The Journal of The Southern African Institute of Mining and Metallurgy APRIL 2015 �vii

closed off my March President’s Corner by making the point that we are in the ‘Age ofUnicorns’ and that without strong math and science skills our future engineers andmanagers will not be adequately equipped to meet the needs and expectations of the

national economy or for managing business complexity in the future. By way ofexample, a 2014 Department of Higher Education and Training report indicated that sixof the top ten occupations in high demand were for graduated/certificated engineers

(see the Table below, and of the remaining four occupations, three required highertechnical training in engineering.

It will take significant efforts at all levels ofour society to rectify ongoing imbalances in theeducation system that seems ill equipped tomotivate more learners to excel in maths andscience and go on to study engineering at asenior level. However, it is all too easy to lookto the education system as having the soleresponsibility to do this. Perhaps we should firstlook at how we, as parents, guide our childrenand nurture their inquisitive and analyticalminds from an early age. More should and canbe done to educate and inform parents abouttheir educational responsibilities and change theculture which seems to imply that; once we havepacked the kids off to school, that is the end ofour educational responsibility as a parent.tt

This edition of the Journal showcases some of our industry’s young engineering talent, who have risen to the abovechallenge, and publishes the work they are doing to better understand some of the technical issues facing both our miningand extractive sectors. Most of the authors either have started, or are about to start, their working careers. The SAIMM hastwo main initiatives through which it strives to contribute towards the ongoing development of engineers, specifically forthe mining industry:

1. The SAIMM Scholarship Trust Fund. This Trust channels financial assistance to underprivileged and talentedundergraduates. We already have numerous case studies of lives being transformed. The SAIMM desperately needsthe support of its member’s contributions to this cause, simply because the results are so immediate andmeasureable

2. The Young Professionals Council: Many of you reading this article will recall with mixed memories and emotionsyour first two to three years working for your first boss who was both task-driven and not a particularly good coach… There are so many young engineers that do not handle this transitional period well, and our Young ProfessionalsCouncil has been tasked specifically to find ways of staying close to graduates and diplomates at this critical start totheir careers, in order to offer friendly advice, support, and guidance.

Technology continues to drive the demand for engineers. Safety and efficiency are drivers of mechanization andautomation in mines and manufacturing around the world; smartphones, cheap sensors, and cloud computing haveenabled a raft of new internet-connected services that are infiltrating the most tech-averse industries—Uber is roiling thetaxi industry; Airbnb is disrupting hotels. Perhaps ongoing research towards continuous mining systems will also re-invent the mining industry? Certainly, technology entrepreneurs are exploiting the new technology opportunities.

So what about the ‘Age of Unicorns’?This also relates to the pace of technology change that is driving the demand for engineers of all disciplines. You will

recall that unicorns are mythical creatures that existed in people’s imaginations? Well, the billion-dollar tech start-up wassupposed to be the stuff of myth; neither Google not Amazon were in the billion dollar league on start-up (Aileen Leecoined the term unicorn as a label for such corporate creatures.). There are now (according to Fortune) more than 80companies with more than this value at start-up. And it is accelerating: in 2013 there was one company with a start-upvaluation of $10 billion but today in 2015 there are eight (including Uber, the on-demand car service worth $41.2 billion.Its valuation is higher than the market capitalization of at least 70% of the companies in the Fortune 500!)!

Why does this all matter? Because these start-ups are not in the ‘classical’ engineering space but are pulling the topmath and science talent from the built environment. We need to look after our own and ensure that even when the miningindustry is in challenging times we do not neglect to invest in our young professional engineers.

J.L. PorterPresident, SAIMM

President’s

Corner

Table I

Top 10 Occupations in high demand in South Africa

No. Occupational Title

1 Electrical Engineer

2 Civil Engineer

3 Mechanical Engineer

4 Quantity Surveyor

5 Project Manager/Engineer

6, 7 Finance ManagerPhysical and Engineering Science Technicians*

8, 9 Industrial and Production Engineers*Electrician

10 Chemical Engineer

�

viii APRIL 2015 The Journal of The Southern African Institute of Mining and Metallurgy

Handover of model stope to Wits School of

Mining Engineering

09 March 2015 – Johannesburg: A life-size mining stope panel was handed over by New Concept Mining (NCM) to the WitsSchool of Mining Engineering on 6 March 2015, to help students learn about stoping activities through a bettervisualization of how a real mine looks.

The stope panel – sponsored to the tune of R250 000 by NCM – is part of a range of simulated facilities sponsored anddeveloped at the School’s premises on West Campus, in partnership with companies active in the mining sector such asAveng, Gold Fields, and Sibanye. These include a mine tunnel, mine shaft steel work, and alamp room.

Professor Cuthbert Musingwini, newly appointed Head of the School of MiningEngineering at Wits, said: ‘We are delighted to add this new facility to our School’sresources and grateful to be partnering with far-sighted stakeholders like NCM who shareour dedication to skills and technology development.’

NCM marketing director Brendan Crompton said the sponsorship of the model stopepanel was driven by NCM’s commitment to safety, efficiency, and productivity in SouthAfrican mines. The SA-based company is a market leader in narr0w-reef stope supportproducts, and has expanded into a number of countries worldwide.

‘As a quality-focused company rooted in South Africa, we recognize that the future ofour mining sector is built on the calibre and skills of graduates from institutions like WitsUniversity,’ said Crompton. ‘Partnering with the School of Mining Engineering at Wits is oneof the ways that we contribute to sustainability and safety in mining, especially as we bothprioritize technological innovation as a key factor in the success of the sector.’

Measuring some seven metres in length, the model stope was constructed from a metalframework, mesh, and concrete. Sculptor Russell Scott used various materials and techniques,including hand-packed cement and layers of paint, to achieve the realistic effect of a workingstope face in an underground platinum mine.

The panel dips at 10 degrees, has a stoping width of one metre, and extends some threemetres on strike. It has been equipped with various items of support infrastructure to demonstrate to students the variety oftechnologies employed underground. These include timber props, timber packs, rockbolts, and safety nets suspended nearthe working face.

NCM has contributed roof support equipment both from its own range of products and from other sources. It is alsomaking available some of its electronic monitoring and warning devices in the stope, augmenting the School’s focus ondigital remote monitoring technologies to enhance safety on mines.

Like the recently completed model mine tunnel, the stope panel is situated in the basement of the School of Minespremises, where it incorporates one of the building’s beams as a geological feature.

Professor Fred Cawood, former Head of the School, initiated constructionof the stope panel as part of his digital mine research at Wits Mining. Thesesimulated facilities form part of the ‘digital mine’ environment which isproviding invaluable tools for learning and research, bringing a real mineexperience to mining engineering students at Wits.

‘Most of the 200 first-year students we welcome each year are straightfrom school and have never been in a mine before,’ said newly appointedHead of School Professor Cuthbert Musingwini. ‘Although mine visits arearranged from time to time, this facility gives easy access to students – sothat they can visualize and test what they are studying theoretically.’

‘While the facility is invaluable for our teaching work, it will also bemade available to our research students as they push the boundaries ofproductivity with digital and other technology in mining,’ said ProfessorCawood. ‘Now more than ever, South Africa needs to encourage and facilitateresearch that can stimulate our mining sector; through facilities like these,Wits School of Mines is showing its commitment to doing that.’

S. Braham

View looking up-dip showing the edge of the gully,mechanical props, and gully pack

Professor Cuthbert Musingwini, Head ofthe School of Mining Engineering, Wits,opening the official handover

The Journal of The Southern African Institute of Mining and Metallurgy APRIL 2015 �ix

9 March 2015 –Johannesburg: ProfessorCuthbert Musingwini hasbeen appointed head ofthe University of theWitwatersrand’s Schoolof Mining Engineering.

Having lectured atWits since 2004, ProfessorMusingwini has over 20 yeaof experience in the mining sector –including mine production management andplanning, consulting, and academia. He beganhis career in the Zimbabwean gold miningindustry as a research fellow – and later alecturer – at the University of Zimbabwe.

He is Senior Vice-President and HonoraryTreasurer of the Southern African Institute ofMining and Metallurgy (SAIMM), a Fellow ofthe SAIMM, a registered professional miningengineer with the Engineering Council ofSouth Africa (ECSA), and holds a PhD inMining Engineering from Wits. He is aManaging Editor of the International Journalof Mining, Reclamation and Environmentpublished by the Taylor and Francis Group(UK). He was awarded a National ResearchFoundation (NRF) C3 rating in 2014, and haspublished and presented extensively bothlocally and abroad.

S. Braham

arssector –

New head of Wits

mining school

announced

Realising possibilities...

...from mine to market.

WorleyParsons adds value through our full scope of services f servicesfrom pit to port including studies, mine planning, impact assessments, permitting and approvals, project management,t,construction management and global procurement.

www.worleyparsons.com

37,50043 166 peoplecountries offices

Environment & Approvals

Non-Process Infrastructure

MineePlanning

Mining & Mine gininM einM&Development

MaterialsHandling

Resource ceEvaluation

Mineral alProcessing

Tailings & Waste etsas & WgniliaTManagement

Smelting emS ginlte& Refining

Transportto Market

g & i ei i gg

Smelting engs & WastTaili emSetsas & WgniliTaT ginlte

g & i ei i gg

Smeltingengs & WastTaili emSetsas & WgniliaT ginlte

g & i ei i gg

Smelting engs & WastTaili emSetsas & WgniliaT ginlte

g & i ei i gg


g & i ei i gg


g & i ei i gg

Smeltinge ngs & WastTaili emSetsas & WgniliaT ginlte

g & i ei i gg


g & i ei i gg


g & i ei i gg


g & i ei i gg

Smelting engs & WastTaili emSetsas & WgniliaT ginlte

g & i ei i gg


g & i ei i gg


ts, permitting and aassessmentudo port including som pit tfr

oalue thrarsons adds vyPorleW

niotaualvEceruosRe

olevDegininM

nginnalPeinM

eR&& emS

tenemganaMetsas & WgniliaT

gnissceorPalreinM

i i g & i egg

eltinglings & Waste SmeeSSeettssaas & Ws & WggiiailTa lliiaa ginlte

t,oject managemenals, prvoappries, mine planning, impact

f servicesope oough our full sc

eturucstrarfnIsescorPn-oN

lsaovrppA&ntemnroivnE

ginldnHalsiaretaM

tneopmeinM&

tekrao MttporsnarT

nginfieginlte

i i g & i egg

eltingggininltltlings & Waste SmeeSSeettssaas & Ws & WggiiailTa lliiaa ee

i i g & i egg


ffiotriesounc 61643

omyparsons.corlewww.w

t andtruction managemenonscp g

i i g & i egg


ffices people,500377,t.emenocurd global pr

j gppp

i i g & i egg


�

x APRIL 2015 The Journal of The Southern African Institute of Mining and Metallurgy

Wits-SRK link boosts rock engineering skills17 March 2015 – Illovo, Johannesburg: Collaboration between Wits University’s Schoolof Mining Engineering and consulting engineers SRK Consulting is nurturing scarcerock engineering expertise, benefiting the mining and other sectors in Africa andbeyond.

For the past decade, SRK has partnered with Wits through providing financialsupport for selected students in the Wits School of Mining Engineering’s postgraduaterock engineering research programme, as well as internship opportunities within thefirm.

’The bursary programme has allowed some of our top students to specialize inrock engineering, which is a key discipline for mining but which for various reasonsattracts relatively little postgraduate interest among graduates,’ said ProfessorCuthbert Musingwini, Head of the School.

The scheme was initiated by Professor Dick Stacey, then Centennial Professor ofRock Engineering at Wits, who approached a number of companies and organizationsin the mining sector to seek their help in dealing with numerous requests from brightbut under-resourced students wanting to undertake MSc studies.

‘I was delighted when SRK took up this challenge, and also offered to take in someof the students as interns,’ said Professor Stacey, who spent 25 years of his careerworking at SRK and is today Professor Emeritus at the Wits School of MiningEngineering. ‘Internships are hugely valuable for postgraduate students, giving themreal-life work experience and practical mentoring while leaving space for them tocomplete their studies.’

SRK partner and principal consultant William Joughin has been integrally involved with the bursary students who haveinternships at the firm.

‘This partnership helps us identify the best MSc students to assist us with many of our projects,’ said Joughin, ‘and it isheartening to see how they develop their skills during their time with us.’

He said that ten students had gone through the bursary-intern route at SRK, while another six SRK employees have completed –or are busy with – an MSc at the Wits School of Mining Engineering. Two more have been employed after they earned their MScdegrees.

Highlighting the quality of the candidates who have benefitted from the scheme, Joughin said: ‘Three of the previous studentshave received medals from the Southern African Institute of Mining and Metallurgy (SAIMM) for papers based on their MSc projectwork carried out at SRK. Some have also received awards from the South African National Institute of Rock Engineering (SANIRE)for top marks in the Chamber of Mines Rock Mechanics certificate.’

Mining engineer Joseph Mbenza Muaka worked on mining operations and at Mintek before completing his MSc at Wits – withsupport from another bursary provider, Coaltech – while working as an intern at SRK. He recently presented a paper at the SouthernHemisphere International Rock Mechanics Symposium (SHIRMS) on numerical modelling.

SRK’s status as a leading consulting firm with strong roots in technical excellence also allows interns to be exposed to cutting-edge investigations. Intern Prince Mulenga, currently busy with his MSc at Wits, will be involved in a project funded by the Safety inMines Research Advisory Committee (SIMRAC) at the Mine Health and Safety Council – also an important partner of the Wits Schoolof Mining Engineering.

The internship system has allowed some of the MSc graduates to progress within SRK and to become mentors to the newerinterns. Philani Mpunzi, who completed his studies in 2011 under Professor Stacey, is now a specialist 3D modeller for SRK andhelps interns to make the most of their time while optimizing their contribution. Mpunzi and another SRK/Wits student, TazibanaMoyo, co-authored a paper on their MSc research at the SHIRMS conference.

‘Having worked in Zimbabwe’s mining sector for six years as a production supervisor and mineplanning engineer, I appreciate being able to share my experience while contributing to thedevelopment of young rock engineers,’ said Mpunzi.

While SRK does not have capacity to absorb all its interns, there is considerable opportunity toprogress through the ranks. One of the first interns, Robert Armstrong, joined SRK as a researchstudent in 2001 and a full-time engineer in 2005; last year, he was promoted to associate partner.

‘Perhaps one of the most valuable aspects of studying and doing research while engaged by SRK isthe ability to get relatively easy access to highly experienced experts in fields like rock engineering,’said Armstrong.

Another positive element of the Wits-SRK partnership is the role played by SRK’s rock engineeringexperts in the postgraduate courses themselves, according to Professor Stacey.

‘At least seven of SRK’s best technical minds have contributed to our MSc courses as guestlecturers,’ he said. ‘There has also been considerable time invested by SRK experts Peter Terbrugge andWilliam Joughin as external examiners for these courses.’

Professor Musingwini said the partnership indicates the way forward for the mining sector in SouthAfrica and beyond our borders; it has already contributed well-qualified rock engineers to companiesoutside SRK and to countries across Africa and abroad.

’It is vital that academia, industry, and the public sector work closer together if we are to successfully overcome the skillschallenge that mining faces and invite other consulting companies to partner with us in other areas of specialization on modelssimilar to the Wits-SRK link,’ he said.

S. Braham

Standing: Philani Mpunzi Rock Engineer SRKConsulting (SA), Prince Mulenga Student Intern SRKConsulting (SA), William Joughin PrincipalGeotechnical Engineer and Partner SRK Consulting(SA), Prof Emeritus Dick Stacey of the Wits School ofMining EngineeringSeated: Joseph Mbenza Muaka, Rock Engineer SRKConsulting (SA) and Prof Cuthbert Musingwini, Headof the Wits School of Mining Engineering

Robert Armstrong geologist andassociate partner SRKConsulting (SA)

These papers will be available on the SAIMM websitehttp://www.saimm.co.za

Student PapersRe-aligning the cutting sequence with general support work and drafting a support sequence at Simunye Shaftby K. Lombard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251Roof support awaiting time (RSAT) is the potential production time lost while waiting for roof support to catch up with the continuous miner. The causes of excessive RSAT were investigated, and a number of solutions were identified, of which the use of hard roof drill bits as standard was shown to be the most effective.

Mining through areas affected by abnormal stress conditions at Syferfontein Collieryby C. Legote . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265This paper investigates the conditions leading to the indefinite termination of production in four critical primary panels at an underground coal mining operation. The observed shortcomings in the mining approach were identified, and a strategy is proposed for mining through the affected panels.

A critical evaluation of the water reticulation system at Vlaklaagte Shaft, Goedehoop Collieryby R. Lombard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271The various factors that contributed to the high water-related downtime, which seriously affected production, were investigated. The water reticulation system was reviewed, and the current and future underground pipe layout and water requirements were determined for the shaft.

Optimization of shuttle car utilization at an underground coal mineby P.R. Segopolo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285The purpose of the project was to convert the current shuttle car utilization on an underground coal mine to best practices by focusing on change-out points and tramming routes, which have a major influence on shuttle car away times. Shuttle car utilization can be improved by balancing the number of cars with the number of open splits and the mining sequence.

Explosives utilization at a Witwatersrand gold mine by M. Gaula. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297This investigation examined the properties of explosives, mine standards and recommendations for usage along with the historic relationship between the quantities of explosives used and the production output. This was then compared to the expected quantity of explosives required per unit of production. The biggest contributor to the apparent under-utilization of explosives was found to be the limitations of the system that tracks the usage of explosives underground.

These papers will be available on the SAIMM websitehttp://www.saimm.co.za

Critical investigation into the problems surrounding pillar holing operationsby J.P. Labuschagne, H. Yilmaz, and L. Mpolokeng . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307An investigation into pillar cutting practices was carried out at a platinum mine in order to improve the compliance for pillar cutting. The findings suggest that the pillar strength problem lies with the implementation of the design rather than the pillar design itself.

LHD optimization at an underground chromite mineby W. Mbhalati . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313The factors preventing the load haul dump machines (LHDs) from tramming the target tonnages at an underground chromite mine were investigated. Simulations showed that production improvements of more than 100% could be obtained by reducing the one-way tramming distances and optimizing LHD utilization.

The viability of using the Witwatersrand gold mine tailings for brickmakingby M. Malatse and S. Ndlovu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321This work examines the use of gold mine tailings, in various ratios with cement and water, for brickmaking. The bricks were tested for unconfined compressive strength, water absorption, and weight loss. The results indicated that gold mine tailings have a high potential to substitute for the natural materials currently used in brickmaking.

Evaluation of some optimum moisture and binder conditions for coal fines briquettingby P. Venter and N. Naude . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329The optimum binder and moisture additions to produce a mechanically strong briquette from coal fines were investigated using two different binders.

Air drying of fine coal in a fluidized bed by M. Le Roux, Q.P. Campbell, M.J. van Rensburg, E.S. Peters, and C. Stiglingh. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335The rate of moisture removal from coal particles in a fluidized bed under a range of operating conditions was investigated. It was found that the relative humidity of the drying air has a greater effect than temperature on the drying rate, even at temperatures as low as 25°C. The energy efficiency of the fluidized bed compared favourably with other thermal drying methods.

Mine background and generalinformationSimunye Shaft is an Anglo Thermal Coalunderground mining operation. The mine issituated approximately 41 km fromEmalahleni, Mpumalanga Province. SimunyeShaft has five bord and pillar sections, four ofwhich are continuous miner (CM)-shuttle carsections and one is a CM-FCT (flexibleconveyor train) section. The No. 4 Seam ismined at an average seam height of approxi-mately 3 m.

At Simunye Shaft, the floor, comprisinglargely sandstone or a sandstone/siltstone

combination, is expected to be reasonablycompetent. The immediate roof, however,consists of an interlaminated unit of shale andsiltstone and may present roof stabilitychallenges. The roof encountered duringmining varies from soft to hard, and hard roofconditions may result in support challenges(Mathetsa, 2013).

Project background‘Roof support awaiting time’ (RSAT) is a termused by Goedehoop Colliery to describe thepotential production time lost due to the CMstanding idle waiting for roof support to catchup. There are two types of RSAT, namelyoperational RSAT and engineering RSAT.

Operational RSAT (Figure 1) is driven byoperational processes e.g. the roofbolter fallingbehind the CM due to adverse geologicalconditions, damaged roofbolts, materialshortage, cutting out of sequence (which maylead to logistical problems that will prevent theroofbolter from finishing support in time), etc.Engineering RSAT is potential production timelost due to roofbolter breakdowns. Supportawaiting time has proven to be a majorbottleneck in production at Simunye Shaft. Asillustrated in Figure 1, support awaiting timeamounted to 1700 and 1400 hours in 2012and 2013 respectively. This means anadditional 280 000 t could potentially havebeen produced in 2013. On average, almost14% of available in-section production timewas lost due to operational RSAT. The minelost a potential R125 million in revenue, andalthough this was less than in 2012, it wasstill an enormous loss. The RSAT of the No. 4Seam CM-shuttle car sections amounted toapproximately 72% of the total RSAT, andtherefore this project covers only No. 4 SeamCM-shuttle car sections at Simunye Shaft.

Re-aligning the cutting sequence withgeneral support work and drafting asupport sequence at Simunye Shaftby K. Lombard*The work presented in this paper was carried out as partial fullfilment for the degreeBEng (Mining Engineering)

Synopsis‘Roof support awaiting time’ (RSAT) is a term used at Goedehoop Colliery’sSimunye Shaft to describe the potential production time lost due to thecontinuous miner (CM) standing idle waiting for roof support to catch up.Investigations revealed that in 2013, Simunye Shaft had approximately1400 hours of RSAT, which suggests that the mine could have potentiallyproduced an additional 280 000 t of coal. This project consisted of twoparts. Firstly, the causes of the high RSAT and means to improve thesituation were investigated. Secondly, as insisted by mine management, theCM cutting sequence was investigated as a possible cause of high RSAT.Machine-related challenges due to the roofbolter installing support tooslowly, geological conditions (mostly hard roof conditions and slips),logistical challenges pertaining to the CM cutting sequence, man-relatedchallenges related to operator fatigue, re-support, operator inexperience,and the absence of support targets were identified as main contributors toRSAT. Furthermore, results showed that the roofbolters in the sections atSimunye Shaft are slower than the CMs. A target of 28% reduction in RSATwas set. Experts from Kennametal and Fletcher were consulted to findsolutions for the identified causes. In total, eight solutions for RSAT wereidentified, but the solution that contributed most significantly to reducingRSAT was to use hard roof drill bits as a standard product at SimunyeShaft. Calculations showed that by using hard roof drill bits, RSAT can bereduced by 43%, which is more than the specified 28% target.

The cutting sequences at Kriel, Greenside, and Simunye Shaft, togetherwith three newly developed cutting sequences, were simulated using theUCMS (Underground Coal Mining Simulation) program. A re-aligningprinciple was incorporated into the newly developed cutting sequences toalign the cutting sequences to general support work and to reduce RSAT. Adecision matrix revealed that a cutting sequence in which boxing takesplace in R3 (third road to the right of the belt road) and in which the re-aligning principle has been incorporated will be the best option for SimunyeShaft. The recommended cutting sequence will lead to a 5% increase inproduction.

Keywordsroof support awaiting time, CM cutting sequence, simulation, hard roof drillbits, support sequence.

* Faculty of Engineering, Built Environment andInformation Technology, Department of MiningEngineering, University of Pretoria.

© The Southern African Institute of Mining andMetallurgy, 2015. ISSN 2225-6253. Paper receivedJan. 2015

251The Journal of The Southern African Institute of Mining and Metallurgy VOLUME 115 APRIL 2015 �

ISSN:2411-9717/2015/v115/n4/a1http://dx.doi.org/10.17159/2411-9717/2015/v115n4a1

Re-aligning the cutting sequence with general support work

fSupport is a major component of the production process.If three faces are left unsupported (according to AngloAmerican Thermal Coal standards) the CM has to wait forsupport before production can commence. Therefore, toimprove section productivity, it is vital to make the timeousinstallation of roofbolts a priority.

The question arises as to what may be the causes of highRSAT. It was the author’s goal to identify the main causes ofthe high RSAT at Simunye Shaft and to suggest strategiesthat will make the timeous installation of roofbolts a priority.

Based on underground logbook data, the causes of RSATcan be divided into five main categories (Figure 2):� Problems related to the roofbolter being too slow or on

breakdown (machine problems)� Geological conditions: hard roof conditions, conditions

in which slips are encountered and in which oslo strapshave to be installed

� Logistical challenges (when the roofbolter is blockingthe CM) due to the cutting sequence not taking theinteraction between the CM and roofbolter into account

� Man-related challenges – operators supporting tooslowly or arriving late for work. Re-support andmaterial shortage also fall into this category. Operatorsneed to re-support when roofbolts are damaged duringthe roofbolting process or when the spacing betweenthe roofbolts is inadequate

� Infrequent events, includes when support is updated, atemporary support jack has to be installed, and whenthe roofbolter has to wait for the LHD to completesweeping.

As indicated in Figure 2, machine, geology, and logisticalchallenges are the main contributors to RSAT, contributing82% of the problem.

It should be noted that the logbook data was veryincomplete and that more than half of the RSAT could not beaccounted for. Only 190 data points out of total of 800 wereused as a result. Owing to the incompleteness of the data, asurvey was conducted among underground workers to obtaina better understanding of the causes of the high RSAT.Twelve surveys were completed (results depicted in Figure 3).The installation of oslo straps and hard roof conditions wereidentified as main causes of the high RSAT. Most of the shiftbosses raised the concern of a lot of new inexperiencedwworkers in their sections, and one shift boss mentioned that60% of his operators were new and had not received

ff fsufficient on-the-job training from the retiring workers. Mostof the workers mentioned that the CM is faster than theroofbolter and that this is the cause for the high RSAT.Engineering breakdowns were also mentioned as a problem,and may be attributed to the fact that maintenance onroofbolters is not seen as a big priority.

It is important to identify and investigate the main aspectsthat contribute to RSAT so that RSAT can be reduced. Bytaking the analysis in Figure 2 and the survey results in Figure3 into consideration, the following will be investigated:� Machine considerations – means by which to increase

the speed of roofbolt installation. Slow roofbolting isthe main contributor to RSAT

� Geological conditions (including hard roof conditionsand slips) – this is the second highest contributor toRSAT

� Logistical issues – means by which to improve the CMcutting sequence

� Man-related challenges – operator fatigue as acontributor to RSAT

� Other challenges – re-support� Operator inexperience� The implementation of support targets (bolts installed

per shift).Engineering breakdowns will not be investigated, because

these are machine-related and can be prevented only bymeans of a better maintenance plan (increase in maintenancetime, maintenance staff, and more reliable equipment).

�

252 APRIL 2015 VOLUME 115 The Journal of The Southern African Institute of Mining and Metallurgy

Figure 1 – Total downtime hours awaiting roof support – Simunye Shaft Figure 2 – Underground logbook RSAT data analysis 2013

Figure 3: Survey results: causes of RSAT

f fInvestigating the engineering breakdowns of the roofbolter asa contributor to RSAT’s is, however, suggested as a topic forfurther work. It should also be noted that purchasing newroofbolters (so that there are two roofbolters available persection) was eliminated as a solution, seeing that SimunyeShaft does not have the capital to purchase new roofbolters.

The high RSAT called into question the current CMcutting sequence, which was developed with a main focus onthe CM. Will a new cutting sequence with a support approachimprove RSAT? A deeper investigation into the cuttingsequence currently employed at Simunye Shaft will beperformed as the mine sees it as a priority to optimize itscutting sequence by re-aligning it to general support workand thereby reducing RSAT.

Objectives and methodologyThe objectives that were identified during the course of theproject as well as the methodology used to meet theobjectives are set out in Table I.

Literature surveyAn extensive investigation into work that has already beendone to reduce RSAT at underground coal mines, yieldedonly limited information. This may be attributed to the factthat, in general, underground data – for example the numberof roofbolts installed and amount of drill steels used per shift– is not recorded in an organized or accurate manner.Therefore, the causes of the RSAT cannot be pinpointedeasily. The analysis of underground logbook data in order todetermine the causes of RSAT is a time-consuming process.

f fThe literature study consisted of five parts. Firstly, astandard support sequence was described. This assisted inunderstanding the roof support process and provided astarting point from which improvements could be made.Secondly, the changes brought about in the support processwhen slips are encountered were described. Slips are one ofthe causes of RSAT – support spacing is reduced when a slipis encountered, which increases the time required forinstalling support. Thirdly, the main causes of RSAT at KrielColliery were investigated to strengthen the motivation forthe study. Fourthly, solutions that may reduce RSAT wereinvestigated. Finally, the cutting sequences employed atSimunye Shaft, Greenside Colliery and Kriel were described toillustrate where improvements in Simunye Shaft’s cuttingsequence could be made.


The Journal of The Southern African Institute of Mining and Metallurgy VOLUME 115 APRIL 2015 253 �

Figure 4 – Plan view of the ideal roofbolt installation sequence for adouble-boom roofbolter (Van Staden, 2014)

Table I

Identified objectives and methodology to attain the objectives

Objective Methodology

Conduct a root cause analysis of the high RSAT at Simunye Shaft • A survey was undertaken among underground workers to obtain theiropinions regarding the causes of the high RSAT

• Underground logbook data was analysed to determine the main causes ofthe high RSAT

• The underground logbook data analysis and the survey results enabled themain areas for improvement to be identified.

Reduce RSAT by 28% • It was necessary to determine the time it takes to support a 9 m heading inorder to determine what effect new technology may have on RSAT

• Experts from Fletcher and Kennametal were consulted regarding newtechnology that could be implemented to reduce RSAT

• The cost of the various solutions and initiatives to reduce RSAT was takeninto consideration to ultimately make recommendations.

Re-align the CM cutting sequence to general support and formulate • Underground observations and interviews with underground workers a support sequence assisted in identifying the main areas of concern regarding to the current

cutting sequence at Simunye Shaft• Experts from other mines were consulted to obtain information with regard

to their cutting sequences• Rock mechanical and ventilation standards were taken into consideration

when cutting sequences were developed• A UCMS simulation program was used to simulate the different cutting

sequences (three proposed cutting sequences and cutting sequences ofother mines) in order to select the best cutting sequence for Simunye

• A logical analysis of the optimal developed CM cutting sequence assistedin drafting a support sequence.

Support targets • Support targets were set up for various underground scenarios (when ahard roof or slips are encountered and for normal conditions)

• The advance rates of the roofbolters were determined through time studiesand industry data

• The effective production time per shift (approximately 3 hours per shift) andthe support advance rate for the various scenarios were used to determinesupport targets.


Standard roofbolt installation sequenceIn Figure 4, the red dots indicate where roofbolts areinstalled, and the numbers indicate the sequence in which theroofbolts are installed (mining takes place from left to right inthe figure). Looking in the direction of mining, support startsat the beginning of the heading at the far left-hand and farright-hand side simultaneously (two roofbolts are installed atthe same time) and then proceeds to the inner left-hand andright-hand side. The same installation sequence is employedat Simunye Shaft.

SlipsIf slips are encountered at Simunye Shaft, the spacingbetween two consecutive lines of support is reduced from 1.5m (normal conditions) to 1 m as illustrated in Figure 5. Itshould be noted that, in the figure, mining takes place fromthe bottom upward. If multiple slips are encountered, oslostraps need to be installed in addition to the reduced spacing.

RRSAT at Kriel CollieryBy analysing the RSAT at another colliery, the significance ofthe problem can be emphasized and the motivation for thestudy strengthened. Kriel Colliery had 1401 hours (almostexactly that of Simunye Shaft) of operational RSAT in 2012.This illustrates that other collieries have the same types ofproblems and a mind-shift is needed to overcome the problem– roof support needs to become a higher priority. Figure 6illustrates the breakdown of the RSAT at Kriel Colliery for2012. Machine-related challenges are documented as the

highest contributor to RSAT, mainly due to the roofbolterbeing too slow. Logistical issues related to the CM cuttingsequence are 10% lower than that of Simunye Shaft. Itshould, however, be noted that almost 64% of the data pointsin this analysis could not be used as they were recorded as‘n/a’, or unaccounted for. It is clear that RSAT is not only aproblem at Simunye Shaft, but also at the other collieries inSouth Africa. It is of great importance to solve this problem toultimately enable collieries to increase their efficiencies.

Possible solutions

The oslo strap holderAccording to Steyn (2013), a consultant at Fletcher, oslostrap holders (which are attached to the roofbolters atGreenside Colliery) have the potential to reduce the time toinstall an oslo strap by more than one minute. The oslo strapholder (Figure 7) assists the roofbolter operator to positionthe oslo strap (which normally takes a considerable amountof time) and in doing so increases safety significantly.

Standardizing the drill bits used at Simunye Shaft to hardroof drill bitsA hard roof decreases the penetration rate and bit life (bitswill have to be changed more frequently). Both the reductionin penetration rate and the excessive replacement of drill bitslead to an increase in support time. Hard roof drill bits areavailable and are used by the mine for hard roof conditions.The hard roof drill bits should increase the life of a single drillbit and may increase the penetration rate.

Table II contains information with regards to thepenetration rates and costs of the different drill bits that canbe supplied to Simunye Shaft and the different roofconditions encountered. The figures are estimated and may

�


Figure 5 – Plan view of the change of roofbolt spacing from normal conditions to when a slip is encountered

Figure 6 – RSAT breakdown at Kriel Colliery for 2012 (derived fromunderground logbook data)

Figure 7 – Oslo strap holder (Steyn, 2013)

vary with operator skill, geology, and the consistency ofvary with operator skill, geology, and the consistency ofadjusting the roofbolter’s settings if roof conditions change.

The KCV4 1 RRWT (Table II) is the current drill bitemployed at Simunye Shaft for normal roof conditions. If ahard roof is encountered, the SV119AE K3012EX02 drill bit(Table II) is used. The PROBORE1 HSVSL (not used atSimunye Shaft) has very good heat resistant properties,wwhich increases drill bit life, but its high price makes itunsuitable for use as a standard product (Bosch, 2013).WWhen used in hard roof conditions, the KCV4 1 RRWT drillbits (which are designed for normal roof conditions) and drillsteel heat up rapidly and melt into the adapter. Removing thedrill steel from the adapter can easily take 30–60 seconds(Bosch, 2013). By standardizing to hard roof drill bits, theinstallation time per roofbolt can be decreased by approxi-mately 30 seconds (Table II). It should be noted that if KCV41 RRWT drill bits are used for hard roof conditions, the timeto install a roofbolt may increase to 8 minutes (Table II).

The torque indicating systemAccording to Sinden (2013), the installation quality of asupport system is directly related to the performance of themachinery used to install the bolts. Statistics shows that only20% of all bolters have torques set within the correct

Newton-metre range (200–350Nm). Sinden also mentionsthat the operator torques the roofbolt according to what hethinks or sees is right, and that that is the reason whycommon faults such as over-torque (flattened washers) andunder-torque (loose washers) conditions occur. In the eventof over-torque, the bond between the resin and bolt may bedamaged, the washers may be deformed, the nuts may berounded, and roofbolt threads may be damaged. Under-torque results in loose washers, incorrect mixing of resin, notbreaking the shearing pin, and not flattening the torqueindicator (Sinden, 2013).

The torque indicating system was designed to avoidsubstandard torque (which may require re-support of roofbolts). The torque indicating light is clearly visible duringoperation and ensures accurate bolting torque. The operator,miner, and technician can also see when the torque of theroofbolter drill chuck is not optimal (Sinden, 2013).

The torque indicating system has a data logging featurethat can capture data such as the torque, time, date, thenumber of roofbolts installed, spinning time of resin, andholding time before a roofbolt reaches torque (to helpestimate accurate roofbolt and resin usage). The torqueindicating system will therefore result in the bettermanagement of roofbolting crews, as managers will knowhow the crew has performed (number of bolts installed pershift). The system will also reduce RSAT resulting from re-



Table II

Relationship between type of drill bit used and the installation time of a roofbolt for various roof conditions(Bosch, 2013)

Drill bit type Cost (R) Installation time of Installation time of Life of drill bit Life of drill bit Life of drill bit Life of drill bitone roofbolt in one roofbolt in penetrating coal penetrating penetrating coal penetratingstandard roof hard roof (number of holes sandstone with intrusions quartzsite

conditions (min) conditions (min) drilled before (depends on (depends on or harder roofreplacement) hardness) (number hardness and (number of holes

of holes drilled the intrusion) geometry ofbefore (number of holes drilled before

replacement) drilled before replacement)replacement)

KCV4 1 RRWT 37.20 2.5 8 200 0-14 3-7 1-4SV119AE K30 43.71 1.6-2.5 6-8 100 5-14 2-5 1-512EX02 PROBORE1 60.16 1.6-2.5 6-8 300 25 0-10 0-7HSVSL

Table III

Average time (actual stopwatch time) to support a9 m heading (Van der Merwe, 2012)

Activity Total Average Averageaverage LH RH(h:min:s) (h:min:s) (h:min:s)

TRS up (re-position time ) 00:00:36 Total Drill 00:00:50 00:00:48 00:00:53 Insert extension drill rod 00:00:21 00:00:14 00:00:29 Change to bolting equipment 00:00:49 00:00:50 00:00:47 Bolting 00:00:17 00:00:16 00:00:18 Re-position for next bolt 00:00:50 00:00:47 00:00:53 TRS down (time per row ) 00:07:58Bolts per heading 25 (9 m) Time per heading 00:50:46 Relocation time 00:20:25

Figure 8 – Current cutting sequence at Simunye Shaft. Green blocks(numbers 1– 28) are the first sequence and the repetition of thesequence is indicated by the red blocks (numbers 29– 56). B/R – beltroad, FB – feeder breaker, L1 – first left road, R1 – first right road etc.Triangular shapes in R3 indicate where boxing takes place


support being required. The torque indicating system is a newproduct on the market and its advantages can be summarizedas follows (Sinden, 2013):� The torque indicating light is highly visible� If the system fails to operate correctly, the operator can

identify and report immediately� Accurate torque on every roofbolt is ensured� The system helps to overcome human errors and faulty

hydraulic systems� The system is cost-effective� It reduces machine working hours� Time on re-installation is saved� Labour and re-installation costs are reduced� The system eliminates the re-occurrence of over- and

under-torque.

The MCS roofbolter monitoring systemMCS offers two options for data recovery (operating time,tramming time, downtime, and number of roofbolts installed)with the roofbolter system. The first system, which iscurrently in use at Anglo Thermal Coal sites, is a flash cardsystem. The operator is responsible for inserting the flashcard at the beginning of the shift and returning it to thecontrol room for processing at the end of shift. The second isa Wi-Fi system, where the onboard data collection unitcommunicates to the node which is integrated into the mine’scommunication backbone, allowing file transfer to the controlroom. This can be done in two ways – firstly, by means oftwo nodes, and secondly by means four nodes. Theadvantage of installing more nodes is increased coverage.The MCS system and torque indicating system are similar.The biggest difference between the two systems is cost – thetorque indicating system is more cost-efficient.

The auto-bolterAccording to Steyn (2013) a roofbolter operator handlesapproximately 1.5 t of steel per shift and makes approximately14 lever movements per roofbolt installed. The operation of aroofbolter involves strenuous tasks, and the time to install aroofbolt can increase from 2.5 minutes at the beginning of ashift to approximately 10 minutes at the end of a shift as aresult of operator fatigue. The weighted average time to install abolt is 6.75 minutes. This was calculated by increasing the timeto install a bolt linearly every 30 minutes over a 3 hour(effective operating time) period.

