DOCTORA L T H E S I S
Department of Civil, Environmental and Natural Resources EngineeringDivision of Operation, Maintenance and Acoustics
Mine Production Assurance Program: Development and Application
Amol Arun Lanke
ISSN 1402-1544ISBN 978-91-7583-787-1 (print)ISBN 978-91-7583-788-8 (pdf)
Luleå University of Technology 2016
Am
ol Arun Lanke M
ine Production Assurance Program
- Developm
ent and Application
Operation and Maintenance Engineering
Mine Production Assurance Program:
Development and Application
Amol Arun Lanke
Division of Operation and Maintenance Engineering
Luleå University of Technology
Luleå, Sweden
Printed by Luleå University of Technology, Graphic Production 2016
ISSN 1402-1544ISBN 978-91-7583-787-1 (print)ISBN 978-91-7583-788-8 (pdf)
Luleå 2016
www.ltu.se
Q = n ×Q
D =N
∑i=1
d
= × ×
� � � �
� � � �
• • • �
• • • �
• • • �
• • • �• • • �• • • �
• • • �
• • •
•= ,�=
= a × b × c
a+b+ c = 1 a,b,c > 0
= 0.30 × 0.34 × 0.36
= 860.30 ×780.34 ×280.37 = 55.05
= 80.220.29 ×48.870.31 ×63.780.41 = 51.19
= 72.210.24 ×50.590.40 ×20.910.36 = 40.26
= 231.2−4× −1.1× −6.7×Ca
× +0.14× +0.05×Ca(shovel−actual)
Ca(shovel−actual)
= 36.25+1.24× +0.7×
= 71.22+76× +2.57× +0.28×
Int. J. Productivity and Quality Management, Vol. 19, No. 3, 2016
Int. J. Productivity and Quality Management
Production improvement techniques in process industries
A. Lanke et al.
Production improvement techniques in process industries
3.1 Method used in a food industry
A. Lanke et al.
KPI Owner Objective
Production improvement techniques in process industries
3.2 MQW: method used in an automobile component manufacturing industry
A. Lanke et al.
3.3 EFQM: method used in various types of process industries across Europe
Production improvement techniques in process industries
A. Lanke et al.
Source:
Source:
Production improvement techniques in process industries
3.4 PAP: method for oil and gas industry
A. Lanke et al.
Production improvement techniques in process industries
Source:
4.1 Comparison of methods
A. Lanke et al.
Similarities and disparities
Method in food industry MQW EFQM PAP
Production improvement techniques in process industries
4.2 Requirements in mining industry
A. Lanke et al.
Planning Extraction transport Processing distribution
Requirement in mining Food industry method MQW EFQM PAP
Production improvement techniques in process industries
A. Lanke et al.
5.1 Limitations of the research and future research directions
Production improvement techniques in process industries
BritishFood Journal
Supply Chain Management: AnInternational Journal
Reliability Engineering
Reliability Engineering
Facilities
International Journal of Productivity and Quality Management
Implementing Six Sigma: Smarter Solutions using Statistical Methods
SPE Annual Technical Conference and Exhibition
Journal of PetroleumTechnology
World Mineral Production 2007–11
Journal of GlobalStrategic Management
Quality Assurance in Education
SPE Reservoir EngineeringInternational
Journal of Occupational Safety and Ergonomics
Journal of Petroleum Technology
Journal of Quality in Maintenance Engineering
A. Lanke et al.
International Journal of Industrial EngineeringMining Equipment Reliability Maintainability and Safety
BP Statistical Review of World Energy
The EFQM Excellence Model
Offshore Technology ConferenceEstimating the Short-term Producibility of Oil and Gas
Occupational Injuries and Sick Leave in the Swedish Mining and Mineral Industry 2012
The US Food Marketing System, 2002: Competition, Coordination, and Technological Innovations into the 21st Century
SPE Annual Technical Conference and Exhibition
The TQM Magazine
Reliability Engineering & System Safety
International Journal of Operations &Production Management
Quality Control Story
Reliability Engineering & System Safety
Integrated Manufacturing Systems
Self-assessment for Business Excellence
International Journal of Productivity and Quality Management
Programming Industrial Control Systems using IEC 1131-3
Mahindra Quality Policy
National Institute Economic Review
SPE Annual Technical Conference and Exhibition
Production improvement techniques in process industries
Journal of Productivity Analysis
Business Information Systems
International Journal ofProductivity and Quality Management
International Journal of Productivity and Quality Management
International Journal of Productivity and Quality Management
European Journal of Operational Research
Journal of Pharmaceutical InnovationSPE Rocky Mountain Regional
Meeting
AIHAJ-American Industrial HygieneAssociation
Implementing the European Foundation for Quality Management Excellence Model
International Journal of Mining Reclamation and Environment
Measuring Business Excellence
International Journal of Production Research
InternationalJournal of Productivity and Quality Management
A. Lanke et al.
Mine production index (MPI)-extension of OEE for bottleneck detectionin mining
Amol Arun Lanke ⇑, Seyed Hadi Hoseinie, Behzad Ghodrati
Division of Operation and Maintenance Engineering, Luleå University of Technology, Luleå 97752, Sweden
a r t i c l e i n f o
Article history:
Received 30 November 2015
Received in revised form 10 January 2016
Accepted 18 March 2016
Available online 20 June 2016
Keywords:
Mine production index
Bottleneck seeking
FDAHP
Production analysis
a b s t r a c t
Although mining production depends on various equipments, significant amount of production loss canbe attributed a specific equipment or fleet. Bottleneck is defined not only by production loss but also by
our satisfaction from the equipment. The user satisfaction could be measured as machine effectiveness.
Mining literatures on performance improvement and optimization of equipment operations assertimportance of availability, utilization and production performance as key parameters. These three param-
eters are useful for evaluating effectiveness of equipment. Mine production index (MPI), which can rep-
resent the effect of these factors, has been applied for continuous operation in mining. MPI uses Fuzzy
Delphi Analytical Hierarchy Process to determine importance of each three parameter for individualequipment. A case study in a Swedish open pit mine was done to evaluate the field application of MPI.
The results reveal that crusher is the bottleneck equipment in studied mine. As a methodical approach,
an algorithm which uses MPI and detects bottleneck in continuous mining operation has been proposed.� 2016 Published by Elsevier B.V. on behalf of China University of Mining & Technology.
1. Introduction
Mining is a complex process; output in mining is affected byvarious uncertainties relative to operation, management, and envi-ronment. Production assurance and thus business objective is acombined effect of these parameters in mining. It is a complicatedtask to evaluate and analyse all these elements without a struc-tured methodology.
Demand for mining products increases every year as the worldmoves towards more industrialisation [1]. This phenomenonpushes mining industry to produce more and more and promptsneed to locate the lagging operations. In order to initiate improve-ment it is necessary to follow a methodical approach which consid-ers the dynamic situation in mining operations. Generally, thelimitation causing loss of production traces back to one or twoequipment/operation. These are commonly called as bottlenecks.Bottleneck from the classic system perspective is the point in sys-tem which slows down the whole operation. However, in advancedand complex systems bottleneck could be defined as the pointwhich reduces the system effectiveness and management’ satisfac-tion from operation or machineries. One of the advantages ofdetecting bottlenecks is that once the bottleneck is detected, focuson system improvement becomes easier. Methodical approaches
which detect the performance of equipment and connect it to rea-son of specific factors are a few in mining researches. The proposedstudy aims to detect the bottleneck in core mining operations. It isalso essential to know which uncertainty is leading cause of bottle-neck. In order to have better improvement, the origin of uncertain-ties are also needed to be known.
The overall equipment effectiveness (OEE) has been used forsystem performance measurement in various industries includingin mining. Equipment with lower OEE is bottleneck in terms ofeffectiveness of equipment. However since OEE has limitations toovercome these limitations, a new measure termed as MPI wasdeveloped and tested for shovels performance in a mining opera-tion [2,3]. The core questions which will be answered in this studyare:
(1) How MPI can help to locate the bottleneck in miningoperations?
(2) How the analysis of MPI can be further used for identifyingthe underlying cause of bottleneck formation?
The article is structured as follows: literature review sectionwhich reviews the elements of availability, utilization and perfor-mance efficiency. These elements contribute to effectiveness ofmining equipment. Methodology section gives depth idea aboutMPI development and its application. Case study section involvesanalysis of the trucks, shovel and crusher for determining the bot-
http://dx.doi.org/10.1016/j.ijmst.2016.05.050
2095-2686/� 2016 Published by Elsevier B.V. on behalf of China University of Mining & Technology.
⇑ Corresponding author. Tel.: +46 920491586.
E-mail address: [email protected] (A.A. Lanke).
International Journal of Mining Science and Technology 26 (2016) 753–760
Contents lists available at ScienceDirect
International Journal of Mining Science and Technology
journal homepage: www.elsevier .com/locate / i jmst
tleneck in a Swedish open pit mine. Once after identification ofbottleneck, we discuss about the method and characteristics whichcould be involved in formation of this bottleneck. In the final part,concluding remarks and improvement potential is presented.
