The functional design of surface mine haul roads - SAIMM · The maintenance aspect of haul road design cannot be considered separately from ... if local mine material can be used

▲169The Journal of The South African Institute of Mining and Metallurgy MAY/JUNE 2000

Introduction

Surface mining in South Africa accounts forover 70% of copper, 80% of ferrous metals and95% or 65 million tons of the land’s industrialmineral mining. In the coal mining industry in1997, over 40% or 106 million tons run-of-mine coal which requires inter alia, thetransport of raw coal from the pit to theloading or transfer point, was produced byopencast methods. In any surface miningoperation, the transport of ore, and to a lesserextent waste, is accomplished by large haultrucks running on haul roads that have, atbest, been empirically designed with little or norecognition of the consequences of inadequatedesign on cost per ton hauled, operationalefficiency or safety. Considering that truckhaulage costs can account for up to 50% of the

total costs incurred by a surface mine, it is ofparamount importance that these costs areminimized. This becomes all the more criticalas tonnage increases and larger haul trucksare deployed. Not only do the maintenancecosts of existing roads of inadequate designincrease but, vehicle operating andmaintenance costs also increase prohibitively.

The operating performance of a pavementcan be subdivided into three distinct designcategories as defined in Figure 1. Thestructural design of surface mine haul roadshas been addressed previously (Thompson andVisser)1,2. Just as important as the structuralstrength of the design, is the functionaltrafficability of the pavement. This is dictatedto a large degree through the choice,application and maintenance of wearing coursematerials. The current functional performanceanalysis methods are subjective and localisedin nature and any deterioration in pavementcondition is consequently hard to assess. Poorfunctional performance is manifest as poor ridequality, excessive dust, increased tyre wearand damage and an accompanying loss ofproductivity. The result of these effects is seenas an increase in overall vehicle operating andmaintenance costs.

The maintenance aspect of haul roaddesign cannot be considered separately fromthe structural and functional design aspectssince the two are mutually inclusive. Designand construction costs for the majority of haulroads represent only a small proportion of thetotal operating and maintenance costs. Whilstit is possible to construct a mine haul road thatrequires no maintenance over its service life,this would be prohibitively expensive—aswould the converse, the latter in terms ofoperating and maintenance costs. An optimal

The functional design of surface minehaul roadsby R.J. Thompson* and A.T. Visser†

Synopsis

The functional design of a pavement relates to the selection ofappropriate wearing course materials which optimise the traffica-bility of a pavement. Trafficability is dictated to a large degree bythe choice, application and maintenance of locally availablewearing course materials. Poor functional performance results indecreased safety and operational efficiency and increasing roadand vehicle maintenance requirements. An optimal functionaldesign will include a certain amount and frequency of maintenance(watering, grading etc.) and thus maintenance itself can beplanned, scheduled and optimized within the limits of required roadperformance and minimum vehicle operating and road maintenancecosts. Since no wearing course material selection guidelines,tailored to the specific needs of surface mines, were available, therewas the need to identify appropriate selection guidelines for therange of materials commonly encountered for road building onSouth African mines.

The aim of this paper is to present an overview of thefunctional design method for mine haul roads and, through ananalysis and quantification of existing wearing course materialperformance, to recommend appropriate wearing course materialselection guidelines. The functional design technique has beenapplied at a number of mines and is illustrated by an applicationcase study which demonstrates the potential reduction in totaltransportation operating costs associated with the optimalfunctional design.

* Department Mining Engineering, University ofPretoria, 0002.

† Department of Civil Engineering, University ofPretoria, 0002.

© The South African Institute of Mining andMetallurgy, 2000. SA ISSN 0038–223X/3.00 +0.00. Paper received Feb. 1999.

The functional design of surface mine haul roads

functional design will include a certain amount andfrequency of maintenance (watering, grading etc.) and thusmaintenance can be planned, scheduled and optimised withinthe limits of required road performance and minimum vehicleoperating and road maintenance costs. The major problemencountered when analysing maintenance requirements forhaul roads is the subjective and localised nature of theproblem—levels of functionality or serviceability being user-and site-specific. No guidelines exist for maintenancemanagement and scheduling for specific levels offunctionality, nor for the cost implications thereof, both interms of vehicle operating and road maintenance.

Geometric design refers to the layout and alignment ofthe road, in both the horizontal (curve radius, etc.) andvertical (incline, decline, ramp gradients, cross-fall, super-elevation etc.) plane, stopping distances, sight distances,junction layout, berm walls, provision of shoulders and roadwidth variation, within the limits imposed by structural,functional and maintenance design parameters. The ultimateaim is to produce an optimally efficient and safe geometricdesign and considerable data already exists pertaining goodengineering practice in geometric design. Suffice it to say thatan optimally safe and efficient design can only be achievedwhen sound geometric design principles are applied inconjunction with the optimal structural, functional andmaintenance designs (Thompson et al)3.

The aim of this paper is to present an overview of thefunctional design method for surface mine haul roads,specifically the development of wearing course materialselection criteria. The results of a functional performancesurvey of existing wearing course materials are presented,from which modelling and categorization of materialperformance and the derivation of optimal selectionparameters follows. The functional design technique hasbeen applied at a number of mines and is illustrated by anapplication case study which demonstrates the potentialimprovements in wearing course trafficability and theassociated reduced maintenance requirements.

Current State of Mine Haul Road Functional Design

Compacted natural gravel and crushed stone and gravelmixtures have been widely used in strip coal mines for haulroad construction, especially for base and wearing course

layers. The functional design of a haul road is the process ofselecting the most appropriate wearing course natural gravelor crushed stone and gravel mixtures that are commensuratewith safety, operational, environmental and economicconsiderations.

Most mines use cost per ton material moved as animmediate measure of haulage efficiency and in generalterms the contribution of haulage costs to total miningworking costs may vary between 10-50%. When consideringthose factors influencing the cost per ton hauled and thetruck/road interaction, those with most significant impact onthe functional performance of the road are rolling resistanceand trafficability. These two factors can have a significantimpact on both immediate and long term performance andcost. Recent work by Paige-Green4 has illustrated that thechoice of wearing course material is critical to optimalfunctional performance, not only in terms of rollingresistance and trafficability, but also in terms of numerousother defects which, in combination, will greatly affect usercosts or the cost per ton hauled.