The autobolter technology, which is currently beingimplemented at Greenside Colliery, may eliminate thisproblem. The autobolter has the following advantages (Steyn,2010):

Safety:� Reduces the number of accidents related to the

handling of roofbolts and the operation of theroofbolter

� Moves the operator to a safer position.Productivity:� Reduces operator fatigue� Ensures the consistent installation of bolts.Reliability:� Removes the human factor� Records roof mapping information� All bolts are installed to the same standards and

procedures.

Engineering:� The autobolter has various pressure settings for feed

and rotation� It is difficult to tamper with the machine setup� Pressures and sensors are displayed on a display

screen.The autobolter ensures that each bolt is installed to the

correct standard (re-support is eliminated) and operatorfatigue is eliminated completely as the roofbolter is remote-controlled and the operators do not have to handle the heavyroofbolts. However, the cost of the autobolter – approximatelyR14 million –eliminates it as a solution as the mine’s budgetdoes not cater for the purchase of new roofbolters.

Current cutting sequence employed at Simunye ShaftIt is necessary to analyse and evaluate the effectiveness ofthe cutting sequence currently employed at Simunye Shaft, asit was developed with a main focus on the CM (a supportsequence was never developed). A support sequence can bedescribed as the sequence in which the roofbolter supportsthe headings and splits cut by the CM. The current supportsequence employed at Simunye consists of the roofbolterfollowing the CM sequentially as far as possible. Theshortcomings of the cutting sequence, with regard to supporthave to be identified in order to identify solutions to theproblem of RSAT. The current cutting sequence employed atSimunye Shaft is illustrated in Figure 8. The green blocks(numbers 1–28) are the first sequence and the repetition ofthe sequence is indicated by the red blocks (numbers 29–56).The belt road (B/R), the first left road (L1) to the fourth leftroad (L4) and the first right road (R1) to the fourth right road(R4) are indicated. The triangular shapes in the R3 indicatewhere boxing takes place. Boxing is when a triangular shapeis cut into the coal face to make it easier for the CM tomanoeuvre when it is cutting cuts number 4 and 1 of eachsequence. The feeder breaker (FB) is also indicated in thefigure. Simunye Shaft’s cutting sequence ends in R2, which isclose to where the following cutting sequence starts. Boxingtakes place in R3 so that through ventilation is established asquickly as possible.

Cutting sequences at other collieriesBy investigating different cutting sequences employed byother mines (with more or less the same pillar sizes asSimunye Shaft), an optimal cutting sequence for SimunyeShaft can be developed.

GreensideFigure 9 illustrates the cutting sequence employed atGreenside Colliery. Boxing takes place in R1 and the cuttingsequence ends at the far left-hand side.

KrielFigure 10 illustrates the cutting sequence followed at Kriel.Boxing takes place in R1 and the cutting sequence ends closeto where the following cutting sequence starts.

In summary the following concepts can be incorporatedinto a cutting sequence:� To box in R1 (reduce cable handling efforts and time)� To box in R3 (establish through ventilation as soon as

possible)� To end the cutting sequence at the far left-hand side

�


� To end the cutting sequence close to where the nextcutting sequence will start.

A combination of these concepts can be used to developvvarious cutting sequences for Simunye Shaft. The developedcutting sequences can then be simulated to determine whichone would be the best for Simunye Shaft.

DDeveloping an optimum cutting sequenceAccording to Shaw (2013), the following factors need to betaken into consideration when a cutting sequence isdeveloped:� The tramming and cable handling of the CM must be

minimized� Through ventilation must be established as soon as

possible� The roofbolter should not be in the way of the CM� The tramming routes for all three shuttle cars should be

optimized.Hirschi (2012) conducted a study on Identifying Optimal

MMining Sequences for Continuous Miners. In this study hementions the following guiding policies and practices whendeveloping a cutting sequence:� Mine crosscuts should be in the direction of ventilation

airflow� A buffer should be maintained between the continuous

miner and roofbolter.Rock mechanical and ventilation standards also need to

be taken into account. In terms of rock mechanical standards,not more than three faces may be left unsupported. If three

f f ffaces are left unsupported, the CM has to wait for support tocatch up. The most important ventilation standard that needsto be taken into consideration is that an air speed of 1 m/sneeds to be maintained in the last through road.

The UCMS simulation programUCMS can be used to simulate cutting sequences. Bychanging input variables such as shift length, pillar sizes,probability of equipment breakdown, speed of the CM, andspeed of the roofbolter, production rates and tramming timevalues can be obtained. This program was used to simulatethe cutting sequences used at Kriel, Greenside, SimunyeShaft, and other developed cutting sequences.

ResultsThe results of the investigation are presented under thefollowing topics:� Time study on supporting a 9 m heading at Simunye

Shaft. Calculations that follow will be based on thistime study

� The effect of geological features on the time to supporta 9 m heading. Both the time study and the effect ofgeological features on support time will be used to setup support targets

� A comparison of the advance rates of the CM and theroofbolter. This will help to determine whether theproblem (of RSAT) lies with the roofbolting process

� The effect of implementing hard roof drill bits as astandard product

� Improving the cutting sequence currently employed atSimunye Shaft. Two developed cutting sequences andGreenside and Kriel’s cutting sequences are simulated,and the performance of the four cutting sequencesevaluated. A third cutting sequence is developed toimprove on the results of the first four cuttingsequences.

Time studyThe time it takes to install a line of support is required todetermine whether the roofbolter is too slow and whereimprovements can be made. If the time taken to install a lineof support is known, RSAT logbook data can be used toquantify the benefits of various systems that can be used toreduce RSAT. A summary of a time study, conducted by Vander Merwe (2012), is set out in Table IV. The roofbolt instal-lation sequence, illustrated in Figure 4, was used. The time



Figure 10 – Cutting sequence at Kriel Colliery (Odendaal, 2014)

Figure 11 – Roofbolter exposed to water from the scrubber fan of theCM and dustFigure 9 – Cutting sequence at Greenside Colliery (Odendaal, 2014)


study was carried out in Ubhejane section at Simunye Shaft.The bolting took place in normal roof conditions.

A few terms need to be understood to interpret the timestudy correctly. Total time per row can be described as thetime from when the TRS (temporary roof support) is up untiljjust before the TRS is lowered. Table III summarizes theaverage time it takes to install a roofbolt on each side (LHS orRHS) as well as the total time to support a 9 m heading andthe relocation time of the roofbolter. Relocation time is thetime from when the roofbolter starts moving to the nextheading to the time that the TRS of the roofbolter is up at thenew heading, and repositioning time is the time to repositionthe bolter to the next hole that needs to be drilled and bolted.Repositioning time also includes the time to lift and lower theTRS.

The average time to insert an extension drill rod isusually around 14–29 seconds (Table III). It was found thatthe time to insert an extension rod can be extended by almost2 minutes if a drill bit has to be changed. Drill bit changescan therefore have a major impact on the time to install a lineof support.

The changeover to bolting equipment usually takes 49seconds (Table III), but during the time study it was foundthat if resin stock runs out it can add approximately 1 minuteto the time. Material shortage or poor planning can thereforealso contribute to a slower installation time.

The time to support a 9 m heading was determined to beapproximately 50 minutes (Table III). Therefore, if a supporttarget has to be set for normal conditions, a time of 50minutes can be allocated to supporting a 9 m heading.

The effect of geological features on support timeIf hard roof conditions or slips are encountered, supportspacing is reduced, as described in the literature survey. Thiswwill result in an increase in support time due to the fact thatmore roofbolts will have to be installed. The increase in timeneeds to be established so that a realistic support target can

f fbe set for situations when such features are encountered. As indicated in Table IV, the time to support a 9 m

heading can increase from approximately 50 minutes innormal conditions (Table III) to approximately 75 minuteswhen a feature is encountered.

Performance of CM vs. roofbolterIt is necessary to determine whether the CM or roofbolteradvances faster, as this will indicate where the problem ofRSAT lies and where improvements can be made. Table Vindicates that the CM is faster than the roofbolter in allpossible scenarios. The roofbolter can therefore not keep upwith the CM in normal conditions. This is a contributingfactor to the high RSAT. If the roofbolter advance rate isincreased, RSAT may be reduced.

It should be noted that the relocation time (20 minutes)of the roofbolters was taken into account to determine theadvance rate. In this report, the advance rate of theroofbolters will be improved to increase the CM advance rate(which already has RSAT intrinsic to it) and reduce RSAT.The increase in the average roof support advance rate willcontribute directly to the increase in production.

Solutions to RSATEight solutions to the challenge of RSAT were identified, butthe most significant solution was to implement hard roof drillbits as a standard product at Simunye Shaft. Only thissolution and the use of support targets are discussed in thisreport.

If normal KCV4 1 RRWT drill bits are replaced with hardroof SV119AE K3012EX02 drill bits (Table II), the life of thedrill bits will increase as well as the roofbolt installation rate.Since the roofbolters at Simunye Shaft are slower than theCMs, the increase in installation rate can contribute directly toreducing RSAT. The hard roof drill bits will prevent operatorsfrom continuing to support using KCV4 1 RRWT drill bits inhard roof conditions while they wait for the hard roof drillbits to arrive. Using normal drill bits in hard roof conditionsmay cause the drill steel and drill bit to expand and becomestuck in the adapter, which can result in time wastage.

To determine the benefit of standardizing on hard roofdrill bits, the average roof support advance rate has to bedetermined. For this it is necessary to take all the factors thatcan reduce the speed of support into account. These include:� The installation of oslo straps� The reduction in support spacing when slips are

encountered� Support in hard roof conditions� Re-support� Material shortage

�


Table IV

Increase in support time when a geological featureis encounteredA Time to support a 9 m heading in normal conditions (min) 50B Lines of support in normal conditions (9 m heading/1.5 m 6

spacing)C Lines of support if spacing is reduced to 1 m 9

(9 m heading/1 m spacing)D Time to support 9 lines of support (ratio calculation) (min) 75E % increase in time 50F Relocation time (Table III) 20G Roof support advance rate (m/min) (9 m heading/(D+F)) 0.095

Table V

CM and roofbolter advance rates

Section Mining height Average production CM advance Roof support advance Roof support advance Roof support(m) rate (t/h) rate (m/min) rate in normal rate when slips are advance rate in hard

conditions (m/min) * encountered (m/min) roof conditions[Table IV] (m/min) *

Imvubu 3.1 465 0.23 0.13 0.095 0.076Ubhejane 3 322 0.17 0.13 0.095 0.076Khomonani 3 436 0.22 0.13 0.095 0.076

� Engineering breakdowns.The support advance rate under the conditions specified

needs to be determined in order to ultimately calculate theweighted average support advance rate throughout the year(Table VI). The weightings (indicating the ranking of eachcondition) of the respective support advance rates wereestimated from underground logbook data. The weightedaverage roof support advance rate was calculated to be 0.15m/min.

When hard roof drill bits are implemented as a standardproduct, the time to install a roofbolt can be reduced by 30seconds (Table II). This means that the time to install a lineof support can be reduced by one minute. A new supportadvance rate for each condition can be calculated. The newweighted average support advance rate (Table VII) can bedetermined by allocating the same weights (as in Table VII)to the respective advance rates.

The increase in the support advance rate can be addeddirectly to the CM advance rate, as the CM advance ratealready includes the effect of RSAT. The average CM advancerate for Imvubu, Khomonani, and Ubhejane will increase to0.221 (Table VIII), which is a 5% improvement in produc-tivity.

As indicated in Table IX, an additional 53 000 t of coalcould have been produced in 2013, which is equivalent to440 hours of RSAT. The result is a 43.13% reduction in

RSAT. The additional cost per year if hard roof drill bits areimplemented as a standard product at Simunye Shaft (forthree sections) will amount to R1.8 million. The data in TableIX shows the additional roofbolts that will have to beinstalled if production is increased annually by 53 000 t forthe three sections.

Support targetsTo increase awareness of the importance of the timeousinstallation of roofbolts, support targets can be set. Metretargets for the CMs are always visible in the sections and helpto formulate goal-orientated tasks for crews underground.The same effect can be created by setting support targets.

Table X indicates support targets for the No. 4 Seamsections at Simunye Shaft. The different conditions that mayarise during roofbolting have to be considered when the targetsare set. Therefore, targets for normal conditions, hard roofconditions, and conditions where slips are encountered areindicated in Table X. No cost is associated with this solution.

Cutting sequence developmentThe support challenges arising from the cutting sequenceemployed at Simunye Shaft can be summarized as follows:� While the CM is cutting cut number 9 (Figure 11), the

roofbolter will be supporting cut number 8 from theleft-hand side. This means the roofbolter will beworking against ventilation. Workers (roofbolteroperators) will be exposed to a lot of dust from the CMand water from the scrubber fan on the CM. This isrepeated when the CM cuts numbers 11, 13, 15, 17,and 19. This scenario is illustrated in Figure 11

� When the roofbolter is supporting cut number 8 fromthe left-hand side, shuttle cars will be moving in R2



Table VI

Roof support advance rate – weighted average(2013)

Condition in which support takes Roof support Weightingplace advance rate in (out of 1000)

stated condition(m/min)

Normal 0.13 806Multiple slips (oslo strap installation 0.08 40and support spacing reduction)Slips (support spacing reduction) 0.095 20Hard roof 0.076 34Re-support 0 20Material shortage 0 30Engineering breakdowns 0 50Roof support advance rate – 0.11weighted average (m/min)

Table VII

Improved roof support advance rate – weightedaverage

Condition in which support takes Roof support Weightingplace advance rate in (out of 1000)

stated condition (m/min)

Normal 0.14 806Multiple slips (oslo strap installation 0.09 40and support spacing reduction)Slips (support spacing reduction) 0.105 20Hard roof 0.082 34Re-support 0 20Material shortage 0 30Engineering breakdowns 0 50Roof support advance rate – 0.12weighted average (m/min)

Table VIII

Increase in average CM advance rate due to anincrease in the support advance rateA Average CM advance rate in three sections (m/min) 0.21B Increase in roof support advance rate (m/min) 0.01

[0.12 - 0.11]C % increase in production [B/A * 100] 5D CM advance rate after standardisation (m/min) [A +B] 0.221

Table IX

Benefit of implementing hard roof drill bits as astandard at Simunye ShaftA Average yearly production for 3 sections 1 800 000

(Imvubu, Ubhejane, Khomonani) (t)B 5% of the average yearly production (tons) [A * 5%] 90 000C Equivalent RSAT (h) 440D Contribution to reduced RSAT (%) [C/1020] 43.13E Potential saleable tons 53 000F Potential revenue (R million) 40.12G Shifts/year (3 sections) (Mphasha, 2014)1 2050H Average lines of support installed per shift per 30

section (Van der Merwe, 2013) (includes extra bolts for production increase)

I Additional costs (R per roofbolt) 7J Additional cost per year (R million) [I * 30 * 4 1.8

bolts per row *G]2

(1) Three production shifts, 5.33 days a week. Therefore: 3 shifts*227days*3 sections≈2050 shifts(2) A 5% contingency was applied


f ftowards the CM. The temporary roof support of theroofbolter may not be able to reach the end of the splitthat needs to be supported or may block the shuttlecars that are approaching the CM. This scenario isillustrated in Figure 12.

In response to the challenges presented by the currentcutting sequence, the re-aligning principle was developed.The re-aligning principle refers to aligning the CM cuttingsequence with general support work. The concept is explainedin Figures 13 and 14. In Figure 13, cuts number 8 and 9create the challenges. In Figure 14, where the cuttingsequence is aligned with support, there is a buffer betweenthe CM and roofbolter. Consecutive cuts (numbers 8 and 9,10 and 11, 12 and 13, etc.) are further apart. This results inthe following advantages:� Safer condition, because the roofbolter and CM will be

further apart (lower risk of collision) and the roofbolterwill not have to work against ventilation

� The roofbolter operators will not be exposed to dustand water from the CM

� The CM will not obstruct the path of the roofbolter andRSAT will be reduced.

The following cutting sequences were developed forsimulation.

CCutting sequence 1As illustrated in Figure 15, boxing takes place in R3 and thecutting sequence ends at the far left-hand side (a lot oftramming time is expected). The re-aligning principle wasincorporated into the cutting sequence.

Cutting sequence 2

As illustrated in Figure 16, boxing takes place in R1 and thecutting sequence ends at the far left-hand side (lesstramming time than cutting sequence 2 is expected). The re-aligning principle has been incorporated into the cuttingsequence.

It should be noted that both these cutting sequences wereapproved by the ventilation department as well as the rockengineer at Goedehoop Colliery.

Simulation resultsIn total five cutting sequences were simulated:� Cutting sequence 1� Cutting sequence 2� Kriel’s cutting sequence� Goedehoop’s cutting sequence (the cutting sequence

employed at Simunye Shaft)� Greenside’s cutting sequence.

�


Figure 12 – Temporary roof support of bolter blocking shuttle carentrance to CM

Figure 13 – Cutting sequence not aligned to support

Figure 14 – The re-aligning principle – cutting sequence aligned tosupport

Figure 15 – Cutting sequence 1

Table X

Support targets, Simunye Shaft

Section Metres target per shift Support target – normal roof Support target – roof with slip Support target– hard roofconditions (no. of roofbolts)* (no. of roofbolts )* conditions (no. of roofbolts)*

Imvubu 42 94 70 55Ubhejane 43 94 70 55Khomonani 42 94 70 55

* Calculated using an effective cutting time of 3 hours per shift and information in Table III

Figure 17 illustrates that cutting sequence 1 had thehighest tramming time. It can therefore be confirmed that thegreater the distance from the last cut of the sequence to thefirst cut of the following sequence, the greater the trammingtime. Goedehoop’s cutting sequence, which is the cuttingsequence employed at Simunye Shaft, had the lowesttramming time. This may be because of the followingreasons:� Goedehoop’s cutting sequence ends close to the first cut

of the following sequence (Figure 18 shows theopposite scenario – a sequence ending far from the firstcut of the following sequence)

� When boxing takes place in R1 tramming time isadded, because the CM has to move all the way fromR4 (cut number 11) to R1 (cut number 12), asillustrated in Figure 19, to continue the sequence. Thisadditional tramming time is eliminated if boxing takesplace in R3 (Figure 20).

Kriel’s cutting sequence came out top in the comparisonof tons booked (Figure 21). A cutting sequence thatincorporates the re-aligning principle and has a high enoughproduction output still needs to be developed, as neithercutting sequence 1 nor cutting sequence 2 resulted in a betterproduction output performance than Kriel.

A third cutting sequence was therefore developed to

improve on the results obtained, combining all the conceptsthat resulted in the highest production and lowest trammingtime.

Cutting sequence 3:As illustrated in Figure 22, boxing takes place in R3 (throughventilation will be established soon) and the cutting sequenceends close to the start of the following cutting sequence. Itcan be seen that the re-aligning principle was incorporatedinto the cutting sequence.

This cutting sequence was approved by the ventilationdepartment as well as the rock engineer at Goedehoopcolliery.

Figures 23 and 24 illustrate the simulation results. Cuttingsequence 3 has a lower tramming time than cutting sequence 1



Figure 17 – Simulation results: comparison of the total tramming timesof the simulated cutting sequences

Figure 18 – Increase in tramming time due to the cutting sequenceending far away from the first cut of the following sequence

Figure 19 – High tramming time as a result of boxing in R3 and endingthe sequence far from the first cut of the following sequence

Figure 20 – Low tramming time due to the last cut of the sequenceending close to the first cut of the following sequence



and average shift production is greater than all the othercutting sequences. Cutting sequence 3 had a 5% greateraverage shift production than the current cutting sequenceemployed at Simunye Shaft. This means the mine canpotentially earn an additional R40 million in revenue per year.

With any of the cutting sequences that incorporate the re-aligning principle, the support sequence will be exactly thesame as the cutting sequence of the CM, as the roofbolter willbe following the CM sequentially.

The results relating to the cutting sequence developmentand simulation are summarized in Table XI.

Analysis and evaluation of resultsIn this section, the economic viability of the identifiedsolutions are discussed, the simulation results analysed, andthe optimum cutting sequence for Simunye Shaft identifiedby means of a decision matrix.

Economic viability of solutions to RSATThe costs of mining a ton of coal is set out in Table XII.

Standardizing the drill bits used at Simunye Shaft to hardroof drill bitsThe break-even analysis (Table XIII) shows that the cost ofthe hard roof drill bits will be recovered in only 2 months. Itshould, however, be noted that the cost will be incurred

�



Figure 23 – Simulation results: comparison of the total tramming timeof the simulated cutting sequences (including cutting sequence 3)

Table XI

Summary: optimum cutting sequence development

Description Cut 1 Cut 2 Kriel Greenside Goedehoop Cut 3

Average tramming time/ 29.8 25.9 24.6 25.1 23.7 27.8shift (min)Average shift production 2 410.5 2 322.4 2 423 2 374.1 2 394.8 2 432.7(t)Is the re-aligning Yes x x xprinciple incorporated No x x xinto the sequence?Is there a buffer between Yes x x xthe roofbolter and CM? No x x x

Figure 21 – Simulation results: comparison of the tons booked for eachsimulated cutting sequence

Figure 24 – Simulation results: comparison between the total bookedtons for the simulated cutting sequences (including cutting sequence 3)

annually. The capital outlay for the hard roof drill bits is lowand this is therefore a viable option to implement.

Optimum cutting sequenceIn order to select the optimum cutting sequence, a decisionmatrix was set up. The cutting sequences are evaluatedagainst seven criteria, namely:� Does the sequence incorporate the re-aligning

principle?� Does the sequence result in a low or high tramming

time?� Does the sequence give rise to a high production rate?� Does the sequence allow the roofbolter to work against

ventilation?� Is there a buffer between the CM and roofbolter?� Is through ventilation established sooner or later?� Is the effort of cable handling high or low?

The cutting sequence with the highest score will berecommended. The means of rating the cutting sequences isdescribed to illustrate how the decision matrix was puttogether.

Does the cutting sequence incorporate the re-aligningpprinciple?Seeing that the re-aligning principle increases safety and willhelp reduce RSAT, a high weight has to be attached to it inthis selection phase. If the cutting sequence incorporates there-aligning principle, a score of 5 is allocated, and if not, a 0is allocated.

Does the sequence result in a high or low tramming time?Excessive tramming time is inefficient. The higher thetramming time, the lower the potential production. Six cuttingsequences were simulated and therefore the sequence withthe lowest tramming time is awarded a score of 6 and thesequence with the highest tramming time a 1.

Does the sequence give rise to a high production rate?Production is directly related to profit. A score of 6 isawarded to the cutting sequence with the highest productionoutput and a 1 to the sequence with the lowest.

Does the sequence allow the roofbolter to work againstventilation?Working against ventilation is not good practice and workersafety is enhanced if the roofbolter does not work againstventilation. If the cutting sequence allows the roofbolter towork against ventilation a score of 0 is allocated to thecutting sequence, and if not a 3 is allocated.

Is there a buffer between the CM and the roofbolter?A buffer between the CM and roofbolter will increase safetydue to the fact that roofbolter operators will not be exposed tothe dust from the CM and water from the scrubber fan of theCM. If a buffer between the CM and roof bolter is maintained,a score of 4 is allocated to the cutting sequence, and if not, a0 is allocated.

Is through ventilation established sooner or later?If through ventilation is established at the start of thesequence, ventilation needs will be met in a more effectivemanner. If through ventilation is established later, additionalfans will have to be installed to maintain an air speed of 1m/s in the LTR. If through ventilation is established soonerrather than later a score of 2 is allocated to the cuttingsequence, and if not, a 0 is allocated.

Is the effort of cable handling high or low?If cable handling requires less effort, fewer problems canarise to reduce production time. Worker morale will alsoimprove. Cable handling efforts can be divided into categoriesas indicated in Table XIV. When boxing takes place in R3 thecable has to be suspended all the way to R3 from R1 (wherethe transformers are), and when the sequence moves towardsR1 again, cable handling is doubled. However, if boxingtakes place in R1 (where the transformers are) a lot of cablehandling effort is eliminated, and flexibility is increased.

The cutting sequence with the highest score is sequence 3(Table XV). This cutting sequence will ensure safer workingconditions, increase production, and reduce RSAT.

ConclusionIn completing a root cause analysis of the high RSAT atSimunye Shaft, underground logbook data was analysed andunderground workers were interviewed. Seven maincontributors to RSAT were identified: machine-relatedchallenges relating to the roofbolter installing support tooslowly (the greatest contributor to RSAT), geologicalconditions (mostly hard roof conditions and slips), logisticalchallenges pertaining to the CM cutting sequence, man-related challenges (operator fatigue), re-support, operatorinexperience, and the absence of support targets.

A 28% reduction in RSAT was set as the target. The useof hard roof drill bits as a standard at Simunye Shaft was



Table XII

Total cost of producing a ton of coal

Description Cost (R/ton)

Mining cost 110Plant washing cost 60Rail cost 150Total 320

Table XIII

Break-even analysis: standardizing to hard roofdrill bitsA Potential extra tons produced per year (ROM) 90 000B Cost per ton (R) 320C Total cost (R million) [A * B] 28.8D Potential revenue (R million) 40.12E Potential profit per year (R million) [D - C] 11.32F Cost of standardizing to hard roof drill bits (R million) 1.8

[Table IX]G Payback period (months) [F/E] 2

Table XIV

Weight allocation to cable handling efforts

Magnitude of effort Timing Weight allocated

High Boxing in R3 0Low Boxing in R1 1


f fidentified as the best method to address the root cause ofhigh RSAT (slow roofbolt installation). By implementing thissolution, RSAT can be reduced by 43.13%, which is morethan the 28% target. The total cost of implementing thesolution will be R1.8 million per year, with a maximumpayback period of 2 months. Simunye Shaft will potentiallyincrease its revenue by R40 million by implementing thissolution.

Logistical issues with regards to the CM cutting sequencewwere also identified as a cause of RSAT, and were thoroughlyinvestigated as suggested by mine management at SimunyeShaft. Challenges with regard to the shaft’s cutting sequence(that may lead to RSAT) were identified and the cuttingsequences of Kriel and Greenside Colliery were analysed toimprove the situation. A re-aligning principle that increasesthe buffer between the CM and roofbolter and prevents theroofbolter from working against ventilation, was developed.Three cutting sequences incorporating the re-aligningprinciple were developed and simulated together with thecurrent cutting sequence of Simunye Shaft, Kriel, andGreenside Colliery. A trade-off study revealed that cuttingsequence 3 had the most promising outcome, and althoughthe effect on RSAT could not be quantified, it was verifiedthat by implementing this cutting sequence Simunye Shaftcould increase production by 5%. The support sequence forcutting sequence 3 is equivalent to the cutting sequence itself– the roofbolter can follow the CM sequentially.

Support targets were set up by using the calculatedroofbolter advance rate in the various scenarios (hard roofand normal conditions, and when slips are encountered). Theadvance rate was then multiplied by the effective productiontime per shift. Findings showed that on average 94 roof boltshave to be installed per shift in normal conditions, 70 boltswwhen slips are encountered, and 55 when hard roofconditions are encountered. Targets for three scenarios wereset, ensuring that the targets are fair and do not demoralisethe work force.

Recommendations It is recommended that Simunye Shaft should adopt hard roofdrill bits as standard and that support targets for each shiftbe set. It is also recommended that cutting sequence 3 shouldbe implemented.

SSuggestions for further workThe following topics are suggested for further work:� An investigation into reducing engineering-related

RSAT� An in-depth study of the cable handling logistics

surrounding the implementation of cutting sequence 3in this study

� A study on optimizing logbook data recordings, whichwill avoid inaccurate data recording

� An in-depth study pertaining to operator fatigueexperienced by a roofbolter operators

� Quantifying the effect of changing operators mid-shift,and setting support targets, on RSAT

� A feasibility study on introducing an auto-bolter intosections at Simunye Shaft

� An investigation into change management aspects thatneed to be taken into account when a new cuttingsequence is to be implemented at a colliery.

AcknowledgementsI would like to thank L.M. Mphasha, my mentor at the mine,and J.A. Maritz, my supervisor at the University of Pretoria,for their guidance and support.

ReferencesBOSCH, C. 2013. Personal communication.HIRSCHI, J.C. 2012. A Dynamic Programming Approach to Identifying Optimal

Mining Sequences for Continuous Miner Coal Production Systems.Dissertation. Southern Illinois University Carbondale.

MPHASHA, L.M. 2014. Personal communication.MATHETSAMM , S. 2013. Personal communication.MCS. 2013. MCS Roof Bolter Monitoring Proposal for Goedehoop Colliery.

Witbank: MCS.ODENDAAL, A. 2014. Personal communication.Sinden, J. 2013. Consultant: Fish Eagle Mining Solutions, Personal communi-

cation.STEYN, J. 2010. Introducing the Fletcher Twin Boom, Automated Roof Bolter.

Witbank: J.H. Fletcher & Co.SHAW, E. 2013. Mining Engineer: Khutala Colliery, Personal communication.STEYN, J. 2013a. Material Handling. J.H. Fletcher & Co., Witbank.STEYN, J. 2013b. Personal communication.VAN DERVV MERWE, B. 2014. Personal communication.VAN DERVV MERWE, B. 2013. Personal communication.VAN DERVV MERWE, B. 2012. Simunye Shaft Ubhejane section – 9m heading.

Witbank: MCSVANVV STADEN, M. 2013. Department of Mining Engineering – Underground Coal

Mining Methods. University of Pretoria �

�


Table XV

Decision matrix: optimum cutting sequence

Cutting Does the Does the Does the Does the Is there a Is through Is the effort of Total scoresequence sequence sequence sequence give sequence buffer between ventilation cable handling

incorporate the result in a low rise to a high allow the the CM and established high or low?re-aligning or high production roofbolter to the roofbolter? sooner orprinciple? tramming time? rate? work against later?

ventilation?

Cutting 5 2 4 3 4 2 0 24sequence 1Cutting 5 3 1 3 4 0 1 20sequence 2Goedehoop 0 6 3 0 0 2 0 13Greenside 0 4 2 0 0 0 1 8Kriel 0 5 5 0 0 0 1 13Cutting 5 1 6 3 4 2 0 26sequence 3

IntroductionSyferfontein Colliery is an underground coalmining operation situated in Trichardt,Mpumalanga Province, within the Highveldcoalfields. At an average depth of 90 m belowsurface, the mine exploits the 4-Lower coalseam. The bord and pillar method used is fullymechanized, producing on average 2100 t percontinuous miner (CM) per shift. Mineexpansion required the sinking of a secondaryventilation shaft, the commissioning of whichwas synchronized with the mining of the fourprimary panels. The primary panels, onceintersected (Figure 1), would serve as the mainintake and return airways serving the newventilation shaft. The panels would also serveas access to millions of tons in proven reserveslocated in the upper eastern block (Figure 2).Commissioning of the ventilation wascompleted, but mining through the panels washalted due to abnormal stress conditions. Thenature of the conditions required investigationin order to formulate a strategy to realign themine with its planned objectives.

The long-term mine plans were based onnumerous factors, chiefly geological in nature.Syferfontein is riddled with large geologicalstructures including dykes, sills, burnt coal,jointed zones, paleo-lows, and downthrowfaults. One such structure is the 13 m widedolerite dyke that separates the mine’sRiversdale and Weltervreden operations. Thefocus of this study is on Riversdale.

Study focus

Events surrounding the termination ofproduction in the primary panelsBoth sections were mining concurrently in theirrespective panels. Section 3 was mining to thenorth, while section 6 was mining to the east.Mining parameters were aligned with thestandard of 24 m × 24 m pillar centres, 4.1 mmining height, a 7.2 m road width, andadvancing 15 m before installing permanentroof support. The factor of safety wasdetermined as 2.28. Omitting the effect ofvertical loading, however, a potential high k-ratio above 2.5 was thought likely (Steenkamp,2013).

Investigation of the conditions that led tothe halt in production within all four primarypanels yielded the following results.

Section 3� Extensive guttering between pillars,

varying in thickness from minor skinningto 45 cm thick chunks, both in supportedand unsupported areas within the section,was first observed in panel R31 NorthIntake, where bolting density wasincreased from the standard four bolts perrow to six, with the row spacing reducedfrom 2 m to 1 m

� The cutting distance prior to installationof roof support was reduced (sometimes

Mining through areas affected byabnormal stress conditions atSyferfontein Collieryby C. Legote*Paper written as project work carried out in partial fulfilment for BEng(Mining Engineering) degree

SynopsisThis paper investigates the conditions leading to the indefinite terminationof production in four critical primary panels at an underground coal miningoperation, the observed shortcomings in the mining approach, and theproposed strategy to mine through the affected panels. Initial assessmentof the abandoned panel conditions indicated time-dependent strata failure,(i.e. bolted roof failure overrunning intersections), which occurred frommere minutes to up to four weeks post-production, with and without priorwarning of failure. This prompted the constant re-supporting of back areas,which raised safety and productivity concerns. Investigation of the initialmining conditions revealed that the failures were due to a criticalcombination of factors, the chief of which was isolated horizontal stress.Other factors that were initially overlooked by the mine (i.e. influence ofhydraulic stress, misinterpretation of borehole data), resulted in theconditions being described as abnormal. Remedial actions were determined,and in so doing, a new strategic approach was formulated to thoroughlyaddress all failure concerns. The four panels were explicitly planned toserve as the main intake and return airways for the recently commissionedsecondary ventilation shaft, as well as providing access to millions of tonsin proven coal reserves. It is thus imperative to mine the panels. Afeasibility study showed that the proposed strategy set for implementationwould be financially viable.

Keywordscritical primary panels, time-dependent strata failure, horizontal stress,proven reserves.

* University of Pretoria.© The Southern African Institute of Mining and

Metallurgy, 2015. ISSN 2225-6253. Paper receivedFeb. 2015



Mining through areas affected by abnormal stress conditions

to less than 5 m)� Large volumes of strata water entered the working face,

even with the drilled drainage holes� Observed roof failures were time-dependent, occurring

during production and up to four weeks post-production,driving the need to re-support back areas with wiremesh plus additional roofbolts. This had an adverseimpact on planned rates of productivity

� Production was first halted in this panel due to theresulting reduction in productivity

� The section moved to the adjacent R30 North Returnpanel (Figure 3). Similar conditions were anticipated andthe increased support strategy was implemented fromthe onset. Similar failure conditions as in the adjacentpanel persisted, but were more serious. These includedbolted roof failures that overran intersections, falls of

ground within splits, and increased guttering betweenbolts

� As indicated in Figure 3, the nature of the failureconditions drove the section to minimize the number ofmining roads from seven to three. This becameunproductive and posed further strata and ventilationchallenges

� Production in this panel advanced only 250 m from theadjacent panel before it was abandoned due to lowproductivity and safety concerns.

Section 6 � Exactly the same failure conditions as in section 3 were

observed in section 6. Section 6, however, was mining inan eastern direction more than 1 km from section 3.Adverse conditions became apparent in the K22 B&P

�


Figure 1 – Detailed locality of panels on mine plan [scale 1:100 000]. Source: Syferfontein survey department (2013)

Figure 2 – Areas of interest (Riversdale): green indicates mined-out and unmineable areas [scale: 1:100 000]. Source: Syferfontein survey department (2013)

fEast panel, before the section relocated to the adjacentK23 B&P East panel.

� Water-driven falls of ground resulting from bursts ofinrushing water occurred before drainage holes could bedrilled. Occurrence was close to the face and before roofsupport could be installed

� Mining span was also reduced to counteract the falls ofground without increasing the pillar centres.

� Similar to section 3, it was found that the failures in allpanels followed a NE-SW direction and were aggravatedby reactivated joint planes

� Production was eventually abandoned citing similarreasons as for section 3.

Results

Observed shortcomings in the mining approachThe normal mining approach failed to address the root causesof the abnormal conditions due to the following shortcomings.� Long development ends (approx. 15 m) prior to instal-

lation of roof support, particularly underlaminated/layered roofs, allowing parting thatcompromised strata competency upon installingpermanent support, as emphasized by Steenkamp(2013)

� Trapped water in roof (hydraulic pressure) driving fallsof ground even in supported and competent roof areas.This was aggravated where strata was laminated, asfailure under these conditions is violent

� Strata sag or movement was addressed through theinstallation of telltales. These were, however, onlyinstalled in intersections. Installing telltales after rooflayers have parted is ineffective. The section telltaleswere mechanically activated and with the application ofstone dust some would be masked and thus overlookedif activated by further sag or movement in the strata.This made it difficult to control time-dependent failures

� Drainage holes (2 m deep) were drilled only in the

fintersections, thus the sources of strata water were notproperly identified

� Dyke structures have a propensity to concentratestresses around them (Khumalo, 2012). Both sectionswere approaching a 1 m thick dolerite dyke, but itsinfluence on the stress conditions was overlooked asboth section halted production approximately 250 maway from this dyke (Figure 4). The magnitude ofconcentrated stress is not necessarily related to the sizeof the feature but rather to the nature and source energyat deposition, as alluded to by Muaka (2013). This wasevident with similar cases around the mine, yieldingdifferent conditions

� The impact of isolated horizontal stress was notaccounted for in the initial approach. According to Bird tetal. (2006), the principal horizontal stress direction overllthe mine region is oriented NW-SE. This conjecture wassupported by the observed falls of ground, which werepredominantly in a NE-SW direction. Other horizontalstress lead indicators such as buckled bolt plates,reported bumping and spitting in the roof duringproduction, and floor heave (De Clerq, 2013) were noted.The approach to addressing horizontal stress is notalways premeditated as the location and size of thestress source varies throughout the mine

� The survey cross-section over the two sections’ panelsindicated that the immediate strata consisted ofcompetent sandstone. Examination of the rockcomposition of the failures revealed that the immediatestrata in fact consisted of laminated shale and mudstone,which has a major impact on the roof support strategyrequired

� Instrumentation used at the time to monitor movementsin the strata was by means of mechanical telltales. Thesewere seemingly ineffective due to limited visibility whenactivated, more so after the application of stone dust

� The size of the area affected by these abnormal stressconditions cannot be determined due to a lack of detailed



Figure 3 – Map indicating conditions in the section 3 panels leading to abandonment of production


geological data required to prepare the mine stress maps(Van der Westhuizen, 2012).

Suggested strategic approachThe strategy formulated addresses the shortcomings initiallyoverlooked and critical elements that did not form part of themining COP at the time of failure. Safety took precedence,together with uninterrupted productivity such as to avoiddelays caused by time-dependent failures.

A staggered approach with similar dimensions to thenormal modified approach (Figure 5) was investigated as itwwould ideally address the predominant failures in the splits.The mining direction cannot be altered. This approach would

frequire the use of battery haulers and a change in cuttingsequence, which will pose ventilation constraints.

Strategy feasibilityThe feasibility of the proposed strategy was determined using amodel designed in conjunction with Sasol Mining personnel todetermine the productivity and costs to be incurred until thepoint of intersection. This model has inputs for miningparameters as well as the required mining consumables,maintenance, and labour.Although the model does not accountfor operational costs such as electricity, it is nevertheless aneffective planning tool. The results are as follows, based on thenormal-modified strategy parameters.