2. Literature review
2.1. Bottleneck seeking and equipment improvement methods in
mining industry
Based on the observation from literature in mining the perfor-mance of each equipment used in mining operation is typicallyanalysed in isolation. Very little consideration is given for its effecton either downstream or upstream processes. However in practicalsituation any optimization using such assumption would not beuseful. In this literature review, article tried to review methodolo-gies and techniques which consider multiple equipment in miningsystem and propose common method to improve. These methodsconsider either partial continuous operations.
Simulation has been demonstrated to assess the performance oftransport system in mining industry [4]. This study’s conclusionranking of equipment has been given by comparison. The equip-ment ranking was done by balance of cost and performance formining system. This is one of the earlier methods for selectingthe best equipment for better production.
Analysis of Kiruna iron ore mine for determining the bottleneckconsidering system aspect was done by Huang [5]. The study wasable to determine the optimal fleet size along with improvementin ore quality suggestions.
Boulder size and its effect to determine the bottleneck opera-tion have been showed in literature by Kumar [6]. Study revealedthat an ore pass gate becomes bottleneck in system when the boul-der size is out of limit. It also shows the boulder size result of fromvarious other processes in operation.
A method to analyse the maintenance data and reliability todetermine the failure trends for all underground mining equip-ment have been shown in study by Vagenas et al. [7]. This studyuses MTTR and MTBF data to carry out analysis underlining impor-tance of availability as metric to determine effective equipment.Bottleneck can be determined by the theory of constraints. How-ever bottleneck seeking would require not only hard (data driven)but also soft (interviews and questionnaires) as proposed by Pau-ley and Ormerod [8]. In this direction Çelebi used equipment eval-uation model, which was bounded by various constraints such asdiggability, number of passes, and production amount [9].
Simulation of surface mining system which includes shovels,trucks, drilling machine and employees was carried out by Charl-ton [10]. Study shows that with little assumption, measurementof time spent on various resources and operations, how the perfor-mance can be improved. The focus of case study was effect of over-burden, resources and coal seams on the total mine productionprediction.
High productivity of mining equipment is based upon availabil-ity, maintainability and reliability analysis of the equipment asshow in a study by Paraszczak [11]. Study was also able to showthat how the data collection and analysis can be used to achievehigh productivity. Discussion on data analysis to determine equip-ment availability, utilization and production performance can helpmine to achieve expected returns [12]. The study concluded thatmines should move from reactive maintenance practices to proac-tive maintenance practices with analysis of such data.
In equipment selection process, one of the criteria is productiv-ity, which is influenced majorly by age and availability [13]. Theperformance is key factor for productivity of equipment [13]. Alter-native methods to determine the failure priorities can facilitate
root cause analysis and chronic failures in all mine equipment. Thisalternative method of analysis includes logarithmic scatterplots[14]. The case study explored uses measures such as downtimeand failures of shovel. Hall and Daneshmend have discussed thatequipment reliability is key factors in loss of production and con-cluded that using reliability analysis technique will be crucial indetermining which part or equipment is cause of competitivenessloss [15].
A model built with production capacity constraint along withgeological uncertainties by Dimitrakopoulos and Ramazan [16].The authors claim that the model supposed to produce an optimalmining schedule which could meet the performance criteria.
Simulation model for a Turkish coal mine, which consideredloaders, crusher and conveyer belts, has been demonstrated [17].In this study equipment capacity, velocity and performance figureswere used to determine the bottleneck operation. Ore storage sim-ulation which when subjected to different conditions, determinesthe bottleneck in mining downstream [18]. This simulation showsthat considering the excavation and processing capacities, eithermine delivery system or concentrator plant can be bottleneck. Lin-ear programming and simulation model for truck, shovel andcrusher show that crusher is bottleneck in system [19]. In orderto evaluate the performance of each equipment criteria used werecapacity, utilization of equipment and fleet size.
In conclusion of literature review, it can be said that: to detectand remove the bottleneck in mine operations, researchers havestressed need to evaluate equipment availability, utilization andproduction performance. Various models and methods use thistype of data to achieve the higher productivity and operation anal-ysis. There has been huge research on individual equipmentimprovement, however, researchers which consider multipleequipment in mining, for bottleneck seeking and elimination isscarce.
The effort with this study is to propose methodology which canbe used for bottleneck detection in mining and can be applied tomost of the equipment in mining operation. For this purpose, basedon the literature, three main parameters used in OEE, namely avail-ability, utilization and production performance, have beenselected. OEE has been used in many industries as a point of refer-ence to measure the performance. However for mining industrythe traditional OEE cannot be used as it is. The limitation of OEEas a metric for mining industry is explained further followingterms.
Quality parameter in traditional OEE is ratio of processed piecesand defective pieces. In mining industry such distinction is notpossible to use. The defective ore or unusable ore is usually thewaste, which is eliminated during the operation. The remainingore is always processed to give the required final product. Consid-ering this limitation a modified OEE equation for mining wasdeveloped where quality rate is replaced by the utilization ofequipment [21].
Depending upon the delivery schedule, types and number ofavailable machine, age of machinery and production performancecan change. Studies on truck optimization for mining have shownthat cycle time for truck is important. The cycle time for trucksinvolves time spent in loading, hauling, dumping, standby time.Since the main purpose of shovel operation is to move material,the payload and digging rate are key performance measures. Totalabove mentioned parameters and restrictions affect the productionperformance. To take account for these considerations it is neces-sary to modify the OEE equation for mining applications. For exam-ple the payload or capacity factor for shovel can directly relate toperformance efficiency in equation rather than availability of sho-vel. Cycle time requirement for truck can be directly attributed toneed of higher utilization. Equipment with high critically for per-formance index may be hampered in performance due to less
754 A.A. Lanke et al. / International Journal of Mining Science and Technology 26 (2016) 753–760
availability during the operation. Each mining equipment isselected during mine design process for a specific purpose.
Literature review in mining industry has specified componentswhich impacts equipment availability, utilization and productionperformance. For the evaluation purpose, the availability, utiliza-tion and production performance defined by equations are pre-sented in next part.
2.2. Basic definitions
2.2.1. Availability
In presented index, the availability is taken into considerationas the operational availability. Operational availability is ratio oftime of the equipment available for the operation to the total oper-ational time [20]. Availability is calculated by Eq. (1) [21]:
Availability ¼ðTH � DTÞ
THð100Þ ð1Þ
where TH is total available hours; and DT the downtime hours.Assumed total hours are total calendar hours i.e. 24 h or what-
ever which is decided by mine operation and production managers.Based on the literature, mining equipment availability is affectedby following factors: reliability of subsystems, time planning,availability of spare parts, early detection of failure and environ-mental conditions [12,13,15,22–36].
2.2.2. Utilization
Utilization of mining equipment can be defined as the quotient,expressed in hours, of the consumption within a specified period(e.g., year, month, day, etc.), and the maximum or other specifieddemand occurring within the same period [37]. In other words, uti-lization is the time consumed by an equipment for producing therequired tonnage production within available time and it is calcu-lated using Eq. (2) [21].
Utilization ¼TH � DT � SH
TH � DT
� �ð2Þ
where SH is the standby hours (idle time).The factors which affect the utilization of mining equipment are
mine planning, production scheduling, location of equipment,logistic planning, and idle time due to legislative/incidental rea-sons [12,16,38–46].
2.2.3. Production performance
The performance of equipment can be defined as the ability ofan item to meet a service demand of given quantitative character-istics; this performance is based on the capability and availabilityof equipment [37]. The production performance is given by Eq.(3) [21].
Production performance ¼AP
TH�DT�SH
RC
� �ð3Þ
where AP is the total actual output by equipment; and RC the ratedcapacity of the equipment. Production efficiency in terms of timecan be given by Eq. (4) [47].
Production efficiency ¼APT
ATT
� �ð4Þ
where APT is the actual productive time; and ATT the actual totaltime.
Based on the literature, production performance is affected byfollowing factors in mines: (1) shovel, including bucket capacity,cycle time, bucket fill factor and operator/job efficiency; (2) trucks,including capacity and cycle time; and (3) crushers, including
crusher volume, material characteristics and speed of crusher[48–60].
3. Methodology
Taking the mentioned operational constraints into considera-tion, the OEE for mining application can be modified with introduc-tion of weight for each factor. Since assigned weights can beapplied to all equipment and can give impact of each factor onentire mine production, it is termed as mine production index(MPI). The MPI equation can be written as follows:
MPI ¼ Availabilityw1 � Utilizationw2 � Performancew3 ð5Þ
The main problem in clarifying the mentioned idea is to assign areasonable weight for each parameter (finding w1 to w3). For thispurposes, it was decided to use the experts’ opinions because thereis no mathematical or regression method to find this kind ofunknown parameters in such equations. Multi-criteria decisionmaking (MCDM) methods are powerful tools for solving this kindof problems. Among the available MCDM methods, Fuzzy DelphiAnalytical Hierarchy Process (FDAHP) was used.
In OEE measure, equipment with higher score is most effectiveequipment and thus contributes most to system performance. Sim-ilarly MPI value can indicate not only effectiveness of equipmentbut also contribution of equipment for overall mining system avail-ability, utilization and production performance. Analysis of the fac-tor which mostly affects MPI will lead to find the root cause ofdetected bottleneck for that specific equipment.