Kaufman and Ault5 provide an early insight into haulroad functionality through a limited consideration of generalroad performance. They stated that the primarycharacteristics to be considered were road adhesion androlling resistance and the most practical constructionmaterials recognised were asphaltic concrete, crushed stoneor gravel and stabilized earth. The concept of functionalitywas not specifically introduced but rather alluded to in termsof some of the defects reported with these variousconstruction materials. Large rocks were seen to lead toexcessive tyre replacement costs, whilst excessive fines orpoor compaction led to dust problems. The impact of the dustproblem on haulage operations was related to excessivevehicle maintenance costs and reduced visibility. Dust controlby watering was associated with adhesion problems anderosion of the road surface, especially where poorlycompacted or unstabilized earth was employed. Inconclusion, they recommended crushed stone or good qualitynatural gravel as wearing course materials, together withspecifications for gradation and Atterberg limits.

Off-highway vehicles were until recently considered‘rugged’ and the quality and condition of a mine haul roadwere not sensitive factors in the application of surface minetransport. Recently, due to the increasing size and variationin the design of haul trucks and the changing economicclimate (altering the balance in the trade-off between haulroute quality, productivity and haul truck maintenance costs)more attention has been given to these factors. Kondo6

suggested that haul trucks are more sensitive to haul roadconditions when travelling at speed than are standardvehicles with a more responsive suspension. This has beenattributed, in part, to the generation of harmonics in thevehicle frame. Combined with high impact stresses producedby irregularities in the road, these vibrations can lead tometal fatigue, often manifest as failure of the goose neckconnections on bottom dump trucks. More recent work byDeslandes and Marshall7 recognised haul road surfacequality as being an important factor influencing structuralfatigue damage of haul truck frames. Recommendations weremade with regard to road maintenance and constructionpractices generally in geometric terms, but also including

▲

170 MAY/JUNE 2000 The Journal of The South African Institute of Mining and Metallurgy

Figure 1—Three components of a total haul road design strategy

reference to the reduction of road surface roughness whereladen travel occurs and at bends and intersections.

Wearing Course Materials

Work by Kaufman and Ault5 concerning the choice ofwearing course materials highlighted the most appropriatematerial characteristic design parameters, namely rollingresistance and adhesion. The most common wearing coursematerial for haul roads remains compacted gravel or graveland crushed stone mixtures. In addition to their low rollingresistance and high coefficient of adhesion, their greatestadvantage over other wearing course materials is thatroadway surfaces can be constructed rapidly and at relativelylow cost. As with structural designs, if local mine materialcan be used for construction, the costs are all the morefavourable. This cost advantage is, however, not apparent inthe long term if the characteristics of the wearing coursematerial result in sub-optimal functional performance.

Wearing course gravel characteristics have beendescribed additionally by Fung8, McInnes9, Atkinson andWalton10 and Taylor and Hurry11. These specify, in general,a good quality natural gravel or crushed stone. Whilst theselimited characteristics broadly define the suitability ofmaterials used for wearing course construction, they aregenerally lacking in their ability to predict the functionalperformance of haul roads. Numerous material selectionguidelines for unpaved public roads have been developed,including those of the Committee of State Road Authorities(CSRA) TRH1412 and TRH2013, illustrated in Figure 2, whichare based on performance or defect-related specifications.The suitability of these guidelines needed to be investigatedin terms of the required functionality of the haul road and theperformance of existing pavements. As a first step inisolating typical haul road functional performance defects, itis necessary to review ideal wearing course requirements.

Ideal Wearing Course Requirements

Whilst immediate measures are useful to a mine in assessingshort-term road functional performance, the definitiveeconomic analysis of haul road functionality is based on thecomparison of the benefits and costs of providing otheralternatives. Benefits are seen as overall cost savingsthrough increased productivity and reduced fuel, tyre and

maintenance costs. Improved functional performance impliesa reduction in pavement defects and since functionalperformance is based almost entirely on qualitative measure,it is useful to review typical unpaved road defects.

McInnes9 introduced the concept of wearing coursetrafficability in which a number of ideal requirements werealluded to. Building on and updating this approach, an idealwearing course for mine haul road construction can also beconsidered by modifying public unpaved road requirementsfollowing Netterberg14 and Paige-Green4;

• The ability to provide a safe and vehicle-friendly ridewithout the need for excessive maintenance.

• Adequate trafficability under wet and dry conditions.• The ability to shed water without excessive erosion.• Resistance to the abrasive action of traffic.• Freedom from excessive dust in dry weather.• Freedom from excessive slipperiness in wet weather.• Low cost and ease of maintenance.

The relative importance of the various characteristicscomprising overall functional performance needs to beassessed, as they apply to mining operations. The effect ofhaul road functional performance and maintenance on mineeconomics and safety is not well defined at present. However,it is clear that a strong relationship exists between roadstructural and functional performance and safe, economicallyoptimal mining operations.

For existing operations, which may not have optimallydesigned and maintained systems, the problem of identifyingexisting deficiencies, quantifying their impact and assigningpriorities within the constraints imposed by limited capitaland manpower, is problematic. Assessing the impact ofvarious haul road functional deficiencies in order to identifythe safety and economic benefits of taking corrective actionssuch as more frequent maintenance, regravelling orbetterment, is hampered by the lack of a problem-solvingmethodology which can address the complex interactions ofvarious components in a haulage system. This is reflected inthe fact that most surface mine operators agree good roadsare desirable, but find it difficult to translate this intoproposed betterment activities.

Functional performance survey

A functional performance survey of existing wearing coursematerials was undertaken to establish the trafficability ofexisting wearing course materials used for haul roadconstruction and to ascertain the validity and applicability ofpublished (unpaved public road) selection guidelines. Themost efficient approach entailed the analysis of a number ofin-service mine roads which covered the greatest range of themajor factors influencing functional performance. This wasachieved through use of a designed factorial experimentwhere a number of dependent (D) and independent (I)variables (factors) were analysed at various levels. Table 1summarizes these variables.