�


Figure 5 – Mining parameters and layout for proposed strategy

Section 3 Section 6Productivity (tons/shift) 1100 1000Mining duration until intersection 9 weeks 10 weeksExtracted coal (tons) 180 000 200 571Estimated mining OPEX (R mil) 2.59 2.85*Approx. revenue (R mil) 36.00 40.11*Avail. working capital (R mil) 33.40 37.21

*Estimated at a selling price of R200 per ton

Figure 4 – Map of section 6 highlighting conditions in the panels

ConclusionProduction was terminated in four primary panels due to theinability of the mining method to proactively address theabnormal stress conditions, which led to an array of failures.This had a negative impact on productivity, and was furtherhampered by ineffective monitoring techniques and misinter-pretation of geological data. The abnormal conditions were dueto a combination of failure factors that were initiallyoverlooked as a whole. All critical factors have been identifiedand can be addressed using the proposed strategy. Prior tomining the panels a detailed risk assessment will need to beconducted to ensure that safety standards are aligned. At thecompletion of the study, the strategy had not yet beenimplemented, although it was being given strong consid-eration.

Recommendations for further workThe application of the staggered pillar method for similarconditions and parameters, along with its feasibility in smalland large panels as well as the proactive use of extensometers,should be investigated.

References

BIRD, P., BEN-AVRAHAM, Z., SCHUBERT, G., ANDEREOLI, M., and VIOLA, G. 2006.

Patterns of stress and strain in southern Africa. Journal of GeophysicalResearch, vol. III. pp. 1–14.

CAIRNCROSS, B. 2013. Guide to borehole core in the Karoo Basin Coalfields South

Africa (1st edn). Struik Nature, Cape Town.

DE CLERQ, A. 2013. Strata control specialist, Rock Engineering Department:

Sasol Mining. Personal communication.

KHUMALO, S. 2012. Senior Geologist, Syferfontein Colliery: Sasol Mining.

Personal communication.

MUAKA, J.J.M. 2013. Investigation into the magnitude and direction of ground

stresses in the coalfields and their impact on safety and productivity. MSc

dissertation, University of the Witwatersrand, Johannesburg, South Africa

OOSTUIZEN, P. 2006. Geotechnical aspects of the development of the Sigma

Mooikraal underground colliery. ARQ Consulting Engineers (Pty) Ltd.

SALAMON, M.D. and MUNRO, A.H. 1967. A study of the strength of coal pillars.

Journal of the South African Institute of Mining and Metallurgy,

September 1967. pp. 56–67.

STACEY, T.R. and WESSELOO, J . 1998. In situ stresses in the South African

mining areas. Journal of the. South Africa. Institute of Mining andMetallurgy, vol. 98, no. 7. pp. 365–368.

STEENKAMP, M. 2013. Strata control specialist, Rock Engineering Department:

Sasol Mining. Personal communication

VAN DER WESTHUIZEN. 2012. Underground manager: Production Services.

Syferfontein Colliery: Sasol Mining. Personal communication

ZHANG, J., LUI, T., ZHANG, Y., PENG, S., and MENG, D. 2002. Strata failure and

mining under surface and groundwater.

http://www.link.springer.com/chapter/10.1007/978-3-540-73295-

2_9?no-access=true [Accessed 23 Mar. 2014].

SYFERFONTEIN SURVEY DEPARTMENT. 2013. Sasol Mining. �



Table I

Normal-modified layout strategy parameters

Normal-modified layout strategy

Support requirements 1.8 m long rod, 4 bolts in a row, 1.5 m spacingBreakerline installed around each intersection Rapid installation of support (proactive), wire mesh to act as area support Systematic sidewall support (2 bolts at each pillar corner)W-straps in areas of closely spaced joints , upgrade to electronic 2 m long telltales, proactive drilling of drainage holes 2m into the roof (not only at intersections as per COP).

Mining parameters Pillar centres: 24 m × 35 m, bord width 5 m, advance 10 m. Longer pillars in direction of mining, ensure a beam of coal isleft beneath the laminated roof (beam to seam height-mining height > 0.6 m minimum). Drill additional inspection holeswithin roadways during roofbolt installation (not only in intersections) and proactively monitor the strata composition.

Equipment requirements Refurbished CM with a designated maintenance plan. Two refurbished single-boom roofbolters (addresses high supportneeds), designated LHD, 3 × 16 t shuttle cars.

Advantages Adaptable with current cutting sequence, reduced production tempo allows for safer production through adequate time toobserve strata behaviour. Ease of ventilation control. Larger pillars account for effects due to horizontal stresses. Stratastability.

Disadvantages Limited pit room increases equipment congestion thus reducing effective productivity. Rerouting of cables over longerpillars increases relocation time, Routing through ventilation in the LTR will take longer due to reduced advance and longerpillars. Reduced percentage extraction (32.1%) compared to >40% under normal circumstances

Benefits A reduced road width allows for the creation of stress relief zones, thus less area for stress to act. The creation of twoindependent teams (possibly from stoneworks teams) will remove the need to use teams and equipment from high-production sections. Improves stability given the panels will be used as main airways. Using refurbished CM and shuttlecars, which are near their scheduled full overhaul, eliminates the need to buy new machinery for both sections.

ExcellentMineralsSolutions

Copyright © 2012, Weir Slurry Group, Inc. All rights reserved. WARMAN is a registered trademark of Weir Minerals Australia Ltd and Weir Group African IP Ltd; MULTIFLO is a registered trademarks of Weir Minerals Australia Ltd; FLOWAY is a registered trademark of Weir Floway Inc.; GEHO is a registered trademark of Weir Minerals Netherlands bv; WEIR is a registered trademark of Weir Engineering Services Ltd.

Weir Minerals is the world leader in the design and manufacture of pumps, valves, hydrocyclones and wear resistant linings, including WARMAN® centrifugal slurry pumps, ENVIROTECH® dewatering pumps and LINATEX® rubber products for the global mining and minerals processing industries. Our reputation is based on engineering excellence applied to innovative, customer focused solutions for processing minerals and aggressive materials.

In line with our customer driven focus, Weir Minerals Africa also offers a pump rental concept as an attractive alternative to an outright purchase.

For more information contact us on +27 (0)11 9292600

Weir Minerals…Expertise where it counts.

www.weirminerals.com

Mine backgroundGoedehoop Colliery is situated approximately 40km east of Witbank in Mpumalanga Province.Currently Goedehoop has two undergroundshafts – Vlaklaagte Shaft, which is situated inthe southern part, and Simunye Shaft, which issituated in the northern part – which consist of11 sections. The bord and pillar mining methodis employed for coal extraction and each sectionis equipped with one double-boom Fletcherroofbolter, one feeder breaker, three 20 t shuttlecars, and one continuous miner (CM).

Goedehoop produces 8.7 Mt of run-of-mine(ROM) yearly, of which 5 Mt are saleable.Ninety-nine per cent of the coal from

Goedehoop is exported through Richards Bay(Becht, 2010).

Vlaklaagte Shaft is currently mining onlythe No. 4 Seam, as the No. 2 Seam has beenmined out. The shaft produced approximately320 634 t of coal per month in 2013 and madea profit of R120 per ton due to the high-qualitycoal (on average 27.5 MJ/kg) that is extracted atthis shaft (Du Buisson, 2013). The shaftconsists of six sections: Section 1 (Simunye),Section 2 (Magwape), Section 3 (Siyaya),Section 4 (Ngwenya), and Block 7 (Section 5/6and Section 9/10).

The main water source for Vlaklaagte is theKomati Dam. Recycled water from surface issupplied from the return water dam (RWD) tounderground sections 1 to 4 via a pipelinerunning alongside the conveyor belt. Sections 2and 4 have been developed more than 8 kmaway from the RWD.

Water requirementsWater is utilized for many purposes, includingdust suppression, cooling, and cleaning (Table I).

Current water reticulation system atVlaklaagte ShaftAs indicated in Figure 1, clean water wassupplied to the No. 4 Seam undergroundsections (via 200 mm galvanized steel pipes)from the surface water cleaning plant until 27July 2013. The raw water dam received waterfrom the Blesbok reservoir, and the water wasthen pumped to the water cleaning plant toprocess the water to drinking quality. However,the pipes that supplied clean water tounderground workings from the raw water damcorroded. As a result, recycled water from theRWD (via 200 mm galvanized steel pipes) wasused as a substitute. A filtration systemconsisting of 2 µm sieves was installed toremove solids (which cause blockages in the CMand belt sprayers) from the recycled water.

A critical evaluation of the waterreticulation system at Vlaklaagte Shaft,Goedehoop Collieryby R. Lombard*Paper wwritten oon pproject wwork ccarried oout iin ppartial ffulfilment oof BB. EEng. ((Mining EEngineering)

SynopsisWater is a very important component in the production process atunderground coal mines. Current unfavourable economic conditions haveforced the coal mining industry to identify and address every possiblebottleneck preventing optimal production. An increase in water-relateddowntime was identified as one of the bottlenecks at Goedehoop Colliery’sVlaklaagte Shaft. The purpose of this project was to identify the variouscauses that contributed to the high downtime (501 hours in 2013, which ledto a potential profit loss of R12.9 million) and to suggest possible solutions.

After a thorough investigation the main causes of water-relateddowntime were identified as low water pressure and low water flow causedby pipe leakages and bursts. The main root cause for the low water flowand pressure was identified as being the low pressure resistance (1600kPa) of the thin-walled galvanized steel pipes used in the undergroundinbye water reticulation system. The pipes were selected according to theprevious 1000 kPa pressure requirement for the continuous miner.However, the pressure requirement changed to 1500 kPa, which resulted inthe pipes being exposed to much higher pressures than designed for.

The water reticulation system was reviewed and current and futureunderground pipe layout and water requirements were determined for theshaft. The time frame in which the water consumption would be thehighest was determined to be between 1 January 2014 and 7 September2014. Machine and sprayer specifications were used to determine thewater consumption at the shaft.

Three different solutions were considered to solve the water-relateddowntime problem and to ensure the efficient supply of water to the newlyopen sections. Permanent underground concrete dams, semi-mobile dams,or new pipe columns with a higher pressure resistance of 3200 kPa wereconsidered. A trade-off study (taking into consideration cost, time tocompletion and ease of implementation, maintenance requirements, safety,and flexibility) was completed to determine which of these solutions wouldbe most viable.

Keywordswater reticulation, down time, pipe bursts, leakages, cascade dam system,permanent dams, portable dams.


Metallurgy, 2015. ISSN 2225-6253. Paper receivedJan. 2015



A critical evaluation of the water reticulation system at Vlaklaagte Shaft

Since 28 July 2013, water has been supplied to the No. 4Seam from the RWD. The water cleaning plant therefore onlysupplies water to the change houses on surface, as recycledwwater is now being used to supply the underground workings.

Surface pump and pipe layoutThe surface pump and pipe layout consists of a centrifugalpump (pump 2) which pumps water into a 23 000 litre tank.The water from the tank is pumped by a five-stage, 65 kWmulti-stage pump (pump 1) to the underground sections.Figure 2 shows the surface pump and pipe layout. Standard200 mm pipes are used on surface.

Figure 3 is a schematic illustration of the surface tounderground pipe layout, including dimensions that arerequired to calculate the available head.

Underground pipe and pump layoutFigure 4 indicates the underground pipe layout and positionsof different water users in the different underground sections

at Vlaklaagte. Recently a 150 mm standard galvanized steelpipe size was selected and these pipes were tested to withstanda maximum pressure of 1600 kPa (Louw, 2013).

Summary of water requirements at Vlaklaagte ShaftTable II is a summary of the water consumption at sections 1,2, 3, and 4 of Vlaklaagte Shaft (31 December 2013).

Water problems experienced at VlaklaagteThe water-related problems that led to downtime, may beattributed to the following facts.� Water is pumped over very large distances, which means

that major pipe friction losses need to be overcome. Thepressure that is required at the CM has changed over thepast years. Previously the CM required only 1000 kPa ofpressure to operate. Pipes were selected according to thispressure requirement, and thin-wall galvanized pipes,which can withstand only 1600 kPa, were chosen.

�


Table I

Water users and requirements

Water users Requirements

CM sprays Requires water for the following purposes: dust suppression, cooling and cleaning. The CMs operate approxi-mately 9 hours per day. According to Richard Lottering (2013), a Barloworld consultant, CMs requires a flowrate of between 120-135 liters/min and a pressure of 1500 kPa. Failing to adhere to the required flow rate andpressure will result in the CMs tripping which will cause downtime.

Feeder breaker and conveyor belt sprays Three water sprays are fitted on every feeder breaker for dust suppression. A water spray is also required onevery transfer point on the conveyor belt for dust suppression. All the sprays require a water flow rate of 15liters/min at a recommended pressure of 1500 kPa (Pieterse, 2013)

Dust suppression for roads Approximately 60 liters/min is required for road dust suppression (Louw, 2013).Benicon (mini-pit) Benicon is a mini-pit near Vlaklaagte that makes use of the water from the RWD and requires approximately

2.1 liters/min.Cleaning Cleaning requires approximately 120 liters/min (Horac, 2013)

Figure 1—Overview of water reticulation system at Vlaklaagte

However, the pressure requirement at the CM changed to1500 kPa, which exposed the pipes to much higherpressures than they were designed for. No action hasbeen taken so far to change the water reticulation systemto adapt to this higher pressure requirement

� Vlaklaagte is an old shaft and therefore has an ageinginfrastructure, including pipelines. The old infrastructureand increased pump pressures are the main causes offrequent pipe damage and leakages leading to low waterflow and low pressure (or no water flow and nopressure) at the face

� Changes made to the water reticulation system over thepast years (such as changes in the pipe sizes in

underground sections, the change from clean water torecycled water, and changes to pump settings and theinstallation of new pumps) were not well documented



Figure 2—Partially flooded suction currently employed at Vlaklaagte

Table II

Water requirements at Vlaklaagte Shaft (31 Dec 2013) for Section 1-4

Different activities Number of Flow of water DOH Quantity (l/day) Quantity (l/s) Quantity Optimal requiring water required (l/min) (hours/day) (m3/month) Pressure

required (kPa)

CM (Joy) 2 120 9 129 600 4.0 3 888 1500 in pipe but 2000

at CMBucyrus (CAT) 2 135 9 145 800 4.5 4 374Conveyor sprays 25 15 21 472 500 6.3 14 175 1000-1800Feeder Breaker sprays 12 15 9 97 200 3.0 2 916 1000-1800Cars for dust suppression/ 15 000 1.0 450 N/Afire hydrants*Cleaning (4 sections) 4 2 57 600 2.0 1 728 N/ABenicon (Mini pit) 60 000 0.0 1 800 N/ATotal + 10%* 107 5470 22.9 32 264

*10% was included to compensate for losses

Table III

Downtime hours summary (2013)

Section Downtime hours Related lost shifts*

Section 1 50 6.3 Section 2 82 10.3 Section 3 182 22.8 Section 4 187 23.4 Total 501 62.6

*Note: 8 hours represents 1 shift

Figure 3—Surface pipe layout

Figure 4—Underground pipe and pump layout at Vlaklaagte Shaft

� New underground mining blocks, such as the extensionin Block 10, for which the current water reticulation wasnot designed, are being accessed further away from theshaft and the RWD.

Data from the water-related downtime logbook was sortedand analysed to determine the extent of the problem and toidentify possible root causes leading to the high downtime.Block 7 (Section 5/6 and 9/10) was excluded from this investi-gation as Block 7 has a separate water reticulation system inplace.

Table III indicates the total hours of production lost by eachsection from 1 January to 31 December 2013 due to water-related downtime. Sections 3 and 4 contributed the most to thetotal downtime of 501 hours. Solving the problems causing thehigh downtime in these two sections can eliminate 74% of thewwater-related downtime. Sections 3 and 4 were thereforeselected for further investigations.

A summary of the combined impact of the different causeson both Section 3 and Section 4 is shown in the pie diagram(Figure 5). The chart clearly indicates that low water flow andlow water pressure are the two main causes for downtime inthese two sections.

PProduction losses due to downtimeEvery time production stops the mine loses potential profit. The

ftotal potential profit lost in 2013 due to water-relateddowntime was calculated as indicated in Table IV and totalledR12.9 million (Du Buisson, 2013). An intervention wasrequired to stop losses due to water-related problems and toensure that the water requirements over the life of the shaft aremet so that water problems do not occur in the future.

Objectives and methodologyThe objectives and methodology are presented in Table V.

ResultsThe current water reticulation was reviewed to quantify thereasons for the pipe bursts. The future water reticulationsystem was also reviewed in order to determine the final pipelayout and underground dam placement.

Analysis of current water reticulation systemThe pipe layout in Figure 5 can be analysed thoroughly byusing the Bernoulli steady-state energy equation (White, 2011):

[1]


�


Table IV

Water-related downtime cost (1 January 31 December 2013)

Section Hours on stop Cutting rate Potential ROM tons Yield Sales tons Potential Profit loss*(tons/hour)

1 50 313 15625 0.59 9219 1.12 82 323 26486 0.61 16156 1.93 182 341 61971 0.71 43999 5.34 187 347 64796 0.59 38229 4.6Total 501 1323 168878 2.50 107604 12.9

*Potential profit loss = Hours on stop x Yield x Cutting rate x Profit

Table V

Objectives and methodology

Objective Methodology

Quantify the problem The downtime logbook was thoroughly investigated to:• Determine the total production hours lost due to water-related issues• Determine the potential profit that was lost due to water-related down time• Identify sections with the highest downtime; and • Determine the main causes of the high downtimeThe company, MCS, was consulted to determine the DOH of the CMs as well as the cutting rateof the CMs. Information on the yield and profit per ton was retrieved by consulting the financialdepartment.

Review current water reticulation system On-site investigations were conducted including: walking the pipelines, observing the differentwater consumers, manifolds, bends and pumps and where they were located.

Investigate and quantify water consumption for the The water consumption was calculated by investigating machine and sprayer specifications andcurrent water reticulation system also consulting with the Mine Overseer, Shift Boss and Pump Crew at Vlaklaagte Shaft. Determine the life of mine (LOM) water requirements The LOM mining plan (obtained from the planning department) for the shaft was investigated and(to prepare for the future) the Mine Planner and Mine Overseer were consulted in order to determine the LOM water

requirements.Investigate different methods for supplying water to Information gathered from the mine was used. Various suppliers were also consulted including:newly opened sections and solving the water-related • Lectropowerdowntime problem • Eljireth

• IncledonDraw conclusions and make recommendations Recommend the best method for supplying water to the current and newly opened sections.from the results of the investigation

Each term in the equation is a length or a head. α = Kinetic energy correction factor (in problems

common to assume that α = 1)

P2 = Pressure required at the end of the pipe system (atthe CM)

P1 = Pressure at the inletV1 = Velocity of the fluid entering the pipe (zero because

static water is pumped out of the dam) V2 = Velocity of the fluid required at the end of the pipe

system (at the CM)Δz = Height difference/ elevation difference (m).

Equation [2] can be used to correlate the head loss to pipeflow problems (White, 2011).

[2]

wheref = Friction factorD = Inner diameter (m)K = Minor losses (read off from the table in Appendix

H)g = Gravitational acceleration (m/s2)V = Velocity of medium flowing through the pipe (m/s).Every pipe section has a different flow rate because of the

location of the different water users, which results in differentfrictional losses within each pipe section. A number wasallocated to each pipe section in order to differentiate betweenthem (as indicated in Figure 6).



Table VI

Friction factor calculation by using Bernoulli’s equation

Section Length (m) Component K factor Total flow u (m/s) Re Friction factor* Hloss (m)**plus 10%

wastage (l/s)

surface pipe 398 - 0 0.0 0.0 0 0.01964 0.01 100 Standard elbow 0.45 0.0 0.0 0 0.02157 0.02 780 Standard elbow 0.45 9.6 0.5 81 700 0.2263 17.83 600 Standard elbow 0.45 9.4 0.5 79 365 0.0227 1.34 1800 Standard elbow 0.45 9.1 0.5 77 031 0.02277 3.75 700 Standard elbow 0.45 8.8 0.5 74 697 0.02285 1.46 600 Standard elbow 0.45 8.5 0.5 72 362 0.02293 1.17 1320 T piece 0.9 8.3 0.5 70 028 0.02302 2.38 570 Standard elbow 0.45 0.6 0.0 4 669 0.03921 0.09 910 - 0 0.3 0.0 2 334 0.04787 0.010 540 T piece 0.9 7.4 0.4 63 025 0.02331 0.811 460 Standard elbow 0.45 0.6 0.0 4 669 0.03921 0.012 50 Standard elbow 0.45 0.3 0.0 2 334 0.04787 0.013 90 - 0 0.0 0.0 0 0.014 500 - 6.6 0.4 56 023 0.02366 0.615 420 Standard elbow 0.45 6.3 0.4 53 688 0.02379 0.416 140 T piece 0.9 6.1 0.3 51 354 0.02393 0.117 1280 Sharp exit 1 2.3 0.1 19 099 0.0282 0.218 700 Standard elbow 0.45 3.3 0.2 28 011 0.02628 0.219 140 Standard elbow 0.45 3.0 0.2 25 677 0.02668 0.020 760 Sharp exit 1 2.8 0.2 23 343 0.02714 0.221 100 T piece 0.9 6.3 0.4 53 688 0.02379 0.122 100 Standard elbow 0.45 2.5 0.1 21 221 0.02763 0.023 85 Sharp exit 1 2.3 0.1 19 099 0.0282 0.024 150 Standard elbow 0.45 3.3 0.2 28 011 0.02528 0.025 325 Standard elbow 0.45 3.0 0.2 25 677 0.02668 0.126 100 T piece 0.9 2.8 0.2 23 343 0.02714 0.027 400 - 0 2.5 0.1 21 008 0.02768 0.128 285 T piece 0.9 2.5 0.1 21 008 0.02768 0.129 60 Sharp exit 1 2.2 0.1 18 674 0.02832 0.0

Total 30.5

*The friction factor was calculated by using the Moody diagram. A friction factor calculator, which can be easily downloaded, was used to accuratelydetermine the friction factor.**The total head loss for each pipe section was calculated by using Equation [2].

Figure 5—Main causes for water-related downtime in Section 3 and 4


Table VI details how the friction losses within each pipesection were calculated using Bernoulli’s steady state energyequation. For all the calculations in Table VI it was assumedthat e = 0.15 mm and µ = 0.001.

As seen in Table VI the friction losses within the systemamount to approximately 31 m. The required head of the pumpcan now be determined by using Bernoulli’s equation(Equation [1]). Taking into consideration that:� The static head available (as indicated in Figure 4) is 40

m� Pressure in the pipes should not exceed 1600 kPa (or

163.2 m)� The allowable head for the pump can be calculated as

123.2 m (163.2 m – 40 m)� The frictional head loss in the total length (21 460 m) of

pipe is 31 m

� P1PP = pgh (h = 2 m, as indicated in Figure 4 the waterlevel in the tank is approximately 2 m above the pipelineexiting the tank)

� P2PP = 1500 kPa (the pressure required at the CM is 1500kPa)

� V1VV = 0 m/s� V2VV = 0.13 m/s (derived from the required flow rate of

135 l/min for the Bucyrus CM)

It can therefore be concluded that the pump pressurerequired for supplying water at the required pressure and flowrate to the four underground sections will cause pipe breaks

�


Table VII

Future water requirements (section 1, 2, 3, and 4) at Vlaklaagte Shaft (1 Jan 2014 – 7 Sept 2014)

Different activities Number of Flow of water DOH (hours/day) Quantity (l/day) Quantity (l/s) Quantity Optimal requiring water required (l/min) (m/month) Pressure

required (kPa)

CM (Joy) 2 120 9 129 600 4.0 3 888 1600 in pipe but 2000 at CM

Bucyrus (CAT) 2 135 9 145 800 4.5 4 374Conveyor sprays 33 15 21 623 700 8.3 18 711 1 600Feeder Breaker 12 15 9 97 200 3.0 2 916 1 600spraysCars for dust 15 000 1.0 450 N/AsuppressionCleaning 4 2 57 600 2.0 1 728 N/ABenicon (Mini pit) 60 000 0.0 1 800 N/ATotals 112 8900 22.8 33 867Total + 10% 37 254

*10% was included to compensate for losses

Figure 7—Future pipe layout

Figure 6—Pipe layout with the numbering of each pipe section(Dec2013)

and bursts. The required pump head (142.21 m) exceeds theallowable head of 123.2 m. No pump will therefore be suitablein this application. Three solutions to this problem wereconsidered:� To replace all the thin-walled pipes with thick-walled

pipes with a higher pressure-holding capacity� An underground cascade dam system using permanent

underground dams � An underground cascade dam system using semi-mobile

underground dams.The solutions needed to be implemented to satisfy the life-

of-mine (LOM) water requirements. Therefore the LOM pipelayout and maximum future water requirements needed to bedetermined.

Summary of maximum future water consumption atVlaklaagte ShaftThe maximum future water requirement for the shaft wasdetermined to be during the period when sections 2 and 4moved to block 10 and Section 1 had not been closed yet. Asummary of the future water consumption for these foursections is given in Table VII.

Future underground pipe layoutThe final pipe layout, including final pipe distances for theLOM of Vlaklaagte Shaft, is illustrated in Figure 7. In Figure 8,each pipe section was numbered to facilitate the analysis of thelayout.



Table VIII

Pipe friction calculation using Bernoulli's equation

Section Length (m) Component K factor Total Flow u (m/s) Re Friction factor* Hloss (m)**plus 10%

wastage (l/s)

surface pipe 398 19.75 0.69 103 729 0.0221 1.43 1 100 standard elbow 0.45 19.75 1.23 184 423 0.02114 1.12 2 780 standard elbow 0.45 19.5 1.21 182 088 0.02116 8.31 3 600 standard elbow 0.45 19.25 1.20 179 754 0.02117 6.24 4 1800 standard elbow 0.45 19 1.18 177 419 0.02119 18.20 5 700 standard elbow 0.45 18.75 1.17 175 085 0.02121 6.92 6 600 standard elbow 0.45 18.5 1.15 172 750 0.02123 5.78 7 1320 t piece 0.9 18.25 1.14 170 416 0.02125 12.39 8 570 standard elbow 0.45 7.5 0.47 70 034 0.02302 0.98 9 440 t piece 0.9 7.25 0.45 67 699 0.02311 0.71 10 200 - - - 11 470 t piece 0.9 7 0.44 65 365 0.02321 0.71 12 2320 standard elbow 0.45 3.25 0.20 30 348 0.02592 0.84 13 200 sharp exit 1 3 0.19 28 014 0.02628 0.06 14 900 t piece 0.9 3.5 0.22 32 683 0.0256 0.37 15 400 - - - 16 90 t piece 0.9 3.25 0.20 30 348 0.02592 0.03 17 1120 standard elbow 0.45 3 0.19 28 014 0.02628 0.35 18 440 Sharp exit 1 2.75 0.17 25 679 0.02668 0.12 19 120 standard elbow 0.45 0.25 0.02 2 334 0.04787 0.00 20 400 - - - 21 540 t piece 0.9 5.75 0.36 53 693 0.02379 0.57 22 460 standard elbow 0.45 0.5 0.03 4 669 0.03921 0.01 23 50 standard elbow 0.45 0.25 0.02 2 334 0.04787 0.00 24 90 - - - 25 500 t piece 0.9 5 0.31 46 689 0.02425 0.40 26 600 - - - 27 420 standard elbow 0.45 4.75 0.30 44 355 0.02443 0.31 28 140 standard elbow 0.45 4.5 0.28 42 020 0.02462 0.09 29 700 standard elbow 0.45 4.25 0.26 39 686 0.02483 0.42 30 140 standard elbow 0.45 4 0.25 37 351 0.02506 0.08 31 760 t piece 0.9 3.75 0.23 35 017 0.02532 0.36 32 600 t piece 0.9 3.5 0.22 32 683 0.0256 0.25 33 100 - - - 34 110 standard elbow 0.45 3.25 0.20 30 348 0.02592 0.04 35 660 Sharp exit 1 3 0.19 28 014 0.02628 0.21 36 200 t piece 0.9 4.5 0.28 42 020 0.02462 0.14 37 250 standard elbow 0.45 4.25 0.26 39 686 0.02483 0.15 38 325 standard elbow 0.45 4 0.25 37 351 0.02506 0.17 39 100 t piece 0.9 3.75 0.23 35 017 0.02532 0.05 40 95 t piece 0.9 3.5 0.22 32 683 0.0256 0.04 41 600 Sharp exit 1 2.75 0.17 25 679 0.02668 0.16 42 90 t piece 0.9 0.5 0.03 4 669 0.03921 0.00 43 200 - - - 44 100 t piece 0.9 0.25 0.02 2 334 0.04787 0.00 45 60 - -

Total 68.03

*The friction factor was calculated by using the Moody diagram. A friction factor calculator, which can be easily downloaded, was used to accuratelydetermine the friction factor.**The total head loss for each pipe section was calculated by using Equation [2].


AAnalysis of future pipe layoutTable VIII shows details of how the friction losses within eachpipe section were calculated with the use of Bernoulli’s steady-state energy equation. The total frictional losses werecalculated to be approximately 68 m. Table VIII can be used todetermine where the underground dams should be placed andhow many dams would be required. The placement was

determined by calculating the distances over which the pipe’smaximum pressure rating will be exceeded.

The pipe layout (Figure 7) is too complex to analyse as asingle network. The network was therefore divided into fivedifferent legs in order to determine how many dams will berequired and where the dams need to be placed. The logicbehind determining when a dam will be required is simple: thepump needs to supply 153.22 m head at each outlet (spray),but the pipes can only withstand a maximum of 163.43 m,therefore whenever the pump needs to overcome frictional

�


Table IX

Calculation of how many dams will be required in leg 1 and where they are to be placed

Section Friction Pressure required to overcome friction Commentloss (m) losses and still give the required 153.22 m

head at the outlet

Surface 1.43 154.65pipes1 1.12 155.772 8.31 164.08 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required2-damA 7.66 163.43damA-3 0.65 153.873 6.24 160.114 18.2 178.31 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required4-damB 3.32 163.43damB-5 14.88 168.1 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is requireddamB- 10.21 163.43damCdamC-5 4.67 157.895 6.92 164.81 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required5-damD 5.54 163.43damD-6 1.38 154.66 5.78 160.387 12.39 172.77 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required7-damE 3.05 163.43damE-8 9.34 162.568 0.98 163.54 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required8-damF 0.78 163.43damF-9 0.11 154.29 0.71 154.9111 0.71 155.6214 0.37 155.9916 0.03 156.0217 0.35 156.3718 0.12 156.49

Figure 9—Dam placement for leg 1, 2, 3, 4 and 5

Figure 8—Future pipe layout with the numbering of each pipe section



Table X



head at the outlet

Surface 1.43 154.65pipes1 1.12 155.772 8.31 164.08 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required2-damA 7.66 163.43damA-3 0.65 153.873 6.24 160.114 18.2 178.31 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required4-damB 3.32 163.43damB-5 14.88 168.1 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is requireddamB- 10.21 163.43damCdamC-5 4.67 157.895 6.92 164.81 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required5-damD 5.54 163.43damD-6 1.38 154.66 5.78 160.387 12.39 172.77 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required7-damE 3.05 163.43damE-8 9.34 162.568 0.98 163.54 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required8-damF 0.78 163.43damF-9 0.11 154.29 0.71 154.9111 0.71 155.6212 0.84 156.4613 0.06 156.52

Table XI



head at the outlet

Surface 1.43 154.65pipes1 1.12 155.772 8.31 164.08 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required2-damA 7.66 163.43damA-3 0.65 153.873 6.24 160.114 18.2 178.31 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required4-damB 3.32 163.43damB-5 14.88 168.1 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is requireddamB- 10.21 163.43damCdamC-5 4.67 157.895 6.92 164.81 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required5-damD 5.54 163.43damD-6 1.38 154.66 5.78 160.387 12.39 172.77 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required7-damE 3.05 163.43damE-8 9.34 162.5621 0.57 163.1325 0.4 163.53 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required25-damG 0.3 163.43damG-27 0.1 153.3227 0.31 153.6328 0.09 153.7229 0.42 154.1430 0.08 154.2231 0.36 154.5832 0.25 154.8334 0.04 154.8735 0.21 155.08


losses exceeding the difference (163.43 m – 153.22 m = 10.21m), the maximum head that the pipes can handle is reached

and a dam is required. The calculation for legs 1–5 arepresented in Tables IX – XIII. As seen in the tables, seven damswill be required in order to ensure that the maximum pressureof 1600 kPa is not exceeded. The locations of the dams on theunderground pipe layout, for all five legs, are shown in Figure9.

Trade-off studyThe three possible solutions were traded off, using five criteria:cost, time to completion and ease of implementation,maintenance, safety, and flexibility.

Based on their importance and the preferences ofVlaklaagte Shaft, the criteria were weighted as set out inTableXIV. The solution that scores the highest in the criteria will berecommended for Vlaklaagte.

�


Table XII



head at the outlet

Surface 1.43 154.65pipes1 1.12 155.772 8.31 164.08 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required2-damA 7.66 163.43damA-3 0.65 153.873 6.24 160.114 18.2 178.31 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required4-damB 3.32 163.43damB-5 14.88 168.1 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is requireddamB- 10.21 163.43damCdamC-5 4.67 157.895 6.92 164.81 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required5-damD 5.54 163.43damD-6 1.38 154.66 5.78 160.387 12.39 172.77 Exceeds the maximum 163.43 m that the pipes can withstand - a dam is required7-damE 3.05 163.43damE-8 9.34 162.5621 0.57 163.1322 0.01 163.1423 0 163.1424 0 163.14

Table XIII



head at the outlet

Surface 1.43 154.65pipes1 1.12 155.7736 0.14 155.9137 0.15 156.0638 0.17 156.2339 0.05 156.2840 0.04 156.3241 0.16 156.4842 0 156.4844 0 156.4845 0 156.48

Table XIV

Weighing of criteria for trade-off study

Criterion Weighting (%)

Cost and payback period 40Time to completion and ease of implementation 20Maintenance 10Safety 25Flexibility 5Total 100

S f h l i f d i hSummary of how solutions performed against thecriteriaA summary of how the three solutions performed against thecriteria is given in Table XV. This table forms the basis forrating the solutions.

fAfter taking Table XV into consideration, the solutionswere rated according to the evaluation rubric that was drawnup as indicated in Tables XVI–XVIII. According to theevaluation rubric, building permanent underground damsscored the highest with a value of 73.8.



Table XV

Summary of how solutions performed against the criteria

Solution 1 Solution 2a Solution 2b

Co

st R 3 875 100 R 438 397 R 2 250 618

Pipes are installed by Vlaklaagte’s operationalteam. It takes approximately 1 week to install1km of pipes, therefore to reinstall 14.5 kmlength of pipe will take approximately 14.5weeks, which adds up to 102 days. This

includes delivery and transport of the pipesand accessories. The implementation of thissolution will be time consuming and more

labour intensive that the other two solutions.

It will take a maximum of 1 week to build oneU/G permanent dam. This includes transportof the material. Therefore it will take approxi-mately 7 weeks to build the 7 permanent U/G

dams. This amounts to 49 days. Eljireth MiningServices are building the U/G dams; therefore

the implementation will be very easy forVlaklaagte, because minimum labour will be

required from Vlaklaagte’s side.

It takes Lectropower approximately 3 weeks tobuild one underground portable dam and todeliver it to the mine. Therefore it will take

approximately 21 weeks to build and deliver 7dams. This amounts to 147 days. Lectropowerare building the dams, therefore the implemen-tation will be very easy for Vlaklaagte, because

minimum labour will be required fromVlaklaagte’s side.T

ime

to c

om

ple

tio

nan

d e

ase

of

imp

lem

enta

tio

n

Low maintenance requirements Higher maintenance requirements thansolution 1. Maintenance of permanent U/Gdams is moderate. Vlaklaagte makes use of

recycled water U/G and therefore silt willaccumulate in the dams. If the silt accumu-

lation becomes too high the dams will have tobe cleaned.

Lower maintenance requirements than solution2a. Maintenance of the semi-mobile U/G dams

is less intensive than permanent U/G damsbecause it has a valve attached to drain the silt

if it accumulates.

Mai

nten

ance

High safety If well maintained, high safety. If the dams arewell built and maintained there should be no

safety hazard.

If well maintained, high safety

Saf

ety

Flexible. Most of the pipes can be re-used forother projects after the Vlaklaagte closes.

Poor flexibility. The U/G semi-mobile dams willbe not re-usable after Vlaklaagte closes.

Flexible. The semi-mobile U/G dams can bere-used for other projects after Vlaklaagte

closes.

Flex

ibili

ty

Table XVI

Evaluation Solution 1

Criterion Weighting factor 100 75 50 25 0 Total

Time tocompletion andease ofimplementation

Cost 40% <R1mil R1mil-R3mil R3mil-R7mil R7mil-R12.9mil >R12.9mil 20R3mil-R7mil

20% 0-1 month tocompletion. Very

easy toimplement

2-3 months tocompletion. Easy

to implement.

3-4 months tocompletion.

Fairly easy toimplement.

4-5 months tocompletion.Difficult toimplement.

>5 months tocompletion. Very

difficult toimplement.

10

Maintenance 10% No maintenancerequired

Lowmaintenance

A fair amount ofmaintenance

required

Highmaintenance-

intensive

Highmaintenance-

intensive

7.5

Safety 25% Very safeCompletely safe Fairly safe Low safety Unsafe 18.8

Flexibility 5% Completelyflexible.

Equipment canbe moved aroundunderground with

ease and allequipment canbe fully re-usedafter closure of

Vlaklaagte

Flexible.Equipment can

be moved aroundunderground withrelative ease and

some of theequipment canbe re-used after

closure ofVlaklaagte

Relativelyflexible.

Equipment canbe moved aroundunderground butwith difficulty andvery little of theequipment canbe re-used after


Low flexibility.Equipment might

be moveableunderground but

with extremedifficulty and very

little or none ofthe equipmentcan be re-usedafter closure of

Vlaklaagte

Inflexible.Equipment

cannot be movedaround

underground andnone of the

equipment canbe re-used after


3.8

Total 100% 60


CConclusionsThe water-related downtime problem at Vlaklaagte Shaft wasquantified through a thorough investigation of the downtimelogbook. The main causes of water-related downtime wereidentified as low water pressure, and low water flow caused bypipe leakages and bursts, the main root cause being the lowpressure resistance of the thin-walled galvanized steel pipesused in the underground inbye water reticulation system,

which cannot withstand the increased pressure now requiredby the CM. The ageing infrastructure and increased pumppressures are also contributory factors.

The current water reticulation system was reviewed and anunderground pipe layout was drawn up for the shaft after on-site investigations. The water consumption of the current waterreticulation system was determined from machine and sprayerspecifications. The LOM plan was used to determine the

�


Table XVII

Evaluation Solution 2a



Cost 40% <R1mil R1mil-R3mil R3mil-R7mil R7mil-R12.9mil >R12.9mil 40<R1mil


easy toimplement


to implement.






15


Lowmaintenance


required

Highmaintenance-

intensive

Highmaintenance-

intensive

5

Safety 25% Completely safe Fairly safeVery safe Low safety Unsafe 12.5




Vlaklaagte





Relativelyflexible.







Vlaklaagte






1.3

Total 100% 73.8

Table XVIII

Evaluation Solution 2b



Cost 40% <R1mil R1mil-R3mil R3mil-R7mil R7mil-R12.9mil >R12.9mil 30R1mil-R3mil


easy toimplement


to implement.