In general, preferences in analytical hierarchy process (AHP) areassigned by linguistics variables. These terms can be imprecise anddubious for application and calculation. To deal with such fuzzi-ness, AHP was appended by incorporating fuzzy characteristicsand developed a new process known as FDAHP [61]. FDAHP thushelps decision makers in dealing with imprecision and subjective-ness in pair wise comparison easily. The FDAHP method provides astructured framework for setting priorities using pair-wise com-parisons that are quantified. As demonstrated by many research-ers, FDAHP process has been used successfully in mining sectorfor example [62,63].
General steps performed in this study for building the FDAHPare as follow:
(1) Polling and interview by the experts: in this stage, the opin-ions of the experts were asked by questioners using thequantitative or qualitative parameters. It was asked fromexperts to mark the importance of each parameter in ques-tionnaires in a very simple way. In order to use the question-naires data in FDAHP method, for each importance level anintensity number from 1 to 9 was assigned. In total, 12 com-pleted questionnaires; five from industrial experts (from thecase study mine) and seven academic experts were received.From the polling the importance levels were obtained foravailability, performance and utilization. For clarifying somedetails in mine operation, the industrial experts wereinterviewed.
(2) Establishment of pair-wise comparison matrixes: in order toestablish the main pair-wise comparison matrix usingFDAHP method, it is essential to have comparison matrixof parameters based on each expert’s opinion. For this pur-pose, according to questionnaires, a comparison matrix isestablished for each machine view point of each expert’sopinion.
Let C1, C2, . . ., Cn denote the set of elements, while aij representsa quantified judgment on a pair of elements Ci and Cj. The relative
A.A. Lanke et al. / International Journal of Mining Science and Technology 26 (2016) 753–760 755
importance of two elements is obtained from division rate of Ci onrate of Cj based on questionnaire. This yields an n � n matrix A asshown in Eq. (6).
ð6Þ
(3) Establishment of major pair-wise comparison matrix: forestablishing of major pair-wise comparison matrix in FuzzyDelphi method and calculation of the relative fuzzy weightsof the decision elements, three steps should be done asfollow:
(a) Computation of triangular fuzzy numbers (TFNs); ãij. In thiswork, the TFNs (Fig. 1 and Eq. (7)) that represent the pes-simistic, moderate and optimistic estimate are used to rep-resent the opinions of experts about each parameter.
aij ¼ ðaij; dij; cijÞ ð7Þ
aij ¼ MinðbijkÞ ðk ¼ 1; . . . ; kÞ ð8Þ
dij ¼Ykk¼1
bijk
!1=k
ðk ¼ 1; . . . ; kÞ ð9Þ
cij ¼ MaxðbijkÞ ðk ¼ 1; . . . ; kÞ ð10Þ
where aij 6 dij 6 cij is obtained from Eqs. (7)–(10); aij the lowerbound; cij the upper bound; bijk the relative intensity of importanceof expert k between parameters i and j; and k the number of expertsin the decision making.
(b) Following above outlines, a fuzzy positive reciprocal matrixA can be calculated [64]:eA ¼ ½~aij�; ~aij � ~aij � 1; 8i; j ¼ 1;2; . . . ;n ð11Þ
or
eA ¼
ð1;1;1Þ ða12; d12; c12Þ ða13; d13; c13Þ
ð1=c12;1=d12;1=a12Þ ð1;1;1Þ ða23; d23; c23Þ
ð1=c13;1=d13;1=a13Þ ð1=c23;1=d23;1=a23Þ ð1;1;1Þ
264375
ð12Þ
(c) Calculation of relative fuzzy weights of the evaluation fac-tors using Eq. (13) [64].eZi ¼ ~aij � . . .� ~ain� �1=n
; fWi ¼ eZi � ðeZi � . . .� eZnÞ ð13Þ
where ~a1 � ~a2 ¼ ða1 � a2; d1 � d2; c1 � c2Þ; � the multiplication of
fuzzy numbers;� the addition of fuzzy numbers; and fWi a row vec-
tor in consist of a fuzzy weight of the ith factor (fWi = (x1, x2, . . .,xn), i = 1, 2, . . ., n). The defuzzification (changing the fuzzy number
to a usual number) is based on geometric average method as Eq.(14) [64]:
Wi ¼Y3i¼1
xj
!1=3
ð14Þ
In this research, major pair-wise comparison matrixes (Tables1–3) were established by the procedure described above and totalweights were determined. Total weights of all factors are listed inTable 4.
Using the weights presented in Table 4, Eq. (5) can be built upfor each machine as presented in Eqs. (15)–(17).
MPIShovel ¼ A0:30 � U0:37 � P0:33 ð15Þ
MPITrucks ¼ A0:28 � U0:41 � P0:30 ð16Þ
MPICrusher ¼ A0:24 � U0:36 � P0:40 ð17Þ
where A is the availability; U the utility; and P the production per-formance. The flowchart for applying the MPI for bottleneck detec-tion is represented in Fig. 2. To have impact of decision made forimprovement sufficient time must be given for full implication.Considering such factor, authors recommend to evaluate the MPIevery three months.
4. A case study
For testing the applicability of MPI measure, a case study car-ried out in a Swedish open pit mine operated in Sweden. For thispurpose fleet of mine production chain consisted of machineriesconsists of shovels, dump trucks and crushers are studied. The sizeof fleet and basic operation data are presented as follows: theyhave 4 shovels each with 40 m3 capacity; fleet of 31 trucks is pre-sent; 21 trucks has rated capacities of 217 tonnes whereas remain-ing 10 trucks has better capacity of 313 tonnes. The three crusherspresent across the mine has average capacities of 7000 tonne/h.
The average loading cycle time is 2.3 min. Since the wastedumping sites are placed strategically along with crusher sitesthe unloading time is also very low for these trucks. The dispatch-ing of trucks is done manually with help of GPS system guiding thetrucks. The data for study is gathered for a period of five months in2013. This data included availability, performance and utilizationpercentages. For shovels and trucks idle time and downtime dataare also considered. The total crusher processing and total shoveloutput were also collected. Trucks availability and utilization werecalculated using this data since direct data for trucks was not avail-able. The mine is operated for 24 h with seven days a week. Eachshift is 8 h except during the weekends in which it is 12 h. The sho-vel operators are changed periodically every two hours by theoperator on trucks and wheel loaders. The truck and shovel opera-tors have a break after replacement from shovel operation for15 min. Crusher is manned by two operators per shift.
The production performance is total output by each fleet ofequipment. In given organization, performance is measured interms of activity i.e. for efficiency the formula used by organizationincludes productive time (actual loading time) over the available
0
1
( )xμ
ijα ijδ ijγ
Fig. 1. Representation of fuzzy numbers.
Table 1
Fuzzy matrix for shovel formed from expert opinions.
A P U
A 1, 1, 1 0.77, 1.8, 1.05 0.77, 1.8, 1.05
P 0.55, 1.28, 0.94 1, 1, 1 0.77, 1.28, 0.95
U 0.55, 1.28, 0.94 0.77, 1.28, 1.04 1, 1, 1
756 A.A. Lanke et al. / International Journal of Mining Science and Technology 26 (2016) 753–760
time. Considering the output as function of the productive time, weassume the performance in terms of efficiency.
The field data from the case study mine is presented inFigs. 3–8. It is seen that even though shovel availability is 86% inaverage and the utilization is averaging at 78% where production
performance with high availability and utilization is very low withaverage of 28%. For crusher the average availability of all crushersis 72%, utilization is 51% and performance is 21%. Trucks have hadthe average availability of 77%, average utilization of 86% andaverage performance of 26%. The shovel has high availabilitycompared to truck and crusher. However, trucks utilization hashigher values than the shovel except for month of December.
Comparing the performance at equipment level, it is seen thatshovels have highest production performance. They have also com-parable utilization with truck in most of the months; however theirperformance is only 8% and 2% better than crushers and trucksrespectively. Utilization and performance of trucks have differenceof 60% whereas for crusher difference between utilization andperformance is 30% and for shovels it is more than 50%, leading
Table 2
Fuzzy matrix for trucks formed from expert opinions.
A P U
A 1, 1, 1 0.71, 1.8, 0.98 0.77, 1.8, 1.1
P 0.55, 1.4, 1.01 1, 1, 1 0.77, 1.8, 1.21
U 0.55, 1.28, 0.83 0.77, 1.28, 0.82 1, 1, 1
Table 3
Fuzzy matrix for crushers formed from expert opinions.
A P U
A 1, 1, 1 0.71, 3, 1.4 0.55, 1.8, 1.04
P 0.33, 1.4, 1.01 1, 1, 1 0.77, 1.8, 1.15
U 0.55, 1.28, 0.83 0.55, 1.8, 1 1, 1, 1
Table 4
Weights obtained using FDAHP method.
Availability Utilization Performance
Shovel 0.30 0.37 0.33
Truck 0.28 0.41 0.30
Crusher 0.24 0.36 0.40
Production and machineries
data
Miningoperation
Calculating the utilisation
Calculating the availablity
Calculating the production
performance
Is actual production desired production
Continuation of operation
Decisionmaking for
improvment
Bottelneck detection
No
MPI calculation & evaluation
Prtoduction and fleet databae: total operation hours, total production,
number of active machineries, TBF, TTR, downtimes, idel times.