Choice of wearing course materials was limited to thoseweathering products (as defined by Weinert15) typicallyencountered in strip coal mining areas and includedpedocretes, argillaceous, arenaceous, basic crystalline and



Figure 2—Wearing course gravel material selection guidelines forpublic roads (after CSRA TRH2013)


acid crystalline, together with mixtures of these. Climate wasdiscounted as an independent variable since most strip coalmines were situated within the same physiographical (N-value) region (following Weinert15) and thus the weatheringproducts used as levels for the independent variable ofwearing course material were unique and limited to thatphysiographical region (N=2 to N=5). Sixteen test sites whichcovered the greatest number of factors were established andthe functional performance of these sites was monitored overa 12-month period.

Since existing mine haul road functional performanceanalysis methods were subjective and localized in nature, anew visual assessment methodology was developed, basedon recorded defects on unpaved public roads13,16 and theStandard Visual Assessment Manual for PavementManagement Systems17, suitably modified to accommodatethe requirements of mine haul road operators.

The condition of the pavement was considered from thepoint of view of the road user and incorporated appraisal interms of those characteristics that affect the quality of travel.The assessment is entirely qualitative and to reduce theamount of subjectivity involved, distress characteristics arerecorded in terms of degree and extent. The degree of aparticular type of distress is a measure of its severity. Degreeis indicated by a number where Degree 1 indicates the firstevidence of a particular type of distress and Degree 5 verysevere distress. The extent of distress is a measure of howwidespread the distress is over the test section. Extent is

indicated by a number where Extent 1 indicates an isolatedoccurrence and Extent 5 an extensive occurrence of aparticular type of distress. The descriptions of extent are notassociated with a specific functional defect and the rating ofextent was applied only to those defects related to thewearing course material. Defects relating to formation andfunction (drainage, erosion and skid resistance) are analysedonly in terms of degree. The general descriptions of degreeand extent are given in Addendum 1.

By summing the product of each wearing course defectdegree and extent, formation and function (degree only), atotal functional defect score could be found for each surveyedsection of road. Figure 3 shows how individual and totalfunctional defect scores were characterised for one mine testsite.

Modelling haul road functional performance

A functional performance assessment of the mine test sites interms of individual defect score variations with time, doesnot enable predictions to be made regarding the effect oftraffic volume, wearing course material type, materialproperties or maintenance intervals on the functionalperformance of the haul road; nor can the propensity of aparticular material property to contribute to a particular haulroad defect be analysed.

The development of a predictive model for defect scoreprogression with time is critical both in terms of thedevelopment of a maintenance design model for mine haulroads and as a measure of pavement condition that can bedirectly associated with vehicle operating costs. The defectscore at a particular point in time is a reflection of the type ofwearing course material used and its engineering properties,the level of maintenance, season and traffic volumes. Whilsta slight seasonal fluctuation in functionality was observed,the comparatively frequent watering and blading activities onmine haul roads obscured any significant seasonalvariations. Thus in the analysis of the effect of maintenanceon defect scores which follows, a combination of defectscores and maintenance interval data over both seasons wasadopted and seasonality ignored.

A defect score progression model was hypothized asshown in Figure 4, based on four distinct traffic,maintenance and wearing course material interactions;

(A) Immediately following maintenance there will be atraffic induced reduction of loose material and dustdefect scores such that the post-maintenance defectscores decrease overall.

(B) A minimum defect score will be acheived where theprogression changes from decreasing to increasing.

(C) The increasing traffic volumes and dynamic loadingsimposed on the road, together with an increase inabrasion result in an increase in the defect scores untiltraffic speed slows and wheel paths change to avoiddamaged sections.

(D) At this point the defect score would remain essentiallyconstant.

In the selection of a model for defect score progression, apiecewise combination of two exponential curves was chosento represent the decreasing and increasing rate of change of

▲


Figure 3—Typical functional performance assessment results over a 12-month monitoring period

Table I

Summary of dependent and independent variables and measuring systems

Variable Measurement System

Traffic volume (D) Production statisticsWearing course material (I) Laboratory classificationDays since last maintenance (D) Mine recordsMoisture conditions of Functional assessment surface layer (D) methodologySurface drainage conditions (D) Functional assessment methodologySurface erosion (D) Functional assessment methodologyFunctional performance (D) Functional assessment methodologyRoad geometry (I) Survey plansRut depth and corrugation geometry (D) Straight edge

defect score with time (or traffic volume). Using a logarithmictransformation of defect scores, a regression function wasdeveloped based on a linear combination of the independentvariables for the rate of defect score decrease (LDDD) andincrease (LDDI). In addition, an expression for the minimumdefect score after maintenance (DSMIN) was sought togetherwith its location in terms of days since maintenance (DM),both assumed to be a linear combination of the independent

variables, as illustrated in Figure 4. The rate of change indefect scores was calculated over a maintenance cycle andthese values used as the dependent variables in a multiplecorrelation analysis in order to identify the significant factorsaffecting defect progression. The independent variables listedin Table 2 were evaluated.

The following models were proposed to describe thedefect score progression18;

LDD = 1,261 + DM(0,000121.CBR.KT-0,02954.GC + 0.009824.SP.DR) [1]

LDDI = 1,7929 + D(0,002276.KT + GC(O,01029.DR - 0,010887)) [2]

DSMAX = 35,0249 + 26,7827.M - 0,5672.KT +1,6508.GC + 0,4464.SP - 10,9393.PI [3]

Since no statistically reliable model could be derived forDM (days since last maintenance), recourse was made to themodal value of DM=2 days to locate the position of DSMIN.The regression of DSMIN on the independent variablesrendered the following model:

DSMIN = 37,9146 - 0.15799.KT + 12.7093.M +1,3836.GC - 0,08752.SP [4]

These predictive models, together with the assumption ofthe modal time since maintenance for the location of theminimum defect score, enable the functional response of amine haul road to be modelled in terms of rates of decreaseand, more importantly, rates of increase in defect score withtime and traffic volumes. Figure 5 compares the predictionmodel with typical mine site defect score progression data,using Equations [1] and [2] bounded by Equations [3] and[4], whilst Figure 6 illustrates the effect of traffic volume (ktper day) variation on defect score progression for oneparticular set of material property and minimum defect scorevalues. As can be seen, if an intervention level (or maximumacceptable defect score) of 70 is used, given a monthlyproduction of 230 000t a maintenance interval of 13 days isadvocated. When the monthly tonnage hauled increases to 1150 000t a maximum maintenance interval of seven days isimplicated for the given wearing course material parametersused in the model.