5


Lowmaintenance


required

Highmaintenance-

intensive

Highmaintenance-

intensive

7.5

Safety 25% Completely safe Fairly safeVery safe Low safety Unsafe 12.5




Vlaklaagte





Relativelyflexible.







Vlaklaagte






1.5

Total 100% 56.5

fmaximum LOM water requirements, and the time frame inwwhich the water consumption would be the highest wasdetermined.

Three different solutions were considered to solve thewwater-related downtime problem and to ensure the efficientsupply of water to the newly opened sections. Permanentunderground concrete dams, semi-mobile dams, and new pipecolumns with a higher pressure resistance of 3200 kPa wereconsidered. The dam placement was determined by calculatingthe friction loss within each pipe section using Bernoulli’senergy equation. The conclusion was that seven undergrounddams should be placed to ensure that the maximum pressureof the pipes (1600 kPa) is not exceeded.

The solutions were compared using an evaluation rubric.Building permanent underground dams was determined to bethe cheapest solution (R438 397) and can be implemented inthe shortest time (49 days). Cost and time to completion werecritical for the solution to be a viable option. The paybackperiod for the cost associated with building undergroundpermanent dams was determined to be 0.035 years, and thesolution will save the mine R12.9 million. Building permanentunderground dams was therefore identified as the bestsolution for implementation.

RecommendationsIt is recommended that seven permanent underground damsshould be built at Vlaklaagte Shaft to solve the water-relateddowntime problem and ensure the efficient supply of water tothe newly opened sections.

SSuggestions for further work

� A sensitivity analysis should be done on the weightingfactors of the different criteria used to trade off the threepossible solutions. This will give an indication of howchanges in the weighting of each criterion would affectthe outcome of the trade-off study

� Studies can be done on a more effective recordingsystem for water-related downtime and for recordingchanges made to the water reticulation system.

Acknowledgement

I would like to thank Prof. R.C.W. Webber-Youngman, mysupervisor at the University of Pretoria, and Charl DuBuisson, my mentor at Goedehoop Colliery, for their guidanceand support.

ReferencesBECHT, E. 2010. General Manager, Goedehoop Colliery. Presentation.

DU BUISSON, C. 2013. Shaft Manager, Goedehoop Colliery. Personal communi-cation.

HORAC, T. 2013. Foreman, Goedehoop Colliery. Personal communication.

LOTTERING, R. 2013. Consultant, Barloworld. Personal communication.

LOUW, N. 2013. Mine Overseer, Goedehoop Colliery. Personal communication.

PIETERSE, C. 2013. Section Engineer, Goedehoop Colliery. Personal communi-cation.

WHITEWW , F.M. 2011. Fluid Mechanics. McGraw-Hill, New York. �



OPTIFLEX 2200 C/F is the answer.The OPTIFLEX 2200 C/F – TDR level meter is available with

either a horizontal or a vertical housing to fit into installations

of many industries and applications: It can measure in nearly

all process applications with temperatures up to +300 °C/

+570 °F and in tanks of maximum 40 m/130 ft height. The

remote converter can be installed and operated as far as

100 m/328 ft from the probe. Intelligent software eliminates

false reflections caused by environmental disturbances

and product build-up – we call it DPR (Dynamic Parasite

Rejection). To ensure safe operation, the TDR level meter

is SIL2-compliant (HFT=0) according to IEC 61508 for

safety-related systems.

So, who needs the OPTIFLEX 2200 C/F? For one, it’s all

about chemicals. The device often replaces RF capacitance,

displacer, float and pressure meters, but also older TDR

meter generations. The OPTIFLEX 2200 C/F is easy on the

budget while ensuring economic value that meets the

highest requirements.

You can count on us: Our robust level meters feature high

accuracy and reliability to suit your needs.

KROHNE – Process engineering is our world.

KROHNE South Africa

8 Bushbuck Close

Corporate Park South

Randtjiespark, Midrand

Tel.: +27 113141391

Fax: +27 113141681

Cell: +27 825563934

John Alexander

[email protected]

Flexibility is the main principle

IntroductionZZibulo, meaning first borng g in Zulu, is the firstnew mine in the Anglo American Inyosi Coal(AAIC) joint venture. It was formerly knownas the Zondagsfontein coal project. The projectis majority-owned (73%) by Anglo Americanand the remaining 27% by Inyosi, the blackeconomic empowerment company formed in2007 as part of Anglo Coal’s second wave ofempowerment in South Africa (AngloAmerican, 2007). The colliery is situated inOgies in Mpumalanga Province. With a life ofmine of 20 years, the project comprises of twooperations; an opencast and an undergroundoperation. This project was carried out at theunderground operation.

Zibulo Colliery is sited within the WitbankCoalfields, which are usually comprised of fiveseams numbered (from the base upwards) No.1 to No. 5 seam. The colliery extracts No. 2seam. The disturbed and relatively shallow

(depth approximately 100 m) coal seam ismined using the bord and pillar method due tothe low capital investment and operating costsrequired, together with its level of selectivityand safety.

Project background

Coal hauling background at ZibuloCoal is hauled by means of both batteryhaulers (BHs) and shuttle cars (SCs) at ZibuloColliery’s underground operation. Thebackground of the coal hauling equipment atthe eight sections of the mine is illustrated inFigure 1.

A total of 21 new coal haulers wereinitially purchased; 9 SCs and 12BHs. The SCswere employed in sections 1, 2, and 3.Sections 4, 5, 6, and 7 were using BHs. Duringthis start-up phase, three redundant BHs werepurchased from Goedehoop colliery, an AngloAmerican Thermal Coal undergroundoperation. These machines had to beoverhauled in order to get them intooperational condition; they were then put intoproduction in section 8 as a temporary solution(Anglo American, 2013). In 2013, they weredeemed to have reached the end of their lifecycle. They displayed low availabilities andthus had to be replaced with new SCs (AngloAmerican, 2013). Six new SCs came on streamin 2013. Sections 4 and 6 are currently usingthree SCs each. The BHs from sections 4 and6, however, were split among sections 5, 7,and 8. The initial BHs from section 8 are nolonger in production; section 8 currently usesthe three BHs from section 4 and an additionalhauler from section 6. A BH from section 6was added to the three haulers in section 5.The other hauler from section 6 (added tosection 7’s fleet) is not currently operated.

Optimization of shuttle car utilization atan underground coal mineby P.R. Segopolo*Paper written on project work carried out in partial fulfilment of BSc. Eng. (MiningEngineering)

SynopsisThe purpose of the project is to convert current shuttle car utilization on anunderground coal mine to best practice by focusing on change-out pointsand tramming routes, which have a major influence on shuttle car awaytimes. Time studies were an integral part of the project as these enabled thedetermination of shuttle car away times. An indirect proportionalrelationship between shuttle car away times and productivity isestablished. Through the time studies, it is deduced that a third shuttle carwill make an insignificant contribution to production when there is onlyone split open. During this time, maintenance on the third car can beoptimized. In order to satisfy the mine’s key performance indicator ofkeeping shuttle car away times less than 75 seconds, a belt extension mustbe scheduled after the third split is open. It is established that at any giventime, a minimum of two shuttle cars should be used. When cutting on theleft-hand-side of the belt road with only two shuttle cars available, thecentre and left (left of the feeder breaker) shuttle cars should be used forcoal hauling. When cutting on the right-hand-side, the centre and theright-hand-side cars should be used. If only one shuttle car is available, thecentre car is the most efficient to use. Alternative anchoring configurationscan be employed so as to enable cars (left or right, especially) to reach theopposite extremities of the panel and hence minimize cable lengthrestrictions.

Keywordscoal mining, underground transport, coal hauling, tramming, scheduling,optimization.

* University of the Witwatersrand, Johannesburg,South Africa.

© The Southern African Institute of Mining andMetallurgy, 2015. ISSN 2225-6253. Paper receivedrrFeb. 2015



Optimization of shuttle car utilization at an underground coal mine

f fIt is clear that SCs have found favour at the operation.Zibulo Colliery currently employs a total of 15 Joy 10SC22-56C machines in five (sections 1, 2, 3, 4, and 6) of its eightsections. The SCs have the following specifications (AngloAmerican, 2013).

Supplier Joy Mining MachineryDrive Dual conveyor motor driveControl VFD OPTIDRIVE traction systemPump capacity 25 kW pump motorLubrication Auto lubrication systemTraction power 2 × 85 kW traction motorsCapacity 20 tMinimum seam height 1.96 m

Cutting sequence and roadsA typical cutting sequence from section 4 is illustrated inFigure 2. The road on which the feeder breaker (FB) islocated is referred to as the belt road (BR). Roads on the leftof the BR are the left roads (referred to as L1, L2, L3 to thefurther left of the belt road), to the right of the BR, roads areR1, R2, R3, R4. This particular sequence is characterized byeight roads; other sections have a different number of roads,depending on the panel width. Ventilation in a section isdirected from right to left; it is for this reason that cutting ineach section generally takes place from the right to the left ofthe section. Once the continuous miner (CM) has made itsfirst cut between R3 and R4 it is trammed to the second cuton R3, thereby making way for roofbolting to be carried outwwhere the first cut took place.

Through roadsThrough roads are also referred to as splits; these areillustrated in Figure 3. These are open roads between the FBand faces to be cut. Zibulo Colliery maintains a maximum ofthree through roads in each section. When three splits areopen, the belt is extended (i.e. the FB moves towards thefaces) two splits ahead.

Shuttle car change-out points, tramming routes,aanchor points, and switchesThe green, blue, and red circles in Figure 4 represent threeshuttle cars in a section. Their respective dashed linesrepresent their trailing cables and thus the way in which theshuttle cars tram towards and back from the CM at R3. Thesecables are anchored on the three points adjacent to the FB,wwhere the SCs all tip from their three distinct points. About

f f20 m of the total SC cable slack is from the anchor point tothe switches, which are placed parallel to the FB. The pointswhere SCs interact with each other (indicated by squares 1and 2 in Figure 4) are the change-out points. At any one ofthese points, a SC waits on the next before it proceedstowards the CM to avoid running over the next SC’s cable. Atsquare 1 in Figure 4, the red car waits on the blue car.

�


Figure 1—Background of coal hauling equipment at Zibulo

Figure 2—Typical section 4 cutting sequence and roads

Figure 3—Through roads

Problem statement and aim of projectEach section is equipped with a FB on which coal can betipped from three distinct points. The mine took advantage ofthis by employing three SCs (each anchored at a distinctpoint) in each of five sections, with the aim of maximizingproductivity. However, the overall SC utilization hasdecreased from 2010 to 2013. Furthermore, data gatheredfrom CM operational reports reveals that CM waiting times(or SC away times) in all five sections employing SCs for coalhaulage are rather long compared to the mine’s keyperformance indicators (KPIs). The decrease in SC utilizationand the longer SC away times leads to lower production rates.The project is therefore aimed at increasing productivitythrough the optimization of SC utilization.

Results and analysisIn each of the five SC sections, a considerable amount of timewwas spent near a CM recording SC away times as well as SCloading times. Away time in this study was taken as the timebetween completion of loading one SC and the arrival of next(or the same car if only one is being used) to be loaded.Before each recording session (or each time the CM had totram to another cut) the numbers of SCs being operated in

the section as well as their tramming routes was established.On a few occasions, the time it took for a SC to tip onto theFB was recorded. The results obtained were predominantlyfrom direct time measurements. From these, the trammingdistances and average tramming speed were obtained.

As a means of confirming the consistency and reliabilityof the results, they were compared to the Joy CM system thatthe mine uses to monitor whether the set production KPIs arebeing met. Although a great degree of similarity wasobserved from the comparison, data calculated from thetramming distances and the average tramming speed couldnot be confirmed. In an attempt to clear this hurdle, theArena simulation software program was used. The programwas able to confirm the findings from the studies.

Shuttle car away times

From observations, SC tramming routes were similar in allfive sections. Long tramming routes lead to longer awaytimes; longer away times lead to lower production rates. Theinverse relationship between the tramming route distancesand productivity is thus established. The main objective,therefore, is to keep tramming routes as short as practicallypossible with respect to the CM position, CM cables, andventilation.

Tables I–IV represent the average away times that wereobtained from the data. The mining height is approximately3.5 m; the pillar and bord widths are 12 m and 7.2 m respec-tively. Approximate distances between the FB and CM,between the FB and the main change-out point (COP), andfrom the COP to the CM were determined using these bordand pillar widths. In Tables I, II, and III, S/C columns denotethe following:

� 1 S/C: SC X� 2 S/C: SCs X and Y� 3 S/C: all three SCs (X, Y, and Z).

For the columns ‘FB to CM’, ‘FB to main COP’, and ‘COPto CM’:� 1st S/C: SC X� 2nd S/C: SC Y� 3rd S/C: SC Z



Table I

Average away times for one split

Average away time (seconds) Distances (metres)

FB to CM FB to main COP COP to CM

Cut Road 1 S/C 2 S/C 3 S/C 1st S/C 2nd S/C 3rd S/C 2nd S/C 3rd S/C 2nd or 3rd S/C

1 R3 & R4 114 25 26 99 99 137 78 116 212 R3 115 26 25 99 99 137 78 116 213 R4 128 35 35 118 118 156 78 116 404 R2 & R3 117 45 44 137 147 175 78 116 595 R4 137 40 42 130 130 168 78 116 526 R3 126 31 31 111 111 149 78 116 337 R1 & R2 128 34 35 118 118 156 78 116 408 R3 129 35 35 117 117 155 78 116 399 BR & R1 115 31 30 99 99 158 59 97 4010 R2 111 20 20 80 80 139 59 97 2111 R1 88 17 17 61 61 120 57 116 2112 R2 109 25 26 92 92 151 59 97 3314 R1 121 26 26 73 73 111 40 78 3316 BR 73 21 17 80 42 80 21 59 2118 BR 82 28 26 92 54 92 21 59 33

Figure 4—Shuttle car change-out points, tramming routes, anchorpoints, and switches


�


Table II

Average away times for two split




3 R4 146 46 36 139 139 177 99 137 405 R4 155 49 43 151 151 189 99 137 521 R3 & R4 132 37 26 120 120 158 99 137 212 R3 133 37 26 120 120 158 99 137 216 R3 142 41 30 132 132 170 99 137 338 R3 147 45 37 141 141 179 99 137 424 R2 & R3 163 52 50 158 158 196 99 137 5910 R2 118 30 21 101 101 139 80 118 2112 R2 126 34 31 113 113 151 80 118 337 R1 & R2 146 45 36 139 139 177 80 118 5911 R1 102 22 21 82 82 120 61 99 2114 R1 111 29 23 94 94 132 61 99 339 BR & R1 131 46 44 120 120 158 61 99 5918 BR 97 36 25 113 75 113 42 80 3316 BR 89 29 17 101 63 101 42 80 21

Table III

Average away times for three split





Table IV

Calculated away times for four split

Calculated away time (seconds) Distances (metres)




ffThe different routes to be trammed by the SCs areillustrated in Figure 5. Note that cycle time study results aretabulated for only the belt road (BR) to the R4, and not fromthe BR to the left-hand-side extremities of the panel. This isbecause mirrored results for cuts on the roads would beobtained for the BR to the L3 such that away times from the2nd cut would be the same as that of the 22nd cut; awaytimes from the 14th cut would be the same as that of the 21stcut. Mining is carried out from the right-hand extremities ofthe panel to the left for ventilation purposes.

Tramming routes for cuts 9, 6, and 8 are illustrated inFigure 5. These are shown by the dashed lines with therespective shuttle car colours. The main COPs are indicatedby the transparent spheres, at these points, all three SCsinteract with each other. Note that there is no distance, forthere is no change-out-point when the no. 1 SC is operatedsolely; it does not interact with the next car.

Zibulo Mine standards allow for a maximum of only threesplits before a belt extension. Although contraventions of thestandard were not observed in practice, away times whenthere were four splits between the FB and the face were notdirectly recorded. These were calculated from the data alreadyobtained. The average speed at which SCs tram wasdetermined; the tramming route distances to each cut werecomputed. Away times were then determined from dividingthe route distances by the SC average tramming speed. Thedetermination of away times when using two or three SCs,however, was rather complex. This involved manualsimulations that were only carried out on paper.

With maximum SC cables lengths of approximately 230m, it is evident from the table that SCs may not be able toreach certain cuts from their anchor points. The 3rd SC wouldnot be able to reach cuts 1 to 9 as well as cut 12 because themaximum SC cable lengths are specified in the minestandards.

AAnalysis of shuttle car away timesAs a means of analysis, three random cuts were selected. Theeffects of the number of SCs being operated for a number ofsplits were analysed from Figure 6, 7, and 8. A general trendfrom all three graphs was observed.

When using a second SC as opposed to just one, awaytimes from all four split numbers decrease significantly. Asecond SC therefore adds significant value; a production rateincrease is realized when using two cars rather than one.WWhen a third SC is added, however, away times from splits 2,3, and 4 decrease. From Figure 6, away times with split do

not change when a third car is added to the two that arealready being operated.

A similar trend is observed from Figures 7 and 8. Addinga second car considerably increases production rates, as itcan be observed that away times decrease. A third SC addsvalue at splits 2, 3, and 4. In the case of a single split, theaverage away times between SCs 2 and 3 does not change.This suggests that adding a third SC when two are alreadyoperating in the section will not increase the production rate.It can therefore be deduced that maximum practicalproduction can be achieved by using only two SCs, whenthere is a single split between the FB and the face being cut.



Figure 5—Different routes to be trammed by the shuttle cars

Figure 6—Average away times when cutting the second cut at distinctsplits using 1, 2, and 3 shuttle cars

Figure 7—Average away times when cutting the 7th cut at distinct splitsusing 1, 2, and 3 shuttle cars

Figure 8—Average away times when cutting the 11th cut at distinctsplits using 1, 2, and 3 shuttle cars


It is to be noted, on the other hand, that the average awaytimes in the case of a single split decrease when a third SC isadded to the operation only if the CM is cutting at the beltroad. Average away time decreases from 21 seconds to 17seconds when cutting the 16th cut by adding on a third SC atthe first split. Therefore, a third SC adds production valueonly when the CM is cutting at the BR (cuts 16 and 18).

To check if the deduction is valid, Arena (miningsimulation program) demonstrations were made available.According to Olivier (2014), ’the shuttle car on the righttakes up two thirds of every 3 shuttle car loads when the CMis cutting on the right-hand-side of the belt road’. When theCM is cutting on the belt road, each of the three cars takes upa third of every three SC loads. Olivier’s statement thusconfirms that a third SC is rather insignificant when there is asingle split between the FB and the face.

Safety implications

At Zibulo colliery, safety is a core value. It is commonpractice to have the change-out points as close as possible tothe face being cut. Figure 9 illustrates change-out points Aand B when the CM is cutting the 12th cut. The stars on bothdiagrams represent where the CM operators would standduring operation. Having a change-out point directlyadjacent to the CM operators’ position brings about the riskof a SC running into the operators as it attempts to make atight turn to position itself in place for loading. For safetyreasons, designing the change-out point to be at A istherefore not recommended. With the common practice bornein mind, it was deemed necessary to carry out a furtherinvestigation on the implications this may have on safety inrelation to production.

Time studies of the away time differences betweenchange-out points A and B were conducted. The difference in

fthese away times implied a 3.48% production loss fromchange-out point A to B. Siyanqoba section (section 4) had aproduction target of 1 Mt for 2014. Designing for change-outpoint B would lead to a production loss of only 34 800 t; thesection may only produce 965 200 t. A 3.48% production lossis not of great consequence if it promotes safety.

Shuttle car configurationsSince each SC section employs three SCs, it has becomegenerally accepted that all three cars should be running at alltimes in order to meet production requirements. Not only isthis not necessarily the case, but utilizing all three cars at alltimes is somewhat impractical, due to factors including cablelengths and operator availabilities. It therefore becomesnecessary to factor in SC configurations that will lead tomaximum productivity. These should be used at all times ifproduction targets are to be met.

Optimal car configurations determined fromsimulations

Table V highlights the effects of SC configurations onproduction, utilizing simulations run on the UndergroundCoal Mining Simulation (UCMS) program. Different numbersof cars and their configurations were simulated to runthroughout the entire panel length from the 1st to the 68thcut, illustrated in Figure 10. The cutting sequence in Figure 10 was initially input as Zibulo’s cutting sequence intothe program. It is important to note, however, that thiscutting sequence varies from that generally followed at ZibuloColliery’s underground sections. According to Olivier (2014),the data obtained from the program can still be relied on ascars are trammed to all different cut locations, only in adifferent sequence.

�


Figure 9—Different possible change-out points when excavating cut 12

Table V

The effect of shuttle car configurations on productivity

3SC No RHS No centre n=No LHS Only centre Only RHS Only LHS

Cycle time 114.79 128.73 131.01 128.54 197.72 234.78 238.11

Av. production rate (TPH) 556.01 496.82 488.27 497.35 324.23 274.32 271.37

Production time (min) 280.84 272.87 274.65 272.65 310.24 322.63 322.82

Tons (booked) 2602.5 2259.47 2235.01 2260.07 1676.46 1475.04 1460.05

-36% -43% -44%-13% -14% -13%0%

No of Cars 3 2 2 2 1 1 1

According to UCMS, 2 602.5 tons can be booked when allthree SCs run for a production time duration of approximately4 hours and 41 minutes. The benchmark was thereforeassumed to be when running all three SCs, hence the zeroproduction decrease indicated in Table V. Variations on thiswwere then investigated in order to see how the results differfrom the benchmark.

Using only the centre car together with either the right orleft SC results in a 13% production decrease. When usingboth the right and left cars with no centre car, UCMS suggeststhat the tons booked will decrease by 14%. It does happenthat a section has only one SC running at a given time due tomaintenance, breakdowns, or operator availabilities. Thiscase was also investigated. According to UCMS, a maximumof 44% of the benchmark tons booked can be lost whenrunning only one SC. This is when the left-hand side (LHS)car is solely used. The least production loss when employingonly one car is obtained when only the centre car is used.

From the 1st cut all the way to the 68th, all six configu-rations are applied (or rather as a result of breakdowns, carmaintenance, or labour and cable management) at any giventime.

PProduction improvements offered by different carconfigurationsAs summarized by UCMS, the maximum improvement inproductivity is achieved when employing all three SCs insteadof only one. This is illustrated in Figure 11; a summary of theimprovement in production when comparing the usage of

ff fdifferent numbers of cars running in the section. A 76%production improvement can be obtained when using threecars instead of just one. Employing two SCs instead of onlyone offers a 55% production improvement. The leastproduction improvement is achieved when transitioning fromusing only two SCs to three; this offers a 15% productionimprovement. Although using all three SCs offers the greatestpractical productivity, using two (even though to a lesserextent) cars is also viable.

Applicable shuttle car configurations determined fromaway time studiesIt is one thing to understand the value that different carconfigurations add to production; it is another to determinewhen and which SCs to apply when extracting coal at anyparticular cut. SC average away times when the CM cuts atseven distinct points, illustrated in Figure 12, were analysedso as to identify which car configurations to apply.

It has already been established that the effect of a thirdSC is rather insignificant when there is a single split betweenthe FB and cut. With two splits, however, using all three carsoffers maximum production. Using two SCs is the second-best option. Table VI shows the SC average away times when



Figure 10—Zibulo Colliery's cutting sequence according to UCMS

Figure 11—Summary of production improvement when using differentnumbers of shuttle cars

Figure 12—Cuts of interest in a typical section 4 cutting sequence

Table VI

Average away times for six shuttle car configurations when being loaded at seven distinct cuts

Cut Average away times

Road RHS only LHS only Centre only RHS+centre LHS+centre LHS+RHS

5 R4 155 184 155 49 71 11249R2 117 146 117 30 67 443010

BR/R1 131 161 117 46 49 51469BR 126 126 97 34 34 3418L1 161 102 102 52 22 572219

L2/L1 161 131 102 52 37 593715L3 190 131 131 66 37 683722


cuts 5, 9, 10, 15, 18, 19, and 22 are being excavated. It canbe seen that minimum average away times are obtainedwwhen using the right-hand side (RHS) car together with thecentre one; this is when the cuts on the RHS of the BR) arebeing mined. Average away times when using the LHS andcentre cars are somewhat moderate; those of using both theLHS and RHS cars are the highest. This suggests that whencutting on the RHS of the BR, a configuration (involving theuse of two SCs only) of the RHS and centre car is the mostvviable for meeting production targets.

When excavating cut 18, which is positioned on the BR,any pair of SCs is viable. This is due to the constant averageaway time (34 seconds) offered by any pair. When cutting onthe LHS of the BR (cuts 19, 15, and 22), on the other hand,minimum average away times are achieved when using theLHS and centre SCs. Again, the average away times whenemploying both the RHS and the LHS cars are the highest.

It is not always possible to have either two or three SCsrunning at any given time in a section. It is thereforenecessary to determine which single car should be employed,as well as when it should be applied. When cutting on theRHS roads of the BR, minimum average away times areachieved when using either the RHS or the centre cars, asboth their average away times are equal. When cutting on theRHS split of the BR, such as cut 9, minimum average awaytimes are obtained when using only the centre car. This isattributed to the fact that SC tramming routes when cutting atthis point are not as straightforward as when cutting on theroads. As expected, the least average away times whencutting at the BR are achieved when using the centre caronly. Using either the LHS or centre car offers minimumaverage away times when cutting on the LHS roads of theBR. Similarly, when cutting on the LHS splits of the BR, suchas cut 15, the least average away times are achieved whenthe centre car only is operating. When cutting at this point,using the RHS car only offers the highest average awaytimes.

Challenges and opportunitiesOf course, the utilization of SCs in most of Zibulo colliery’sunderground sections presents more advantages than thecars’ counterparts, the battery haulers. However, as De Lange(1988, p. 151) previously postulated about the future ofunderground transport on large coal mines, ‘coal mining is amajor transport business and hence there will always be newchallenges to meet in underground transport’. SC operations

f finvolve the consideration of various factors, both technicaland non-technical. A challenge presents an opportunity toimprove or employ new techniques. The challenges of SCutilization and the associated opportunities they present arediscussed in the following sections.

Tramming routesA direct relationship between the tramming route distancesand the average away times has already been established; itis therefore important to keep the tramming routes as shortas practically possible. Not only should the distances be keptminimal; the following factors should be considered whendesigning or determining tramming routes.

Avoiding turns as far as practically possibleFigure 13 illustrates the possible tramming routes that thecentre SC (SC X) can follow when the CM is cutting at the10th cut. The tramming route illustrated on the LHS diagramof Figure 13 shows that car X would have to make four turnsto and from the CM to the FB . As stated by Smit (2014), ‘atramming route should have as little turns as possible’.Tramming routes with more turns are both unsafe andineffective, such that car X on the LHS diagram will havelonger average away times than car X on the RHS diagram ofFigure 13 (Smit, 2014). Like any other trackless equipment,the tramming (or hauling) speed is reduced when turning. Onthe LHS diagram, the average shuttle car speed of 2.7 m/s(obtained from SC time studies) will be reduced. As theaverage away times consequently increase, the productionrate decreases.

For any equipment, although training is offered, ease ofoperation should be the main objective. A tramming routewith more turns compromises the ease of SC operation. Thecars can easily collide with pillars. An operator may becomefatigued relatively quicker compared with the operatoroperating car X on the RHS diagram.

Tramming route obstructionsIf the shortest tramming routes are to be used at all times, itis important to make sure that they are free of anyobstructions. Figure 14 shows a SC cable and a brattice thatwere placed on the area indicated by the star on the RHSdiagram. This implies that car X used the tramming routeillustrated on the LHS diagram of Figure 13. This constitutesgood housekeeping as well as thorough and effectiveplanning. Brattices should be installed such that they do notobstruct the desired tramming routes. For example, the

�


Figure 13—Different tramming routes for shuttle car X when the CM is excavating cut 10

finstallation of the brattice indicted by the dashed line on theRHS diagram of Figure 14 should be delayed until cuts 3, 4,and 5 are excavated so as to clear the shortest, mostpractically possible and viable tramming routes to thesepoints.

Floor conditionsBad floor conditions can be as a result of an uneven floor(attributed to geological conditions or floors that are notswept), poor water drainage, and steep gradients. Theseconditions may significantly reduce the life of the cars’components and consequently cause premature failure, or thecost per ton of the operation may increase due to losses inefficiency and productivity (Callow, 2006, p. 821). At ZibuloColliery’s underground sections, coal extraction is carried outon a relatively flat gradient. Tramming routes are relativelyflat and so shuttle cars tramming at high gradients is not aconcern. The section floors are generally kept in goodcondition. During the rare cases of floor flooding, however,corrective measures to drain water should be taken as quicklyand efficiently as practically possible.

BBelt extensionTo maintain overall short tramming distances, it is importantto schedule a belt extension effectively. This means that thereshould be a maximum number of splits open before each beltextension. To determine this, the overall average away timeswwhen using all three SCs for five split scenarios wereobtained. The graph in Figure 15 is a result of this analysis.According to Zibulo’s KPIs, the maximum average away timethat should be obtained at any particular time to reach the setproduction targets is 75 seconds. This KPI is indicated by thehorizontal line on the graph. When there are one to threesplits between the FB and the CM cut position, the averageaway times are below the KPI. During the transition fromthree to four splits, however, the KPI is reached and exceededbefore the 4th split is entirely open. Thereafter, the averageaway times remain higher than the KPI.

Figure 15 therefore implies that if the average away timesare to be kept below the set KPI, then a belt extension is to becarried out before the 4th split is open. This should be donejjust after the 3rd split is fully open. At this point, the beltwwill be extended over two splits as illustrated in Figure 16(by the red dashed line) such that after the belt extension,one split will be open, allowing for the effective use of only

two SCs. This then becomes a cycle; the belt is extended twosplits ahead with one split open between the FB and line ofcut, the 2nd split is open and as soon as the 3rd is entirelyopen another belt extension should be carried out. The initial3rd split will therefore become the 1st split after the belt isextended.

Cable managementAt Zibulo Colliery, cable management is a major concern. Themine is relatively new and so the effective management



Figure 14—A shuttle car cable and brattice placed in the section

Figure 15—Average away times vs. the number of open splits

Figure 16—Illustration of a belt extension with respect to the splits


techniques to be employed are still being determined. Unlikethe more established Anglo American Thermal Coalunderground operations such Goedehoop Colliery, Zibulo stilloutsources some of the crucial tasks that are directly linked tooperational efficiencies. Cable repairs and maintenance arecarried out by the contracting company, Lectropower (Smit,2014). The company runs an underground mine workshop(referred to as the cable shop) to which cables are deliveredfor repairs.

The cable management system, according to the authorand a student who conducted a project on cable management,is somewhat ineffective. The basic ideal handling of damagedcables at Zibulo Colliery is as follows:

(i) When a cable is damaged in a section, it is reportedto the cable shop

(ii) The cable is then manually loaded onto a load hauldumper (LHD) at the section; the LHD transports thedamaged cable from the section to the cable shop

(iii) At the cable shop, the cable is offloaded. A sparecable is then manually loaded onto the LHD, whichreturns to the section

(iv) Once the cable has been repaired and tested, thesection is notified and the cable is then kept in thespare cable zone (Horstmann, 2014).

Some shortfalls were identified. The following are some ofthe findings that compromise the ideal procedure togetherwwith the operational system.

� Damaged cables are not always handled correctly; thisconsequently often leads to them not being repairable(Horstmann, 2014)

� No particular operator in the section is responsible forthe transportation of damaged cables; this means thatin the case of a damaged cable, any one of the section’soperators (one of the two CM operators, one of thethree SC operators, or any of the two roofboltoperators) is pulled out and assigned to the task ofdelivering the damaged cable to the cable shop. Normaloperation is therefore disrupted

� The LHDs are shared between sections. No section hasits own LHD. A section with a damaged cable usuallyhas to wait for a relatively long period for the nextsection to deliver the LHD. This implies that valuableproduction time is not used effectively while the sectionwaits on an LHD to deliver the damaged cable to thecable shop

� Cables are manually handled; the removal and instal-lation of a new cable can take up to an hour

� In most cases, there are no spare cables in the sectionand so a section had to wait for a spare from the cableshop

� Old cables that can still be used are left in the oldworking sections during section moves (Horstmann,2014)

� The cable shop floor space is insufficient; there is nospace for the CM cables. These cables are then keptoutside the shop, thus leaving them vulnerable todamage (Horstmann, 2014).

It is suggested that Zibulo investigates how the moreestablished Anglo American Thermal Coal undergroundoperations manage their cables and implement cablemanagement initiatives.

AnchoringOwing to continuous repairing of cables, cables lengths areusually shortened. From a maximum cable length of 230 m, amaximum of only 200–210 m remains for SC tramming fromthe anchor point at the FB to the loading zone and back.Therefore SCs may not be able to reach loading zones whenthere are three or more splits are between the FB and thecuts.

According to Smit (2014), SC anchoring requires a greatdeal of experience to redesign. In particular, due to shortercable lengths (which consequently lead to cars not being ableto reach certain cuts), the location of anchor points in thesection can be altered so as to enable effective tramming evenfor the cars with relatively short cable lengths. All the AngloAmerican Thermal Coal underground operations employ asimilar anchoring method to that used at Zibulo Colliery.Figure 17 illustrates the possible alternative anchoring pointsthat can be explored.

If car Z was anchored on the original position (on theLHS of the FB), it may not be able to reach the 2nd cut fromthe anchor point. To avoid this situation the car can beanchored a road away from car Y, on the RHS of the FB. Thiswould significantly reduce the SC’s tramming route distanceto the 2nd cut. The reduction in the tramming distanceimplies a reduction in the shuttle car average away times andthus increased productivity rates.

SC Z can also be anchored midway between the FB andcut 2. In this case the SC uses an effective cable length that ishalf the cable length required to tram the car to the 2nd cut ifit was anchored where car Y is anchored, for example. Smit(2014) suggests that the major concerns with the differentanchoring systems are:

�


Figure 17—Alternatives to the traditional shuttle car anchor points

� fThe running over of a car cable by the next car to passthat point

� The increase in change-out points� Increase in shuttle car away times.

It has been established that change-out points should bekept as close as practically and safely possible to the loadingzone; they should also be kept as close as possible to eachother. At a change-out point, one SC has to wait on the nextto make way for it to tram to the loading zone or the FB.Moving an anchor point midway between the loading pointand FB not only reduces the cable length requirements; itintroduces more change-out points in the section asillustrated in Figure 17. This may lead to increased awaytimes as cars are to wait on each other at more points.However, a constant number of SC anchor points can bemaintained if all are anchored midway of their respectivetramming routes.

As a SC moves away from the anchor point, the cable andthus the anchor point are under tension. This is illustrated inFigure 18. This tension leads to pillar damage as illustrated.Pillar damage may extend to the anchor point and thuswweaken the anchoring to the extent that the anchor isultimately pulled off from the pillar. This becomes a safetyhazard; it can also lead to cable damage.

To reduce anchor tension that is caused by the cabletension as the shuttle car moves towards the CM, a ‘springeffect’ can be introduced. A used tyre, for example, can beused to connect the anchor point and cable as illustrated inFigure 19. The tension in the cable will be absorbed by thetyre and thus the anchor point will not be greatly affected. Aspring can also be used as an alternative.

LLabour managementAn average of 77.94% of the actual labour complement wasutilized per section from April 2013 to December 2013. Froma full complement average of 10.4, a labour complement ofonly 8.1 was achieved. Even though the full complement wasreduced to 10 from 11 in August 2013, the labourcomplement target is still not being reached. A typical shiftper section comprises the following:

� 2 CM operators� 2 roofbolt operators� 3 SC operators� 1 miner (shift supervisor)� 1 electrician� 1 fitter.

Absenteeism is attributed to sick leave, absence withoutpay (AWOP), and planned absence. For a section to be barelyproductive, eight people (two SC operators less) are required.However, production targets may not be reached becauseonly one SC will be operated. Thus, absenteeism is to bemanaged and provisioned for. Zibulo Colliery is currentlyrunning a project on discipline enforcement by frontlinesupervisors.

ConclusionWWhen there is only one split between the FB and CM, the useof a third SC has no effect on the production already obtainedby using two SCs. Three SCs are collectively effective onlywwhen there is more than one open split between the FB and

CM. However, due to the Zibulo underground operationstandard, a maximum of three splits should be open before abelt extension. This is also due to cable length restrictions;shuttle cars may not be able to reach some cuts when foursplits are open.

When only two SCs are used at any given time, it iseffective to use the centre and either the right or left car,depending on whether the cut being excavated is on the rightor left of the BR. Therefore, the centre car should always bein operation, even when only one car is being used. However,the use of at least two cars at any given time should bemaintained.

The procedure for handling damaged cables should berevisited so as to improve cable management. Other cableanchoring options can be explored. The number of change-out points can be maintained if all three SCs are anchoredmidway between the FB and CM on their respective trammingroutes. Different anchoring configurations can reduce SCcable restrictions to enable the cars to reach certain cuts. Theaverage labour complement of 8.1 is sufficient to run asection; however, if production targets are to be met thenabsenteeism will have to be managed.



Figure 19—Anchoring through a tyre to introduce a spring effect to theanchor point and reduce tension

Figure 18—Cable at anchor point under tension


RecommendationsIt is recommended that when only one split is open, two SCsshould be used. The third can be scheduled for maintenance.WWhen two splits are open, all three cars should be used.However, in cases where this is not possible, a minimum oftwo SCs should be used at any given time in the section. Thebelt should be extended after three splits are open. Whencutting on the right-hand side of the BR and using only twoSCs, use the right-hand and centre cars. When cutting on theleft-hand side of the BR and running only two cars, use theleft and centre cars. For safety reasons, pull the change-outpoint back to avoid shuttle cars turning onto CM operators.Especially in cases such as a left-hand side car not being ableto reach the RHS extremities of the panel, other anchoringconfigurations should be employed. The car should beanchored midway between the FB and CM cutting point; thiswwill reduce the cable length initially required for the car toreach such extremities.

Designated equipment and operators, particularly forcable transportation, are to be employed so as to avoidhaving operators being pulled out from a section to transporta damaged cable. It should be ensured that each car has threecables; one in use, one spare in the section, and one at thecable shop. Zibulo should continue with discipline

f fenforcement by frontline supervisors. Consequencemanagement should be timely, definite, and consistent.

ReferencesANGLO AMERICAN. 2007. Anglo American announces approval of the

Zondagsfontein coal project. Press release:http://www.angloamerican.com/media/releases/2007pr/2007-12-07.aspx[Accessed 20 March 2014].

ANGLO AMERICAN. 2013. Zibulo Colliery: Application for Capital Funding For thePurchase of Three New Replacement Shuttle Cars. Internal Report, AngloAmerican Thermal Coal Operations.

CALLOW, D.J. 2006. The impact of mining conditions on mechanized miningefficiency. Proceedings of the II Underground Operators Conference,Nacrec, Johannesburg, South Africa, 11-12 September 2006. SouthernAfrican Institute of Mining and Metallurgy, Johannesburg. pp. 821–830.

DE LANGE, M.J. 1988. Underground Transport on Large South African CoalMines. PhD thesis, University of the Witwatersrand, Johannesburg, SouthAfrica.

OLIVIER, J. 2014. Techincal manager, Zibulo Colliery, Mpumalanga. Personalcommunication.

HORSTMANN, E. 2014. Cable management at Zibulo Colliery. Undergraduatereport. University of Pretoria.