Repeat the analysis every 3 months
Effectiveness measures
Improvement solutions and workorders
Fig. 2. Flowchart for applying the MPi for bottleneck detection in mining operation.
95%
90%
85%
80%
75%
70%
65%
60%
Decembe
r
Janua
ry
Februa
ryMarc
hApri
lMay
Shovel Truck Crusher
Fig. 3. Availability of studied fleets.
A.A. Lanke et al. / International Journal of Mining Science and Technology 26 (2016) 753–760 757
to conclusion that trucks utilization is effective than any otherequipment.
Considering such variability of mine equipment parameters, itbecomes necessary to evaluate the effect of these parameters alto-gether. Output by each equipment will not be correct considerationfor evaluation of this equipment together. Performance is rated
highly with leading weights for crusher and shovels, whereas uti-lization is rated high for trucks. Availability is rated behind utiliza-tion and performance for each equipment. It seems thatperformance is criteria favored by experts over utilization andavailability. Considering the weight from Table 1 and averagevalues for all equipment, MPI is calculated using Eq. (5). Replacingthe values for availability, utilization and performance inEqs. (15)–(17) will result in MPI for each equipment.
MPIShovel ¼ 860:3 � 780:37 � 280:33 ¼ 55:05% ð18Þ
MPITrucks ¼ 77:40:28 � 26:10:41 � 860:3 ¼ 51:19% ð19Þ
MPICrusher ¼ 72:20:24 � 50:60:36 � 20:90:4 ¼ 40:26% ð20Þ
5. Discussion
Although there are various methods for calculating and deter-mining the mining equipment productivity, very few studies focuson considering all equipment together. The methods discussed inliterature review show that for evaluating the productivity of anequipment or fleet, availability, utilization and performance areimportant criteria. Considering Nakijima’s OEE equation, this effectcan be calculated [65]. However, original OEE equation has limita-tion in terms of application for mining industry. Prompting theneed to modify this equation, OEE for mining considers qualityparameters, which becomes irrelevant in mining context as perits original definition. The effect of each parameter in OEE equationfor mining has to be evaluated. Therefore, this study proposes MPI.Using the MPI it is easy to evaluate not only effective equipmentbut also the effect of parameters on the effectiveness. As it is seen,only depending upon these three criteria independently, bottle-neck is not easy to find.
In the case study, application of MPI reveals that, crusher is thebottleneck. Crusher is thus by far lagging behind the shovel andtrucks in terms of effectiveness for production. With the weightsobtained it can be said that: without increase in availability or uti-lization from current level, effectiveness of crusher could beenhanced with an increase in the performance. However, theweight evaluation suggests that availability must be focused uponmore closely for improvement. Although crushers’ availability iscomparable to that of trucks and its utilization is less than bothother equipment, performance is the criteria that must be focusedupon.
With satisfaction of condition actual production being less thandesired production, the data that should be collected from the sys-tem includes downtime of equipment, the actual output by equip-ment and capacities, etc. This data can be utilized to calculateavailability, utilization and performance for each equipment. Thisdata along with importance of the equipment can be thus utilizedfor MPI evaluation and calculation.
100%
90%
80%
70%
60%
50%
40%
30%
Decembe
r
Janua
ry
Februa
ryMarc
hApri
lMay
Shovel Truck Crusher
Fig. 4. Utilization of studied fleets.
35%
30%
25%
20%
15%
10%
Decembe
r
Janua
ry
Februa
ryMarc
hApri
lMay
Shovel Truck Crusher
Fig. 5. Performance of studied fleets.
Availability Utilization Performance100%
90%
80%
70%
60%
50%
40%
30%
Decembe
r
Janua
ry
Februa
ryMarc
hApri
lMay
20%
10%
Fig. 6. Effectiveness measures for shovel fleet.
Availability Utilization Performance100%
90%
80%
70%
60%
50%
40%
30%
Decembe
r
Janua
ry
Februa
ryMarc
hApri
lMay
20%
10%
Fig. 7. Effectiveness measures for dump truck fleet.
Availability Utilization Performance80%
70%
60%
50%
40%
30%
Decembe
r
Janua
ry
Februa
ryMarc
hApri
lMay
20%
10%
Fig. 8. Effectiveness measures for crushers.
758 A.A. Lanke et al. / International Journal of Mining Science and Technology 26 (2016) 753–760
Improvement solutions and work orders are taken based ondetected bottleneck and dominant factor through MPI comparisonand evaluation. The decision will help reduce or eliminate the dis-crepancy between actual produced output and the desired output.
6. Conclusions
As discussed in the article, the proposed index, MPI, is an oper-ational measure which helps to detect the bottleneck in mine pro-duction process accurately. It is an extension of OEE concept whichcan be to apply it in field more clearly. Assigning different weightsfor the OEE components helps the engineers to detect the bottle-neck and find out the root cause of it in a proper and reliable way.
The literature review shows that there is strong correlationbetween three parameters of OEE and MPI and the bottlenecksperformance.
In the case study part, validation of MPi was tried by applicationin real production process. The analysis in a Swedish open pit minereveals that crusher is the production bottleneck machine fromview point of MPI evaluation. The validity of crusher being bottle-neck for the period of study was confirmed with mine productiondepartment and management. According to weights assigned andMPI value calculated, performance of crusher can be strongly asso-ciated with reason for it being bottleneck. It seems that crusherperformance is associated with its design parameter along withscheduling and planning.
Exact relationship between type of uncertainties mentioned inliterature review and bottleneck equipment characteristic is partof further research of this study. For effective decision making,dynamic situation must be considered during the evaluation ofthe process effectiveness.
References
[1] Brown T, Shaw R, Bide T, Petavratzi E, Raycraft E, Walters A. World mineralproduction 2007–2011. Brit Geol Surv; 2013. p. 76.
[2] Muchiri P, Pintelon L. Performance measurement using over-all equipmenteffectiveness (OEE): literature review and practical application discussion. Int JProd Res 2008;46(13):3517–35.
[3] Lanke A, Hoseinie H, Ghodrati B. Mine production index (MPI): new method toevaluate effectiveness of mining machinery. In: Proceedings of internationalconference on mining and mineral engineering (ICMME).
[4] Mutagwaba W, Hudson J. Use of object-orientated simulation model to assessoperating and equipment options for underground mine transport system.Trans Inst Min Metall 1993;102:89–94.
[5] Hunag Y. The production system analysis of Kiruna iron ore mine Licentiatethesis. Luleå: Luleå Tekniska H; 1993.
[6] Kumar U. Study of problems caused by oversized boulders in a mineproduction system: a case study. Int J Surf Min Reclam Environ 1997;11(2):69–73.
[7] Vagenas N, Runciman N, Clement SR. A methodology for maintenance analysisof mining equipment. Int J Surf Min Reclam Environ 1997;11(1):33–40.
[8] Pauley GS, Ormerod RJ. The evolution of a performance measurement projectat RTZ. Interfaces 1998;28(4):94–118.
[9] Nes’E Çelebi. An equipment selection and cost analysis system for open pit coalmines. Int J Surf Min Reclam Environ 1998;12(4):181–7.
[10] Charlton G. Simulation for performance enhancement of systems: a case studyof an open-cut coal mine operation. Int J Syst Sci 1999;30(4):417–26.
[11] Paraszczak J. A failure and downtime reporting: a key to improve mineequipment performance. CIM Bull 2000;93(1044):73–7.
[12] Lewis MW, Steinberg L. Maintenance of mobile mine equipment in theinformation age. J Qual Maint Eng 2001;7(4):264–74.
[13] Samantha B, Sarkar B, Mukherjee S. Selection of opencast mining equipmentby a multi-criteria decision-making process. Min Technol 2002;111(2):136–42.
[14] Knights PF. Rethinking Pareto analysis: maintenance applications oflogarithmic scatterplots. J Qual Maint Eng 2001;7(4):252–63.
[15] Hall RA, Daneshmend LK. Reliability modelling of surface mining equipment:data gathering and analysis methodologies. Int J Surf Min Reclam Environ2003;17(3):139–55.
[16] Dimitrakopoulos R, Ramazan S. Uncertainty based production scheduling inopen pit mining. SME; 2004. p. 316.
[17] Simsir F, Ozfirat MK. Determination of the most effective long wall equipmentcombination in longwall top coal caving (LTCC) method by simulationmodelling. Int J Rock Mech Min Sci 2008;45(6):1015–23.
[18] Berton A, Jubinville M, Hodouin D, Prévost CG, Navarra P. Ore storagesimulation for planning a concentrator expansion. Miner Eng 2013;40:56–66.
[19] Faraji R. A comparison between linear programming and simulation modelsfor a dispatching system in open pit mines; 2013.
[20] Macheret Y, Koehn P, Sparrow D. Improving reliability and operationalavailability of military systems. In: Proceedings of aerospaceconference. IEEE; 2005. p. 3948–57.
[21] Dhillon BS. Mining equipment reliability, maintainability, and safety. SpringerScience & Business Media; 2008.
[22] Samanta B, Sarkar B, Mukherjee S. Reliability modelling and performanceanalyses of an LHD system in mining. J S Afr Inst Min Metall 2004;104(1):1–8.
[23] Koellner WG, Brown GM, Rodríguez J, Pontt J, Cortés P, Miranda H. Recentadvances in mining haul trucks. Ind Electron 2004;51(2):321–9.