Whilst a model of defect score progression is useful topredict and compare the functional performance of aparticular wearing course material (in terms of its



Figure 4—Schematic illustration of the development of functionaldefects on a haul road

Figure 5—Estimation characteristics of prediction model for rate ofincrease in functionality defect score

Table II

Independent variables used in the Defect ScoreProgression Model

Independent DescriptionVariable

KT Average daily tonnage hauled (kt)

M Wearing course material type;0=ferricretes1=mixtures of materials

P075 Percentage of material passing 0,075mm sieve

DR Dust ratio, defined as;

where P425=percentage of material passing the 0,425mm sieve

PI Plasticity index

CBR 100% Mod. California Bearing Ratio of wearing course material

GC Grading coefficient, defined as;

where P265=percentage of material passing the 26,5mm sieveP2 =percentage of material passing the 2,0mm sieveP475=percentage of material passing the 4,75mm sieve

SP Shrinkage product, defined as;

where LS= Bar linear shrinkage

PL Plasticity limit

D Days since last maintenance

DM Days between last maintenance and minimum cycle defect score

DSMIN Minimum defect score in cycle

P075P425

(P265 - P2)xP475100

LSxP425


engineering properties and the effect of traffic volume on theroad) with the acceptability requirements of the road-user, itis also useful to determine the propensity of a particularmaterial to form specific functional defects, throughconsideration of the material’s engineering properties. One ofthe major objectives of defect score prediction was to comparethe proclivity of various types of materials to deteriorate overtime. Results of the study are presented by Thompson18,from which it is seen that, in keeping with other suchanalyses, such deterioration models showed low R-squaredvalues with statistically significant correlations by virtue ofthe large sample size, with the exception of materialproperties which were defined over a much smaller inferencespace than the previous studies, which may limit theapplicability of the models where materials significantlydifferent from those encountered during testwork are to beassessed.

Despite these limitations it may be concluded thatwearing course geotechnical properties, especially the particlesize distribution and plasticity, together with traffic volume,are the most important material parameters with regard tothe prediction of deterioration progression.

Acceptability criteria for haul road functionalperformance

Functional defect progression models and wearing courseparameter properties cannot be used to assess theapplicability of current material selection guidelines forunpaved road construction without some measure ofacceptability limits for the various material, formation andfunctional defects previously analysed. In this assessment ofacceptability the following areas were assessed;

1. Road-user assessment of desirable, undesirable andunacceptable limits of performance for each functionaldefect previously identified

2. The impact of these defects on the safety andeconomics of a mining operation.

The first area was assessed by using the standarddescriptions of degree and extent for each wearing course,formation and function defect referred to in Addendum 1.Respondents classified the lower limit of desirability and the

upper limit of unacceptability for each defect.The second area was quantified using an approach

developed by the United States Bureau of Mines MineralsHealth and Safety Technology Division (USBM)19, suitablymodified to accommodate those conditions or characteristicspreviously identified as important in the functionalperformance of wearing course materials. Respondents wereasked initially to decide if a given condition or characteristiccan affect either the truck, the tyres or the operation’sproductivity and the degree to which this occurs was scored.The safety impact was estimated by scoring the accidentpotential of each condition and characteristic. Respondentswere asked to consider each item in a broad sense, ie. scoringin terms of its impact on average or typical daily operatingconditions on the haul road. Full details are presentedelsewhere18.

In order to quantify these parameters various mines wereinvited to complete a functionality rating questionnaire inwhich both production and engineering personnel hadinputs. To further quantify the limits of acceptabilityrespondents were also invited to categorize each defect interms of its impact on the components of the hauling system,namely the truck, tyres or operation. In addition, haul truckmanufacturers were also invited to respond so as to qualifymine operators’ functionality requirements with those of themanufacturers. The results of the questionnaires representedover 169 years of operating experience with mine haul roads.In addition, the data was representative of nearly 70% of thetotal coal tonnage transported on mine haul roads in SouthAfrica.

Road-user Assessment of Functional PerformanceLimits

Figure 7 presents a summary of the responses in terms of theaverage acceptability limits for each functional defect, uponwhich is superimposed results from one mine test siteshowing the average range of defect score variation. Fromthe figure it is evident that potholes, corrugations, loosematerial, dustiness, loose stones and wet and dry skidresistance are considered undesirable when the defect scoreexceeds 3-4 (for skid resistance wet and dry, the assumedextent of 5, representing conditions affecting the whole road,artificially exaggerates the lower limit of desirability). It is

▲


Figure 6—Effect of increasing traffic volume on defect scoreprogression rates and associated maintenance interval

Figure 7—Range and annual average values for mine test site functionalperformance in relation to established performance limits

also seen that the defects of loose material, dustiness andstoniness (fixed and loose) seldom exhibit desirableperformance, demonstrating the need for improved wearingcourse material selection parameters.

Whilst establishing the acceptability limits for eachfunctional defect provides an insight into the ideal levels ofperformance expected for a wearing course material, anappraisal of the impact of these defects on the haulingoperation is necessary to qualify the extent to which defectsmay affect economics and safety.

Road User Assessment of Defect Impact andAccident Potential

The impact of a particular functional defect was quantified onthe questionnaire using the impact ranking scale whichreflects that common functional defects, resulting from a lessthan optimal wearing course material (or maintenanceprogram) are not catastrophic. Results were compiled foraverage annual functionality of a mine’s road. The resultsechoed the road user assessment of functionality, withdustiness and wet skid resistance perceived as being primarydefects affecting the operation, each accounting for an 11-15% reduction in productivity. Impacts on the truck centredon the defects of potholes, corrugations and skid resistance,accounting for downtimes of less than one shift. Impacts onthe tyre were similar, including, in addition, loose materialand stoniness, which account for a 5-10% decrease in tyrelife. Cracks were considered almost irrelevant in terms oftheir impact on the hauling components.