SMIT, R. 2014. Section foreman, Zibulo Colliery, Mpumalanga. Personalcommunication. �

�


IntroductionThe South African gold mining industry isbased predominantly in the WitwatersrandBasin. The gold reefs found in this basin aregenerally less than 2 m thick and extend todepths in excess of 3 km below surface, withapproximate dips ranging between 20 and 25°from surface (MRM, 2012, p.20). Mines thatextract deposits of this nature are narrow-reefmines. The project site is one of these mines,and conventional drill-and-blast miningmethods are employed. Hand-held pneumaticrock drills are used for face drilling, explosivesare used to fragment the rock, and electric-powered scraper winch systems clean theworking areas by removing broken rock fromthe face and tipping it to the orepass systemthrough a system of in-stope tipping points.

The nature of the orebody and miningenvironment necessitates the use of explosivesas a rock-breaking mechanism, thus makingexplosives an integral part of the mining cycle.Without them, production cannot take place.

Explosives utilization is the usage of

explosives in a manner that yields the desiredresults and that exploits every aspect of theirability to break rock. In order to optimize theuse of explosives, a thorough understanding oftheir properties, characteristics, rock-breakingmechanisms, and application is necessary. Anunderstanding of the basic operationalfunctions of explosives will encourage theimplementation of techniques that lead tooptimal utilization of explosives.

ObjectivesThe project is aimed at investigating the typesof explosives in use at the project site andimproving their utilization by at least 10% bydetermining the following:� Factors contributing to explosives

utilization� Whether explosives are currently being

optimally utilized� The relationship between explosives and

production� Mine standard pertaining to explosives

utilization� Possible causes and consequences of

over- or under-utilization of explosives� How explosives utilization can be

improved by at least 10%.

Explosives consumptionThe mine has an expected broken rock outputper unit of explosives used (de Sousa, 2013).The ratio of explosives used to production(centares) is obtained empirically using thefollowing parameters:� Length of drill steel: 1.2m� Length of drill steel chuck: 0.3 m� Length of hole: 0.9 m� Drilling density: 4 holes per m2

� Shock tubes: 4 tubes per m2

� Panel length: 30 m� Panel width: 1 m

Explosives utilization at aWitwatersrand gold mine by M. Gaula*Paper wwritten oon pproject wwork ccarried oout iin ppartial ffulfilment oof BBSc. ((Mining EEngineering)

SynopsisGold bearing deposits of the Witwatersrand basin are generally less than2m thick and require conventional narrow-reef mining methods forextraction and employ explosives as a means of rock breaking. Optimalutilization of explosives is dependent on the overall design of the blast. Theunder-utilization of explosives arises when shot-holes are drilled inconsis-tently, overcharged, and when tamping is absent. This can be rectified byemphasizing the importance of good drilling practices as part of inductionprogrammes and refresher courses. The project was aimed at determiningwhether or not explosives are being optimally utilized at project site. Thiswas investigated through a study of the properties of explosives, minestandards, and recommendations for usage. Underground observationswere made to determine whether or not mine standards were being adheredto. Historic data was obtained to establish the historic relationship existingbetween the quantity of explosives used (kg) and the production output(m2). This was then compared to the quantity of explosives the mineexpects to use per unit of production. The results obtained were analysed todetermine the presence and extent of over- or under-utilization. It wasfound that explosives are being under-utilized at the mine. More explosivesare ordered than expected per unit of production. The explosives’ propertiesare not thoroughly exploited during blasting, thereby requiring the use ofmore explosives than prescribed.

Keywordsblasting practices, explosives utilization, blast design.

* University of the Witwatersrand, Johannesburg..© The Southern African Institute of Mining and

Metallurgy, 2015. ISSN 2225-6253. Paper receivedMar. 2015



Explosives utilization at a Witwatersrand gold mine

� Burden spacing: 0.6 m.One case of explosives (25 kg) breaks rock over a span of

10 m2. A 30 m panel would therefore require 75 kg ofexplosives and 120 shock tubes.

The quantity of explosives required per panel in kilogramswwould then be obtained from the sum of the explosives used inshot-holes and those used in preconditioning holes:Shot-holes:

1 case = 10 m2

3 cases = 3 m2

Therefore 75 kg explosives would be required for 30 m2

= 2.5 kg/m2

Preconditioning holes:Nine preconditioning holes are expected and there are three

cartridges per hole. A 25 kg box of explosives contains 100cartridges, each with an approximate mass of 0.25 kg.

The mass of explosives in preconditioning holes for theentire panel is

The total explosives mass required for a panel is the sum ofthe mass for the shot-holes and of the mass for the precondi-tioning holes, which equates to 2.725 kg/m2.

It is important to note that this method of calculating theapproximate quantities of explosives required to produce theexpected output is based on the following assumptions:� Face preparation, drilling, charging, and timing are per

mine standard� Panel length is maintained at 30 m and stoping width

kept constant� Secondary blasting is neglected� Blasting of the gullies is not accounted for.

BBlast designOptimal explosives utilization is dependent on the overall blastdesign (de Beer, 2013). It is important to ensure that facepreparation, drilling, and charging are done correctly.

Face preparationBlast designs may vary for various reasons, one major reasonfor this being the stope width. The distance between blast-holes, also known as the burden spacing (G), can be obtainedGas follows:

[1]

wwhereMcMM = mass of explosive per metre of blast-hole (kg/m)K= powder factor (kg/mKK 3).The explosives in use have a density of 1.15 g/cm3. Using

the expression M=ρV, the mass of explosives contained in aVVhole and subsequently, a panel can be obtained. Undergroundobservations carried out on the western panel of the 16th level,31st crosscut are used below to derive the burden spacing of60 cm as per mine standard.

The panel has on average 66 shot-holes and 7 precondi-tioned blast-holes. The total mass of explosives contained inthe panel is obtained as follows.For the shot-holes:

[2]

wherel = length of priming cartridge (cm)ρ = density of cartridge (g/cm3)R = effective radius of shot-hole (cm).

wherel = length of column charge cartridge (cm)

For the preconditioned holes:

Therefore, the total mass of explosives in the panel is

In order to determine the burden spacing, one needs totake into account the powder factor. This is the mass ofexplosive required to break one cubic metre of rock, and iscalculated using the expression:

[3]

The burden spacing is given by

�


McMM fis the mass of explosives contained per blast-hole. Thevvalue has been derived by dividing the total mass contained inthe shot-holes by the number of shot-holes,

The length of the blast-hole that actually contains theexplosive is then found. Since the holes are 1.2 m long and thetotal length of the combined cartridges is 580 mm, only 1.2 ×0.580 m is fitted with explosives. As a result, the mass ofexplosive contained per blast-hole becomes

Substituting this and the K value into the burden spacingequation yields:

This is the maximum burden that the explosives caneffectively handle.

DrillingThe mine has regulatory policies (mine standards) for allactivities carried out during the ore extraction process, whichshould be adhered to at all times. The mine standards fordrilling are as follows:� All drill-holes must be drilled on the position marked on

the face and aligned underneath the direction line � Holes are to be drilled to the full length of the drill steel� All the holes marked on the face should be drilled

ensuring that each hole has the same burden to break� Holes must be drilled at an angle no less than 75° to the

face� Temporary support is to be installed prior to

commencement of drilling.

ChargingThe mine standards prescribe the following when charging upand blasting (de Beer and Ross, 2012):� The primer is prepared by inserting the metal end

halfway into the cartridge. This should be done in a safe,approved priming bay away from the blast site tominimize the risk of accidental firing, which could becaused by stray currents or electromagnetic radiation

� Blast-holes are to be de-sludged using an aluminium 3-way blowpipe and an approved scraper wire. Safetygoggles are to be worn at all times when de-sludgingblast-holes

� Explosives should then be transported to the workingface in elephant bags. The cartridges and accessoriesshould be transported separately in approved containers(elephant bags)

� The primer should be inserted into the hole first andpushed to the bottom of the hole using a square-endedcharging stick

� The column charge is then inserted into the blast-hole.

Proper coupling should be ensured by pushing thecolumn charge as far into the hole as is possible withoutdamaging it

� The remainder of the hole should be tamped to containgases inside the hole using clay tamping provided by themine

� Shock tubes should then be carefully connected to eachother. The connector blocks should be more than 10 cmapart

� Excessive slack between the shock tubes should beavoided in order to prevent whiplash and damage

� Lastly, the shock tube starter is connected to the chargedface, and this is connected to the central blasting system,which is controlled from the control room on surface.

ResultsThe results presented include historic results obtained from theexplosives supervisor and observations recorded undergroundduring the project. The expected explosives utilization iscalculated based on the ratio used by the mine – 2.725 kg perm2. Ordered explosives are calculated based on order anddelivery forms obtained from the mine, and the ratio obtainedby dividing the mass of explosives used by the productionthroughput for the month.

Historic dataThe data here enables a direct comparison to be made betweenthe planned and actual explosives consumption, based on theplanned production output (obtained from the mineral resourcemanagement (MRM) department) and the actual productionoutput (obtained from the production personnel at the shaft)for the period from September to December 2013. The graphswere constructed by comparison of the total planned and actualproduction in relation to the explosives quantities used.

Underground observationsObservations were made in two panels, on levels 18 and 16, togain an understanding of the quantity of explosives used perblast and to determine whether blasting was conducted as per



Figure1 – Mine standards for charging and blasting


�


Table I

Stoping production results for September 2013

Panel Total production Miner Explosives Explosives Shock Shock tubes(m2) expected (kg) ordered (kg) tubes expected ordered

V1 127 A 346 525 508 300V2 190 B 518 300 760 300V3 102 C 278 500 408 900V4 145 C 395 200 580 0V5 203 D 553 1150 812 1010V6 99 E 270 750 396 400V7 37 E 101 0 148 0V7 0 F 0 275 0 300V8 89 G 243 500 356 0V9 86 G 234 250 344 700

Table II

Stoping production results for October 2013

Panel Miner Total production Explosives Explosives Shock Shock tubes (m2) expected (kg) ordered (kg) tubes expected ordered

V1 A 190 517.8 150 760 0V10 B 128 348.8 300 512 100V11 C 95 258.9 300 380 0V11 C 153 416.9 600 612 300V13 E 0 0 600 0 0V5 D 185 504.1 0 740 0V14 D 0 0 600 0 600V6 E 93 253.4 0 372 0V7 E 156 425.1 0 624 0V7 F 130 354.3 275 520 200V8 G 202 550.5 0 808 0V15 G 51 139.0 450 204 300V16 H 0 0.0 125 0 100

Figure 3 – Comparison of expected and actual number of shock tubesordered for September 2013

Figure 2 – Comparison of expected and actual quantity of explosivesordered for September 2013

Figure 5 –Comparison of expected and actual number of shock tubesordered for October 2013

Figure 4 – Comparison of expected and actual quantity of explosivesordered for October 2014



Table III

Stoping production results for November 2013

Panel Miner Total production Explosives Explosives Shock Shock tubes(m2) expected (kg) ordered (kg) tubes expected ordered

V1 A 0 0 550 0 300V17 A 0 0 0 0 0V7 B 97 264.33 275 388 0V11 C 148 403.3 200 592 200V11 C 228 621.3 750 912 300V5 D 243 662.18 250 972 200V18 F 0 0 0 0 0V7 F 107 291.58 250 428 200V19 6 16.35 0 24 0V8 G 164 446.9 0 656 0V15 G 77 209.83 300 308 100V16 H 0 0 125 0 0V20 H 40 109 125 160 100V10 B 102 277.95 50 408 100V21 I 0 0 125 0 0V14 D 0 0 800 0 300V22 J 0 0 600 0 0V23 B 0 0 50 0 0

Figure 6 – Comparison of expected and actual quantity of explosivesordered for November 2013

Figure 7 – Comparison of expected and actual quantity of shock tubesordered for November 2013

Table IV

Stoping production results for December 2013

Panel Miner Total production Explosives Explosives Shock Shock tubes (m2) expected (kg) ordered (kg) tubes expected ordered

V1 A 86 235 0 344 200V17 A 39 106 0 156 0V7 B 105 286 225 420 0V3 C 68 185 0 272 0V4 C 152 414 0 608 0V5 D 0 100 0 0V6 B 180 491 0 720 0V18 F 32 87 0 128 0V7 F 98 267 125 392 100V24 D 83 226 332V8 G 0 0 50 0 100V25 G 0 0 0 0 0V15 G 0 0 200 0 100V16 H 0 0 100 0 0V20 H 75 205 0 300 0V22 J 274 747 0 1096 0V11 C 0 0 200 0 100V21 I 0 0 200 0 200V11 C 0 0 250 0 100V14 D 0 0 400 0 600V22 D 0 0 275 0 100


�


Figure 8 – Comparison of expected and actual quantity of explosivesordered for December 2013

Figure 9 – Comparison of expected and actual number of shock tubesordered for December 2013

Figure 10 – Comparison of expected and actual explosives used percentare

Figure 11 – Comparison of expected and actual number of shock tubesused per centare

Table V

V2 breast panel

Panel characteristics Week 1 Week 2 Week 3 Average

Panel length (m) 20 20 19 19.7Stoping width (m) 1.2 1.1 1.1 1.1Number of marked holes 77 76 70 74Number of preconditioned holes 7 7 6 7Average burden spacing (cm) 58 60 60 57Number of cartridges used 200 200 180 193Number of shock tubes used 100 100 100 100Advance (m) 0.8 0.8 0.8 0.8

Table VI

V20 wide raise

Panel characteristics Week 4 Week 5 Week 6 Average

Panel length (m) 11 11 15 13Stoping width (m) 1.2 1.2 1.2 1.2Number of marked holes 42 45 40 43Number of preconditioned holes 3 3 3 3Average burden spacing (cm) 58 60 55 58Number of cartridges used 100 100 100 100Number of shock tubes used 50 50 50 50Advance (m) 0.8 0.76 0.8 0.79

ff fmine standard, as well as to investigate the effects of notfollowing the mine standard. Underground observations werelimited to two panels because monitoring of the input andoutput parameters and subsequent analysis was to be doneover a series of blasts to increase the accuracy of the results.

Observations were recorded for each shift spent in therespective working places. The weekly averages were thencalculated and from these Tables V and VI were compiled. Thenumber of marked holes is inclusive of the holes marked toblast the gullies in both cases, but excludes the contribution ofsecondary blasting.

AAnalysis of results In September 2013, the expected explosives utilization wasexceeded by 51.3%. This was calculated by direct comparisonof the explosives ordered and the production throughput(Table I). The general trend for the month was that moreexplosives were ordered than expected. Eleven cases ofexplosives and 300 shock tubes were ordered for panel V7, yetthere was no production from that panel. The mine records thisas explosives unaccounted for (wasted). Upon investigation, itwwas found that the 16th level is seismically active with badground conditions. This particular panel had been badlyaffected by a seismic event and had been closed, and the minerwwas assisting in panels V2 and V7, since all three panels are onthe same working level. Panel V2 ordered only 58% ofexpected explosives, and V7 ordered no explosives. Possibly,the first miner was placing explosives orders for the two panelshe was assisting in. An explosives order may only be placed bya miner for a workplace officially assigned to him (de Sousa,2013). Therefore, 11 cases and 300 shock tubes can beaccounted for. The remainder of the panels ordered moreexplosives than expected, and the possible reasons for this arediscussed in detail later.

The results obtained for October indicate that moreexplosives were ordered than expected. There were againpanels that received explosives yet showed no production. Inthis case, V5 and V14 were under the administration of thesame miner who received 24 cases of explosives for one panelthat were actually intended for another panel. The same appliesto panels V13 and V6.

During the month of November, six panels orderedexplosives with no production throughput confirming wherethey have been used. No relationship can be establishedbetween panels that ordered explosives without producing andthose that produced without ordering explosives. A total of 90cases of explosives and 600 shock tubes were ordered andthese remain unaccounted for. Nothing can be said about theirutilization and these explosives can be concluded to have beenwwasted. December shows the same trend- explosives wereordered yet nothing produced.

Occurrences of November and December are, for thepurposes of this report, extreme cases that have requiredextensive research and enquiries about exactly what happenedduring that period. The remainder of the cases are those wheremore explosives were ordered than were expected by the mine.

The ratio of explosives (kg) to production output (m2)expected by the mine is 2.725:1, and 4:1 for shock tubes.Figures 10 and 11 indicate the performance of the mine inrelation to the expected figures. Variations in the ratios areevident, indicating cases of both over- and under-utilization of

fexplosives. Over-utilization occurred when fewer explosiveswere used than expected, and under-utilization when moreexplosives were used than planned. The contributing factors toboth over- and under-utilization of explosives, based on minestandards and underground observations, are discussed indetail below.

Inconsistent blast-hole length and drilling angleUnderground observations made revealed that at times, theblast-holes are drilled to a shorter length than specified in themine standard. The impact of shorter blast-holes isdemonstrated using the following simple example.

The ideal case (according to mine standard), assuming a30 m long panel with a 1m stoping width, is as follows:� Blast-hole length: 0.9 m� Advance per blast: approx. 0.8 m� Explosives used per blast: 29 307.21 g� Advance over 20 blasts: 16 m.

The effect of short blast-holes can be seen from thefollowing calculation:� Blast-hole length: 0.85 m� Advance per blast: 0.75 m� Explosives used per blast: 29 307.21g� Advance over 20 blasts: 15 m

When blast-holes are drilled shorter than prescribed by themine standards due to incorrect drilling angles, the advance isreduced although the same quantity of explosives is used asfor the full-length blast-holes. This results in under-utilizationof explosives because the full potential of the explosives is notused. The calculation above (case 2) is exaggerated slightlybecause it assumes that all holes in the panel are drilled at0.85 m length. However, this calculation demonstrates theeffect of shorter blast-holes on the utilization of explosives. Inaddition, if blast-holes are drilled to insufficient lengths, 4.8cm of face advance is lost per blast (de Beer, 2013).

Incorrect burden spacingFor every 10 cm increase in burden spacing, 10% face advanceis lost per blast (de Beer 2013). A burden spacing of 60 cmensures optimal fragmentation, due to the interaction betweenadjacent charges.

When the burden spacing is increased, the explosiveenergy needs to travel further than 0.3 m to effectively breakrock from the adjacent blast-hole. Thus the explosive energy isdepleted before optimum fragmentation is achieved. Thisresults in poor hangingwall and footwall conditions and anuneven face shape, as well as over-utilization of explosives.

If the burden spacing is reduced, the explosive energyfreleased is more concentrated, leading to finer fragmentation of

the rock mass, but also to overbreak of the hangingwall. Anydeviation from the prescribed burden spacing results inapproximately 10% overutilization of explosives, an unevenface shape, and poor fragmentation.

OverchargingMuch of the explosives energy concentrated in the blast-hole isnot evenly distributed but is concentrated within the confinesof the surrounding rock mass. As has been observedunderground, there is a misconception that overcharging isbeneficial to the advance achieved. However, when morecartridges are placed in a blast-hole than the quantity required,more energy is released into the blast-hole. This energy, if




ff f ftamping is sufficient, causes both overbreak and fine fragmen-tation, as were as over-utilization of explosives.

DDrill bit deteriorationThe drill bits used in the working places are 34 mm indiameter. According to Jijingubo (2013) deterioration due towwear and tear results in the gradual reduction of the drill bitdiameter, thus causing a reduction in the diameter of the blast-hole. Jijingubo suggested that this reduces free movement ofthe cartridge inside the blast-hole, thus rendering explosivesless effective than they would be when using fairly new drillbits.

PPoor or no tampingThe importance of tamping should not be underestimated.Underground observations showed that adequate tamping ofblast-holes is often neglected when charging up, especiallyclose to the end of the shift. Figure 12 illustrates theimportance of tamping.

Explosive energy released into the blast-hole uses twoprimary mechanisms for rock fragmentation: shock and heave.For effective fragmentation, the explosive energy should becontained in the blast-hole long enough to cause expansion ofthe cracks induced by the shock mechanism. Tamping aids inthis regard by enabling the explosive itself and the energy itreleases to remain in the blast-hole and cause expansion as thegaseous products from detonation penetrate the inducedcracks. The absence of tamping or even poor quality instal-lation of tamping allows the gas to escape and hence theenergy is released into the surrounding environment. Thissometimes causes damage to permanent support elements andoverbreak, because the energy is not fully released into theblast-hole but is allowed to escape to other areas where it isnot desired.

Figure 13 illustrates the effect of tamping on face advance.Because the absence of tamping allows gases to escape, theend of the blast-hole is often not blasted, leaving socketsbehind and subsequently reducing the advance achieved perblast. For 0.9 m holes, no tamping results in 12 cm loss perblast (de Beer, 2013).

Unused cartridges and shock tubes remaining at thefaceThe mine standards require that unused explosives andaccessories be returned to the explosives box and locked away.

Strict explosives control policies are employed at the mine – allexplosives should be accounted for. The miners keep a recordof the quantity of explosives and accessories in storage, andupon receipt of a new batch the quantities are adjustedaccordingly. A record of explosives used is to be kept as well.Underground observations proved non-compliance to thisrequirement, since in both panels observed, no unusedexplosives were returned to the explosives boxes, and it wasassumed that all explosives and accessories taken into the facewere used.

Blasting of gullies and secondary blastingunaccounted forThe mine standards require that gullies be blasted such thatthey lead the face. This is to ensure that the ore blasted has afree face to break into. The centre gully should always lead theface, while following the survey line pegs (Figure 14). Inpractice, the quantity of explosives used to blast an entire panelincludes the explosives used to blast the gully, as well as theface. However, the means of determining the quantity ofexplosives required per square metre does not distinguishexplosives used for blasting gullies. Thus the results over-estimate the utilization of explosives to blast the face, whereassome of these were used to keep the gully ahead of the face.Blasting of the gullies is such an important aspect ofproduction that this usage should be allocated an explosivesconsumption factor.

The blasting of gullies in a 20 m panel entails five blast-holes and would consume approximately 12–15 cartridges, 5–7shock tubes, and detonating cords. This may appearinsignificant, but it increases the amount of explosives usedwhile not contributing to production. The resulting higher-than-expected explosives utilization factor can be corrected byincluding the explosives used to blast gullies (centre andstrike) in the planning of the quantity of explosives expected tobe used in a panel.

Secondary blasting is usually unaccounted for whenallocating explosives to working places. This is done whenremoving obstructions such as large rocks from grizzlies andwhen blasting bad hangingwall conditions, including brows.Unlike the case of blasting gullies, secondary blasting is usedonly irregularly and is considered a result of poor primaryblasting. However, it is a contributing factor to the apparentover-utilization of explosives.

�


Figure 12 – The effect of tamping on explosives effectiveness (de Beer,2013)

Figure 13 – The effect of tamping on face advance (de Beer, 2013)

LLimitations of the record-keeping/monitoring systemThe mine has a record-keeping system in place in which allminers order their explosives and accessories for specificwworkplaces. As can be seen from the results obtained, somewworkplaces have placed orders for explosives while there is noproduction to account for the usage. However, it is common fora miner in charge of multiple panels that are relatively close toeach other to order explosives for panel A but use them to blastpanel B. Comparisons of the expected and actual quantity ofexplosives used per panel exaggerate the extent of theproblem, since the trade of explosives between panels is nottaken into account.

ConclusionsExplosives are a vital component of hard-rock miningoperations using conventional mining methods for oreextraction. Mine standards are in place to ensure that allactivities involved in the production process are carried out in awway that ensures employee safety and maximizes productionoutput. This study indicates that explosives are not beingutilized to their full capacity at the mine. The biggestcontributor to the apparent under-utilization of explosives isthe limitations of the system that tracks the usage ofexplosives underground.

The system does not allow a miner to order explosivesunless they are for a specified panel officially assigned to him.There are currently no means of determining how much of theordered explosives is actually used underground and howmuch is returned to the explosives boxes. Other factorscontributing to under-utilization of explosives are directlyrelated to the overall blast design. These include, but are notlimited to; overcharging, incorrect drilling lengths and drillingangles, secondary blasting not being accounted for, as well asthe somewhat impractical expectation of explosivesconsumption that the mine currently has. Under-utilization ofexplosives also leads to poor ground conditions and increasedcosts because the mine has to purchase more explosives thanrequired yet the production output remains unchanged.

The utilization of explosives can be improved byimplementing changes in the explosive ordering process andproviding a means of tracking whereby ordered explosives areused. By so doing, no explosives will be unaccounted for and theutilization problems encountered in the stopes can be addressedwwith a realistic picture of the extent of under-utiliation.

RecommendationsThe results of this project indicate that the current system usedby the mine to calculate the amount of explosives that shouldbe used per square metre has the following limitations:� It is based on a panel length of 30 m, which is not the

average panel length for the shaft� Blasting of gullies and secondary blasting is not

accounted for when calculating the expected explosivesconsumption

� Once explosives are delivered to the miner, no furtherrecords are kept of their distribution among the variousworking places

� Miners are permitted to order explosives only for thepanels officially assigned to them. The system assumesno trading of explosives takes place between miners.

These limitations exaggerate the extent of explosivesunaccounted for and the extent of under-utilization. In order toimprove the utilization of explosives, it is important thatexplosives are used to obtain the best results and not under-estimated. The mine can apply the following measures toimprove the utilization of explosives.

The explosives usage calculatorThe explosives usage calculator can be introduced into thesystem to aid in monitoring of explosives usage underground.This form (Figure 15) would be made available together withthe explosives order form. After blasting, the form should beinserted into the communication book at the end of the shift.The availability of this information is aimed at encouraging theminer to directly monitor explosives usage and compare it tothat which is expected.

Adjustment of consumption parametersPlanning for explosives consumption at the mine is somewhatunrealistic. The benchmark of 2.725 kg/m2 is based on a panellength of 30 m and constant stoping width. This is not a truereflection of the mining conditions, since pillar extraction is thepredominant mining method and the panel lengths areconstantly adjusted owing to ground conditions andintersection of geological structures (Tsibuli, 2013). Instead ofa fixed benchmark, the mine can employ a consumptioncalculation method that allows for flexibility due to changinglocalized conditions and accounts for the blasting of gullies aswell as secondary blasting. This will present a practical modelfrom which consumption parameters can be calculated andreduce the apparent extent of under-utilization, therebyimproving utilization in future.

Training

Formal training

Scientific details of rock-breaking should be included ininduction programmes and refresher courses to broaden theknowledge of explosives handling personnel and help themunderstand the importance of a 60 cm burden spacing. Minersshould constantly be reminded that overcharging is in no waybeneficial to mining operations. In addition, employees shouldbe informed and constantly reminded about the financialimplications of face advance loss per blast and how this affectsthem.



Figure 14 – Marking of the centre gully


I f l i iInformal trainingDiagrammatic representations in the form of clearly visiblelaminated posters at waiting places and in the change housesinforming employees about the impact of poor drilling practiceson the centares they produce monthly and their inability toreach set targets.

Models made of rubber, clay, or any recyclable materialdisplayed at various places in the shaft. These should bedesigned such they show the goal (reaching the mine callfactor) and all the factors that prevent the set targets betingreached, such as incorrect burden, shorter shot-holes, poortamping, overcharging etc. These factors could be representedccby e.g. parasites feasting on the target – something everyonecan relate to and work together against.

Introduction of light, flexible 60 cm long strings made ofrecyclable materials that can be folded into 10 cm or 5 cmportions. These would be made available to all stoping crews.The aim here is to involve the crew in adhering to a consistent60 cm burden spacing, and holding the miner accountable forany inconsistencies, which can then be raised by the crewinstead of production supervisors. This is an example of thebottom-up management approach.

A k l dAcknowledgementsEurIng C.R Beaumont: Project SupervisorMr C.K Tsibuli: Production Supervisor, Project SiteMr D. Setshoantsho: Production Supervisor, Project SiteMr R. de Beer: Blasting Design Engineer, Project SiteMr B. Prout: Lecturer.

ReferencesDE BEER, R. 2013. Blasting Design Engineer, Project Site. Personal communi-

cation, December 2013.

DE BEER , R. and ROSSRR , A. 2012. Project Site, Charge up and Connect Blast. pp.2–5.

DE SOUSA, J. 2013. Explosives wSupervisor, Project Site. Personal Communication.

DE SOUSA, J. 2013. Explosives Supervisor, Project Site. Underground Visit ProjectSite Report. p. 3.

PROJECT SITE MINE MINERAL RESOURCERR MANAGEMENTMM TEAM. 2012. Competent Person’sReport on the Project Site. p.20.

JIJINGUBO, X. 2013. Miner, Panel V20, Project Site. Personal communication,December 2013.

TSIBULI, C.K. 2013. Production Supervisor, Project Site. Personal communication,December 2013. �

�


Figure 15 – The explosives usage calculator

IntroductionThe investigation was undertaken at aplatinum mine on the western limb of theBushveld Complex (BC), just outside the townof Rustenburg in South Africa’s North WestProvince (Figure 1).

The mine, designated ‘mine X’, exploitsthe Merensky and UG2 reefs, which arecurrently the only reefs of economicimportance.

The BC consists mainly of alternatinglayers of norite, pyroxenite, and anorthosite.The general stratigraphy of the Complex isshown in Figure 2.

The mine employs the conventionalnarrow-reef breast mining method. Theworkings are served by footwall waste rockdevelopment. The mine infrastructure consistsof a main vertical downcast shaft that extendsdown to 15 level, 598 m below surface, and adecline that extends from 15 level down to 28level at a depth of 927.1 m below surface. TheMerensky Reef workings are accessed from thedecline, whilst the UG2 workings are accessedfrom the vertical shaft. The support methodconsists of a regional pillar and chain crushpillar combination in order to support thehangingwall. The method of stoping andsupport is illustrated by Figure 3 (note thepositions of the regional and chain crush(yielding) pillars.

The mine has a history of substandardpillar cutting practice. A few of thesurrounding mines that have experiencedpillar bursts due to substandard pillar cuttingwere taken as case studies in order to gaugethe effect of noncompliance and identify thepossible causes that can lead to a pillar burst.

The objectives of this investigation were toreview the current design of crush pillars aspractised at the mine, and to compare thecurrent design with an alternative method.This further entailed the identification of thepractical limitations experienced undergroundduring pillar cutting. Finally, the recommen-dations to rectify the problems identified areprovided.

Mine standards and complianceThe workings at mine X are divided into twosections, the Merensky section and the UG-2section. The Merensky section is minedbetween 598 m and 927.1 m below surface,and the UG-2 section is mined down to 598 mbelow surface. The two sections have differentsupport standards. Pillar cutting compliancerefers to the percentage of crush pillars thatare cut to the mine standards.

Merensky sectionCrush pillars for the Merensky section aredesigned to be 4 m in length and 2.5 m wide,with a pillar height of approximately 1.1 m.The layout of a Merensky section panel can beseen in Figure 4.

Critical investigation into the problemssurrounding pillar holing operationsby J.P. Labuschagne*, H. Yilmaz†, and L. Mpolokeng‡Paper written on project work carried out in partial fulfilment of BSc. Eng.(Mining Engineering)

SynopsisAn investigation into pillar cutting was carried out at a platinum mine onthe western limb of the Bushveld Complex. The focus was on crush pillardesign and implementation in order to ultimately improve the compliancepercentage for pillar cutting. The major findings from the investigationsuggest that the pillar cutting problem lies with the implementation of thedesign rather than the design itself. Observations of the practical issuesunderground that prevent good pillar cutting were made. After these issueshad been identified, recommendations to rectify these problems and a fewother issues identified during the investigation were provided. Therecommendations are aimed at improving the pillar cutting compliance andreducing the likelihood of pillar bursts or pillar runs, which will ultimatelycreate a safer mining environment.

Keywordscrush pillar, design, practical issues, implementation.

* University of the Witwatersrand, Johannesburg,South Africa, Anglo American Platinum BursaryHolder/Trainee.

† School of Mining Engineering, University of theWitwatersrand, Johannesburg, South Africa.

‡ Anglo Development Centre, HRD Co-ordinatorMining.

© The Southern African Institute of Mining andMetallurgy, 2015. ISSN 2225-6253. Paper receivedrrJan. 2015



Critical investigation into the problems surrounding pillar holding operations

The position of the crush pillar (1) can be seen in relationto the pillar reference line (4), the holing between pillars (3),the panel face (5), the advanced strike gully (ASG) (6), thedip (8) and strike (7) direction. Note that the distancebetween the pillar reference line and the side of the crushpillar (2) should be 0.5 m.

The pillar compliance percentages for 3 months werecompiled and are shown in Table I.

UG2 sectionThe crush pillars for the UG2 section are designed to be 3 min length, 3 m wide, and approximately 1.1 m in height.Figure 5 illustrates the layout of a UG2 section panel. The explanation is the same as for the Merensky section (Figure 4), with the only difference being the crush pillardimensions.

The compliance percentages for 3 months were compiledand are shown in Table II.

From Table I and Table II it can be seen that thecompliance percentages are well below the acceptablestandard of 80%, and this situation is indeed a cause forconcern.

�


Table I

Compliance percentages for three months for theMerensky section

Year Month Compliance, %

2013 September 632013 November 492013 December 42

Figure 1—Location of Rustenburg in the Bushveld Complex (Watson etal., 2007)

Figure 2—General stratigraphy of the Bushveld Complex (Jager andRyder, 1999)

Figure 3—Stoping and support method (Watson et al., 2007)

Figure 4—Merensky Reef crush pillar layout (not to scale). (1) Crushpillar, (2) distance to pillar reference line, (3) holing between pillars, (4)pillar reference line, (5) panel face, (6) advanced strike gully, (7, 8) dipand strike directions

Figure 5—UG2 Reef crush pillar (not to scale)

Case studiesTwo case studies were conducted at mines that are in closevvicinity to mine X and where pillar burst incidents hadoccurred. One involved a confirmed pillar burst at mine Y,and the other one involved a suspected pillar burst at mine Z.Both these incidents occurred during the project investi-gation. These case studies shed light on some importantissues pertaining to the pillar cutting practice.

MMine Y confirmed pillar burstThe pillar burst at mine Y generated a 2.9 magnitude seismicevent. As a result, six employees were injured; fortunately,no fatalities occurred. This pillar burst was the result of anoversized pillar that was left inside the stope as miningprogressed. According to the mine standards the pillar shouldhave been roughly 3 m wide by 3 m in length. The actual sizeof the pillar was 17.5 m in length by 6.5 m in width, asillustrated in Figure 6. The blue blocks indicate the size of thecrush pillars that should have been left according to the minestandards. As can be seen from Figure 6, the pillar that burst(hatched in grey) was unacceptably oversize.

According to Watson et al. (2007, 2010), pillars shouldcrush close to the face (preferably within the first 7 m) understiff loading conditions in order for controlled crushing tooccur. Watson et al. (2010) also show that pillar bursts arelikely to occur at 10 m to 14 m from the face under softloading conditions. ‘Stiff’ and ‘soft’ loading conditions aresimilar to the concept of loading done by the stiff and softtesting machines used in rock mechanics laboratories.

It can therefore be concluded that, referring to Figure 6,the smaller crush pillars that surrounded the oversize pillarall failed progressively under stiff loading conditions as theface advanced. The oversize pillar remained intact withoutcrushing and moved well into the back area of the panel,wwhere high stresses were accumulating in the pillar. Thepillar eventually started crushing, and the excess strainenergy that was stored in the foundation rocks of the pillarthen released and caused violent failure.

MMine Z suspected pillar burstNot much is known about the seismic event at mine Z, whichoccurred in December 2013, but it is suspected that the eventcan be attributed to a pillar burst. It is speculated that a largepillar with a width–to-height ratio greater than 10 was leftbehind in one of the panels. This was deemed acceptable,since the pillars with width–to-height ratios greater than 10are known to be virtually indestructible (Ozbay et al., 1995).The problem suspected here was that the pillar consisted oftwo different rock types, one of which was weaker than the

other. The rock types were separated along the vertical plane.The weaker part of the pillar started failing slowly, leavingonly the stronger portion of the pillar intact. At that stage, thepillar was positioned well into the back areas as miningadvanced, where soft loading conditions dominate. Owing toits reduced dimensions, the pillar was then within thebursting range. It is suspected that the pillar then burst,causing a seismic event.

Design and analysisCrush pillars are designed in order to prevent a back-break ofthe hangingwall, while maximizing the percentage extraction.Crush pillars are generally used where mining takes placebetween 600 m and 1000m below surface (Jager and Ryder,1999). The support layout used is a regional pillar and crushpillar combination as shown in Figure 3. This support layoutallows for an increased extraction ratio while still ensuringstability of the hangingwall.

The first issue to focus on when designing a crush pillaris the residual strength of the pillar. The residual strength ofa crush pillar is the strength that the pillar has after crushinghas occurred. The residual strength must be sufficient toprevent a back break in order to be effective. Note that thecrush pillars do not support the entire hangingwall strata tothe surface. The regional pillars left on each side of the stopeare responsible for the hangingwall support to surface. Thismeans that crush pillars have to support only the tensile zonethat exists between regional pillars.

According to Watson et al. (2010), in order to prevent aback break the residual strength of a crush pillar should bebetween 8 MPa and 13 MPa when pillar lines are spaced30 m apart. For the purposes of this design, crush pillars aredesigned with a residual strength of 13 MPa.

The residual strength of a crush pillar can be determinedby using a formula developed by Salamon (Watson et al.,2010):

[1]



Table II

Compliance percentages for three months for theUG2 section

Year Month Compliance, %

2013 June 612013 July 622013 September 45

Figure 6—Oversized pillar (grey hatching) compared with standardpillars (blue blocks)


wwhereh = pillar height (m)w = pillar width (m)Cb = the cohesion of the crushed rock material (MPa)

(Watson et al., 2010, citing Salamon).

The Cb value was taken as 1.6 MPa, and h as 1 m. Thepillar width corresponding to a residual strength magnitudeof 13 MPa can then be calculated by trial-and-error andinterpolation.

The next step is to find the pillar strength using the pillarwwidth, which was obtained using Equation [1]. For designpurposes, the pillars are taken to be square. There are twopillar strength formulae that are used in this design. Equation[2] refers to the slender pillar formula as currently used bymine X. Equation [2] is an adjusted version of the 1972Hedley and Grant formula (Watson, 2010).

[2]

wwherek = the design rock mass strength (DRMS) (MPa)h = pillar height (m)w = pillar width (m)β = 0.75α = 0.5 (Watson, 2010).

The other pillar strength formula that is used forcomparison is explained by Watson et al. (2010) as follows(Equation [3]).

[3]

wwhereh and w are as defined aboveLL = pillar length (m)he = [1 + 0.2692 (w/h)0.08)h.Note that the 136 denotes a strength factor that should bealtered for a different rock type. In the case of this design,136 will be replaced by the DRMS that was established forthe particular rock type.

The next step in the design process is to determine theaverage pillar stress (APS) for the specific scenario. This canbe derived by using the tensile zone thickness that the crushpillars have to support.

The pillar factor of safety can then be determined in orderto ensure that the proposed pillar will fail close to the faceunder stiff loading conditions. The pillar should then stillprovide sufficient support resistance based on the residualstrength of the crushed pillar.

The results that were obtained showed a range of safetyfactors between 0.62 and 0.38, which ensure crushing understiff loading conditions. Both the peak pillar strengthequations [2] and [3] delivered similar results for the pillarstrength. The pillar width–to-height ratio obtained was 2.15,wwhich correlates well with the mine standards, where awwidth-to-height ratio range of 2.0–3.5 is acceptable.