[24] Barabady J, Kumar U. Reliability analysis of mining equipment: a case study ofa crushing plant at Jajarm Bauxite Mine in Iran. Reliab Eng Syst Saf 2008;93(4):647–53.
[25] Hadi HS, Mohammad A, Reza K, Uday K. Reliability modelling ofhydraulic system of drum shearer machine. J Coal Sci Eng 2011;17(4):450–6.
[26] Roy SK, Bhattacharyya MM, Naikan VNA. Maintainability and reliabilityanalysis of a fleet of shovels. Min Technol 2001;10(3):163–71.
[27] Lugtigheid D, Banjevic D, Jardine AKS. System repairs: when to perform andwhat to do? Reliab Eng Syst Saf 2008;93(4):604–15.
[28] Vayenas N, Wu X. Maintenance and reliability analysis of a fleet of load-haul-dump vehicles in an underground hard rock mine. Int J Min Reclam Environ2009;23(3):227–38.
[29] Galar D, Naeem HB, Schunesson H, Berges L, Tormos B. Fusion of operations,event-log and maintenance data: a case study for optimising availability ofmining shovels. Mine Plan Equip Sel 2014:1173–94.
[30] Markeset T, Kumar U. Design and development of product support andmaintenance concepts for industrial systems. J Qual Maint Eng 2003;9(4):376–92.
[31] Ghodrati B, Kumar U. Operating environment-based spare parts forecastingand logistics: a case study. Int J Logis: Res Appl 2005;8(2):95–105.
[32] Ghodrati B, Banjevic D, Jardine A. Developing effective spare parts estimationsresults in improved system availability. In: Proceedings of reliability andmaintainability symposium (RAMS). IEEE; 2010. p. 1–6.
[33] Godoy DR, Pascual R, Knights P. Critical spare parts ordering decisions usingconditional reliability and stochastic lead time. Reliab Eng Syst Saf2013;119:199–206.
[34] Sottile Jr J, Holloway LE. An overview of fault monitoring and diagnosis inmining equipment. Ind Appl 1994;30(5):1326–32.
[35] Balaba B, Ibrahim MY, Gunawan I. Utilisation of data mining in miningindustry: improvement of the shearer loader productivity in undergroundmines. In: Proceedings of the 10th IEEE international conference on industrialinformatics (INDIN). p. 1041–6.
[36] Cui LR, Hayakawa Y, Gustafson A, Schunnesson H, Galar D, Kumar U.Production and maintenance performance analysis: manual versus semi-automatic LHDS. J Qual Maint Eng 2013;19(1):74–88.
[37] IEC. IEC 60050-191: International electrotechnical vocabulary: dependabilityand quality of service Technical report. International ElectrotechnicalCommission (IEC); 1990 [chapter 191].
[38] Suglo RS, Szymanski J, Booth D, Frimpong S. Simulation analysis of miningmodels. In: Proceedings of IASTED international conference on appliedsimulation and modelling (ASM2003), Marbella.
[39] Kear R. Strategic and tactical mine planning components. J S Afr Inst MinMetall 2006;106(2):93.
[40] Angelov A, Naidoo K. Plan compliance process for underground coal mines;2010.
[41] Hustrulid WA, Kuchta M, Martin RK. Open pit mine planning and design, twovolume set & CD-ROM pack; 2013.
[42] Godoy M, Dimitrakopoulos R. Managing risk and waste mining in long-termproduction scheduling of open-pit mines. SME Trans 2004;316.
[43] Guoqing L, Nailian H, Xuchun H. Study on the application of supply chainmanagement (SCM) for Jiaojia gold mine. In: Applications of computers andoperations research in the minerals industries. South African Institute ofMining & Metallurgy; 2003.
[44] Souza MJ, Coelho IM, Ribas S, Santos HG, Merschmann LH. A hybrid heuristicalgorithm for the open-pit-mining operational planning problem. Eur J OperRes 2010;207(2):1041–51.
[45] Topal E, Ramazan S. A new MIP model for mine equipment scheduling byminimizing maintenance cost. Eur J Oper Res 2010;207(2):1065–71.
[46] Hartman H, Britton SG. SME mining engineering handbook, vol. 2. Society forMining, Metallurgy, & Exploration; 1992.
[47] Elevli S, Elevli B. Performance measurement of mining equipments by utilizingOEE. Acta Montanistica Slovaca 2010;15(2):95.
[48] Edwards DJ, Holt GD. Estivate: a model for calculating excavator productivityand output costs. Eng, Const Archit Manage 2000;7(1):52–62.
[49] Widzyk-Capehart E, Lever PJ. Towards rope shovel automation operationinformation system. In: Proceedings of CRC mining conference.
[50] Frimpong S, Hu Y, Inyang H. Dynamic modelling of hydraulic shovel excavatorsfor geomaterials. Int J Geomech 2009;8(1):20–9.
[51] Segarra P, Sanchidrián JA, López LM, Querol E. On the prediction of muckingrates in metal ore blasting. J Min Sci 2010;46(2):167–76.
A.A. Lanke et al. / International Journal of Mining Science and Technology 26 (2016) 753–760 759
[52] Sanchidrián Blanco JA, Segarra Catasús P, López Sánchez LM. On loaderproductivity, rock strength and explosive energy in metal mining; 2011.
[53] Brown GM, Elbacher B, Koellner WG. Increased productivity with ac drives formining excavators and haul trucks. In: Proceedings of industry applicationsconference. IEEE; 2000. p. 28–37.
[54] Bascetin A. A decision support system using analytical hierarchy process (AHP)for the optimal environmental reclamation of an open-pit mine. Environ Geol2007;52(4):663–72.
[55] Burt CN, Caccetta L. Match factor for heterogeneous truck and loader fleets. IntJ Min Reclam Environ 2007;21(4):262–70.
[56] Krzyzanowska J. The impact of mixed fleet hauling on mining operations atVenetia mine. J S Afr Inst Min Metall 2007;107:215–24.
[57] Maerz NH, Palangio TC. Online fragmentation analysis for grinding andcrushing control. In: Proceedings of control symposium, SME annual meeting,Salt Lake City, Utah. p. 109–16.
[58] Rosario P, Hall R, Maijer D. Liner wear and performance investigation ofprimary gyratory crushers. Miner Eng 2004;17(11):1241–54.
[59] Sbarbaro D. Control of crushing circuits with variable speed drives. In: Worldcongress, 6. p. 1691.
[60] Hulthén E. Real-time optimization of cone crushers. Chalmers University ofTechnology; 2010.
[61] Van Laarhoven P, Pedrycz W. A fuzzy extension of Saaty’s priority theory.Fuzzy Sets Syst 1983;11(1):199–227.
[62] Mikaeil R, Ataei M, Yousefi R. Application of a fuzzy analytical hierarchyprocess to the prediction of vibration during rock sawing. Min Sci Technol2011;21(5):611–9.
[63] Karimnia H, Bagloo H. Optimum mining method selection using fuzzyanalytical hierarchy process-Qapiliq salt mine, Iran. Int J Min Sci Technol2011;25(2):225–30.
[64] Ya CL, Chao SC. A new approach for application of rock mass classification onrock slope stability assessment. Eng Geol 2007;89:129–43.
[65] Nakajima Seiichi. Introduction to TPM: total productivemaintenance. Cambridge (Massachusetts): Productivity Press; 1988. p. 129.
760 A.A. Lanke et al. / International Journal of Mining Science and Technology 26 (2016) 753–760
Abstract—OEE has been used in many industries as measure of
performance. However due to limitations of original OEE, it has been modified by various researchers. OEE for mining application is special version of classic equation, carries these limitation over. In this paper it has been aimed to modify the OEE for mining application by introducing the weights to the elements of it and termed as Mine Production index (MPi). As a special application of new index MPishovel has been developed by authors. This can be used for evaluating the shovel effectiveness. Based on analysis, utilization followed by performance and availability were ranked in this order. To check the applicability of this index, a case study was done on four electrical and one hydraulic shovel in a Swedish mine. The results shows that MPishovel can evaluate production effectiveness of shovels and can determine effectiveness values in optimistic view compared to OEE. MPi with calculation not only give the effectiveness but also can predict which elements should be focused for improving the productivity.
Keywords—Mining, Overall equipment efficiency (OEE), Mine Production index, Shovels.
I. INTRODUCTION ITH highly competitive environment, organizations need to improve with losses occurred during the
operations. These losses include losses due to breakdown, low speed, idle time, and defect and rework [1]. A key performance index (KPI) which includes these operational losses for equipment was developed by Nakajima, 1979 [1] termed overall equipment efficiency (OEE). OEE can be calculated by (1).
(1)
where, performance rating includes comparison between ideal time and the operating time of equipment. Availability rating refers to part of total working time and effectively addresses the losses such as breakdowns, setups, adjustments [2]. Quality rating element of OEE presents measure of yield. Quality rating is ratio of total good pieces produced by equipment to total defective pieces produced by equipment.
OEE is considered as a key performance indicator of a company [3]. This KPI can be used to measure and improve
Amollanke.Phd student with the Division of Operations and Maintenance e, Luleå University of Technology, Luleå, Sweden (phone: +46920-49-1586; e-mail: [email protected]).