The accident potential was determined in a similarfashion, in this case irrespective of which component of thehauling it affected. The accident-potential scale assignedprobabilities that the impact will occur, following USBMguidelines. The average accident potential scores are given inFigure 8, from which it is seen that the defects of dust andskid resistance are the functional factors most likely to causeaccidents. The formational defect associated with drainage onthe side of the road was recognised as having a high accidentpotential, which also implicates the functional defect of loosematerial or in more general terms, wearing course materialerodibility, in accidents.

The acceptability criteria were derived by combiningdefect impact and accident potential, thus identifying and

ranking aspects of functional performance that should enjoypriority when considering opposing selection criteria. Themethodology for ranking defects involved summing theproduct of impact and accident potential for each defect togive a cumulative score for each defect. The product ofcumulative impact score, cumulative defect score andaccident potential then gave the overall ranking of the defect.Figure 9 shows the actual ranking scores from which it isevident that wet skid resistance has the greatest impact andaccident potential, followed by dustiness, dry skid resistanceand loose material. The formational aspects of drainage, bothon and off the road, are also significant in the ranking offunctional defects.

Derivation of wearing course material specifications

The derivation of wearing course material selectionguidelines was based on the identification, characterizationand ranking of haul road functional defects, as previouslydiscussed. Prior to the development of the specifications, areference framework was developed within which suitablespecifications should fall.

Two approaches were adopted in deriving suitablespecifications. Initially, the important material propertyparameters controlling both functional performance andindividual defect score progression rates, were assessed inrelation to the overall haul road functional performanceclassification in order to identify likely trends and limits forindividual parameter values. Secondly, the suitability of thewearing course material selection guidelines proposed inTRH2013 as a source for mine haul road materialspecification were analysed. This enabled specifications to bedeveloped which, whilst stipulating individual parameterlimits, also have predictive capabilities which contribute to anunderstanding of the consequences when materials outsidethe specified ranges are used as wearing course materials.

Haul Road Wearing Course SpecificationRequirements

The development of suitable specifications for wearing coursematerials should ideally encompass both individual wearingcourse material parameter specifications and a broader



Figure 8—Accident potential of haul road functional defects

Figure 9—Ranking of haul road functional defects from the combinedaccident potential and impact scores


indication of likely functional defects associated withdeparture from the established guidelines. Ideal specificationrequirements following Paige-Green4 are described below,modified for mine haul road operation and design;

(i) They should be simple with as few requirements or testmethods as possible.

(ii) They should be inexpensive, reproducible, necessitatethe minimum of sophisticated equipment and operatortraining.

(iii) The limits should not be restricted to a narrow range ofa significant property, but must also be adequatelycomprehensive in order to recognise and reject unsuitable materials.

(iv) The specifications should not be unduly restrictive andaccommodate mine haul road construction cost andmaterial volume considerations. An indication of the likely consequences of employing local minematerial which falls outside the recommendedparameter range is useful.

Material selection guidelines must thus take cognisanceof the road-user functional performance requirements andthe limitations imposed by material availability, cost andvolume considerations. Since some defect/material propertytrade-off is inevitable when local mine construction materialsare used, it is important to establish a performance rankingsystem in which material properties associated with criticaldefects enjoy priority over less significant defects, especiallywhere opposing material selection parameters areencountered.

In the road-user assessment of defect acceptabilitycriteria, a number of defects which critically affectfunctionality were identified and considered to represent thecritical defects which should be addressed in the derivation ofmaterial specifications. Limits of acceptability were alsodetermined in terms of desirable, undesirable andunacceptable levels of defect score. These acceptability limitsare categorized in Table 3. Since those mine sites exhibiting areasonable level of functional performance were notadequately differentiated from noticeably poorer sites, afurther sub-division of performance classification wasdeveloped in order to adequately differentiate between thesesites and defects (upper B1 and C1) and lower (B2 and C2)sub-groups.

Using these acceptability levels (A-C2) it was possible toinvestigate material property and performance relationships,both in terms of overall test site performance and the

individual defect contribution to overall performance. Inaddition, the utility of existing guidelines was assessed interms of the extent to which such guidelines accommodateand reflect the various overall and individual defect rankings.

Assessment of Material Property and PerformanceRelationships

From the statistical analysis and modelling of overall roadand individual defect functional performance, the materialparameters of plasticity and grading were identified asprimarily controlling the functional performance of a haulroad. Specifically, the grading coefficient (GC), dust ratio(DR), shrinkage product (SP), plasticity index (PI) and liquid(LL) and plastic (PL) limits were found to contribute to therate of defect score increase or decrease. Accordingly, thesewearing course material property values were classifiedaccording to the overall road or individual defect acceptabilitylevels (between A-C2) in an attempt to determine wearingcourse material property limits.

The relative significance of each critical defect analysed isimportant when an overall classification of performance andassociated material properties is attempted. Greaterimportance should be attached to those material propertiesassociated with the more critical functional defects. This wasachieved by incorporating the defect weighting factorsderived from the assessment of acceptability criteria anddefect accident potential. In this manner, overall performancewas related to the criticality of a defect, those with high-ranking scores contributing proportionally more to the overallranking.

Derivation of Selection Guidelines

The TRH2013 wearing course material selection guidelineswere developed from functional performance considerationsof unpaved public roads as described by Paige-Green4 in hisdevelopment of the guidelines, and thus were selected as themost suitable basis for mine haul road wearing coursematerial selection. Figure 10 illustrates the location of eachmine test site in terms of shrinkage product (SP) and gradingcoefficient (GC) values and the overall functionalperformance ranking (A-C2). From the figure it is clear thatthe majority of the mine sites lie within the recommendedmaterial selection limits (a’b’c’d’ in Figure 2). Of those siteslying outside the recommended limits (R1, R3, N1 and N3),only sites R1 and N3 exhibited the predicted excessiveravelling and corrugation defects.