The results also showed that as depth increases, thetensile zone thickness decreases, which means that larger in-panel pillars are more acceptable at shallower depths than atdeeper levels.

The results obtained in this investigation suggest thatpillar cutting is an application problem and not a designproblem. For this reason, further investigations were carriedout into the practical problems surrounding pillar cutting.

Practical problems in pillar cuttingFour practical problems with pillar holing operations wereidentified. These problems are discussed in the followingsections.

Drilling discipline of rock drill operatorsWhen the rock drill operators (RDOs) drill the stope face forproduction purposes, they drill in the direction of the ASG(strike direction). According to mine standards, holingsbetween crush pillars have to be blasted after every 7 m faceadvance for the Merensky section, and after 6 m face advancefor the UG2 section. This is due to the different pillar lengths(4 m in the Merensky section and 3 m in the UG2 section),with the pillar holings in both sections being 3 m wide. Theproblem occurs when it is time to blast the holings into thesiding to create the crush pillars. The blast-holes are markedon the stope face and on the siding where the holing shouldbe blasted. The RDOs then begin by drilling perpendicularholes into the stope face in the direction of the ASG. When itis time to turn 90° towards the siding of the panel in order todrill the blast-holes for the holing, RDOs tend to turn lessthan 90°. This results in a holing that is not blasted in thecorrect direction, but which is skewed in the direction of theface advance. Owing to this skewness, the holings tend to belonger than usual, hence the pillar width-to-height ratios areaffected. The pillars that result from this poor drillingtechnique are longer than designed for. This in turn affectsthe effective pillar widths, and can hence lead to a pillar burstproblem if not addressed.

Pillar reference line pegs lag behind panel advanceThe pillar reference line has to be parallel to the ASG and adistance of 0.5 m from the side of the in-stope crush pillars.This reference line is painted onto the hangingwall of thestope in line with pegs that are installed by the surveyors.These pegs are normally installed regularly to ensure that thepillar line can be extended in a straight line and does notdeviate from its intended direction. Offsets are taken from thepillar reference line to the siding of the panel, indicatingwhere the pillars have to be situated. When the pegs for thepillar reference line lag the advancing panel, the lineextension cannot be marked accurately. Generally this resultsin the pillar line deviating from its intended course, eithertowards the siding of the panel or towards the panel itself.When the reference line deviates towards the siding of thepanel, undersize pillars can be expected. Conversely, if theline deviates towards the panel, then oversized pillars can beexpected.

Substandard face markingIn some situations, the face marking is substandard due tothe team leaders marking the face, and not the miner. As aresult, the drill-holes are drilled in the wrong position ordirection. Ultimately, the poor marking practice also affectsthe holing size and shape. Oversize or undersize pillars areinevitable, depending on the way the siding is marked. If the

�


spacing between consecutive blast-holes is too large, then thepillars will tend to be undersize, and when the spacing is toosmall, the pillars tend to be oversize.

RRock removal (scraping) difficultiesThe crush pillars are designed roughly 3 m × 3 m on both theMerensky and UG-2 reefs. This means that three blasts arerequired for a holing to be completed since the advance perblast is normally 1 m. This relates into a face advance ofroughly 3 m during this three-blast period, since the face alsoadvances roughly 1 m per blast. Therefore, in order toremove the broken rock from the blasted holing, the facescraper has to be moved back approximately 5–6 m from theface and rigged suitably at the holing after scraping the face.This takes extra time and effort, and the workersunderground in some cases prefer not to lose this time inorder to scrape the small quantity of rock, which eventuallybuilds up in the holing.

Scraping the blasted rock in the holing usually causes amore serious problem – the sticks (elongates) become scrapedout during this practice. The line of sticks needs to beconstantly advanced as the face advances. The maximumdistance that these sticks may be from the face is 4.3 m.WWhen re-aligning the face scraper to remove the blasted rocksafter the second and third blast of the holing, the scraper willhave to be moved back. This means that it is highly likely toscrape out the sticks, which then have to be re-installed. Inaddition to the safety risks, the re-installation process takestime and wastes supplies, and hence costs are also raised.

The solutions applied underground to correct these kindsof problems are often crude in the sense that the solutioncreates another problem somewhere else. For example, thepillar holings could be blasted at an angle to try and avoidthe problems associated with rock removal. This in turn couldlead to incorrectly and unevenly sized pillars, which are indanger of bursting.

ConclusionsWWhen attempting to design in-stope crush pillars, thedetermination of the tensile zone thickness becomesimportant in order to evaluate the demand required fromcrush pillars. As shown in this paper, the tensile zonethickness decreases with increasing depth. This is thepremise upon which crush pillars can be implemented instopes deeper than 600 m with fairly good results. The mostimportant consideration when designing in-stope crushpillars is the residual strength that is required from the pillarin order to arrest a back break. This residual strength ismatched to the required pillar width-to-height ratio. The peakpillar strength can then be computed for the width-to-heightratio required, and thus compared to the average pillar stressto determine whether the factor of safety is adequate. Itshould be noted that in crush pillar design, the safety factorshould be less than 1.0, and optimally around 0.7. The lowfactor of safety is necessary in order to prevent a pillar burst,and to promote pillar crushing close to the stope face understiff loading conditions.

Important considerations in crush pillar design werehighlighted by the case studies of the pillar bursts at mine Yand mine Z. The problem at mine Y was that the pillars werecut with inconsistent dimensions, and one of the oversized

fpillars burst in the back area under softer loading conditions.The suspected pillar burst at mine Z showed that particularattention should be paid to avoid leaving pillars in situ thatconsist of more than one rock type.

The mine standards on crush pillars are found to comparewell to the design results achieved in this investigation.However, the poor pillar cutting track record at mine X wouldlead to pillar burst problems in the near future in view of theincidents of crush pillar failures at the surrounding mines.Pillar cutting compliance therefore has to improve. During theinvestigation it was found that the problems regarding pillarcutting is not due to the design, but rather to implementation.

RecommendationsThe following recommendations are offered to rectify the keyissues identified surrounding pillar cutting that can causepillar bursts and subsequent seismic events.

� The practical problems that were identified during thisinvestigation have to be addressed to improve pillarcutting practice in order to avoid future crush pillarfailures at the mine

� The tensile zone thickness increases as the miningdepth decreases. This means that in-stope crush pillarsat shallower depths carry higher loads than the deeperones. The safety factor will therefore decrease as depthdecreases if the crush pillar size remains the same. Thecrush pillar width-to-height ratio could be increased(so as to create a stronger pillar at shallower depths) tomaintain a safety factor of about 0.7. It is thereforerecommended that the width-to-height ratio of in-stopecrush pillars is determined separately for each mininglevel. Mine management would then need to ensurethat shift supervisors and mine overseers are aware ofthe different sizes of in-stope crush pillars on differentlevels

� Emphasis should be placed on cutting pillars consis-tently according to mine standards. Larger pillarsshould not be left in situ in order to compensate forsmaller pillars left previously. The leaving of pillarswith inconsistent dimensions increases the likelihoodof premature pillar failures. The mine standards shouldrather be applied, even if the pillars were cut oversizeor undersize previously

� Special care has to be taken to ensure that pillars arenot cut in a position where they consist of more thanone rock type. If this cannot be avoided, additionalsupport such as thin spray-on liner or wire mesh andlacing could be applied to improve the crush pillar’syielding capability

� Pillars that have been identified as oversize pillarswithin bursting range should be de-stressed by meansof drilling two parallel blast-holes into the pillar andblasting with a low powder factor in order to inducecrushing. This preferably has to be done before thepillar moves more than 7 m away from the advancingface, where soft loading conditions will start to have aneffect

� Drilling and blasting the holings from the top strikegully of the adjacent panel heaves the rock straight intothe top gully of that panel, which solves the rock




f fremoval problems identified. Any rock left inside thesiding can be removed by a hand-shovel more easilydown-dip

� The RDOs should be re-trained in order to emphasizethat the pillar holings have to be blasted perpendicularto the face. Also, the RDOs should be made aware thedangers of leaving oversize or undersize pillars. Thistraining should preferably be done by rock engineers

� Regular checks should be made to see whether the pegsare installed so that the pillar reference lines can beextended according to plan. More surveyors could beappointed to ensure that the pegs of the pillar referencelines do not fall behind the plan. Better communicationbetween miners and the survey department should beencouraged so that pillar reference lines can be kept upto date

� Disciplinary measures should be taken against repeatoffenders. This action can be justified by the panelsthat actually comply with the mine standards. A newbonus system could also be introduced in order tomotivate employees to cut the in-stope crush pillarsaccording to mine standards

� The substandard face marking problem can beovercome by either appointing more miners to sharethe work load or by training team leaders for face

fmarking. Miners can then focus on marking the pillarholings since the team leaders can mark the face. Thisreduces the work load on the miner and hence leavesno excuses for substandard marking of the face orholing

� Using a burn cut blasting pattern with 3 m long holescan be trialled in order to see whether it can solve therock removal problems. If this succeeds, then a holingcan be blasted in a single blast, and no rock removalproblems will be experienced.

ReferencesJAGER, A.J. and RYDERRR , J.A. 1999. A Handbook on Rock Engineering Practice for

Tabular Hard Rock Mines. Safety in Mines Research Advisory Committee,Johannesburg.

OZBAY, M.U., RYDERRR , J.A., and JAGER, A.J. 1995. The design of pillar systems aspractised in shallow hard-rock tabular mines in South Africa. Journal ofthe South African Institute of Mining and Metallurgy, vol. 95. pp.7–18.

WATSONWW , B.P. 2010. Rock behaviour of the Bushveld Merensky Reef and thedesign of crush pillars. Faculty of Engineering and the Built Environment,University of the Witwatersrand, Johannesburg. pp.201-207

WATSONWW , B.P., KUIJPERSKK , J.S., and STACEY, T.R. 2010. Design of Merensky Reefcrush pillars. Journal of the Southern African Institute of Mining andMetallurgy, vol. 110. pp. 581–591.

WATSONWW , B.P., ROBERTS, M.K., NKWANA, M.M., KUIJPERSKK , J., and VAN ASWEGEN, L.2007. The stress-strain behaviour of in-stope pillars in the Bushveldplatinum deposits in South Africa. Journal of the Southern AfricanInstitute of Mining and Metallurgy, vol. 105. pp. 187–194. �

�


IntroductionThis report focuses on the factors inhibiting theproductivity of load haul dump machines(LHDs) at Conzal Mine from achieving settargets. The aim was to identify the majorfactors contributing to low productivity andrecommend ways of reducing or eliminatingthem.

Background Conzal Mine1 is located near the town ofSteelpoort on the eastern limb of the BushveldComplex, approximately 350 km north-east ofJohannesburg, in Limpopo Province (Figure 1).

The mine exploits the MG1 and MG2chromite seams of the in the RustenburgLayered Suite of the Bushveld Complex. ConzalMine historically consisted of three shafts:

namely, the North, South, and Broken Hillshafts. The main focus of the project was on theSouth and Broken Hill shafts, as the North shaftwas closed down after reaching the end of itsproductive life. Both shafts access theunderground workings by means of declines,and the mining method is bord and pillar. Oretransportation at Conzal Mine is by means of abatch and a continuous transportation system.The batch transportation utilizes LHDs, whichload the broken ore from the face and tram it tothe tip, which then feeds into the continuoustransportation system consisting of conveyorbelts. The mine makes use of a total of 11 low-profile LHDs (illustrated in Figure 2), with threeof the eleven used as standby machines and forunderground construction.

Project objectivesProduction is effectively the heart of all miningoperations. It is therefore essential thatproduction targets are met on a daily basis sothat the mine remains profitable.

Throughout 2013, Conzal Mine wasstruggling to meet production targets. It isbelieved that the main reason for the mine notmeeting its production target was due to theLHDs not being able to tram enough ore fromthe face to the grizzly tips. There was thereforea need to determine main factors that inhibitedLHDs at Conzal Mine from meeting theirtramming target of 2 200 t/d.

The objectives of the project were to:� Identify major factors leading to the

inability of the LHDs to reach theproduction target

� Provide solutions and recommendationson how to overcome these inhibitingfactors.

MethodologyIn order to achieve the above objectives, thefollowing methodology was employed:

LHD optimization at an undergroundchromite mineby W. Mbhalati*Paper wwritten oon pproject wwork ccarried oout iin ppartial ffulfilment oof BBSc. ((Mining EEngineering)

SynopsisConzal Mine was not meeting production targets in 2013, and it wasestablished that this was caused by the inability of the load haul dumpmachines (LHDs) to tram the required tonnages. An investigation of theLHD productivity was therefore conducted to identify the inhibiting factors.This was accomplished by carrying out a literature review on LHDoperations to gain in-depth knowledge and conducting observations in theworking environment. The relevant information and data on the LHD typeused at Conzal was also acquired.

The major inhibitor was found to be excessively long trammingdistances in all the sections of the mine. The one-way tramming distanceswere all significantly greater than the 90 m set in the mine’s code ofpractice (COP), with the Main Shafts section having the longest averageone-way tramming distance of 260 m. The other inhibitor was LHDutilization, which in 2013 was only 47% against a target of 70%.Simulation of the LHD operation, taking these two factors into account,showed that production could be increased by more than 100%. As a result,it was recommended that conveyor belts should be extended regularly inorder to keep tramming distances within the recommended one-waydistance of 90 m. In addition, utilization can be improved by minimizingemployee absenteeism as well as by modifying the travelling routes suchthat LHDs do not encounter unnecessary delays.

Keywordsunderground transport, tramming, load-haul-dump optimization.

* University of the Witwatersrand.© The Southern African Institute of Mining and

Metallurgy, 2015. ISSN 2225-6253. Paper receivedMar. 2015


1 The actual mine name is omitted owing to theconfidential nature of the information in thisinvestigation.


LHD optimization at an underground chromite mine

� A visit underground to inspect the operatingenvironment of LHDs

� A literature review of LHD operations� Interviews with relevant personnel (mine manager, mine

overseers, engineer, GIT, foremen, artisans, mineplanner, miners and operators) to gain additionalinformation on LHD performance and mining conditions

� Underground observation of LHD operation� Collection of statistical data such as monthly tonnages,

LHD availability and utilization, tramming distances, andbreakdowns

� Data analysis and simulations� Draw conclusions and provide recommendations based

on the data analysis and simulation results.

Literature reviewThe invention of LHDs can be said to be a milestone in mining.The LHD has since become a dominant machine in mechanizedmines and plays an important role in overall mine production(Samanta et al., 2004, p. 1). LHDs are very versatile andpowerful machines, capable of working in the most hostile ofconditions while producing the required tonnages provided that

the mine design suits their usage.The literature review was instrumental in highlighting

common problems in the mining industry that were relevant tothe LHD problem faced by Conzal Mine. Some of these commonfactors are discussed below.

MaintenanceMaintenance is a very critical aspect of the availability andproductivity of LHDs as it can reduce machine-relateddowntime (Inductive Automation, 2011, p. 7). It is imperativethat maintenance is conducted on a regular basis to ensure thatLHDs retain their original performance capability and reducethe rate of wear. To mitigate the effects of machine downtime,Conzal makes use of a maintenance structure that can beclassified into three categories, consisting of preventative,periodic, and breakdown maintenance.

AvailabilityThe availability of a machine influences the ability to tram therequired tons from the face to the grizzly tip. Availability is thetotal up-time of an LHD expressed as a percentage of the totalallotted time for the machine. LHD availability decreases withage, therefore the optimization of availability can defer theneed to spend capital on new machinery when the current fleethas reached its end of life (Machine Downtime, 2014). Fromempirical results, the average availability of an efficient LHDsystem in the mining industry should be around 80% to 85%.

UtilizationThe utilization of a machine can be defined as the time duringwhich the machine is in motion (transmission hours)expressed as a fraction of the engine hours. Conzal measuresutilization in the following manner:

Utiliation=Transmission hours

/Engine hours [1]s

Utilization can depend on many factors, such as thetransport sytem design, availability of panels to be loaded, andpresence of operators, to name a few. The major issue affectingutilization is the transport system design. Poor transportsystem design during mine planning can result in too few ortoo may machines in relation to the number of of producingfaces, resulting in underutilization of some of the LHDs(Mathiso, 2013).

Tramming distances and cycle timeThe tramming distance can be defined as the one-way distanceto the dumping point (as in the case of Conzal Mine) or as thetwo-way distance, which includes the distance travelled back tothe loading point. Tramming distances have a direct influenceon the cycle time of a machine, and long tramming distancesreduce the rate of ore delivery (tons per hour) by the LHDs.This has an overall effect on the LHDs’ ability to meetproduction targets. LHD efficiency generally starts to dropwhen the effective one-way tramming distances are greaterthan about 80 m (Leeuw, 2013). Massive production lossesand inefficiencies can occur once the tramming distanceexceeds this general rule–of-thumb distance.

Observations, results, and analysisTramming distances, cycle times, and other underground andsurface observations were used to further understand workingconditions and obtain data for analysis. Further useful

�


Figure 1 – Location of Conzal Mine

Figure 2 – LHD 307 tramming ore to the grizzly tip

finformation like the monthly tonnages, downtime, andavailability and utilization data was obtained from the relevantmine departments.

MMonthly tonnageThe original hoisted ore target at Conzal Mine was 2 000 t/d,wwhich was later increased to 2 200 t/d when the Broken HillShaft was re-opened for production in April 2013. The tonnagestatistics for 2013 are given in Table I.

From Figure 3, it can be seen that Conzal failed to meettarget for of 2013. Only during the months of May, June, andAugust was the mine able to meet its target. As a result, thevvariance in the tonnage for the entire year amounted to a valueequivalent to one month’s production (31 117 t).

It can be clearly observed in Figure 3 that there was asudden reduction in the production in December 2013. This isbecause the mine produces for only half the month due topublic holidays and the customary Christmas break in SouthAfrica. When considering tonnages in isolation, it isinconclusive whether the main cause of not meeting target isdue to LHDs, as the tonnages produced depend on many otherfactors, such as:� Poor or substandard mining operations in the form of

bad shot-hole drilling and blasting practices� Availability of blasted panels for loading as a result of

poor ground conditions or even a loss of panels

� Stoppages of mining as a result of the Department ofMineral Resources (DMR) issuing Section 54 notices fordangerous or unsafe working conditions, which meansthe loss of several production days while the concernsare addressed.

AvailabilityThe LHD availability was calculated based on the effective shifttime. At Conzal the effective LHD shift time for the day andnight shifts is 7 hours, and 6 hours for the afternoon shift;hence a total of 20 hours per day is available for tramming.Table II shows the average monthly availability of all LHDs atConzal from 2011 to 2013.

Figure 4 shows the availability of LHDs from 2011 to2013. It can be seen that the availability of the machines hasincrease by 19%, from 63% in 2011 to 82% in 2013. Theimprovement was a result of the commissioning of arefurbished fleet in 2012 from Sandvik. The previous machineshad been in continuous operation since 2004. As a result,older machines spent more time being repaired due to the highrate of component failure, which contributed to the low 63%availability in 2011. Although there has been a substantialincrease in availability since then, it is still below the minetarget of 85%.



Table II

Three-year LHD availability statistics for ConzalMine, %

Month 2011 2012 2013

January 78 64 69February 76 73 66March 60 84 76April 50 84 85May 53 67 87June 47 69 88July 41 69 81August 63 66 87September 69 75 86October 53 79 90November 83 64 80December 84 73 83Average 63 72 82

Table I

2013 monthly tonnages for Conzal Mine

Month Actual Target Variance

January 31 660 38 000 -6 340February 32 350 40 000 -7 650March 31 970 38 000 -6 030April 40 490 42 000 -1 510May 42 630 42 000 630June 38 150 38 000 150July 46 770 50 600 -3 830August 39650 46200 -6 550September 41 620 44 000 -2 380October 40 010 50 600 -10 590November 39 755 44 000 -4 245December 17 245 19 800 -2 555Total 442 300 493 200 -31 117

Figure 3 – Monthly tonnages, 2013

Figure 4 – Three-year availability graph for LHDs


UtilizationEffectively, only a maximum of eight LHDs are utilized pershift due to the number of available panels for loading. Theselection of the eight LHDs designated for full-time operationwwas based on the age of the fleet. From the fleet of elevenLHDs, seven have been in operation for more than five yearsand the other four were commissioned in 2012. As a result,three of the older LHDs are used as standby units in case ofbreakdowns. Table III shows utilization values in 2013.

The utilization of LHDs at the mine is a major concern as itis critically low. During 2013 Conzal Mine achieved an averageLHD utilization of 47%, which is far below the mine target of70%.

In order to obtain a better analysis of utilization, data wasacquired from Mbhazo Mine, which utilizes the same miningmethod and machinery as Conzal. Only the months of April toAugust were taken into consideration because of theavailability of data from Mbhazo. It can be seen from Figure 5that the average utilization at Conzal was lower than atMbhazo, despite the former marginally outperforming Mbhazoin July and August. The average utilization for Conzal for theperiod under consideration was 45%, while the average forMbhazo was 58%. However a firm conclusion cannot be drawnfrom this comparison as the data covers only a short period.The situation could be different if data for a longer period, suchas a year or more, could be compared.

Tramming distancesThe tramming distances were measured from the plansprovided by the surveyors. These were measured according tothe sections in each shaft to get a better understanding of theaverage tramming distances. Table IV indicates the one-waytramming distances for the various sections at Conzal Mine.

The average tramming distance for Strike 2 could not bemeasured due to the unavailability of an updated plan at thattime, as well as the mine’s SOP, which restricts personnel fromwalking on designated LHD routes and thus prohibitsunderground physical measurements.

From the data collected one can clearly see that trammingdistances are a serious concern. For LHDs to operate efficientlythe one-way tramming distances should not be greater thanthe COP of 90 m. At Conzal, the average one-way distances forthe sections measured were all significantly greater than 90 m.This resulted in great losses in efficiency and ultimatelyimplied that the mine is not getting its return on investmentfrom these machines because they are unable to tram therequired tonnages to the tips. Figure 6 shows that the MainShafts section had the longest tramming distances, followed byStrike 5 and Broken Hill.

As stated in the literature review, there is a relationshipbetween long tramming distances and the tonnages producedby each section because long distances increase cycle times andconsequently reduce tonnage output. This relationship betweenthese three factors can be seen in Table V, which shows thetramming distances, cycle times, and tonnages produced ineach section in December 2013.

It should be noted that Strike 2 had four working panels,whereas the other sections had eight panels each. This meantthe section was naturally bound to produce less than the othersections.

�


Table IV

Tramming distances of different sections, m

Section Distance

Broken Hill 188Strike 2 -Strike 4 171Strike 5 214Main Shafts 260

Table III

LHD utilization for 2013, %

Month Achieved Benchmark Variance

January 62 70 -8February 54 70 -16March 50 70 -20April 43 70 -27May 43 70 -27June 42 70 -28July 50 70 -20August 47 70 -23September 52 70 -18October 43 70 -27November 35 70 -35December 43 70 -27Average 47 70 -23

Figure 5 – Conzal and Mbhazo utilization comparison graph

Figure 6 –Tramming distances for the different sections

The data in the table shows clearly that trammingdistances had a direct impact on the ability to meet productiontargets. The Main Shafts had the longest tramming distancesand cycle times, which resulted in a lower tonnage than theother sections with an equal number of panels.

Cycle timeSeveral cycle times were recorded by means of a stopwatch overthree days during the afternoon and night shifts across all thesections. The means of the readings are tabulated in Table VI.

One can see the differences in cycle time according todifferent sections in Figure 7. The Main Shafts section has thelongest average cycle times, followed by Broken Hill and thenStrike 4. Lengthy cycle times can be attributed to many factors,

f ffvarying from operator efficiency and attitude to badly designedtramming routes.

Lengthy cycle times have a direct impact on the tonnagestrammed, since fewer cycles will be completed, resulting inpoor production output. From Figure 8, one can see that thereis an inverse relationship between cycle time and productionoutput. As the cycle time increases, the tonnage outputs tendto decrease (apart from Strike 2, which had half the number ofpanels of the other sections).



Table VI –

LHD cycle times at different sections, min

Broken Strike 2 Strike 4 Strike 5 MainHill Shafts

1 11 6 8 7.5 12.52 9.5 5 9.5 7.5 11.53 10 6 8.5 8 10.54 10.5 5.5 8 6.5 12.55 15 5 9 7 126 9 6.5 9 5.5 127 8 6.5 11 7 128 11 6 8 7.5 149 13.5 6 11 7 1210 9 6.5 8 8 13Average 10.7 5.9 9 7.2 12.2

Table V

Tramming distances, cycle times and tonnagesproduced for December 2013

Section Distance (m) Cycle time (min) Tonnages (t)

Broken Hill 188 10.7 2 788Strike 2 - 5.9 1 974Strike 4 171 9 5 135Strike 5 214 7.2 4 277Main Shafts 260 12.2 3 069

Figure 7 – Section LHD cycle time graph

Figure 8 –Effect of cycle time on tonnage output

Figure 9 – Strike 4 footwall conditions


As mentioned, badly designed tramming routes, whichcould manifest in the form of a rolling footwall, have bearingon cycle time. For example it was observed that Strike 4,despite having the shorted tramming distance, experiencedlong cycle times because of several issues. Firstly, the footwallwwas often was rolling and in some parts was unacceptablysteep. Secondly, there was standing water on the footwall,wwhich reduced traction. Figure 9 shows the roadway conditionsat the time of the investigation at Strike 4.

DDowntimeIn December 2013, there were several issues, such as lack ofspares, that led to certain LHDs experiencing lengthydowntimes. Table VII shows the downtime for different LHDsin the month.

From the pie chart (Figure 10), it can be clearly seen thatLHDs 304, 306, 701, and 702 had had the longest downtimesin December 2013. It should be noted that machines 304, 306,and 701 were old and had not being refurbished. Consequentlythese LHDs were used as standby machines, seeing that theywwere already spending a lot of time under repair.

Figure 11 shows LHD 306 undergoing repairs afterbreaking its rear axle. Such breakdowns take approximately awwhole shift to repair and if the spares are not available, it couldtake up to two days to get the machine back in service. For thisparticular machine, the downtime was two full days as themine had to wait for the replacement part from the supplier.

These downtimes have a huge impact on the tonnageshoisted. A loss of 352 t per operational LHD was experiencedfor the entire December period (see Appendix A for thecalculation of this value). This illustrates the need fordowntime to be minimized in order to maximize outputs.

LHD output simulationsLHD simulations using Microsoft Excel® were conducted toquantify possible improvements in production when thetramming distances were reduced to the standard of approxi-mately 90 m and the LHD utilization increased to thebenchmark of 70%. A single month (December) wasconsidered and the simulations were done for all the sectionson the mine. In order to complete these simulations, atriangular distribution was carried out on factors such asoperator efficiency, LHD bucket fill factor, and swell to simulatetheir random nature during loading and hauling operation. The

basic equations that were used to complete the simulation werethe cycle time, LHD payload, and LHD output per month. Thecalculations, as well as input variables, can be found inAppendix B. It should be stressed that the results of thesimulations do not consider other factors such as blastingdelays and interruption from engineering and other servicedepartment. To obtain a realistic potential LHD output subjectto these considerations, more than a thousand randomiterations were done for the two different scenarios:� Scenario 1: the utilization was kept at the then-

prevailing average value of 47% and the trammingdistances were randomized through a triangular distri-bution based on the ideal one-way distance of 90 m

� Scenario 2: the utilization was set to a target value of70% while the tramming distances were randomized asin scenario 1.

Broken HillBroken Hill section achieved only 2 778 t in December 2013.The results of the simulations, tabulated in Table VIII, indicatethat the LHDs at Broken Hill section could achieve up to 7 866 tin the first scenario (approx. 90 m tramming distance and 47%utilization) and up to 12 612 t for the second scenario (approx.90 m tramming distance and 70% utilization). This means thatan increase of up to 179% can be achieved by just reducing thetramming distances to the optimum, and up to 352% byreducing tramming distances and increasing utilization.

�


Table VII

Total LHD downtimes for December, hours

LHD no. Total available time Uptime Downtime

LHD 304 200 137.08 62.92LHD 305 200 188.37 11.63LHD 306 200 144.1 55.9LHD 307 200 182.07 17.93LHD 308 200 167.91 32.09LHD 309 200 165.25 34.75LHD 310 200 184.8 15.2LHD 311 200 191.37 8.63LHD 312 200 181.53 18.47LHD 701 200 150.65 49.35LHD 702 200 134.35 65.65Total 2200 1 827.48 372.52

Figure 10 – LHD downtime pie chart

Figure 11 – LHD 306 undergoing maintenance

Strike 2Strike 2 follows a similar trend to the Broken Hill section.Compared to the actual output of 1 974 t, the simulation resultsindicated that Strike 2 section’s production could be improvedby 358% in scenario 1 and 648% in scenario 2. Table IXshows the summary for this section.

Strike 4This section managed to achieve 5 135 t for the month ofDecember. Based on the simulations, this section has thepotential to improve by 65% and 171% for scenario 1 and 2respectively. One can now see the effect of the trammingdistances, as the increments are not as high as those forBroken Hill or Strike 2, since this section had shorter trammingdistances than the others.

Strike 5This section follows the same trend as the other sections withthe simulations indicating that there would be an improvementin the output tonnages when the two scenarios are applied.The improvements based on Table XI are 83% and 198% forscenarios 1 and 2 respectively.

Main ShaftsThe Main Shafts section performed relatively poorly in that itmanaged to produce only 3 069 t in December 2013. Thesimulated output of the LHDs in this section could be increasedby up to 131% for the first scenario and 276% for the secondscenario. This demonstrates that the extremely long trammingdistances in this section constitute a serious bottleneck in theoverall attainment of the mine’s production target. Such longdistances translate into very low LHD efficiencies.

ConclusionsLHD transportation needs to be designed in such a way that itis efficient and meets the production requirements of a mine.

fConzal’s production for 2013 was 442 283 t, against a target of493 200 t.

Based on the analysis of the tonnages, availability andutilization figures, tramming distances, cycle times,breakdowns, and the LHD simulation results, it was concludedthat the major factors preventing the LHDs from meeting therequired production target were tramming distances andutilization. LHD simulations showed that the productionoutputs of all the sections could be increased considerably bydecreasing tramming distances and increasing LHD utilization.This is because tramming distances have a direct impact on allthe other issues investigated, and if the tramming is reduced tooptimal distances, then all the other issues will mostly likely bealleviated. When the tramming distance is reduced:� The cycle times will be reduced because the LHDs will be

travelling a shorter distance at the same speed, whichcan translate to more ore being trammed in a givenperiod

� Breakdowns are likely to be reduced because themachines will suffer less wear and tear

� The reduction in breakdowns will automatically improveavailability, which will in turn further increase the oretonnage trammed.

RecommendationsBased on the analysis of the results and observations, thefollowing recommendations were made to address the factorsthat resulted in failure to meet production target.� Tramming distances for all the sections must be reduced

to less than 90 m. This can be achieved by extending thebelts on all the sections on a regular basis to ensure thattramming distances are within the optimal distance

� The utilization of the LHDs needs to be improved signifi-cantly. A low utilization value is unacceptable forefficient tramming. Utilization can be improved invarious ways, including reducing absenteeism and



Table VIII

Broken Hill simulation results

Scenario 1 Scenario 2

Tons achieved 2 778Simulated output tons 7 766.30 12 611.55Distribution NormalStandard deviation 442.39 726.96

Table IX

Strike 2 simulation results



Table X




Table XII

Main Shafts simulation results



Table XI





increasing team spirit amongst the employees. Thedesign of the tramming system also needs to be carefullyre-evaluated to address any design errors that could leadto low utilization.

� Maintenance needs to be optimized so that LHDdowntime can be reduced. This can be achieved bymaintaining an adequate stock of spares in the minestore

Cycle time needs to be reduced, and this can be done byimproving roadways and introducing an improvedhousekeeping system to ensure that they are maintained ingood condition.

AAcknowledgementsI would like to offer my special thanks to the following personsand organizations for their assistance with this investigation:� The company referred to as Conzal Mine, for giving me

the opportunity to conduct the investigation and fulfilthe requirements of the project

� My project supervisors, P. Leeuw, D. Pretorius, and B.Mathiso for their generosity and dedication in assistingme to produce this report

� My mother and father, Mr and Mrs Mbhalati, for theirlove and constant support in times of difficulty andstruggle

� The mining and engineering crews at the Mine for theirassistance in acquiring the data and knowledge requiredfor the project.

AAppendix ACalculations of the tons lost per LHD for the month ofDecember 2013:� Total available tramming time for the eight full-time

operational LHDs TtTT = 1600 hours� Total downtime for the eight LHDs TdTT = 204.35 hours� Total average downtime per LHD

� Daily tonnage target =

Effective loading time per day = 20 hours� Required tons per hour:

� Required tons per hour per LHD,

The total tons lost for December due to average downtimeper LHD can be calculated as follows:� Tons lost per LHD for December = Td-aveTT * tLHDtt

= 25.54 h * 13.75 t/h= 351.23 t per LHD

Appendix BTemplate of LHD simulation input variables

LHD output per month:

1. Cycle time (h) =

[2]

2. LHD payload (t) =

∗ Bucket size (m3) ∗ Bucket fill factor [3]

3. LHD output (t/month) =

∗ No. of LHDs ∗ LHD payload ∗ Cycle time ∗ Availability ∗ Utilization ∗ Opperator efficiency [4]

References

INDUCTIVE AUTOMATIONAA . 2011. White Papers on Automation, Process Control &Instrumentation Topics.http://www.automation.com/pdf_articles/Whitepaper-Reduce-Downtime-Raise-OEE.pdf. [Accessed 21 April 2014].

LEEUW. P.K.J.. 2013. Mine Transportation. Notes for Mine Transportation course.University of the Witwatersrand.

MACHINEMM DOWNTIME. 2014. Machine Downtime. http://machine-downtime.com/[Accessed 23 April 2014].

MATHISOMM , B. 2013. Maintenance coordinator [Interview] 9 December 2013.

SAMANCOR CHROMe. 2014a. About Us - Company Overview.http://www.samancorcr.com/content.asp?subID=2 [Accessed 15 April2014].

SAMANCOR CHROME. 2014b. Our Business - Operations and Locations.http://www.samancorcr.com/content.asp?subID=8 [Accessed 15 April2014].

SAMANTA, B., SARKAR, B., and MUKHERJEEMM , S.K. 2004. Reliability modelling andperfomance analyses of an LHD system in mining. Journal of the SouthAfrican Institute of Mining and Metallurgy, vol. 104, no. 1. pp. 1–8. �

�


Effective loading hours per day 14

Number of LHDs 2

Availability 83%

Utilization 70%

LHD bucket size (m3) 2.2

LHD average speed (m/s) 2.2

Two-way tramming distance (m) 177

In-situ density 4.7

Assumptions:

Operator efficiency 79%

Bucket fill factor 94%

Swell 1.49

Loading and dumping time (s) 100

Additional time 60

IntroductionSouth Africa is a mineral-rich country withmetals such as gold, copper, and platinumgroup metals being exploited to a significantextent in the country’s mining history. Mininggenerates large volumes of tailings, withconsequent disposal and environmentalproblems. By far the most gold that has beenmined in South Africa (98%) has come fromthe Witwatersrand goldfields (Messner, 1991).The gold mines in this area are situatedaround an ancient sea (over 2700 millionyears old) where rivers deposited sediments inthe form of sand and gravel that became theconglomerate containing the gold (Messner,1991). The extensive exploitation of the goldresources has led to numerous mine tailingsheaps scattered around the WitwatersrandBasin. As long as mining contributes signifi-cantly to the economic development of SouthAfrica, generation of these tailings isinevitable.

The major environmental impacts fromwaste disposal at mine sites can be dividedinto two categories – the loss of productiveland following its conversion to a wastestorage area and the introduction of sediment,acidity, and other contaminants intosurrounding surface and groundwater (MiningFacts, 2014). The gold mining and processingwastes contain large amounts of sulphideminerals such as pyrite, which generate acidmine drainage (AMD) (Rosner and vanSchalkwyk, 2000). South Africa is currentlyfaced with the challenges resulting from AMDand the government and mining companies areunder pressure to find viable solutions to thisproblem. This, coupled with the increasinglandfill costs, and stricter implementation andenforcement of environmental legislation, hascaused the scientific community to focus onfinding innovative methods of utilizing minetailings. Even though some applications of thegenerated tailings have been exploited, such asin the building of slimes dams and backfill inunderground mines, these uses do not take upmore than a fraction of the total amount oftailings in the Witwatersrand region. There istherefore a significant need to developing otherlong-term, commercially viable uses for minetailings in order to minimize the disposal costsand the impact on the environment.

According to Statistics South Africa(2013), South Africa has a human populationof about 52.98 million. This population isgrowing, and this consequently results in anincreasing demand for housing, which placessevere stress on the natural resources used forconstruction materials. Conventional bricks areproduced from clay fired in high-temperaturekilns or from ordinary Portland cement (OPC)concrete. Clay, the common material used for

The viability of using the Witwatersrandgold mine tailings for brickmakingby M. Malatse* and S. Ndlovu*Paper written on project work carried out in partial fulfilment of BSc. Eng. (Metallurgyand Materials Sciences)

SynopsisThe Witwatersrand Basin is the heart of South Africa’a gold miningindustry. The cluster of gold mines located in the Witwatersrand Basingenerates a significant amount of mine tailings, which have adverse effectson the environment and ecological systems. In addition, disposal costs arevery high. The exponential population growth in the Witwatersrand areahas resulted in pressure on the reserves of traditional building materials.Quarrying for natural construction material is very expensive and damagesthe landscape. This work therefore examines the use of gold mine tailingsin the production of bricks.

Different mixing ratios of gold tailings, cement, and water were used.The resulting bricks were then cured in three different environments – sundried, oven dried at 360°C, and cured in water for 24 hours. The bricks werethen tested for unconfined compressive strength, water absorption, andweight loss. The results showed that the mixture with more cement thantailings had a compressive strength of approximately 530 kN/m2. It wasalso found that the best brick curing system was in a water environment.Bricks made from tailings cost more than conventional bricks because ofthe higher quantity of cement used, but the manufacturing processconsumes less water. Overall, the results indicated that gold mine tailingshave a high potential to substitute for the natural materials currently usedin brickmaking.

Keywordsgold mine tailings, construction materials, brickmaking.

* School of Chemical and Metallurgical Engineering,University of the Witwatersrand, Johannesburg,South Africa.

© The Southern African Institute of Mining andMetallurgy, 2015. ISSN 2225-6253. Paper receivedFeb. 2015



The viability of using the Witwatersrand gold mine tailings for brickmaking

brickmaking, is usually mined in quarries. Quarryingoperations are energy-intensive, adversely affect thelandscape, and generate a high level of waste (Zhang, 2013;Bennet et al., 2013). Furthermore, in many areas of thewworld, there is already a shortage of natural resource materialfor the production of the conventional bricks (Zhang, 2013).To conserve the clay resources and the environment, somecountries such as China have started to limit the use of bricksmade from clay (Zhang, 2013). Thus the depletion of thesenatural resources has created a need need for an alternativesource of construction materials in order to sustaindevelopment.

