MMS MAG NOVEMBER 2008

Real-Time Condition Monitoringfor Mobile Mine EquipmentBy Harinder Hara and A.J. Bartkoske

16

SOLUTIONS MODULAR MINING SYSTEMS

>

IntroductionMachine Condition Monitoring (CM) is avital tool for ensuring high asset reliabilityand plant availability in the ongoing revolution of modern plant maintenancestrategies. However, the mining industryhas lagged in the uptake of this methodol-ogy. Real-Time Condition Monitoring(RTCM) overcomes limitations of conventional CM and plays a key role inthe maintenance and reliability of mobilemine equipment.

The evolution of ConditionMonitoringCondition Monitoring can be traced backto the roots of understanding equipmentfailure and the maintenance strategies thatevolved to deal with these failures. Theearliest and most basic maintenance strat-egy was Breakdown Maintenance. In thisstrategy, assets ran until they failed or nolonger performed their designed functions;teams would then make costly repairs,resulting in excessive downtime and lostproduction before returning these assetsto service.

The subsequent methodology,Preventive Maintenance, required teams toperform maintenance tasks on a timedbasis to avoid failures related to the asset’sage. Equipment was serviced on a fixedtime schedule; maintenance personnelreplaced subsystems or components whenestimated service life had been reached.Service life, calculated by OEM’s, was overly conservative. The costs associatedwith this strategy are high and laden withrepeated downtimes.

Condition Monitoring techniques devel-oped significantly over the last threedecades as improved technology and dataprocessors became available. The mostcommonly used tools in CM include, butare not limited to: oil analysis; wear debrisanalysis (tribology); vibration analysis;infrared thermography; ultrasound; radiography; non-destructive testing; andon-board/online health monitoring.

These tools work to detect degradationor symptoms of degradation in operating

assets; they are integral to ReliabilityCentred Maintenance (RCM) programmesand essential for detecting and preventingrecurring failures.

The P-F intervalFor successful Condition Monitoring,maintenance teams must have enoughtime and resources after detecting animpending failure to effectively performcorrective action on the asset. Equipmenthas similar behaviours and responsesthroughout machine life; assets exhibitsimilar characteristics when failing. Whendeterioration begins (a factor that may berandom in terms of service life), equip-ment undergoes a reduction in the abilityto fulfill its intended function. At somepoint (P), approaching failure is detectablethrough inspection or other CM tech-niques. Unless addressed and dealt with,performance levels can deteriorate to apoint where the asset no longer performsat its demand level (this is called “functional failure”, F).

The shape of the P-F curve (shown inFigure 1) and the time between potentialfailure and functional failure (the P-F inter-val) will vary depending on a number offactors, such as asset type, duty cycle, andfailure mode. Constant failure modes forthe same asset can exhibit different curvesbased on the operating context; this isespecially relevant to mobile equipment(i.e. trucks and shovels) that experiencehighly variable duty cycles.

Figure 2 illustrates an extreme casewhere a CM test (t1) takes place justbefore potential failure. Before the nexttest (t2), sufficient time must be allocatedto conduct planning and schedulingbefore asset failure occurs. Production lossresulting from not running the asset isthen minimised, and an optimal mainte-nance plan can be made, componentsprocured, and scheduling performed.

Planning and scheduling can take sever-al days, especially when standard job plansare not available. If the P-F interval is notlong enough, it can result in very short,impractical CM testing. If the asset is criti-cal, the CM test interval should be select-ed so that maintenance can shut the assetdown well in advance of functional orcomplete failure.

If the P-F interval is too short to allow apractical CM test frequency that wouldprovide sufficient warning, teams must

allow the asset to run-to-failure or per-form a re-design of the system. Thebiggest challenge in mobile mine equip-ment is reliably detecting potential failure(point P) considering the variability of theoperating context.

Real-Time Condition Monitoringand Mobile Mine EquipmentReal-Time Condition Monitoring (RTCM) isdefined as the ability to measure and per-form Condition Monitoring in real-time.Data transfer latency is measured in seconds (not milliseconds as in some “real-time” process control applications).The recent advent of RTCM in the miningmarket has created significant opportuni-ties for savings and enhanced safety. Thesesystems introduce the ability to alertmaintenance and operations of abnormalconditions; alarms are triggered whenmonitored parameters exceed OEM manu-facturer or user defined (UDF) thresholds,or when equipment is operating in anundesired manner. Sites can programalarms to detect simple instantaneousthreshold violations or more complicated,rapid increase violations. Such events arethen extrapolated to predict potential viola-tions that foreshadow failures. Conditions relating to operating context can beapplied to UDF alarms for validation.