▲


Figure 10—Overall mine site functional performance classification inrelation to TRH20 specifications

Table III

Acceptability limits for critical functional defects

Acceptability Acceptability limits of Description of Operabilitylevel defect score acceptability limits limits

• Corrugation • Wet skid• Loose resistancematerial • Dry skid• Dustiness resistance• Loosestoniness

A <5 <5 Desirable OperableB1 5 - 7 5 - 10 Undesirable (upper) Operable

B2 8 - 10 11 - 15 Undesirable (lower) Marginal

C1 11 - 17 16 - 20 Unacceptable (upper) Inoperable

C2 >17 >20 Unacceptable (lower) Inoperable

The overall functional classification shown in Figure 10reveals that most of the test sites exhibited undesirable(lower B2) to unacceptable (upper C1) performance, albeitoperable. Of those sites lying outside the recommended limits(R1, R3, N1 and N3), only sites R1 and N3 exhibitedunacceptable performance (upper C1) and were excludedfrom the modified recommended selection range for minehaul road wearing course materials. Since all the mine testsites lie within or close to the recommended materialselection limits, points outside this area are ideally requiredto confirm the selection parameters. It is apparent that theTRH20 specifications provide a suitable base for materialspecification and in addition, reflect the typical defectassociated with departure from the specifications. If the threemost critical defects are considered in the light of the TRH20specifications, it appears that road-user preference is formuch reduced wet skid resistance, dust and dry skidresistance defects, at the expense of an increase in the otherdefect scores. This alters the focus point of the specificationsto an area bounded by a grading coefficient of 25-32 and ashrinkage product of 95-130 in which the overall and

individual defect performance is optimised (Area 1).Extending this region to encompass poorer (but neverthelessoperable) performance enables an additional area (Area 2) tobe defined as given in Table 4 and Figure 11.

The suitability of the modified TRH20 technique ofwearing course material selection based on gradingcoefficient and shrinkage product parameters has beenestablished, together with a range over which optimalperformance is assured. This approach should be temperedthrough the consideration of the other material propertiesidentified as important in functional performance, but notdirectly assessed in the TRH20 technique. Table 4 presents asummary of these property limits, derived from the statisticalanalysis of wearing course material defect progression rates.

Wearing course material specifications associated withthe structural design of mine haul roads have beenproposed1,2 in terms of TRH1412. In addition, haul roaddesign work often specifies material requirements in terms ofTRH14 and it is thus useful to consider the equivalence ofthe latter to the modified specifications established in Table4. Material available on South African surface coal mines forthe construction of the wearing course, is derived fromborrow pits comprising, generally, ferricrete or similarmaterials and is classified (following TRH14) as G4-G7.Using G4 material specifications, a location range can bedetermined for the equivalent TRH20 specification. The rangeof grading coefficient lies between 12 and 52 and that ofshrinkage product between 30 and 90 (for the full allowablegrading variability specified in TRH14). Whilst the gradingcoefficient parameter encompasses materials liable to erodeand to ravel, the shrinkage product lies in the range ofmaterial types associated with ravelling and corrugation only.If poorer quality materials are considered (G5-G7), althoughno specific grading requirements are given in TRH14, theincrease in allowable linear shrinkage should improve thelocation range of these materials in terms of the optimumhaul road material selection parameter ranges given in Table4. It is clear that TRH14 alone does not provide sufficientdifferentiation between material parameters and haul roaddefects to enable it to be used as a specification for mine haulroad wearing course material selection.

Wearing course material selection case study

The wearing course material as presently used at a miningoperation (M1) is depicted in Figure 12 in terms of theproposed selection guidelines, together with two othermaterials that could be used for blending with the currentwearing course (BS and ASH). When the wearing coursematerial is considered in relation to the recommendedoptimal material selection ranges, besides being prone todustiness when dry, it may also present a skid resistancehazard when wet (typically a slippery surface after rain orwatering).

If the wearing course material is modelled in terms ofhow its functional defect score progression varies with theinterval between maintaining the road, the unsuitability ofthe wearing course material is evident from the high defectscore and rate of deterioration, as shown in Figure 13. Thesedefect scores and optimum maintenance intervals maychange, depending on the specific model parameters adopted.



Figure 11—Optimum mine haul road wearing course material selectionranges (1, 2) and general trends of increasing functional defect scores

Table IV

Recommended parameter ranges for mine haulroad wearing course material selection

Material Parameter Range Impact on FunctionalityMin Max

Shrinkage Product 85 200 Reduce slipperiness but prone toravelling and corrugation

Grading Coefficient 20 35 Reduce erodibility of fine materials,but induces tendency to ravel

Dust Ratio 0,4 0,6 Reduce dust generation but inducesravelling

Liquid Limit (%) 17 24 Reduce slipperiness but prone todustiness

Plastic Limit (%) 12 17 Reduce slipperiness but prone todustiness

Plasticity Index 4 8 Reduce slipperiness but prone todustiness and ravelling

CBR at 98% Mod AASHTO 80 Resistance to ersosion, rutting andimproved trafficability

Maximum Particle size (mm) 40 Ease of maintenance, vehicle-friendly ride and no tyre damage


In this case a daily production of 15kt was used; as tonnagehauled increases, the rate of defect score increase alsoincreases, as illustrated in Figure 6.

In order to improve the functional performance of thewearing course at the mine, some blending of materials wasnecessary. Using the recommended material specifications, inconjunction with the defect score progression model, it waspossible to determine the optimal mix associated with thebest functional performance. In this case, 40% BS and 30%ASH was added to the original wearing course, to achieve amix within the specified selection range. Figure 13 shows themuch reduced predicted functional defect progression rate forthe new wearing course when subjected to the same trafficvolumes as previously. If maintenance is delayed, the defectscore eventually stabilises at a much lower value than for theoriginal wearing courses. In terms of the optimalmaintenance interval required for the roads, it may be seenthat the materials are not as sensitive to over- or under-maintenance as the original materials.