Extensive research has been conducted on the productionof bricks using waste material (Zhang 2013; Saeed andZhang, 2012). These waste materials include mining waste,construction and demolition waste, wood sawdust, cottonwwaste, limestone powder, paper production residues,petroleum effluent treatment plant sludge, kraft pulpproduction residue, cigarette butts, waste tea, rice husk ash,crumb rubber, cement kiln dust, and coal fly ash (Zhang2013; Bennet et al., 2013; Saeed and Zhang, 2012). Themining and mineral processing waste includes miningoverburden, waste rock, mine tailings, slags, granulated blastfurnace slag (GGBS), mine water, water treatment sludge,and gaseous waste ( Zhang, 2013; Saeed and Zhang, 2012;Koumal, 1994; Dean et al., 1968; Bennet et al., 2013).

The extensive research on the utilization of wastematerials to produce bricks can be divided into three generalcategories based on the production methods – firing,cementing, and geopolymerization,

Production of bricks from waste materials through firinguses waste material(s) to substitute partially or entirely forclay and follows the traditional method of kiln-firing. Chen etal. (2011) studied the feasibility of utilizing haematitetailings and class F fly ash together with clay to producebricks. Tests were performed to determine the compressivestrength, water absorption, and bulk density of brick samplesprepared under different conditions.

Bennet et al. (2013) conducted research on thedevelopment of geopolymer binder-based bricks using fly ashand bottom ash. During the synthesizing process, silicon-aluminium bonds are formed that are chemically andstructurally comparable to those binding the natural rocks(Bennet et al., 2013), giving geopolymer binder-based bricksadvantages such as rapid strength gain and good durability,especially in acidic environments. Research into geopolymerbricks has also incorporated copper mine tailings and cement

kiln dust (Bennet et al., 2013). In this process, an autoclavedaerated cement (AAC) material is produced (Koumal, 1994).Ahmari and Zhang (2012) investigated the utilization ofcopper mine tailings to produce geopolymer bricks by usingsodium hydroxide (NaOH) solution as the alkali activator.They produced cylindrical brick specimens by using different initial water contents, NaOH concentrations, forming pressures, and curing temperatures. Copper minetailings bricks have been found to have good physical andmechanical properties such as a water absorption of 17.7%, compressive strength of 260 kg/cm2, and density of1.8 g/cm3 (Be Sharp, 2012).

The method of producing bricks from waste materialsthrough cementing is based on hydration reactions similar tothose in OPC to form mainly C–S–H and C–A–S–H phasescontributing to strength (Zhang, 2013). The cementingmaterial can be the waste material itself or other addedcementing material(s) such as OPC and lime. Again, manyresearchers have studied the utilization of waste materials toproduce bricks based on cementing. The brickmaking processhas involved the use of waste and tailings such as those fromcopper, nickel, gold, aluminium, molybdenum, and zincprocessing as additives replacing some of the cement (Jain etal., 1983). Morchhale et al. (2006) studied the production ofbricks by mixing copper mine tailings with different amountof OPC and then compressing the mixture in a mould. Theresults showed that the bricks had a higher compressivestrength and lower water absorption when the OPC contentincreased. Roy et al. (2007) used gold mill tailings mixedwith OPC, black cotton soils, and red soils in differentproportions to make bricks. The cement-tailings bricks werecured by immersing them in water for different periods oftime and their compressive strengths were determined. Brickswith 20% cement and 14 days of curing were found to besuitable. Gold mine tailings have also been used to produceautoclaved calcium silicate bricks (Jain et al., 1983). Thebricks are cured under saturated steam and in the process,lime reacts with silica grains to form a cementing materialconsisting of calcium silicate hydrate. Some miningcompanies such as Bharat Gold Mines in India have exploredthe idea of brickmaking using gold ore tailings (Be Sharp,2012).

Table I shows the chemical composition of some of thewaste materials used in bricks as well as the composition ofquarry clay material that comprise the conventional feedmaterial (Bennet et al., 2013). The gold mine tailings arefrom a Chinese mine (Yang et al., 2011).

�


Table I

Composition of material used in brickmaking (Bennet et al., 2013; Yang et al., 2011)

Oxide component Fly ash GGBS Bottom ash Clay material Gold mine tailingsMass % Mass % Mass % Mass % Mass %

SiO2 53.3 35.47 56.76 61.8 38.60Al2O3 29.5 19.36 21.34 25 7.06Fe2O3 10.7 - 5.98 8 12.76CaO 7.6 33.25 2.88 - 29.24SO3 1.8 - 0.72 - 3.21FeO - 0.8 - - =-MgO - 8.69 - 1.2 7.85Na2O - - - 0.1 -K2O - - - 2.76 -

From the chemical compositions shown in Table I, it canbe seen that the waste materials have similar major oxides intheir compositions. The compositions are also relativelysimilar to the typical clay material used in brickmaking. Thewwaste materials all have a predominantly high content ofsilica, alumina, and haematite (with the exception of thegranulated blast furnace slag, GGBS, which contains nohaematite), which are important in brickmaking materials.Considering the source of the gold deposits in theWWitwatersrand Basin (river sediments in the form of sandand gravel), it is therefore likely that the tailings from thisarea will also contain a high level of silica.

The purpose of this work is therefore is to ascertain thetechnical and economic viability of using the Witwatersrandgold tailings for brickmaking using the cementing method.The tailings-based bricks will be compared with thecommercial bricks available on the market. The evaluationwwill be based on parameters such as compressive strength,wwater absorption, and weight loss tests. This work has thepotential to unlock large resources of material needed in theconstruction industry that would help conserve the naturalresources commonly used. In addition it would eliminate theland requirements for waste disposal, thus realizing savingson disposal and landfill costs and also lessening environ-mental damage. But above all, this work has the potential toprovide an additional revenue stream for the gold miningsector.

Materials and methodsThe materials used in this test work were gold mine tailings,wwater, and cement as a binding material. Gold mine tailingswwere provided by a local gold mining company, AngloGoldAshanti. The Larfarge 42.5 kN cement was provided by thePlanning, Infrastructure and Maintenance Department at theUniversity of the Witwatersrand, Johannesburg. The cementwwas used on the day of delivery and tap water was used inthe mixing process.

Characterization of gold mine tailingsRepresentative samples used in all experiments wereprepared using a riffle splitter (model 15A, Eriez Magnetics,South Africa). The gold tailings were characterized byinvestigating the phase mineralogy, particle size, and quanti-tative chemical analysis. The particle size analysis was doneby physically screening the samples using test sieves(Fritsch, Germany) of various screen sizes up to 212 μm. Thephase mineralogy analysis was carried out using an X-raydiffractometer (X’Pert, PANalytical, Netherlands) operatedwwith Co-K radiation generated at 40 kV and 50 mA. Thechemical analysis was carried out using wavelengthdispersive X-ray fluorescence (XRF) spectrometry (Axios,PANalytical, Netherlands) operated with a rhodium tubeexcitation source.

The brickmaking processDifferent mixing ratios of tailings, cement, and water wereused in the brickmaking process (Table II). From eachmixture, a number of bricks were cast and dried.

The three feed material (tailings, cement, and water) weremixed in the appropriate ratios in a commercial mixing

fmachine. Dry mixing was done first and then a controlledamount of water was added while continuing to mixthoroughly. The total mixing time was 15 minutes. Themixture was then cast into the brick moulds. The brickmoulds were then placed on a vibrating machine for 5minutes in order to fill the voids in mixture comprehensivelyand thus prevent the formation of air pockets. The brickswere then labelled and allowed to cure for 24 hours. Threecuring methods were used. These included atmosphericdrying under the sun, curing in water, and drying in an ovenat 360°C. After curing, the bricks were de-moulded using anair compressor, weighed, and tested for compressive strength.

Unconfined compressive strength testingThe cast and cured bricks were tested for compressivestrength using a Tinus Olsen compressive strength testingmachine. In the compressive strength testing process, a forcewas applied on the brick until the brick failed and the forcemeasured at failure was documented. The compressivestrengths obtained were then averaged. The mixture ratiothat gave the highest compressive strength was subsequentlyemployed to manufacture bricks for water absorption, weightloss, and leaching rate tests. Unconfined compressive testswere also done on commercial bricks to provide a basis forcomparison.

Water absorption rate and weight loss testsTwo solutions with different pH values, one acidic and oneneutral, were used for these tests. The tailings bricks werefirst prepared from mixture 7 (Table II) and cured in water for24 hours. Tests were conducted on four samples in eachsolution. The bricks were immersed in water baths, onecontaining water at pH 7 and the other an acidic solution atpH 4. The solid–to-liquid ratio was maintained at 15. Thesaturated weight of the bricks (WsW ) was measured every 24hours over a 5-day period. After 5 days, the bricks were driedat 110°C for 24 hours and the oven-dried weight (WdWW )recorded. The bricks were again tested for compressivestrength. The percentage water absorption rate was thencalculated as

Water absorption (%) = [(WsW - WdWW ]/WdWW × 100

The weight loss tests were done in the neutralenvironment only (pH 7). The average weight loss wasmeasured after the bricks had been soaked in neutral waterfor seven days then dried overnight at 110°C.



Table II

Different mixtures used in brickmaking

Mixture number Tailings (kg) Cement (kg) Water (L)

1 2 1 0.62 14 2 2.653 9 6 3.04 7 8 2.55 10 5 2.56 12 3 2.57 5 10 3.38 10 5 3


Results and discussion

PParticle size distribution Figure 1 shows the particle size distribution of the materialused in the brickmaking process. The results are presented incumulative form, in which the total amount of all sizesretained or passed by a single notional sieve is given for therange of sizes.

The results indicate that most of the particles fell into90–200 μm range. 80% of the material passed the 200 μmscreen aperture while about 12% passed the 90 μm screen.The particle size range used in standard commercialbrickmaking includes coarser sand particles as well as fineparticles. The material used in these tests was, in comparison,relatively fine. A cost analysis study done by Roy et al.(2007) showed that cement-tailings bricks are generallyuneconomical compared to the soil-tailings based bricks,therefore future test work will have to consider the additionof coarse particles, possibly from mining overburden.

MMineralogical and chemical analysisTable III shows the mineral phases and the respectivequantities present in the sample as determined by XRD andXRF analysis. The table indicates that the mineralogical andchemical composition of the tailings bear close similaritieswwith the composition of the conventional materials used forcommercial brickmaking, as well as with the waste materialsthat have been tested in the past (see Table I). The resultsindicate that the major oxides in the mine tailings sample aresilica, magnesium oxide, alumina, sulphur trioxide,potassium oxide, calcium oxide, and haematite. The otherconstituents such as uranium oxide are found in tracequantities. Although uranium oxide is present only at0.0064% its presence is worth noting as uranium is a veryradioactive element and therefore can present safetyimplications.

Unconfined compressive strengthThe main mechanical property of bricks that is tested for iscompressive strength. A good brick should be hard andstrong. The compressive strength tests on commercial brickswwere undertaken in order to provide a basis for comparisonwwith the gold mine tailings bricks. Table IV shows the resultsof the compressive strength of the commercial bricks. It wasnoted during the tests that the more uneven and rough thesurface of the brick, the quicker it failed.

f fUnconfined compressive strength of the gold tailingsbricksThe quality and durability of the concrete mix depend notonly on the quality and properties of the ingredients, but alsoon the method of preparation and the curing environment(Ahmad and Saiful Amin, 1998). Proper curing isindispensable in developing optimum properties. Table Vshows the compressive strength for the gold tailings basedbricks cured in different environments.

The average values shown in Table V are depictedgraphically in Figure 2. For mixture 1, high-temperaturedrying in an oven yielded the highest compressive strength.For mixture 2, ambient drying conditions resulted in thehighest compressive strength, followed by oven drying formixture 3, curing in water for mixtures 4 and 5, oven dryingfor mixture 6, and curing in water for mixtures 7 and 8. Theoverall trend reveals that the majority of the mixtures yieldedhigher compressive strength when cured in water (50%),followed by oven drying (37.5%), and lastly drying underambient conditions (12.5%). This can be attributed to the factthat curing the bricks in water contributes to the cementationprocess and hence increases the strength of the bricks. An

�


Figure 1—Particle size distribution of the gold mine tailings

Table III

Major constituents of the gold mine tailings

Number Component Result (%)

1 Na20 0.6132 MgO 1.793 Al2O3 10.24 SiO2 77.75 P2O3 0.0856 SO3 0.9057 K2O 1.198 CaO 1.939 TiO2 0.46910 Cr2O3 0.4511 MnO 0.054912 Fe2O3 4.5113 Co2O3 0.006314 NiO 0.017715 CuO 0.00716 ZnO 0.00817 As2O3 0.0118 Pb2O 0.004119 SrO 0.015120 ZrO2 0.031221 U3O8 0.0064

Table IV

Compressive strength of commercial bricks

Brick Force (kN/m2)Flat face

1 8902 9203 6654 6955 6906 641

adequate supply of moisture is necessary to ensure sufficienthydration for reducing the porosity to such a level that thedesired strength and durability are attained

The results also show that in general, bricks from mixture7 had a higher compressive strength in all three curingsystem used. However, the highest overall compressivestrength was obtained from mixture 7 that was cured inwwater. This mixture had a higher amount of cement comparedto the tailings (2:1 cement to tailings mass ratio), whichresulted in a larger surface area of the tailings being incontact with the cement and hence resulting in a strongermixture. These results also follow for mixtures 3 and 4. Thehigher strength is probably due to the superior plasticity andbinding properties provided by the higher amount of cement.It is also known that cement cures well in water (America'sCement Manufacturers, 2014); hence the mixture with thelargest quantity of cement cured in water resulted in thehighest compressive strength.

Water absorption and weight loss testsCompressive strength and water absorption are two commonparameters considered by most building materials researchersas required by various standards. Water absorption willinfluence the durability and strength of the bricks. Figure 3shows the water absorption rate.

For both solutions, the absorption was highest on thefirst day of the test followed by a more constant rate insubsequent days. It can also be seen from Figure 3 that the

absorption rate was slightly higher in the neutral solutionthan in the acidic solution. The unconfined compressivestrengths after water absorption are shown in Table VI.

The results show that the bricks soaked in the neutralenvironment had a higher compressive strength than thosesoaked in an acidic environment. This can be attributed tothe fact that during the water absorption test, the neutralsolution acts as a natural curing agent and furtherstrengthens the bricks.

The weight loss over the seven day period was quitenegligible at 0.06%. This means that although the bricksshow significant water absorption rate, they regain theiroriginal weight after drying.

Cost analysisIt is important to check if the outcome of the research projectis economically viable for it to be beneficial to society. Inorder to market the bricks, cost comparison with traditionalbricks is essential. The following factors were considered.

Gold tailings are available in abundance and areexpected to be free of cost

Portland cement=R65 per 50 kg bag (OLX, 2014).

Using a base figure, for commercial brickmaking, themasonry cement recipe can be estimated as follows:

8 bags of cement=1000 bricks (Kreh, 2003), or 1 bag ofcement=125 bricks.

For commercial brickmaking, the water addition shouldbe 20 litres per 50 kg of cement (Hydraform, 2014). Theprice of water for industrial companies according to the Cityof Johannesburg’s Mayoral Committee is R20.96 per kilolitre.(COJ: Mayoral Committee, 2013).



Table V

Average compressive strength of bricks cured underdifferent

Mixture Average compressive strength (kN/m2)

Water Oven Ambient

1 141 165 1572 20 25 293 325 359 3184 440 439 3235 262 261 2346 215 235 2307 530 479 4548 149 98 127

Figure 2—Compressive strength of the cement tailings bricks cured indifferent environments

Figure 3—Water absorption rate

Table VI

Average compressive strength after absorption tests

Solution Compressive strength (kN/m2)

pH 4 445pH 7 476


fIn this research project, it was found that the higheststrength was obtained in mixture 7, with 5 kg of tailings and10 kg of cement mixed with 3.3 litres of water, followed bymixture 4 with 7 kg of tailings and 8 kg of cement mixedwwith 2.5 litres of water. Using the option with the higheststrength, it was found that one bag of cement is equivalent to55 bricks (compared with 125 commercial bricks per bag ofcement). However, the water consumption was calculated tobe 16.5 litres per bag of cement, which is less than the 20litres used in commercial brickmaking. Water is an expensivecommodity in South Africa and using tailings to make thebricks saves water. Thus, the more economical option wouldbe the second mixing ratio, since it uses less water andcement but still results in relatively high brick compressivestrengths. Even though the second option is economicallyacceptable, the high cement content is a disadvantage.However, regarding the overall brickmaking process someother factors should be considered. The brickmaking plantcan be close to the tailings dumps in order to cut down oncosts. In addition, it is important to note that most of thetailings material already occurs in fine form, therefore notmuch size reduction (which is an energy-intensive process) isrequired.

Since the use of tailings for brickmaking conservesnatural resources, one could say that the benefit to theenvironment outweighs mere economic considerations. Theuse of tailings would mean that the companies have to spendless on waste management, while at the same time reducinghuman exposure to tailings, consequently reducing the effectthat mine waste has on the health of inhabitants in themining area. The use of gold mine tailings for brickmakingalso constitutes an additional source of revenue for the goldmining companies and in the process creates jobs.

Conclusions and recommendationsThis laboratory-scale study was aimed at utilizingWWitwatersrand gold mine tailings in making bricks. Theresults from XRD and XRF showed that the chemicalcomposition of the Witwatersrand gold mine tailings issimilar to that of the clay material used for commercialbrickmaking. It was then concluded that it would betechnically viable to use the tailings for brickmaking.Following the South African masonry standards forbrickmaking and testing, it was found that the commercialbricks have an average compressive strength of 750 kN andthat the strongest bricks made from the tailings gave anaverage compressive strength of 530 kN.

Results from water absorption tests showed that waterabsorption is higher in neutral solutions compared to acidicsolutions. The rate of absorption is high in the first day, butthen stabilizes. The weight loss over a seven-day period wasnegligible at 0.06%.

It is recommended that more tests be conducted with awwider range of tailings to cement ratios as this might lead toidentifying a ratio that yields a stronger brick than what hasbeen observed in this project. In addition, the sizes of thetailings used as aggregate should be varied to a wider range.This can be achieved by adding overburden to the finetailings material.

As regards the economic considerations, the tailingsbricks were found to utilize more cement than the commercial

fbricks, possibly due to lack of plasticity in the tailingmaterials used. This is a disadvantage since cement isexpensive. It is thus recommended that cheaper alternativeadditives that have a high plasticity or binding properties beexplored in the place of cement. Looking at the bigger picture,the use of tailings as brickmaking material would have greatadvantages in terms of environmental conservation andreduction of waste management costs.

Since the XRD analysis showed that uranium is presentin Witwatersrand gold tailings, extensive research withregard to the chemical properties and the chemical stability ofthe bricks produced from gold mill tailings is required.

Acknowledgments The authors would like to acknowledge the School ofChemical and Metallurgical Engineering, University of theWitwatersrand, Johannesburg for granting the first author theopportunity to complete her Bachelor’s degree in Metallurgicaland Materials Engineering. In addition they would like toacknowledge all the laboratory personnel in the School whoprovided unlimited support during the research work.Further, the authors would like to acknowledge the team ofacademics and laboratory personnel at the School of CivilEngineering for their guidance on masonry standards and forproviding laboratory space and equipment during the project.Lastly, much appreciation is due to AngloGold Ashanti for theidea behind the project and for supplying the tailings used inthis study.

ReferencesAHMAD, S. and SAIFUL AMIN, A.F.M. 1998. Effect of curing conditions on the

compressive strength of brick aggregate concrete. Journal of CivilEngineering, vol. CE 28, no. 1. pp. 37–49.gg

AHMARI, S. and ZHANG, L. 2012. Production of eco-friendly bricks from coppermine tailings through geopolymerization. Construction and BuildingMaterials, vol. 29. pp. 323–331.

ALGIN, H.M . and TURGUT, P. 2008. Cotton and limestone powder wastes as brickmaterial. Construction and Building Materials, vol. 22, no. 6. pp. 1074–1080.

AMERICA'S CEMENT MANUFACTURERS. 2014. Curing. http://www.cement.org/cement-concrete-basics/working-with-concrete/curing [Accessed 2October 2014].

BE SHARP. 2012. Blocks from mine waste and industrial waste.http://besharp.achider.ord/IMG/pdf/blocks_from_industrial_waste.pdf[Accessed 3 April 2014].

BENNET, J.M., SUDHAKAR, M., and NATARAJAN, C. 2013. Development of coal ash -GGBS based geopolymer bricks. European International Journal of Scienceand Technology, vol. 2, no. 5. pp. 133–139.

CHEN, Y., ZHANG, Y., CHEN, T., ZHAO, Y., and BAO, S. 2011. Preparation of eco-friendly construction bricks from hematite tailings. Construction andBuilding Materials, vol. 25. pp. 2107–2111.

CHOU, M.I., CHOU, S.F., PATEL, V., PICKERING, M.D., and STUCKI, J.W. 2006.Manufacturing fired bricks with class F fly ash from Illinois basin coals.Combustion Byproduct Recycling Consortium, Project Number 02-CBRC-M12, Final Report; 2006.

CITY OF JOHANNESBURG (COJ). Mayoral Committee, 2013. 63 Amendment of tariffcharges for water services and sewage and sanitation services:2013/14.Johannesburg.

DAILY TELEGRAPH. 2014. The brick directory words and blog.http://www.brickdirectory.co.uk/html/brickdirectorywords.html [Accessed11 October 2014].

DEAN, K.C., FROISLAND, L.J., and SHIRTS, M.B. 1968. Utilisation and stabilisationof mineral wastes. Bulletin 688. US Bureau of Mines.

ELJHD ADMIN. 2011. An action plan for acid mine drainage.http://www.earthlife.org.za/2011/04/an-action-plan-for-acid-mine-drainage [Accessed 10 October 2014].

�


ENVIRONMENT f. 2011. Acid mine drainage South Africa.http://www.environment.co.za/poisoning-carcinogens-heavy-metals-mining/acid-mine-drainage-amd-south-africa.html [Accessed 10 October2014].

FAST ONLINE. 1995. Global overview of construction technology trends: Energyefficiency in construction. http://www.fastonline.org/CD3WD_40/CD3W/CONSTRUC/H1661E/EN/B856_6.HTM [Accessed 10 October 2014].

UNIVERSITY OF HIROSHIMA. 2001. Geology and non-fuel mineral deposits of Africaand the Middle East. http://www.home.hiroshima-u.ac.jp/er/Rmin_kS_01_A%26ME.html [Accessed 10 October 2014].

HYDRAFORMHH . 2014. Brick making. http://www.hydraform.com/PDFs/vibraform/V2-LE.pdf [Accessed 19 October 2014].

JJAIN, S.K., GARG, S.P., and RAIRR , M. 1983. Autoclaved calcium silicate bricks fromgold mine tailings. Research and industry, vol. 28. pp. 170–172.

JJONES, M.Q. 2003. Thermal properties of stratified rocks from Witwatersrandgold mining areas. Journal of the South African Institute of Mining andMetallurgy, vol. 103, no. 3. pp. 173–186.

KOUMAL, G. 1994. Method of environmental clean-up and producing buildingmaterial using Cu mine tailing waste material. US patent US5286427 A.

KREHKK , R.T. 2003. Masonry Skills. 5th edn. Delmar Learning, USA.

LIU, Z., CHEN, Q., XIE, X., XUEXX , G., DU, F., and NING, Q. 2011. Utilization of thesludge derived from dyestuff-making wastewater coagulation for unfiredbricks. Construction and Building Materials, vol. 25, no. 4.pp. 1699–1706.

MANOHARAN, C., SUTHARSAN, P., and DHANAPANDIAN, S. 2012. Characteristics ofsome clay materials from Tamilnadu, India, and their possible ceramicuses. Ceramica, vol. 58. pp. 412–418.

MATIVON BRICKS. 2012. Brick depot and building supplier.http://www.mativonbricks.co.za/products [Accessed 19 October 2014].

MESSNER, M. 1991. Witwatesrand Basin (RSA). http://www.edu.uni-klu.ac.at/mmessner/sites.rsa/wits.htm [Accessed 18 October 2014].

MINING FACTS. 2014. How are waste materials managed at mine sites.http://www.miningfacts.org/environment/How-are-waste-materials-managed-at-mine-sites [Accessed 25 September 2014].

MORCHHALE, R.K., RAMAKRISHNANRR , N., and DINDORKAR, N. 2006. Utilization ofcopper mine tailings in production of bricks. Journal of the Institute ofEngineers, Indian Civil Engineering Division, vol. 87. pp. 13–16.

MUELLER, H., MAITHY, S., and PRAJAPUTI, S. 2008. Greenbrick making manual.Hillside Press, Nepal.

OLX. 2014. Cement South Africa. http://www.olx.co.za/q/cement/c-910[Accessed 19 October 2014].

ROSNER, T. and VAN SCHALKWYK, A. 2000. The environmental impact of goldmine tailings foot prints in the Johannesburg region, South Africa. Bulletinof Engineering, Geology and Environmetal Studies, vol. 59. pp. 137–148.

ROY, S., ADHIKARI, G.R., and GUPTA, R.N. 2007. Use of gold mill tailings inmaking bricks: a feasibility study. Waste Management Research, vol. 25.pp. 475–482

SAEED, A. and ZHANG , L. 2012. Production of ecofriendly bricks from Cu minetailings through geopolymerization. Construction and Building materials,vol. 29. pp. 323–331.

STATISTICS SA. 2013. South African statistics 2013. Johannesburg. 2013.VERMEULENVV , N.J. 2001. The Composition and State of Gold Tailings. University

of Pretoria ETD. Retrieved 12 April 2014. Witwatersrand gold orecomposition. http://upetd.up.ac.za/thesis/submitted/etd-03102006-122937/unrestricted/02chapter2.pdf [Accessed 12 April 2014].

WENDEL, G. 1998. Radioactivity in mines and mine watere - sources andmechanisms. Journal of the South African Institute of Mining andMetallurgy, vol. 98, no. 2. pp. 87–92.

YANGYY , Y., ZHU, S., LI, Q., YANGYY , B., and CHEN, Y. 2011. Research on making firedbricks with gold tailings. International Conference on ComputerDistributed Control and Intelligent Environmental Monitoring, QingdaoggShandong, China.

ZHANG, L. 2013. Production of bricks from waste materials - a review.Construction and Building Materials, vol. 47. pp. 643–655. �



SEW-EURODRIVE offers innovative drive solutions for all applications in the mining industries.

All SEW-EURODRIVE products and systems make the best use of the space available around the machine and ensure great flexibility and

reliability. Minimum maintenance and simple operation ensure that you will operate machines and equipment efficiently from the very

beginning.

Thanks to the modular design and countless combination options, all drive engineering components can be replaced quickly, if the need

arises. From gravel mining to the excavation of gold, platinum, coal and diamonds - we put the drive into all facets of the mining industry.

SEW-EURODRIVE - Driving the world.

Tel: +27 11 248-7000 Web: www.sew.co.za

We drive the mining industry

Geared Motors \ Drive Electronics \ Drive Automation \ Industrial Gears \ Services

1THREAD_4139_SAI

Introduction

The need for coal briquettingThe coal mining industry in South Africa hasbeen operating for more than 120 years.Annually, 224 Mt of coal is produced, of which25% is exported. The remainder is used to feedSouth Africa's industry, with 53% used forgenerating electricity (Eskom, 2014). Electricitygenerated from coal-fired power stationsaccounts for 77% of South Africa's electricitysupply.

Fines generation in coal mining hasincreased as a result of mechanized mining, andup to 6% of the run of mine product can be inthe -200 µm fraction. Coal-fired power stations

in South Africa do not accept a product of -200µm size because of the high moisturecontent(England, 2013). In this paper, any -8mm material will be considered as fines.Additional problems arising from coal finesgeneration include flow problems fromcontainers, dust formation in plants and firehazards during stockpiling.

An increased moisture content inevitablyreduces the calorific value of the coal, as well asincreasing handling problems. Instead ofpumping these fines to slime dams ordiscarding them in old workings, a means ofeconomical agglomeration can be beneficial tomines and power stations.

ESI Africa (2014) estimates the amount ofthermal-grade fines stockpiled over the past 100years at about 1 billion tons. Only in the lastfew years have methods to utilize thesestockpiles in South Africa started to beexplored. These methods include briquetting,pelletizing, and granulation.

Briquetting is a pressure agglomerationmethod where loose material is compacted intoa dense mass (FEECO International, 2014). It isa more advanced and more expensive processthan pelletization and granulation, but the end-product can withstand the rigorous handlingmethods that export coal undergoes, and insome cases, is water-resistant.

Table I shows that, assuming that Eskomwill pay a similar price per ton of briquettes tothat of coarse coal, it is not financially viable toproduce briquettes using a generous amount ofbinder. Producing briquettes with a modestamount of binder look more promising.According to Sastry et al., binder cost mayrepresent 60% of the total cost of briquettemanufacture. However, since low binderadditions can be detrimental to the mechanicalstrength of the briquettes, a compromise mustbe reached between binder content andmechanical strength.

Evaluation of some optimum moistureand binder conditions for coal finesbriquettingby P. Venter* and N. Naude*Paper wwritten oon ffinal yyear pproject wwork ccarried oout iin ppartial ffulfilment oof BB.Eng ddegree iinCoal BBeneficiation

SynopsisCoal mining is a thriving industry and 53% of the coal mined in SouthAfrica is used for electricity generation. Mechanization has made coalmining more efficient, but fines generation has subsequently increased. Upto 6% of the run of mine material can report to the -200 μm fraction.Common problems associated with fines handling include dust formation,storage problems, and high moisture levels. A method to turn this materialinto a saleable product instead of stockpiling it can add value to acompany.

Briquetting is a pressure agglomeration method where loose material iscompacted into a dense mass (FEECO International, 2014). The briquettesmust be able to withstand rigorous handling and transport operationswithout disintegrating. This study aims to investigate the optimum binderand moisture conditions required to produce a mechanically strongbriquette using two different binders – a PVA powder (binder A) and astarch powder (binder B).

It was found that for binder A the optimum moisture level was 12% to14%. At this moisture level the greatest compression strength gains wereobserved, and low amounts of fines produced in impact and abrasion tests.The minimum amount of binder added while still obtaining a strongbriquette was 0.5% binder A. For binder B the optimum moisture level wasalso 12% and the minimum amount of Binder B to be added was found tobe 1%. Briquettes that were dried outside reached their peak strength afterabout four days, whereas the briquettes that dried inside took about 20days to reach their strength plateau. Hardly any degradation took place onthe surface of the binder A film after exposure of 300 hours of artificialweathering. Thermogravimetric analysis confirmed that neither binder Anor binder B will add to the ash content of the coal fines, as both binderstotally decompose above 530°C.

Binder B yielded stronger briquettes after 15 days and also generatedless fines. It is therefore superior to binder A and would be recommendedfor further use.

Keywordscoal fines, briquetting, binder, moisture level.


Metallurgy, 2015. ISSN 2225-6253. Paper receivedJan. 2015



Evaluation of some optimum moisture and binder conditions for coal fines briquetting

A review of previous work done on briquettingThe earliest coal briquettes were made in hand-filled brickmoulds using clay and cow dung as binders. These bricks hadpoor mechanical properties which made them unsuitable fortransportation over long distances. Only by the 1850s weremechanical methods introduced to briquette brown coal andlignites without the use of binders, while hard coal briquettesrequired binders to stay intact.

Roller presses were first developed in Belgium by Louiseauto address the need for strong briquettes. Pillow-shaped holesin the roller faces compact material into dense briquettes thatweighs no more than 50 g each. The basic principles of thesemachines have remained relatively unchanged over the yearsapart from small improvements to extend the life and reducemaintenance.

The briquetting process is conducted at room temperatureand the use of binders depends on the coal grade being used.The complete briquetting process and binder options arediscussed later in this paper.

Binders for coal briquettingIt is possible to use binderless pelletization for coal; however,most coal fines are not self-agglomerating. The literaturesuggests numerous binders for the pelletization of coal (Altunet al., 2001; Dehont, 2006):� Coking and oil refining residues such as tar and coal

pitch� Residues from paper mills (lignosulphonate)� Molasses with possible additions of lime� Starch � Synthetic resins� Synthetic polymers .

These binders should have adequate binding strength,relatively low cost, and be resistant to weathering (Waters,1969).

Briquetting process The main operations during a briquetting process are asfollows (Waters, 1969): � Screening and drying of the coal (if too wet)� Mixing of coal with binder� Feeding to briquette machine and pressing� Drying� Storing and packaging.

According to Dehont (2006), the size distribution of thecoal fines should be roughly as follows:� 50% from 0 to 0.5 mm� 25% from 0.55 to 1 mm� 20% from 1 to 2 mm� 5% from 2 to 3 mm.

The fines should have a low moisture content, goodcompatibility, small particle size, and a wide particle size distri-bution to facilitate good packing of the particles. Proper mixingis also critical to ensure that the binder is distributed evenlythroughout the mixture.

Physical testing of briquettesAccording to Richards (1990), the most important physicalproperties of briquettes are resistance to crushing, impact,abrasion, and water penetration. These properties are allheavily dependent on the development of strong and durablebonds between particles during the agglomeration stage. It iscritical that briquettes withstand storage, handling, andtransport during which they will be dropped, experienceabrasion on conveyors, and be exposed to the elements. Fourlaboratory tests are recommended to monitor the physicalstrength properties of briquetted fuels either during processdevelopment or commercial production. These are a dropshatter test, a crushing resistance test, a tumbler abrasion test,and an immersion water resistance test (Richards, 1990).

MethodCoal fines from Exxaro's Mafube coal mine near Middelburgwere used in the briquetting process. This study forms part ofresearch by Exxaro into coal fines agglomeration, and thebinders that were used were prescribed by Exxaro. Twobinders were used – a PVA powder (binder A) and starchpowder (binder B). Coal fines from Exxaro's Mafube coal minenear Middelburg were used in the briquetting process.

Ten days of testing were allocated for physical tests,ranging from day 0 to say 35. ‘Day 0’ refers to the day that thebriquettes were manufactured, as these briquettes have notbeen allowed to dry for a full day.

The mechanical properties of the briquettes wereinvestigated by means of following tests.

Compressive strength Compressive strength is the maximum crushing load abriquette can withstand before cracking or breaking. A singlebriquette was placed on the platform of the tensile strengthtesting machine and, with the machine operating in thecompressive mode, a constant load was applied until thebriquette fractured.

The load at fracture can also be converted to a stress usingthe equation:

By expressing the load as a stress (force per unit area) it ispossible to compare briquettes of various sizes and incorpo-rating different binders. For the purpose of this study allbriquettes were of similar size and shape, and therefore onlythe load force was used. A batch of 20 briquettes was tested ata time.

Impact resistance A batch of 20 briquettes was dropped once from a height of2 m, and the particle size analysis of the pellets and brokenpieces conducted. According to Richards (1990), ‘Impactresistance testing is considered to be the best generaldiagnostic of briquette strength’.

�


Table I

Binder costs per ton of briquettesCost (R/ton)

Coal Binder A Binder A Binder B Binder B (selling (0.1 wt%) (0.9 wt%) (0.3 wt%) (3 wt%)price)

Minimum 150 20 178 1 149Maximum 400 28 248 3 298

Abrasion resistanceA charge consisting of 20 briquettes was rotated in a tumblermachine for 100 revolutions at 50 revolutions per minute. Thetumbler drum had dimensions of 278 mm in length by 20 mmin diameter, and a 38 mm wide lifter plate was also weldedalong the length of the drum. The charge was then collectedand the particle size distribution (PSD) of the fines (-8 mm)that were generated was calculated.

Water resistanceSince briquettes may in some cases be stockpiled outside andexposed to the elements, it is important to test for waterresistance. A single weighed briquette was submerged in abeaker of cold tap water and inspected for disintegration byapplying finger pressure at intervals of about ten minutes. Ifthe briquette remained intact after 30 minutes, the surfacewater was wiped off with a cloth and the briquette wasweighed again. To obtain a quantitative comparison, a waterresistance index (WRI) was calculated as follows:

WRI = 100–%water after 30min

Richards (1990) argues that a WRI > 95% should beobtainable after 30 minutes

Artificial weathering of binder AA QUV Accelerated Weathering Tester was fitted with A340 UVlamps. Binder A was exposed to alternating cycles of UV lightand elevated temperatures. The temperature was set to 63°Cand the irradiance at 0.67 W/m2, and the samples were run ona dry cycle.

After exposure of only a few days the QUV tester canreproduce damage that will take months or years outdoors.

The rate of polymer oxidation was measured by conductinginfrared spectroscopy (IR) on the exposed films. By followingthe growth of the carbonyl peak near 1720 cm-1, a carbonylindex was defined by the ratio of this absorption to that at2900 cm-1 and used to quantify the progression ofdegradation.

Thermogravimetric analysisThe sample was heated to a maximum temperature of 900°C,and the residual mass plotted against temperature. Thematerial that remains after the maximum temperature isreached should correspond to the total ash percentage of thecoal sample.


Compressive strengthThe SABS 1399:1999 standard was used to ensure thebriquettes met the minimum compressive strengthrequirements. This standard specifies the requirements forcharcoal made from wood in either lump or briquette form. Thecompression strength of the briquettes was first evaluated as afunction of moisture content (the total moisture of the batchduring the briquetting process). From Figure 1 it can be seenthat the compressive strength increased from day 0 to day 15for each moisture level. At 10% moisture level, briquettes withbinder B, with an initial strength of 37 N, gained hardly anystrength after 15 days. Binder A yielded an increase of onlyabout 40 N after 15 days. This low moisture content does notallow for efficient dispersion of the binder throughout the coal

fines. Briquettes with 12% moisture showed the greateststrength gains for both binder A and binder B. Day 0compression strength results could not be obtained frombriquettes with 18% moisture and using binder B, as they weretoo soft and disintegrated under the load.

In Figure 2 the strong dependence of compressive strengthon binder content is illustrated very clearly. The initialstrength, as well as day 15 strength, increases with increase inbinder content. Briquettes with binder B at levels of 0.3% and0.5% show poor strength, and attain strength values wellabove the SABS standard of 25 N only with binder additions of1% and more. The maximum strength of 358 N was obtainedon day 15 for 3% binder B. However, it may be necessary tomake a trade-off between the strength and binder cost. Loweradditions of binder B result in acceptable briquette strengths,with gains of 61 N and 97 N for 1% and 2% binder B additionrespectively.

Low levels of binder A also resulted in weak briquettes.Only at a level of 0.5% binder A and higher is sufficientcompression strength achieved in the briquettes. The highestbinder A addition of 0.9% also resulted in the greatest strengthgain of 68 N. This is followed by 0.5% binder A with a gain of46 N, which will be a more economical option.

The observed trend of increasing compressive strengthwith increased binder addition is expected because at higherbinder levels more binder is available to be dispersed betweenthe coal particles and ensure bonding between binder and coal.



Figure 1 – Compressive strength as a function of varying moisturecontent (constant binder addition)

Figure 2 – Compressive strength as a function of binder addition(constant 12% moisture)


DDrop and tumble testsThe results of the drop and tumble tests were combined on asingle axis for easier comparison. In Figure 3 the binderaddition is constant, with moisture as the variable. SABSstandard 1399 specifies a maximum of 5% fines. Fines areconsidered to be material that passes through an 8 mm screen.The tumble test produced more fines than the drop test for allmoisture contents. For day 0 testing, when the briquettes are attheir weakest, the briquettes with binder A generated onaverage 24% fines, while for binder B this was significantlyless at 8%.