Real-Time Condition Monitoring helpssolve the dilemma of choosing appropriateCM test intervals. When properly config-ured, OEM and UDF alarms detect poten-tial failures for many failure modes. Theremaining requirement? Sufficient time toplan and schedule corrective work or, atthe very least, to shut down equipment.

At present, examples of RTCM applications for mobile mine equipmentare limited. While there is a growing num-ber of onboard health monitoring systemsfor modern equipment, these require aphysical connection to download data anduse proprietary software for data manipu-lation. This results in non-real-time data(periodic data capture) and, sometimes,inconsistent data handling.

The MineCare® maintenance manage-ment solution from Modular MiningSystems, Inc. links onboard health monitoring across disparate OEM’s and3rd parties. This link—connecting communications infrastructures and cen-tral software applications supported bymine management systems—provides a

“The biggest challenge in mobilemine equipment is reliably detecting

potential failure considering the variability of the operating context.”

17NOVEMBER 2008 MMS MAG<

MODULAR MINING SYSTEMS SOLUTIONS

Figure 1: The P-F CurveThe commonly used industry P/F chart explains potential vs. actual failure.

Figure 2: Sizing the CM Test IntervalThe commonly used industry P/F chart explains potential vs. actual failure.

true RTCM solution. Such systems haveproven their mettle in some of the mostremote mining operations, saving thou-sands of dollars on maintenance costsaround the globe.

The Modular MineCare® system hassaved many customers downtime throughReal-Time Condition Monitoring. Two specific examples are detailed below.

Differential turbo exhaust temperaturesOne site observed a significant differencein turbo exhaust temperature on a largehaul truck. The right turbo exhaust temperature was higher than the left, sug-gesting that the left turbo was in failureand the right bank was working harder tomeet the power demand. Correctiveaction was planned and scheduled toreplace the turbo before complete failureoccurred. While down, the covers werealso removed from the right cylinders,revealing several cylinders that were expe-riencing partial ring failure. The rings werereplaced and the truck was down for onlyeight hours. RTCM prevented a costlyengine/Turbo repair and several days ofdown time.

Low oil pressure alarmReal-Time Condition Monitoring has alsoreclassified alarm priority for low oil pres-sure. In this case, the OEM classified aplugged engine filter as a low level alarm(meaning it would not display to central

system attendants or operators) becausethe filter went into bypass and the enginestill received oil. Reclassifying the alarm ascritical, the site began detecting plugged fil-ter alarms for one of the trucks. In order tomeasure the crankcase pressure, which wasvery high, the attendant used the system toremotely connect to the unit in real-time.The filter was plugged with metal causedby partial failure of several piston rings. Thetruck was shut down for ring repair andreturned to service in a timely manner. ThisRTCM event saved critical engine failure,downtime, and production loss.

RTCM proactively detects potential fail-ures before they are noticed by operatorsor inspection methods. This allows time toplan and schedule corrective actions, or toat least stop the asset before significant orcatastrophic damage is done. By combin-ing RTCM alarms with mitigating process-es, failures can be effectively eliminated.

ConclusionOver the years, mining has incorporatedvarious maintenance philosophies as theyhave evolved from BreakdownMaintenance to Preventive Maintenanceto Real-Time Maintenance. With a betterunderstanding of failure modes and theadvent of Condition Monitoring, overallreliability of equipment and effectivenessof maintenance has improved substantial-ly. Real-Time Condition Monitoring allows

for detection of potential failures, overcoming the constraint of short P-Fintervals that make CM tests impractical.

Real-Time Condition Monitoring holdshuge potential for proactively managingasset maintenance, and it is particularlybeneficial for mobile mining equipment.Users no longer have to shut down equip-ment to board it and download aged dataand removing this download step increasesproductivity. Additionally, this augmentssafety by eliminating the need to physicallyvisit various equipment fleets and onboardsystems. RTCM for mobile equipmentaffords increased reliability and up-time,more effective use of maintenanceresources, and—most importantly—enhanced safety and production.

For more information contact WynandVosser on: 011- 463 5995 or email vosser@mmsi.com

“RTCM proactively detects potential failures before they are noticed byoperators or inspection methods. This allows time to plan and schedule corrective actions, or to at least stop the asset before significant or catastrophic damage is done.”

Time

Failu

reR

esis

tanc

e/

Perf

orm

anc

e

Time

Failu

reR

esis

tanc

e/

Perf

orm

anc

e

CM Test Interval

t1 t1

P-F Interval

*

*

Planning &Scheduling

Complete Failure

Functional Failure

Potential Failure

Start of Failure (not necessarily age related)

Minimum Performance Level

P-F Interval

P P

F F

*

**

*