For the new wearing course material mix, a maximummaintenance interval of two days is recommended, thismaintenance interval giving the lowest average defect score,

commensurate with maximized safety and productivity. Amaintenance interval of 4 days would give the sameminimum defect score as the unimproved wearing coursewhen a 2-day maintenance regime is applied. Whenconsidering total haulage costs, the cost componentsassociated with operating the haul truck (fuel, tyres,maintenance parts and labour) and maintaining the road(grader and water-car operating costs) need to be analysedsystematically in conjunction with the wearing coursematerials.

Conclusions

Functional design aspects refer to the ability of the haul roadto perform its function, i.e. to provide an economic, safe andvehicle-friendly ride. This is dictated to a large degreethrough the choice, application and maintenance of wearingcourse materials. The commonality between typical defectsreported for unpaved public roads and the functionalityrequirements for mine haul roads, indicated that existingspecifications for unpaved public road wearing courseconstruction materials would form a suitable base for thedevelopment of specifications for mine haul roads. Aqualitative functional performance assessment methodologywas developed, based on typical haul road wearing course,formation and function defects, in order to assess the utilityof established performance-related wearing course selectionguidelines and as a basis for revised functional performanceparameter specifications.

From the functionality assessment exercise it was foundthat the major haul road functional defects encountered weredustiness, loose material and fixed and loose stoniness. Astatistical analysis of deterioration and maintenance effectsassociated with these key defects revealed that wearingcourse material properties, especially grading and plasticityparameters, together with traffic volume, could be used toadequately model the functional performance of these keydefects. However, the applicability of the model was limitedby the relatively small inference space of the data and wherematerials are encountered which differ significantly fromthose assessed during the test work, judgement and careshould be exercised when applying the predicted results. Indetermining suitable wearing course material selectionguidelines, this work confirmed qualitative observations thatgrading and plasticity parameters would adequatelyanticipate the functional performance of a wearing coursematerial.

The development of acceptability criteria for haul roadfunctionality fulfilled a requirement for a structured approachto the assessment of mine haul road functionality. Inaddition to assigning acceptability ranges to each type ofdefect, the impact and accident potential of each defect wascategorized and ranked according to the total impact andaccident potential on the components of hauling, namelyoperation, truck and tyre. It was concluded from the rankingexercise that wet skid resistance, dustiness, erodibility andravelling and corrugating are critical defects which controlthe functionality of mine haul roads and that theconsequences, in terms of the possible generation of thesedefects, should therefore be incorporated into any suitableselection criteria established for mine haul road wearingcourse materials.

▲


Figure 13—Predicted variation of functionality for original and optimisedwearing course materials

Figure 12—Location of case-study mine 1 wearing course material incomparison to recommended selection range

The derivation of wearing course material selectionguidelines was based on the identification, characterizationand ranking of haul road functional defects. A referenceframework was developed within which suitablespecifications should fall, based on an assessment of therequirements of good specifications in the light of functionaldefect ranking and acceptability limits. The TRH20 wearingcourse material selection guidelines were found to be asuitable source for the specification of mine haul roadwearing course material parameter requirements. A revisedrange of parameters was derived based on the road-userpreference for much reduced wet slipperiness, dustiness anddry skid resistance defects. The specification included theparameters of shrinkage product and grading coefficient andlimits of 85-200 and 20-35 respectively, were proposed. In

addition, from analysis of the range of material propertyparameters assessed and their association with the functionaldefects analysed, parameter ranges were additionallyspecified for density, dust ratio, Atterberg limits, CBR andmaximum particle size. By analysing the trends evident inthe individual defect rankings, the predictive capability of thespecification was enhanced by depicting the variation infunctional defects which would arise when departures aremade from recommended parameter limits.

AcknowledgmentsThe permission of Anglo Coal to publish this paper, togetherwith their financial support for the research is gratefullyacknowledged, as is the assistance and co-operation receivedfrom mine personnel.



Addendum 1 Classification of haul road functionl defects

Characteristic Description

Degree 1 Degree 2 Degree 3 Degree 4 Degree 5

Potholes Surface is pock-marked Potholes 50-100mm diameter. Potholes 100-400mm diameter Potholes 400-800mm diameter, Potholes >800mm diameter,<50mm diameter. and influence riding quality. influence riding quality and influence riding quality and

obviously avoided by most require speed reduction ortotal vehicles. avoidance.

Corrugations Slight corrugations, difficult to Corrugations present and Corrugations very visible and Corrugations noticeable in haul Corrugations noticeable in haulfeel in light vehicle. noticeable in light vehicles. reduce riding qual;ity noticeably. truck and causing driver to ruck and causing driver to

reduce speed. reduce speed significantly.

Rutting Difficult to discern unaided, Just discernible with eye 20- Discernible, 50-80mm. Obvious from moving vehicle, Severe, affects direction<20mm. 50mm. >80mm. stability of vehicle.

Loose material Very little loose material on Small amount of loose material Loose material present on road Significant loose material on Considerable loose material,road, <5mm depth. on road to a depth of 5-10mm. to a depth of 10-20mm. road to a depth of 20-40mm. depth >40mm.

Dustiness Dust just visible behind vehicle. Dust visible, no oncoming Notable amount of dust, Significant amount of dust, Very dusty, surroundingsvehicle driver discomfort, good windows closed in oncoming windows closed in oncoming obscured to a dangerous level.visibility. vehicle, visibility just vehicle, visibility poor.

acceptable, overtaking difficult.

Stoniness - fixed in Some protruding stones, but Protruding stones felt and Protruding stones influence Protruding stones occasionally Protruding stones requirewearing course barely felt or heard when heard in light vehicle. riding quality in light vehicle require evasive action of light evasive action of haul truck.

travelling in light vehicle. but still acceptable. vehicle.

Stoniness - loose on Occasional loose stone Some loose stone, 2-4/m2 Loose stone 4-6/m2, occasional Considerable loose stone on Large amounts of loose stoneroad (>75mm diameter), <2/m2 discomfort felt. surface, >6/m2, reduce riding causing significant reduction in

quality. riding quality.

Cracks - longitudinal Faint cracks discernible when Distinct, mostly closed, easily Distinct, mostly open, Open cracks, >3mm separation Extensive open cracks, >3mmsurface cleaned. discernible when walking. discernible from vehicle. or wide open cracks >10mm separation together with

separation, in travelling lanes. secondary cracks or extensivewide open cracks >10mmseparation, in travelling lanes.