For both binders, the greatest amount of fines generatedduring both the drop test and the tumble test were from thebriquettes with lowest moisture level of 10%. This is probablya result of the inefficient distribution of the binder throughoutthe mixture.

The amount of fines generated in both tests decreased withincreasing moisture level, but this trend was more pronouncedwwith the drop test. For binder A the optimum moisture levelindicated by drop test and tumble test results is 16% and 18%respectively. For binder B, a moisture level anywhere between12% and 18% will give similar results from drop tests andtumble tests.

In tests where the binder content was varied and themoisture level was kept constant at 12%, the drop test andtumble test results followed similar trends as in Figure 3, withfewer fines being produced in the drop test. The optimum

f famount of binder A was found to be 0.5%. This amountproduced the second lowest amounts of fines during thetumble test and drop test, at 8.8% and 32.6% respectively.From an economic point of view, adding 0.5% binder ispreferable to adding 0.9%.

An amount of 1% and higher of binder B will givefavourable results. Very small amounts of fines were generatedduring the tumble tests for 1%, 2%, and 3% binder additions.This is an indication of how well binder B briquettes canwithstand an abrasive environment like transport on aconveyor belt. Hardly any fines were generated for the threehigher amounts of binder additions, and hence these briquetteswill stay relatively intact when tipped from a transport truck orwhen falling from a conveyor belt.

The optimum binder B content is therefore 1%, being thelevel that will ensure a low amount of fines generated.

Drying conditionsFigure 5 shows the strength gain for both binders over the firstfive days of drying under different conditions. Drying outsidein direct sunlight exposes the briquettes to a higher averagetemperature than those allowed to dry indoors. This results infaster evaporation of the moisture and the different bondscreated by the binders are established much earlier in thecuring process. For both binders, a strength plateau is reachedafter five days of drying outside, and little additional strengthis gained from day 5 to day 35.

For the samples dried indoors, the greatest strength gain isachieved in the first two days of curing. At this stage thesebriquettes do not have the same strength as the samples driedoutside, but from day 5 they continue to steadily gain strengthuntil the final strength on day 35 is similar, or close to, that ofthe samples that were dried outside.

Drying samples outside looks like the obvious choice asthis allows for a much shorter curing process. However, itshould be noted that neither of the binders results in a water-resistant briquette, and therefore when these briquettes arestored outside, they must be under cover to prevent raindamage.

Water resistanceWhen binder B briquettes were submerged in water, theyimmediately disintegrated. It was clear that these briquettes didnot have any water resistance.

�


Figure 4 – Compressive strength of briquettes with binder A and binderB and 12% moisture that were dried inside and outside

Figure 3 – Drop test and tumble test results for briquettes with varyingmoisture (constant 0.5% binder A)

Figure 5 – Infrared spectra of new and weathered binder A film

Binder A was expected to impart water-resistant propertiesto the briquettes, but these also disintegrated when submergedin water. Even after 15 days of curing the binder was not ableto establish water-resistant bonds between the particles.

AArtificial weathering of binder A The results from attenuated total reflection (ATR) spectroscopyindicated that no weathering took place on the surface of thefinder A film. In Figure 5 barely any difference is seen betweenthe curves for the different exposure times, indicating that littleto no degradation had taken place on the surface of the filmafter exposure of 300 hours in the QUV.

Thermogravimetric analysisThe results from TGA are plotted in Figure 6. Hardly anydifference is seen between the pure coal and coal with bindermixtures. The coal, with or without binder, is thermally stableup to 400°C, and between 400°C and 600°C the binder andother volatiles decompose. Above 600°C only about 30% of thematerial remains, and this value is similar to the total ashcontent of 30.4% shown in Figure 6.

Binder A starts decomposing at about 95°C and is totallydecomposed at about 530°C. Binder B shows slow decompo-sition from 25°C to 90°C followed by rapid decompositionbetween 90°C and 490°C. Binder B is totally burned off above500°C.

The binder additions to the coal were the maximumamounts that were used throughout this study: 0.9 % binder Aand 3% binder B. It is safe to say that neither of the twobinders will add to the ash content of the coal as both binderstotally decompose above 530°C.

ConclusionsThe compressive strength of the briquettes depends on thebinder addition and the moisture content. For both binder Aand binder B, the optimum moisture level was 12%. Theminimum binder addition for adequate strength was 0.5% forbinder A, and 1% for binder B. Further additions of binder Bincreased briquette strength, but the higher cost of binder Brenders this option uneconomical.

For binder A, the optimum moisture level was 12% to 14%.At this moisture level the largest compressive strength gainswwere observed, as well as a low amount of fines produced. Theminimum amount of binder to be added to obtain a briquette ofadequate strength was 0.5%.

For binder B the optimum moisture level was also 12%,and the minimum binder addition was found to be 1%.

The briquettes that were dried outside reached their peakcompressive strength after about four days. The briquettes thatdried inside took about 20 days to reach maximum strength.

Neither of the binders resulted in water-resistantbriquettes, as all of the briquettes tested disintegrated whensubmerged in water.

ATR spectroscopy indicated that no degradation of thebinder A film took place after 300 hours of exposure in theQUV.

The TGA results confirmed that neither binder A norbinder B will increase the ash content of the coal fines, as bothbinders totally decompose above 530°C.

The cost of binder B is higher than binder A, but itsstrength after 15 days of curing and the low amount of finesproduced with minimum of 1% binder addition makes it thepreferred binder to use.

References

ALTUN, N. E., HICYILMAZ, C., and KÖKKK , M.V. 2001. Effect of different binders on the

combustion properties of lignite. Part 1. Effect of thermal properties. Journal

of Thermal Analysis and Calorimetry, vol. 65. pp. 787–795.

DEHONT, F. 2006. Coal briquetting technology.

http://www.almoit.com/allegati/applicazioni_particolari/15/COAL%20BRIQU

ETTING%20TECHNOLOGY.pdf [Accessed 13 October 2013].

ENGLAND, T. 2013. The economic agglomeration of fine coal for industrial and

commercial use: A review of past and present work both locally and interna-

tionally.

http://www.coaltech.co.za/chamber%20databases%5Ccoaltech%5CCom_Doc

Man.nsf/0/09EFEE68D9257C164225781B00364F21/$File/Task%204.4%2

01%20-%20Agglomoration%20of%20Fine%20Coal%20-

%20Trevor%20England.pdf

ESI-AFRICA. 2014. The economic possibilities of South Africa’s coal fines.

http://www.esi-africa.com/the-economic-possibilities-of-south-africas-coal-

fines/ [Accessed 17 June 2014].

ESKOM. 2014. Coal Power.

http://www.eskom.co.za/AboutElectricity/ElectricityTechnologies/Pages/Coa

l_Power.aspx [Accessed 17 June 2014].

FEECO INTERNATIONAL. 2014. Briquettes, Granules, and Pellets – What’s the

difference? http://feeco.com/2012/01/11/briquettes-granules-and-pellets-

whats-the-difference/ [Accessed 17 June 2014].

RICHARDSRR , S.R. 1990. Physical testing of fuel briquettes. Fuel Processing

Technology, vol. 25. pp. 89–100.

SABS 1399:1999. Wood charcoal and charcoal for household use. South African

Bureau of Standards. Pretoria. http://www.sabs.co.za [Accessed 20

November 2014].

SASTRY, K.V.S. and FEURSTENAU, D.W. 1977. Kinetics and process analysis of

agglomeration of particulate materials. Agglomeration '77. AIME, New York.77

pp. 318-402.

WATERSWW , P.L. 1969. Binders for fuel briquettes: a critical survey. Technical

Communication 51. CSIRO. �



Figure 6 – Change in mass of coal and binders during thermogravi-metric analysis

For further information contact:Head of Conferencing

Raymond van der Berg, SAIMMP O Box 61127, Marshalltown 2107

Tel: +27 (0) 11 834-1273/7E-mail: [email protected]

Website: http://www.saimm.co.za

Copper CobaltAfrica

Copper CobaltAfrica

In association withThe 8th Southern African Base Metals Conference

6–8 July 2015Zambezi Sun Hotel, Victoria Falls

Livingstone, Zambia

Second Announcement& List of Abstracts

Join us for the inaugural Copper Cobalt Africa Conference in the heart of Africa.To be held at Victoria Falls, one of the Seven Natural Wonders of the World, this prestigious

event will provide a unique forum for discussion, sharing of experience and knowledge, andnetworking for all those interested in the processing of copper and cobalt in an African context, inone of the worldʼs most spectacular settings.

The African Copper Belt has experienced a huge resurgence of activity in recent years followingmany years of political and economic instability. Today, a significant proportion of capital spending,project development, operational expansions, and metal value production in the Southern Africanmining industry are occurring in this region. The geology and mineralogy of the ores aresignificantly different from those in other major copper-producing regions of the world, often havingvery high grades as well as the presence of cobalt. Both mining and metallurgy present someunique challenges, not only in the technical arena, but also with respect to logistics and supplychain, human capital, community engagement, and legislative issues. This conference provides aplatform for discussion of these topics, spanning the value chain from exploration, projects, throughmining and processing.

For international participants, this conference offers an ideal opportunity to gain in-depthknowledge of and exposure to the Southern African base metalsindustry, and to better understand the various facets of miningand processing in this part of the world that both excite andfrustrate the industry.

A limited number of places are available for post-conferencetours to Zambiaʼs most important commercial operations,including Kansanshi, the largest mine in Zambia, with 340 kt/ycopper production and its soon-to-be-completed 300 kt/ysmelter, and Chambishi Metals.

Jointly hosted by the mining and metallurgy technicalcommittees of the Southern African Institute of Mining andMetallurgy (SAIMM), this conference aims to:

• Promote dialogue between the mining and metallurgicaldisciplines on common challenges facing the industry,

• Encourage participation and build capacity amongst youngand emerging professionals from the Copper Belt region,

• Improve understanding of new and existing technologies,leading to safe and optimal resource utilisation.

The organising committee looks forward to your participation.

SPONSORS:

Premium

IntroductionThe increased use of mechanized coal miningmethods has resulted in greater amounts of coalfines being generated. Many operations reportan estimated 6% of their ROM production to bein the -2 mm size range. Fine and ultra-finesize ranges constitute about 11% of the nominalproduct and retain the bulk of the moisture(SANEDI, 2011). Given the problems associatedwith moisture in fine coal, it is important toinvestigate and improve available moistureremoval techniques. Coal constitutes theprimary source of energy in South Africa and isa major contributor to the economy, andtherefore improving coal quality will in effect

maximize the quantity of usable coal (De Korteand Mangena, 2004). With the use of effectivedewatering methods, fine coal can be benefi-ciated and added to the coarse particle circuitswithout compromising the quality of the productand hence substantially increasing the overallplant yield (Condie and Veal, 1998). Condie andVeal (1998) suggest that by rule of thumb, forevery ton of moisture removed the clean coalproduct stream is supplemented with about 4 tof fine coal. This is a powerful incentive fordeveloping advanced dewatering techniques forfine coal particles. Additionally, excessivemoisture adds to the mass-based transportcosts of coal. For this reason, developingadvanced efficient fine coal drying techniques isbeneficial from an economical point of view(Campbell, 2006).

Rowan (2010) states that coal preparationplants generally discard coal fines with sizefractions below 150 μm into waste ponds. Thisposes a danger of spontaneous combustion,acid mine drainage, and dust release as thesurface of the coal is exposed to ambient air andweathering conditions for long periods (DeKorte and Mangena, 2004). Coal fines are moresusceptible to water absorption than coarsercoal and can contain up to 25 wt% totalmoisture after filtration (Le Roux, 2003).Thermal drying methods are more efficient thanmechanical dewatering techniques (De Korteand Mangena, 2004), but the price of coal limitsthe use of these methods (SANEDI, 2011).Studies conducted at North-West Universityshowed that drying of fine coal (-2 mm +1 mm)in a fluidized bed is possible at low temper-atures between 25°C and 40°C.

Work by Le Roux et al. (2012) on vacuumfiltration showed that intentionally damaging afilter cake improved the airflow infiltration,leading to a lower pressure differential acrossthe cake but increasing the dewatering

Air drying of fine coal in a fluidized bed by M. Le Roux*, Q.P. Campbell*, M.J. van Rensburg*,E.S. Peters*, and C. Stiglingh*Paper wwritten oon pproject wwork ccarried oout iin ppartial ffulfilment oof DDegree iin CChemical EEngineering(NWU) —— Pursuing MMasters iin CCoal BBeneficiation

SynopsisThe demand for energy has continued to rise worldwide in line withpopulation growth. The majority of South Africa’s electricity is supplied bycoal-fired power stations. The amount of fine coal (-2 mm) generated atcoal processing plants has increased, due mainly to mechanized miningmethods. Fine coal retains more water, which lowers its heating value.

Drying the coal is costly and it is difficult to achieve the requiredmoisture content. Consequently, coal fines are often discarded. Anestimated 8% of the total energy value of mined coal is lost1.

Fluidized bed technology is often used to dry coal thermally, but thismethod is expensive and has an adverse environmental impact. Theobjective of this study was to investigate the removal of moisture from finecoal (<2 mm) in a fluidized bed operated with dry fluidizing air at moderatetemperatures as the drying agent. The effects of different air temperaturesand relative humidity levels were investigated in a controlled environment.The study further investigated the influence of coal particle size onmoisture removal.

The drying rate was found to increase with increasing temperature. Therelative humidity of the drying air had a more pronounced effect on thedrying rate, even at temperatures as low as 25°C.. It became morechallenging to remove moisture as the particle size decreased. The gain incalorific value was greater than the energy required to dry the coalsamples, showing that a fluidized bed using moderately warm dry air is anenergy-efficient drying technology. The energy efficiency of the fluidizedbed compared favourably with other thermal drying methods.

Keywordscoal fines, drying, fluidized bed, energy efficiency.

* School of Chemical and Minerals Engineering,North-West University, Potchefstroom, SouthAfrica.

© The Southern African Institute of Mining andMetallurgy, 2015. ISSN 2225-6253. Paper receivedJan. 2015


1 Rowan, S.L. 2010. Analysis and scaling of atwo-stage fluidized bed for drying of fine coalparticles using shannon entropy, thermodynamicexergy and statistical methods. PhD dissertation,University of West Virginia, Morgantown WV.


Air drying of fine coal in a fluidized bed

efficiency. This work confirmed that high airflow conditionsresulted in a lower final cake moisture content of 3–5 wt%.Further studies found that an increased airflow rate resulted ina more effective moisture transfer from the coal fines to thedrying air (Le Roux et al., 2013).ll

The moisture content of fine coal particles is made up ofsurface, capillary, and chemically bound moisture (Rong,1993) as depicted schematically in Figure 1. Free moisture isfound on the exterior surface of coal particles (Condie and Veal,1998), and can be removed by mechanical methods such asfilters and centrifugal units. Capillary-bound moisture isabsorbed and held tightly within micro-capillaries and micro-pores of individual coal particles (Rong, 1993). Removal of thismoisture calls for thermal drying techniques for completedrainage (Condie and Veal, 1998). Chemically bound moistureis not included when measuring the total moisture content ofthe coal (Campbell, 2006), and can be removed only bypyrolysis.

The equilibrium moisture content of coal is characterized asthe moisture content at which the coal particles no longer gainor lose moisture, and it varies according to the temperature andrelative humidity conditions of the atmosphere surrounding theparticles. Mechanical methods are insufficient for the removalof this equilibrium moisture, which can be reduced only bymeans of evaporation (Le Roux et al., 2013). The relativellhumidity and temperature act as driving forces that change thephase equilibrium between vapour and liquid, with lowerhumidities and higher temperatures leading to moisture beingabsorbed from the particle by the drying medium (Koretsky,2004).

Experimental methodThe aim of this project was to determine the effect oftemperature, relative humidity, and particle size distribution on

ff f f fthe efficiency of drying fine coal particles in a fluidized bed.The energy consumption during the drying process wascalculated and compared to published data on thermal dryingprocesses.

Sample preparationSouth African bituminous coal from the Waterberg coalfieldwas used for these experiments. The proximate analysis of thistype of coal is given in Table I. The coal was crushed andsieved into three particle size ranges: fines (between 2 mm and1.18 mm) and ultra-fines (between 1.18 mm and 0.5 mm).The samples were drenched in water for a day and the excessfree moisture was removed by pressure filtration. The moisturecontent of each filtered sample was determined(SANS5925:2007) before the coal was fed to the fluidized bedfor dewatering.

ApparatusA fluidized bed column (10 cm inner diameter × 40 cm length)was constructed from polycarbonate (Figure 2). The columnwas connected to a blower, which was used to drawconditioned air at a set temperature and relative humidity froma climate chamber (CTS climate test chamber Type: C-40/100).A packed bed of glass marbles in the bottom section of thefluidized bed acted as airflow distributor. Mesh covers (0.5 mmaperture) were placed at the top and bottom sections of thefluidized bed to retain the bulk coal sample within the cylinder.The outlet air from the column was returned to the climatechamber, and was recirculated to the column after thetemperature and relative humidity values attained the pre-setlevels. For each test, 100 g of fine coal sample with a totalmoisture content of approximately 25-35 wt% (typical of apressure filter product) was fed to the fluidized bed cylinder.The weight of the column was continually monitored duringfluidization to determine the loss of moisture. A number of

�


Figure 1 – Forms of moisture related to coal (after Lemley et al., 1995)

Table I

Proximate analysis (air-dried basis)

Component Percentage by weight

Fixed carbon 39.98Moisture content 2.58

Ash content 22.82Volatile matter 34.62 Figure 2 – Experimental set-up of the fluidized bed (Van Rensburg, 2014)

fselected experiments were repeated in the fluidized bed todetermine the repeatability of the results.


IInfluence of temperatureTo study the effect of air temperature, wet samples of 100 gcontaining about 30 wt% total moisture were placed in thefluidized bed cylinder, and drying air was introduced at asuperficial velocity of 1.5 to 1.7 m/s, which was slightly abovethe predetermined minimum fluidization velocity. Two sets ofexperiments were conducted at air temperatures of 25°C and55°C respectively and a relative humidity (RH) of 30%.Figure 3 shows the moisture loss from the ultra-fine sample(-1.18 mm +0.71 mm) under these conditions. The drying ratewwas quicker at 55°C than at 25°C. The drying time was reducedfrom 28 minutes to 21 minutes. This confirms the observationmade by Rowan (2010) that elevated temperatures lead tohigher dewatering rates. Higher temperatures disrupt the phaseequilibrium and increase the amount of water transported fromthe coal sample into the surrounding air (Condie and Veal,1998).

Duplicate experimental runs proved the repeatability of theresults, the maximum and minimum standard deviation being3.40 wt% and 0.12 wt% respectively.

f fInfluence of relative humidity For the next set of experiments, the fluidizing air wasintroduced at relative humidities of 30%, 50%, and 70% at aconstant temperature of 55°C. The drying curves are shown inFigure 4. A comparison of Figure 4 with Figure 3 shows thatrelative humidity has a greater effect on the drying rate thantemperature. Lower relative humidities counteract the capillaryforces retaining the moisture in the coal particle, leaving adried product in about 14 minutes at 30% RH, compared toover 30 minutes for 70% RH at the same temperature. Higherrelative humidities weaken the moisture transfer mechanism,and therefore the moisture is displaced from the capillarychannels at a lower rate (Condie and Veal, 1998). VanRensburg (2014) stated that the mechanism for the transfer ofwater molecules from the coal to the air is enhanced when thedrying air contains low moisture levels. This leads to a hightransfer rate of the water molecules from an area of highmoisture content to an area of low moisture content.

Influence of particle sizeThree wet filter cake samples, all with an initial moisturecontent of 33 wt%t and different size ranges (-1.18 mm +0.71mm, -0.71 mm +0.50 mm, and a 50/50 mixture of -1.18 mm+0.71 mm and -0.71 mm +0.5 mm), were dried in the fluidizedbed using feed air at 55°C and 50% RH. Figure 5 shows thedrying curves for these experiments.

The coarse fraction (-1.18 mm +0.71 mm) and the 50%mixture showed similar drying responses, reaching a finalmoisture value in less than 20 minutes, while the finer fraction(-0.71 mm +0.5 mm) reached a similar final moisture valueonly after 37 minutes. This shows that it is increasinglydifficult to remove moisture as the size of the coal particlesdecreases. Small particles have large surface areas and moremicropores to absorb water, resulting in a higher degree ofwater retention (De Korte and Mangena, 2004).

Drying ratesFigure 6 shows the different drying rates for fluidizing feed airconditioned at 25°C across the set relative humidity ranges. Adecrease in relative humidity clearly increases the drying ratefor all particle size ranges. The drying rate increased from0.010 wt%/min to 0.015 wt%/min for the -2 mm +1.18 mmparticles at 25°C with 70% and 30% RH respectively. It is alsoapparent that the drying rate increased with an increase in



Figure 4 – Effect of relative humidity on the drying of the mixture -1.18mm +0.71 mm and +0.71 mm -0.5 mm at 55°C

Figure 3 – Effect of temperature on the drying of -1.18 mm +0.71 mmcoal at 30% RH

Figure 5 Effect of particle size on the drying of coal at 55°C and 50% RH


particle size at less than 50% RH. It is noteworthy that at25°C, particle size is not the rate-limiting factor at relativehumidity conditions exceeding 50% RH. The -0.71 mm +0.5mm fraction had the lowest drying rate at 25°C and 70% RH,wwhile the fastest drying rate was for the -2 mm +1.18 mmparticle size range at 30% RH.

It is clear that relative humidity has a more significanteffect on the drying rate of the coal particles than temperature,hence dry air at moderately low temperatures is effective indrying fine coal.

EEnergy consumptionThe energy required during the drying process was calculatedby considering the change in enthalpy of the water, the workdone by the blower, as well as the energy required for theconditioning of the air. A temperature of 25°C and relativehumidity of 50% were chosen as a basis for the calculations,since these were the average ambient conditions in thelaboratory. The calorific value of the coal was upgraded by 8MJ/kg on average using air at these conditions as dryingmedium. Figure 7 shows the calculated maximum andminimum of energy requirements to dry to three differentmoisture levels using the fluidized bed with those of otherexisting thermal drying technologies. It can be seen thatfluidization is more energy-efficient than other thermalprocesses, with the exception of the Fleissner process.

CConclusionThis work has shown that fine and ultra-fine coal particles canbe dried at moderate temperatures and low relative humiditiesin a fluidized bed. The time required for fines is about half ofthat required to dry ultra-fines. The main driving force forremoval of moisture from fine and ultra-fine coal is relativehumidity. Energy calculations demonstrate that fluidization ismore energy-efficient than other thermal drying processes.Using dry air at a moderate temperature in a fluidized bed todry coal particles is thus a promising technique warrantingfurther study and development, since it has a potential energyadvantage as well the ability to increase the calorific value ofthe coal.

AcknowledgementsThe authors would like to acknowledge the followinginstitutions for their contribution towards this project:� Coaltech� NRF (National Research Foundation).

This work is based on research supported by the SouthAfrican Research Chairs’ Initiative of the Department of Scienceand Technology and the National Research Foundation ofSouth Africa. Any opinion, finding, or conclusion, orrecommendation expressed in this material is that of theauthors and the NRF does not accept any liability in thisregard.

ReferencesCAMPBELLCC , Q.P. 2006. Dewatering of fine coal with flowing air using low

pressure drop systems. PhD dissertation, North-West University,Potchefstroom. 130 pp.

CONDIE, D. and VEALVV , C. 1998. Improved fine coal dewatering via modelling ofcake desaturation. CSIRO, Australia. pp. 1–34.

DE KORTEKK , G.J. and MANGENAMM , S.J. 2004. Thermal Drying of Fine and Ultra-fineCoal. Report no. 2004 – 0255. Division of Mining Technology, CSIR,Pretoria. pp. 5–24.

KARTHIKEYANKK , M., ZHONGHUA, W., and MUJUMDARMM , A.S. 2009. Low-rank coaldrying technologies – current status and new developments. DryingTechnology, vol. 27, no. 3. pp. 403–415.

KORETSKYKK , M.D. 2004. Engineering and Chemical Thermodynamics. 2nd edn. JohnWiley & Sons Hoboken, NJ.

LE ROUXRR , M. 2003. An investigation into an improved method of dewatering finecoal. Master’s dissertation, North-West University, Potchefstroom. 96 pp.

LE ROUXRR , M., Campbell, Q.P., and Smit, W. 2012. Large-scale design and testingof an improved fine coal dewatering system. Journal of the Southern AfricanInstitute of Mining and Metallurgy, vol. 112, no. 7. pp. 673–676.

LE ROUXRR , M., CAMPBELLCC , Q.P., and VANVV RENSBURGRR , M.J. 2013. Fine coal dewateringusing high airflow. International Journal of Coal Preparation andUtilization, vol. 34. pp. 220–227.

SANEDI (South African National Energy Development Institute). 2011.http://www.sanedi.org.za/coal-roadmap/ [Accessed 25 June 2014].

RONGRR , R.X. 1993. Literature review on fine coal and tailings dewatering.Advances in Coal Preparation Technology, vol. 2. Project P239A. JKMRC,University of Queensland. Brisbane. Australia. 120 pp.

ROWANRR , S.L. 2010. Analysis and scaling of a two-stage fluidized bed for dryingof fine coal particles using shannon entropy, thermodynamic exergy andstatistical methods. PhD dissertation, University of West Virginia,Morgantown WV. 154 pp.

VANVV RENSBURGRR , M.J. 2014. Drying of fine coal using warm air in a dense mediumfluidised bed. Master’s dissertation, North-West University, Potchefstroom.98 pp.

YUYY , A.B., STANDISH, N., and LU, L. 1994. Coal agglomeration and its effect onbulk density. Powder Technology Journal, vol. 82, no. 1. pp. 177–189. ll �

�


Figure 7 – Energy consumption of different thermal drying methods forcoal in the -2 mm +1.18 mm particle size range

Figure 6 – Drying rates of different size fractions at 25°C

2015� CONFERENCE

Mining, Environment and Society Conference12–13 May 2015, Mintek, Randburg, Johannesburg

� CONFERENCECopper Cobalt Africa Incorporating The 8th Southern African Base Metals Conference6–8 July 2015, Zambezi Sun Hotel, Victoria Falls, Livingstone, Zambia

� SCHOOLProduction of Clean Steel13–14 July 2015, Emperors Palace, Johannesburg

� CONFERENCEVirtual Reality and spatial information applications in the mining industry Conference 201515–17 July 2015, University of Pretoria, Pretoria

� CONFERENCEMINPROC 2015: Southern African Mineral Beneficiation andMetallurgy Conference6–7 August 2015, Vineyard Hotel, Newlands, Cape Town

� CONFERENCEThe Danie Krige Geostatistical Conference 201519–20 August 2015, Crown Plaza, Johannesburg

� CONFERENCEMINESafe 2015—Sustaining Zero Harm: Technical Conference andIndustry day26–28 August 2015, Emperors Palace Hotel Casino, Convention Resort,Johannesburg

� CONFERENCEFormability, microstructure and texture in metal alloys Conference201515–17 September 2015

� CONFERENCEWorld Gold Conference 201528 September–2 October 2015, Misty Hills Country Hotel and Conference Centre, Cradle of Humankind, Muldersdrift

� SYMPOSIUMInternational Symposium on slope stability in open pit mining and civil engineering12–14– October 2015In association with the Surface Blasting School15–16 October 2015, Cape Town Convention Centre, Cape Town

� COLLOQUIUM13th Annual Southern African Student Colloquim 201520 October 2015, Mintek, Randburg, Johannesburg

� CONFERENCEYoung Professionals 2015 Conference21–22 October 2015, Mintek, Randburg, Johannesburg

� CONFERENCEAMI: Nuclear Materials Development Network Conference28–30 October 2015, Nelson Mandela Metropolitan University, North Campus Conference Centre, Port Elizabeth

� SYMPOSIUMMPES 2015: Twenty Third International Symposium on MinePlanning & Equipment Selection8–13 November 2015, Sandton Convention Centre, Johannesburg, South Africa

SSAAIIMMMM DDIIAAAAAAAAARRYYYYY

Forthcoming SAIMM events...

For further information contact:Conferencing, SAIMM

P O Box 61127, Marshalltown 2107Tel: (011) 834-1273/7

Fax: (011) 833-8156 or (011) 838-5923E-mail: [email protected]

F or the past 120 years, theSouthern African Institute ofMining and Metallurgy, has

promoted technical excellence in theminerals industry. We strive tocontinuously stay at the cutting edgeof new developments in the miningand metallurgy industry. The SAIMMacts as the corporate voice for themining and metallurgy industry in theSouth African economy. We activelyencourage contact and networkingbetween members and thestrengthening of ties. The SAIMMoffers a variety of conferences thatare designed to bring you technicalknowledge and information ofinterest for the good of the industry.Here is a glimpse of the events wehave lined up for 2015. Visit ourwebsite for more information.

Website: http://www.saimm.co.za

EXHIBITS/SPONSORSHIP

Companies wishing to sponsor

and/or exhibit at any of these

events should contact the

conference co-ordinator

as soon as possible

�

xii APRIL 2015 The Journal of The Southern African Institute of Mining and Metallurgy

201512–13 May 2015 — Mining, Environment and SocietyConference: Beyond sustainability—BuildingresilienceMintek, Randburg, South AfricaContact: Yolanda RamokgadiTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156 E-mail: [email protected]: http://www.saimm.co.za

14–17 June 2015 — European MetallurgicalConferenceDusseldorf, Germany, Website: http://www.emc.gdmb.de

14–17 June 2015 — Lead Zinc Symposium 2015Dusseldorf, Germany, Website: http://www.pb-zn.gdmb.de

16–20 June 2015 — International Trade Fair forMetallurgical Technology 2015Dusseldorf, GermanyWebsite: http://www.metec-tradefair.com

6–8 July 2015 — Copper Cobalt Africa IncorporatingThe 8th Southern African Base Metals ConferenceZambezi Sun Hotel, Victoria Falls, Livingstone, Zambia Contact: Raymond van der BergTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

13–14 July 2015 — School Production of Clean SteelEmperors Palace, JohannesburgContact: Yolanda RamokgadiTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156 E-mail: [email protected]: http://www.saimm.co.za

15–17 July 2015 — Virtual Reality and spatialinformation applications in the mining industryConference 2015University of PretoriaContact: Camielah JardineTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156 E-mail:[email protected]: http://www.saimm.co.za

6–7 August 2015 — MINPROC 2015: Southern AfricanMineral Beneficiation and Metallurgy ConfereneVineyard Hotel, Newlands, Cape TownContact: Raymond van der BergTel: +27 11 834-1273/7

Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

19–20 August 2015 — The Danie Krige GeostatisticalConference: Geostatistical geovalue —rewards andreturns for spatial modellingCrown Plaza, JohannesburgContact: Yolanda RamokgadiTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

25–27 August 2015 — Coal Processing – UnlockingSouthern Africa’s Coal PotentialGraceland Hotel Casino and Country Club SecundaContact: Ann RobertsonTel: +27 11 433-0063

26–28 August 2015 — MINESafe 2015—SustainingZero Harm: Technical Conference and Industry dayEmperors Palace Hotel Casino, Convention Resort,JohannesburgContact: Raymond van der BergTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

15–17 September 2015 — Formability, microstructureand texture in metal alloys ConferenceContact: Yolanda RamokgadiTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

28 September-2 October 2015 — WorldGoldConference 2015Misty Hills Country Hotel and Conference Centre,Cradle of HumankindGauteng, South AfricaContact: Camielah Jardine, Tel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.z

12–14 October 2015 — Slope Stability 2015:International Symposium on slope stability in openpit mining and civil engineeringIn association with theSurface Blasting School

INTERNATIONAL ACTIVITIES

�xiiiThe Journal of The Southern African Institute of Mining and Metallurgy APRIL 2015

201515–16 October 2015Cape Town Convention Centre, Cape TownContact: Raymond van der BergTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156 E-mail: [email protected]: http://www.saimm.co.za

20 October 2015 — 13th Annual Southern AfricanStudent ColloquiumMintek, Randburg, JohannesburgContact: Yolanda RamokgadiTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

21–22 October 2015 — Young Professionals 2015ConferenceMaking your own way in the minerals industryMintek, Randburg, JohannesburgContact: Camielah JardineTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail:[email protected]: http://www.saimm.co.za

28–30 October 2015 — AMI: Nuclear MaterialsDevelopment Network ConferenceNelson Mandela Metropolitan University, North CampusConference Centre, Port ElizabethContact: Raymond van der BergTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

8–13 November 2015 — MPES 2015: Twenty ThirdInternational Symposium on Mine Planning &Equipment Selection Sandton Convention Centre, Johannesburg, South AfricaContact: Raj SinghalE-mail: [email protected] or E-mail: [email protected]: http://www.saimm.co.za

201614–17 March 2016 — Diamonds Conference 2016BotswanaContact: Yolanda RamokgadiTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

13–14 April 2016 — Mine to Market Conference 2016South AfricaContact: Yolanda RamokgadiTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

17–18 May 2016 — The SAMREC/SAMVALCompanion Volume ConferenceJohannesburgContact: Yolanda RamokgadiTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

May 2016 — PASTE 2016 International Seminar onPaste and Thickened TailingsKwa-Zulu Natal, South AfricaContact: Raymond van der BergTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

9 –10 June 2016 — 1st International Conference onSolids Handling and ProcessingA Mineral Processing PerspectiveSouth AfricaContact: Raymond van der BergTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

17–20 July 2016 — Hydrometallurgy Conference 2016‘Sustainability and the Environment’in collaboration with MinProc and the Western CapeBranchCape TownContact: Raymond van der BergTel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

16–19 August 2016 — The Tenth InternationalHeavy Minerals Conference ‘Expanding the horizon’Sun City, South AfricaContact: Camielah Jardine, Tel: +27 11 834-1273/7Fax: +27 11 838-5923/833-8156E-mail: [email protected]: http://www.saimm.co.za

INTERNATIONAL ACTIVITIES

�

xiv APRIL 2015 The Journal of The Southern African Institute of Mining and Metallurgy

Company AffiliatesThe following organizations have been admitted to the Institute as Company Affiliates

AECOM SA (Pty) Ltd

AEL Mining Services Limited

Air Liquide (PTY) Ltd

AMEC Mining and Metals

AMIRA International Africa (Pty) Ltd

ANDRITZ Delkor(Pty) Ltd

Anglo Platinum Management Services (Pty) Ltd

Anglo Operations Ltd

Anglogold Ashanti Ltd

Atlas Copco Holdings South Africa (Pty) Limited

Aurecon South Africa (Pty) Ltd

Aveng Moolmans (Pty) Ltd

Axis House (Pty) Ltd

Bafokeng Rasimone Platinum Mine

Barloworld Equipment -Mining

BASF Holdings SA (Pty) Ltd

Bateman Minerals and Metals (Pty) Ltd

BCL Limited

Becker Mining (Pty) Ltd

BedRock Mining Support (Pty) Ltd

Bell Equipment Company (Pty) Ltd

BHP Billiton Energy Coal SA Ltd

Blue Cube Systems (Pty) Ltd

Bluhm Burton Engineering (Pty) Ltd

Blyvooruitzicht Gold Mining Company Ltd

BSC Resources

CAE Mining (Pty) Limited

Caledonia Mining Corporation

CDM Group

CGG Services SA

Chamber of Mines

Concor Mining

Concor Technicrete

Council for Geoscience Library

CSIR-Natural Resources and theEnvironment

Department of Water Affairs and Forestry

Deutsche Securities (Pty) Ltd

Digby Wells and Associates

Downer EDI Mining

DRA Mineral Projects (Pty) Ltd

Duraset

Elbroc Mining Products (Pty) Ltd

Engineering and Project Company Ltd

eThekwini Municipality

Evraz Highveld Steel and Vanadium Corp Ltd

Exxaro Coal (Pty) Ltd

Exxaro Resources Limited

Fasken Martineau

FLSmidth Minerals (Pty) Ltd

Fluor Daniel SA (Pty) Ltd

Franki Africa (Pty) Ltd Johannesburg

Fraser Alexander Group

Glencore

Goba (Pty) Ltd

Hall Core Drilling (Pty) Ltd

Hatch (Pty) Ltd

Herrenknecht AG

HPE Hydro Power Equipment (Pty) Ltd

Impala Platinum Limited

IMS Engineering (Pty) Ltd

JENNMAR South Africa

Joy Global Inc. (Africa)

Leco Africa (Pty) Limited

Longyear South Africa (Pty) Ltd

Lonmin Plc

Ludowici Africa

Lull Storm Trading (PTY)Ltd T/A WekabaEngineering

Magnetech (Pty) Ltd

Magotteaux(PTY) LTD

MBE Minerals SA Pty Ltd

MCC Contracts (Pty) Ltd

MDM Technical Africa (Pty) Ltd

Metalock Industrial Services Africa (Pty)Ltd

Metorex Limited

Metso Minerals (South Africa) (Pty) Ltd

Minerals Operations Executive (Pty) Ltd

MineRP Holding (Pty) Ltd

Mintek

MIP Process Technologies

Modular Mining Systems Africa (Pty) Ltd

Runge Pincock Minarco Limited

MSA Group (Pty) Ltd

Multotec (Pty) Ltd

Murray and Roberts Cementation

Nalco Africa (Pty) Ltd

Namakwa Sands (Pty) Ltd

New Concept Mining (Pty) Limited

Northam Platinum Ltd - Zondereinde

Osborn Engineered Products SA (Pty) Ltd

Outotec (RSA) (Proprietary) Limited

PANalytical (Pty) Ltd

Paterson and Cooke Consulting Engineers (Pty) Ltd

Polysius A Division Of ThyssenkruppIndustrial Solutions (Pty) Ltd

Precious Metals Refiners

Rand Refinery Limited

Redpath Mining (South Africa) (Pty) Ltd

Rosond (Pty) Ltd

Royal Bafokeng Platinum

Roymec Tecvhnologies (Pty) Ltd

RSV Misym Engineering Services (Pty) Ltd

Rustenburg Platinum Mines Limited

SAIEG

Salene Mining (Pty) Ltd

Sandvik Mining and Construction Delmas (Pty) Ltd

Sandvik Mining and Construction RSA(Pty) Ltd

SANIRE

Sasol Mining(Pty) Ltd

Scanmin Africa (Pty) Ltd

Sebilo Resources (Pty) Ltd

SENET

Senmin International (Pty) Ltd

Shaft Sinkers (Pty) Limited

Sibanye Gold (Pty) Ltd

Smec SA

SMS Siemag South Africa (Pty) Ltd

SNC Lavalin (Pty) Ltd

Sound Mining Solutions (Pty) Ltd

SRK Consulting SA (Pty) Ltd

Time Mining and Processing (Pty) Ltd

Tomra Sorting Solutions Mining (Pty) Ltd

TWP Projects (Pty) Ltd

Ukwazi Mining Solutions (Pty) Ltd

Umgeni Water

VBKOM Consulting Engineers

Webber Wentzel

Weir Minerals Africa

University of PretoriaDepartment of Mining Engineering

me

dia

ch

ef4

988

As an engineer, you have a set of skills that not many people have. You’re the ultimate problem solver. You connect the dots that connect people and places, and make cities work. Not everyone can do what you do – and earn

#joinourtable

Your skill set is unique.Just like your earning potential.

Saimm 201504 apr

Documents

Transcript of Saimm 201504 apr