Hadi Hoseinie works with the Luleå University of Technology, Luleå, Sweden. He is Post-doctoral researcher in division of operations and maintenance (e-mail: [email protected])
Behzad Ghodratiis with Luleå University of Technology, Luleå, Sweden, He works as Assistant Professor in division of operations and maintenance(e-mail: [email protected]).
the overall performance of an industry [4]. The case studies illustrate that OEE is applicable in variety of industries. Use of OEE as a metric to improve the equipment performance leads to increase in overall performance of system as indicated in many studies.
As a successful case study, OEE was used as a major measure of factory performance indicator and thus an enabler for better operation in polypropylene manufacturing. This was compared with automotive assembly where this measure was absent, causing it to be an inhibitor of manufacturing strategy [5]. OEE is used under umbrella of total production management (TPM), in survival of a government owned bearing manufacturing company as documented by [6]. Reference [7] reveals that OEE is associated with six big losses, leading to loss of revenue. It has been showed that increase in OEE from 62% to 85% of world class manufacturing level decreases the loss by 40% causing increase in revenue [8]. Also achieving more accurate delivery schedule for increased market share and reputation, OEE can be part of maintenance strategy. According to [9] railway infrastructure improvement can also be positively affected by use of OEE.
For evaluation of TPM and thus maintenance performance OEE serves as a metric to evaluate the production capability and impact of quality [10]. One of the most important strength of OEE can be defined as its ability to integrate different aspects of manufacturing into one single measurement tool [11].
While OEE is effective parameter to determine the performance it has limitation. Considering the interaction of parameters in a factory the performance of equipment in its isolation cannot determine its impact thoroughly on the system. OEE considers improvement of system performance by improving individual equipment itself. The characteristics of one equipment may not be same as the next one i.e. Normalization of system performance with respect to OEE measure of single equipment can not enough [12].
The evolution of OEE and its various modifications are well reviewed by [11] as shown in Table I. As seen from table, OEE limitation of application to system level was identified and rectified by various researchers so far.
Mine Production Index (MPI): New Method to Evaluate Effectiveness of Mining Machinery
Amol Lanke, Hadi Hoseinie, Behzad Ghodrati
W
World Academy of Science, Engineering and TechnologyInternational Journal of Environmental, Chemical, Ecological, Geological and Geophysical Engineering Vol:8, No:11, 2014
755International Scholarly and Scientific Research & Innovation 8(11) 2014 scholar.waset.org/1999.6/9999723
Inte
rnat
iona
l Sci
ence
Inde
x, G
eolo
gica
l and
Env
ironm
enta
l Eng
inee
ring
Vol
:8, N
o:11
, 201
4 w
aset
.org
/Pub
licat
ion/
9999
723
TABLE I VARIOUS MODIFICATION OF OEE (ADOPTED FROM [11])
TEEP [13] Can be applied to whole processing plant as a single entity and includes effect of maintenance.
PEE [14] Considers effects of parameter on the elements of OEE [14] thus proposed weights for element of OEE.
OFE [11] Measure gives effectiveness over the factory rather than single equipment.
OAE [11] To consider losses on overall production process Discussed in [11]
OPE [11] To consider losses on overall production process Discussed in [11]
OWEE [15] Uses the weighted approach and stating that OEE neither does nor prioritize the problematic equipment appropriately. [13]
II.OEE IN MINING INDUSTRY Mining industry is characterized by high volume of output
and high capacity of equipment. This industry is deeply dependent upon use of equipment for achieving targets of profitability. High amount of production time is lost due to unplanned maintenance in mining industry [16] i.e. lack of availability. For early return on investment and reduction of production cost equipment utilization is very important [17]. This emphasizes crucial need for higher utilization in mining industry. Standby equipment increases cost of operation, whereas machinery subjected to downtime causes less output. This directly affects the delivery assurance for mining industry. Hence performance of mining equipment is an important factor. Therefore OEE in mining application should involve elements of availability, utilization and performance. According to [19] OEE can be used along with other parameters for improvement of mining performance.
OEE has been used to determine the loaders and trucks performance in Namibian mines with results of suggestions to improve the availability of the equipment [18]. Referring to [19] OEE through TPM is applicable for improvement dragline performance in terms of reliability, cost of operation and productivity. As evident by the literature analysis and application, OEE can be used to determine the performance in mining industry as well. Elevli and Elevli in application of OEE to mining industry have shown benchmark formation for improvement for shovel and trucks performance [17]. They applied quality parameter with respect to defect loss with net operating time. Where the case study in Namibian is mines quality loss as was used as ratio of loaded capacity to full capacity [19].
Since quality parameter is not used as it is defined in original OEE equation, quality rate cannot be used for mining industry in its original definition [19]. The original definition of quality rating includes processed and defect amount. In mining, it is quite difficult to define such a distinction for extracted ore. Considering these limitations, a new OEE was developed which shown in (2) [19], [16].
(2)
where Availability (AV) is given by
(3)
where TH = total hours, DT= downtime in hours, and SH= standby hours.
Where production efficiency PE is given by;
(4) AP = Actual production RC= Rated capacity of equipment in hours
and Utilization U is given by;
(5) For the mining applications, OEE equations elements can
be termed as production efficiency, utilization and availability.
III. METHODOLOGY This modified OEE can be used to determine the
performance of mining production. However mining operation is characterized by high degree of uncertainty. Depending upon the delivery schedule, types and number of available machine, age of machinery, production performance can change [20]. Each mining equipment is selected during mine design process for a specific purpose. Studies on truck optimization for mining have shown that cycle time for truck is important [21]. The cycle time for trucks involves time spent in loading, hauling, dumping, standby time. Since the main purpose of shovel excavation is to move material, the payload and digging rate are key performance measures [22].
In total above mentioned parameters and restrictions affect the production performance. To take account for these considerations it is necessary to modify the OEE equation for mining applications. For example the payload or capacity factor for shovel can directly relate to performance efficiency in equation rather than availability of shovel. Cycle time requirement for truck can be directly attributed to need of higher utilization. Equipment with high criticality for performance index may be hampered in performance due to less availability during the operation.
Taking these operational constraints into consideration the OEE for mining application can be modified with introduction of weight for each factor. Since assigned weights can be applied to all equipment and can give impact of each factor on entire mine production, it is termed as Mine Production index (MPi) for equipment.
The MPi equation can be given as;
MPi = Ava ×PPb× Uc (6)
where Av is Availability, PP is Performance and U is Utilization and
0<a, b, c<=1 and a, b, c=1. In order to calculate and assigned the weights (a, b, c) a
reliable and quantitative analytical method is needed. One the applicable approach is to use the multifactorial decision making techniques. Based on the past experiences of the
World Academy of Science, Engineering and TechnologyInternational Journal of Environmental, Chemical, Ecological, Geological and Geophysical Engineering Vol:8, No:11, 2014
756International Scholarly and Scientific Research & Innovation 8(11) 2014 scholar.waset.org/1999.6/9999723
Inte
rnat
iona
l Sci
ence
Inde
x, G
eolo
gica
l and
Env
ironm
enta
l Eng
inee
ring
Vol
:8, N
o:11
, 201
4 w
aset
.org
/Pub
licat
ion/
9999
723
authors, the analytical hierarchy process (AHP) method can be used for assigning the weights to the main parameters used in the MPi formula.
AHP method was developed by Satty in 1980 [23] that provides a visual structure of complex problems in form of two or more levels of hierarchy [24] and facilitates evaluation of active parameters in decision making process. It can be used for solving the problems with qualitative and quantitative parameters.
The General stages of AHP method are enumerated as follows, 1. Goal objective definition, which takes head for which
evaluation is done. 2. Development of a hierarchy between the criteria related to
the goal. i.e. second or more level of hierarchy. 3. Pairwise comparison of elements and evaluation of factors
impact. 4. Formulate paired comparison of criteria as ratio. This
paired comparison is used to determine the weights of each criterion in terms of its effect on the objective goal.
5. Consistency index is calculated by equation
CI= ( max –n) /( n-1) (7)
where is maximum Eigen value of matrix, n= size of pairwise matrix.
In evaluation, comparison and assigning the weights to each factor involved in MPi the following cases are considered: 1) Cost of operation. 2) Production capacity 3) Production cycle time of the equipment 4) Criticality to production
It should be concluded that MPi is a general index which should be developed for each type of mining machineries individually. It means that the final aim of this index is to present a special MPi for each mining machine for example MPi for trucks, shovels, drilling machines, etc.
In this paper the MPi which has been developed for shovels is discussed and the details are presented in case study part.
IV. CASE STUDY In production process of mines, shovels play a critical role
and have significant impact on whole operation productivity. In order to evaluate its productivity; MPi is applicable as a practical indicator.
To evaluate the weights of parameter of MPi considering the shovel operation, a team of experts in field of mining machinery in Luleå University of Technology were gathered. They discussed on the importance of each parameter of shovel productivity. Some industrial consultation and field visits were also done. It was asked from experts to mark the importance of each parameter in questionnaires in multifactorial decision making software: Expert choice.
Based on the expert decisions and comments, the assigned weights for MPi’s parameter are as shown in Table II. Based on the expert decisions and comments, the assigned weights for MPi’s parameter are as shown in Table II.