Cracks - slip Faint cracks discernible when Distinct, mostly closed, easily Distinct, mostly open, Open cracks, >3mm separation Extensive open cracks, >3mmsurface cleaned. discernible when walking. discernible from vehicle. or wide open cracks >10mm separation together with

separation, in travelling lanes. secondary cracks or extensivewide open cracks >10mmseparation, in travelling lanes.

Cracks - crocodile Very faint cracks in wheel Faint cracks discernible when Distinct cracks upto 2mm wide, Open cracks (>2mm) with Open cracks with severepath. walking, closed. no apparent deformation. some deformation and/or deformation and/or spalling of

spalling of cracked areas. edges.

Skid resistance - wet Wearing course material of Wearing course strength and PI Wearing course strength low, Wearing course strength low, Wearing course strength verygood quality, road properly acceptable, road cambered PI fairly high, unsatisfactory PI high, water standing on low, PI very high, road verycambered, little loose material loose material acceptable. camber and loose material. surface when raining, loose slippery when wet, loosepresent. material influences skid material reduces skid

resistance significantly. resistance unacceptably.

Skid resistance - dry Wearing course material of Wearing course strength and PI Wearing course strength low, Wearing course strength low, Wearing course strength verygood quality, road properly acceptable, road cambered PI fairly high, unsatisfactory PI high, loose material low, PI very high, loosecambered, little loose material loose material acceptable. camber and loose material. influences skid resistance material reduces skidpresent. significantly. resistance unacceptably.

Drainage on road Very little water accumulates Shallow depressions may retain Water may be retained in ruts Water retained over a Water ponding on road toon road, no surface erosion is water for a limited time, most and potholes, some surface significant portion of the road, depths >50mm and erosionevident. water drains away rapidly. erosion evident. surface erosion <50mm deep channels deeper than 50mm.

in channels.

Drainage at roadside Side drains very effective, well Slightly irregular, some loose Drains irregular in shape, Drains irregular or eroded and Side drains deeply eroded orshaped with no obstructions. debris or occasional erosion, blocked or eroded, road above blocked over >25% road non existent along 75% of

road well above side drain side drain level. length, road and side drain at road length or road surfacelevel. same elevation. below side drain.


References

1. THOMPSON R.J. and VISSER A.T. An overview of the structural design ofmine haul roads. Jnl. South African Institute of Mining and Metallurgy,Jan/Feb 1996, Johannesburg, South Africa. pp. 29-37.

2. THOMPSON R.J. and VISSER A.T. Pavement design for ultra-heavy haultrucks. 5th International Conference on the Bearing Capacity of Roadsand Airfields, BCRA ’98, 5-8 July 1998, Trondheim, Norway, pp. 649-661.

3. THOMPSON R.J., VISSER A.T., SMITH R.A.F. and FOURIE G.A.F. Final report forproject SIMOT 308, Safety in Mines Research Advisory Committee, OtherMines Sub-committee, Department of Minerals and Energy, May 1997,Pretoria, South Africa.

4. PAIGE-GREEN, P. The influence of geotechnical properties on theperformance of gravel wearing course materials. PhD dissertation,University of Pretoria, 1989. South Africa.

5. KAUFMAN, W.W. and AULT, J.C. The design of surface mine haul roadsmanual. USDOI Information circular 8758, USBM, 1977, pp. 2–7, 19–30.

6. KONDO, S. Relation between haul road and damage of off-highway dumptrucks. Transactions, Society of Automotive Engineers Inc, paper840778, 1984. USA.

7. DESLANDES, J.V. and MARSHALL, G.A. Factors affecting structural fatiguedamage of mine haul trucks. Australian IMM/Australian IEE Newmancombined group large open pit mining conference, 1986. Mt Newman,Australia.

8. FUNG, R. (Ed.). Surface mine haul road design. Surface mine engineeringtechnology; engineering and environmental aspects, Noyes Data Corp.,1981. USA. pp. 44–57.

9. MC INNES, D.B. The selection of gravels for use on unsealed access roads.Australian IMM conference on off-highway truck haulage, 1982, Mt

Newman, Australia, pp. 23–31.

10. ATKINSON, T. and WALTON, G. Design and layout of haul roads for surfacemines. Institute of Mining and Metallurgy Surface Mining Symposium,1986, Bristol, UK. pp. 369–381.

11. TAYLOR, P.F. and HURRY, P. Design, construction and maintenance of haulroads. SAIMM Symposium on design, planning and operation of openpitand strip mines, Pretoria, 1987, South Africa. pp. 139–149.

12. COMMITTE OF STATE ROAD AUTHORITIES. Technical recommendations forhighways (TRH14): Road construction materials. CSRA TRH14, 1985,Pretoria, South Africa.

13. COMMITTE OF STATE ROAD AUTHORITIES. The structural design, constructionand maintenance of unpaved roads. CSRA TRH20, 1990, Pretoria, SouthAfrica.

14. NETTERBERG, F. Wearing courses for unpaved roads. NITRR technical noteTS/33/85, CSIR, 1985, Pretoria, South Africa.

15. WEINERT, H.H. The natural road construction materials of southern Africa.Academica Press, 1980, Cape Town, South Africa.

16. PIENAAR, P.A. and VISSER, A.T. The evaluation and prioritisation ofunpaved road maintenance in developing areas. Civil Engineer in SouthAfrica, Vol. 34, No 2. 1992.

17. COMMITTE OF STATE ROAD AUTHORITIES. Pavement management systems:Standard visual assessment manual. CSRA TMH9, 1990, Pretoria, SouthAfrica.

18. THOMPSON, R.J. The design and maintenance of surface mine haul roads.PhD Thesis, University of Pretoria, 1996, South Africa.

19. UNITED STATES BUREAU OF MINES (USBM). Study of mine haulage roadwaysurface safety hazards. USDOI, Mines, Minerals, Health and SafetyTechnology Division, Bumines OFR 5-83., 1981, USA. ◆

Documents

The functional design of surface mine haul roads - SAIMM · The maintenance aspect of haul road design cannot be considered separately from ... if local mine material can be used