Based on the resulted weights, the MPi formula for shovels is shown in (5);
MPiShovel= Av0.2944× PP0.3375× U0.3681 (8)
TABLE II
WEIGHTS OBTAINED FOR MPI FACTORS FOR SHOVELS
Parameters Weights obtained Availability 0.2944
Production performance 0.3375 Utilization 0.3681
In the next stage of this research after developing MPiShovel,
in order to check the applicability of this index a case study was done on four electrical and one hydraulic shovel in a Swedish mine. Data for availability, utilization and production performance of these shovels in period of December 2013 to April 2014 was used. Figs. 1 to 3 show the collected data in graphical format for comparison.
Fig. 1 Average availability of studied shovels for 5 months
Fig. 2 Average production performance of studied shovels for 5
months
70%
75%
80%
85%
90%
95%
El1 El2 El3 El4 Hy
Availibility
20%
23%
26%
29%
32%
35%
El1 El2 El3 El4 Hy
Perfroman
ce
World Academy of Science, Engineering and TechnologyInternational Journal of Environmental, Chemical, Ecological, Geological and Geophysical Engineering Vol:8, No:11, 2014
757International Scholarly and Scientific Research & Innovation 8(11) 2014 scholar.waset.org/1999.6/9999723
Inte
rnat
iona
l Sci
ence
Inde
x, G
eolo
gica
l and
Env
ironm
enta
l Eng
inee
ring
Vol
:8, N
o:11
, 201
4 w
aset
.org
/Pub
licat
ion/
9999
723
Fig. 3 Average utilization of studied shovels for 5 months
As it seems availability of shovel 4 is lowest among the
other shovels and rest of the fleet have similar availability level range of 85 to almost 90%. As it is obvious from figures above El4 is critical machine in viewpoint of availability and in rest of parameters El2 is critical with lowest values of performance and utilization. However, low value of utilization of El4 is also evident. It’s obvious that it is too difficult to recognize weakest shovel in this mine in viewpoint of operational aspects. Therefore using a comprehensive index is essential to be able to evaluate loading machinery in open pit mines and MPiShovelcan be a suitable approach for this purpose.
After data analysis, MPiShovel were calculated for each machine (Fig. 4).
Fig. 4 Calculated MPi
Shovel
As it can be seen El2 has lowest MPi value and El4 is
second weakest machine. El1 and El3 are following ones whereas Hy has highest MPi value.
As it was discussed before, MPiShovel is a modified version of classic OEE i.e. (2) for mining application thus the comparison of these two measures will be valuable for future applications. As shown in Fig. 4 the new index gives optimistic values of machinery effectiveness inherent characteristics of MPi equation. Nevertheless, the classic OEE gives very low and pessimistic values which sometimes are not representative of actual effectiveness of equipment. This problem leads mine engineers to underestimate the actual production ability of their fleet and sometimes can add higher cost to mining operations. Application of MPi helps to explore operational condition of studied fleet in an acceptable level
because when effectiveness values are as low as represented by OEE, it means that operational condition is not good at all. For example 13% OEE is almost negligible. This also means that equipment is not up to par with performance and can be considered obsolete. However it is against current condition and reality of case study conducted because this machine works and produces the ore in low level but not as bad as OEE depicts.
V.CONCLUSION In this study a new index termed as mine production index
is proposed and special case of termed MPi shovel was conducted on shovel in a Swedish mine. Following is the list of main results and findings of this study: 1. OEE for mining applications, includes utilization has
limitations, hence needs to be modified with addition of weights to the elements of OEE.
2. The weights in MPi proposed will underline effect of parameters involved on OEE of equipment.
3. MPi will give optimistic values of effectiveness with respected to OEE.
4. MPi with calculation not only gives the effectiveness but also can predict which elements should be focused for improving the productivity.
5. Regarding comments of expert team, utilization is most important factor in calculating the overall effectiveness of shovels in case study and performance and availability follows in the order
6. The case study showed that new developed MPi index is applicable for evaluation of overall productivity of shovels and in future research special MPi’s consisted on different weights can be developed for each mining machine such as trucks, dozers and crusher etc.
7. The calculation of MPi and weights should be done more frequent as per the requirement of mining industry. A simulation approach can be used to determine the impact on intended production assurance
ACKNOWLEDGMENT Authors would like to acknowledge CAMM: centre for
advance mining and metallurgy project for their financial support of this study. Authors would also like to acknowledge the contribution by the mining factory from whom the case study data was gathered. Authors would also like to thank Dr. Jan Lundberg, professor luleå University of technology for suggestions and improvement of this study.
REFERENCES [1] Nakajima, S. (1988). Introduction to TPM: total productive
maintenance. Productivity Press, Inc, P. O. Box 3007, Cambridge, Massachusetts 02140, USA, 1988. pp.129.
[2] Jonsson, Patrik, and Magnus Lesshammar, "Evaluation and improvement of manufacturing performance measurement systems-the role of OEE," International Journal of Operations & Production Management, vol19. no.1pp: 55-78, 1999.
[3] Kumar, Uday, and AdityaParida. "Maintenance performance measurement (MPM) system," in Complex System Maintenance Handbook, London, Springer, 2008,pp: 459-478.
60%
65%
70%
75%
80%
85%
90%
El1 El2 El3 El4 Hy
Utilization
30%
35%
40%
45%
50%
55%
El1 El2 El3 El4 Hy
MPi
sho
vels
World Academy of Science, Engineering and TechnologyInternational Journal of Environmental, Chemical, Ecological, Geological and Geophysical Engineering Vol:8, No:11, 2014
758International Scholarly and Scientific Research & Innovation 8(11) 2014 scholar.waset.org/1999.6/9999723
Inte
rnat
iona
l Sci
ence
Inde
x, G
eolo
gica
l and
Env
ironm
enta
l Eng
inee
ring
Vol
:8, N
o:11
, 201
4 w
aset
.org
/Pub
licat
ion/
9999
723
[4] Parida, Aditya, and Uday Kumar. "Maintenance productivity and performance measurement," in Handbook of maintenance management and engineering, London, Springer , 2009,pp: 17-41.
[5] Harrison, Alan. Enablers and inhibitors to manufacturing strategy, Cranefiled school of management, Crainfield,UK, 1998.
[6] S. Regan, B.G. Dale, "Survival to success: the case of RHP Bearings, Blackburn," The TQM Magazine, Vol. 11,no 1, pp.22 – 28,1999.
[7] Chand, G., and B. Shirvani. "Implementation of TPM in cellular manufacture." Journal of Materials Processing Technology, vol. 103, no. 1, pp: 149-154,2000.
[8] Al-Najjar, Basim, and ImadAlsyouf. "Enhancing a company's profitability and competitiveness using integrated vibration-based maintenance: A case study."European Journal of Operational Research, vol.157, no 3, pp: 643-657,2005.
[9] Thomas Åhrén, AdityaP arida, "Maintenance performance indicators (MPIs) for benchmarking the railway infrastructure: A case study", Benchmarking: An International Journal, Vol. 16,no: 2, pp.247 – 258,2009
[10] Jostes, Robert S., and Marilyn M. Helms. "Total productive maintenance and its link to total quality management." Work study, vol. 43, no.7, pp: 18-20, 1994.
[11] Muchiri, Peter, and LilianePintelon. "Performance measurement using overall equipment effectiveness (OEE): literature review and practical application discussion." International Journal of Production Research, vol 46, no.13 pp: 3517-3535,2008.
[12] Scott, Douglas, and Robert Pisa. "Can overall factory effectiveness prolong Moore's law?," Solid state technology ,vol.41 no. 3,pp: 75-82, 1998.
[13] Ivancic I., “Development of Maintenance in Modern Production”, in Euromaintenance Conference Proceedings, Dubrovnik, Croatia, 1998.
[14] Raouf, A., "Improving capital productivity through maintenance." International Journal of Operations & Production Management, vol.14 no.7, pp: 44-52,1994.
[15] Wudhikarn, R. "Overall weighting equipment effectiveness." In Industrial Engineering and Engineering Management (IEEM), 2010 IEEE International Conference on, IEEE, 2010, pp. 23-27.
[16] Dhillon, Balbir S, Mining equipment reliability, maintainability, and safety. London, Springer-Verlag, pp: 59-60, 2008.
[17] Elevli, Sermin, and Birol Elevli. "Performance measurement of mining equipments by utilizing OEE." Acta MontanisticaSlovaca, vol.15, no.2,pp : 95 ,2010.
[18] Akande, JideMuili, Abiodun Ismail Lawal, and Adeyemi Emman Aladejare." Optimization of the overall equipment efficiency (OEE) of loaders and rigid frame trucks in NAMDEB Southern Coastal Mine Stripping fleet, Namibia."Earth Science, vol.2 no.6 pp: 158-166,2013.
[19] Paraszczak, Jacek. "Understanding and assessment of mining equipment effectiveness." Mining Technology, vol.114 no.3,pp:147-151,2005.
[20] Samanta, Bimal, B. Sarkar, and S. K. Mukherjee. "Selection of opencast mining equipment by a multi-criteria decision-making process." Mining Technology, vol.111 no.2,pp: 136-142,2002.
[21] Burt, C., L. Caccetta, and P. Welgama. "Models for Mining Equipment Selection 1.", 2005, Accessed at http://citeseerx.ist.psu.edu/viewdoc/ summary?doi=10.1.1.375.8249 on June 1st 2013.
[22] Kwame Awuah-Offei, Samuel Frimpong, “Cable shovel digging optimization for energy efficiency,” Mechanism and Machine Theory, Volume 42, Issue 8, Pages 995-1006, 2007.
[23] Saaty TL (1980). The analytic hierarchy process: planning, priority setting and resource allocation, NY, McGraw-Hill,1980.
[24] Kumar, S. Suresh. "AHP-based formal system for R&D project evaluation," Journal of Scientific and Industrial Research, vol.63 issue 11 pp: 888-896,2004.
World Academy of Science, Engineering and TechnologyInternational Journal of Environmental, Chemical, Ecological, Geological and Geophysical Engineering Vol:8, No:11, 2014
759International Scholarly and Scientific Research & Innovation 8(11) 2014 scholar.waset.org/1999.6/9999723
Inte
rnat
iona
l Sci
ence
Inde
x, G
eolo
gica
l and
Env
ironm
enta
l Eng
inee
ring
Vol
:8, N
o:11
, 201
4 w
aset
.org
/Pub
licat
ion/
9999
723
doi: 10.5277/ msc162312
Mining Science, vol. 23, 2016, 147−160 Mining Science
(previously Prace Naukowe Instytutu Gornictwa Politechniki Wroclawskiej, ISSN 0370-0798 in Polish)
www.miningscience.pwr.edu.pl ISSN 2300-9586 previously 0370-0798
Received July 6, 2016; reviewed; accepted September 9, 2016
UNCERTAINTY ANALYSIS OF PRODUCTION IN OPEN PIT MINES – OPERATIONAL PARAMETER REGRESSION ANALYSIS OF MINING MACHINERY
Amol A. LANKE1* Behzad GHODARATI1 Seyed Hadi HOSEINIE2 1 Luleå University of Technology, Luleå, Sweden 2 Hamedan University of Technology, Hamedan, Iran
Abstract: In mining uncertainties related to equipment and operation are major reasons for loss of pro-duction. In order to address this issue, a wide literature review was done in this study. It showed that reliability of equipment, spare part availability, automation of equipment are researched areas focused. However, a methodology which relates operational issues directly to production levels has been not stud-ied with detailed analysis. In order to overcome this issue and propose, a method to achieve production assurance is the objective of this study. A case study with 2.5 years of data from a large open pit mine is carried out. Following the statistical principles, multiple regressions modeling with details analysis, opti-mization of payload and interpretation of analysis are used. It showed that at system level availability, utilization and maximum capacities are important criteria for finding root cause in loss of production. Model for shovel fleet showed that availability is the most important characteristics hindering it to achieve a higher level of production. It was also seen that 3 to 4 number of shovels are optimal for achiev-ing current level of production. For truck fleet model represented that capacities involved are less im-portant factor as compared to the utilization of the fleet.
Keywords: mine production, availability, utilization, regression
INTRODUCTION AND BACKGROUND
In order to increase the output and satisfy increasing market demand, mining re-search is focused on optimization which is branched off in many areas, such as automa-
_________ * Corresponding author: [email protected] (A. Lanke)
Amol A. LANKE, Behzad GHODARATI, Seyed Hadi HOSEINIE 148
tion, uncertainty analysis, equipment design and implementation (Gustafson, 2011; Ramazan and Dimitrakopoulos, 2013; Samanta et al., 2002).
In mining where production of ore is most important outcome, knowledge of oper-ations is essential for the management of risk, maintenance of low costs, and increas-ing of the output. In broad sense uncertainty may be defined as being any deviation from the unachievable ideal of completely deterministic knowledge of a relevant sys-tem (Walker et al., 2003). Uncertainties are inherent in the circumstances in which mines are planned, developed and operated. The uncertainties related to mining pro-ject are represented in Figure 1. Larger the degree of uncertainty, more valuable it becomes to know effect of internal and external factors causing uncertainty. The sources of mining uncertainties could either be planned or unplanned and due to inter-nal or external events.
Equipment production is affected by the factors causing uncertainties. These fac-tors broadly can be divided in; mechanical properties of equipment such as design, shovel bucket size, operational and working plan for equipment, environmental factors such as temperature, rain, ice etc., human factors such as skill, competency, fatigue experienced by operator etc.
Focus of current study relies in area of increasing the production based on availa-ble equipment. To increase the production by equipment in mining, most of research in mining is related to increase the reliability of equipment. Mining equipment output is mainly based on availability, utilization and its rated capacity (Lanke et al., 2016). Temperature and effect of precipitation can also associated with this variation of out-put.
It may be advantageous to acquire additional information about the relationship be-tween these factors for mining production increase. The important information or knowledge here is the relation between the internal and external factors which govern the equipment output uncertainty.
A methodical approach, based on analysis could be beneficial to determine the un-certainties and affecting the equipment operation and their relationship with total out-put. In this paper it is tried to develop a general uncertainty analysis model of mining machinery production in open pit mines.
Cause/s of origin of uncertainties in mining are sometimes difficult to determine, they may be specific to environment as explained by Barabadi (2011), or system (Abdel Sabour et al., 2008) and context under study (Dimitrakopoulos and Sabour, 2007). One of earliest study which related external uncertainties and their cause-effect is presented in a study by Vernon (1984). This study states prominent reasons explained for uncer-tainty are external causes such as supply and demand fluctuations, political instability and lack of vertical integration in process.
Economic uncertainties are another area studied and improved in mining research. Economic uncertainties include market volatility, demand and price fluctuations un-
Uncertainty analysis of production in open pit mines – operational parameter regression… 149
certainty has been addressed with real options approach in many studies (Dehghani and Ataee-pour, 2012; Dimitrakopoulos and Sabour, 2007).
Fig. 1. Sources of uncertainty in mining (Kazakidis, 2001)
Equipment and their operation are one of the uncertainty that is shown in figure 1. In a study by Samanta et al. (2002) author mentions various internal and external fac-tors which could help the optimal production by equipment. It is seen that effects of operational factors on equipment output has been addressed by most equipment selec-tion studies in mining research for example (Haidar et al., 1999; Ekipman et al., 2003; Samanta et al., 2004). The factors or scenarios related to equipment and its output are focused with optimal mine conditions and stabilizing the reliability of equipment. These research topics explain and propose methods and tools to increase the reliability and life of equipment and specify criteria required for high output.
An effect of environmental factors specifically in area with extreme temperature and weather on mining equipment has also been studied. Availability of spare parts and its effect of equipment production have been shown in (Ghodrati, 2005). Throughput capacity and environmental effects using covariate analysis have been explored (Barabadi et al., 2011).
These studies however, do not specifically target factors which cause loss of pro-duction due to uncertainties related to equipment operation. Studies which specifically address operational parameters related to equipment operation and its effect on pro-duction are limited. Considering such limitation an extensive literature review led to conclusion that availability, utilization and performance of mining equipment are key parameters along with environmental factors (Lanke et al., 2016). As lack of knowledge i.e. contribution and effect of operational factors on equipment production
Market price
Sources of uncertainty in mining projects
OPERATING
Legislation/Regulation
Political Risk
Government policy
Environmental issue
Societal issues
Exogenous Endogenous
Labour Force
Managmenet Operating team
Grade Distribution
Ground Condition
Equipment
Infrastructure
Recovery method
Industrial relations
Amol A. LANKE, Behzad GHODARATI, Seyed Hadi HOSEINIE 150
is important uncertainty, solution to reduce the uncertainty is quantification of such uncertainty. This would lead to achievement of planned production beneficial for min-ing organization.
METHODOLOGY
Based on literature review in uncertainty reduction for mining, it can be seen that analytical methods are pervasive and applicable for mining research and application. Many of these analytical methods are based on statistical analysis. Using statistical analysis provides principles and methods for collecting, summarizing, and analyzing data, and for interpreting the results. Applied statistical methods are useful for describ-ing the data and proposing inferences based on analysis. Collection of data for the factors which are relevant for analysis based on requirements and finalizing the analyt-ical method and tools are two main steps that must be followed. Initial step will help in carrying out statistical analysis.
It has been established that operation factors affecting the equipment output are di-vided into three main components; equipment availability, utilization and their per-formance (Lanke et al., 2016). These factors are dependent upon downtime, standby time, and rated capacities. These factors can, in turn, represent as the function of in-volved parameters’ (Lanke et al., 2016; Dhillon, 2008). They are represented as fol-lowing equations ((1) to (3)):
A = f (TH, DT) (1)
U = f (TH, DT, SH) (2)
P = f (RC, AC) (3)
where: A – availability, U – utilization, P – performance, TH – total hours, DT – downtime, SH – standby hours, RC – rated capacity and AC – actual capacity.
The next step was to identify the suitable approach for data analysis and estimation of components characteristics such as availability, utilization, performance, downtime, standby hours, idle times, and their effect on production. In order to propose a formal presentation of theory in terms of the equation; a method must be determined. The chosen method should reply two questions for the study. How much variance in the equipment output is accounted for by the combination of the considered factors? How to represent effect of factor in terms of change in equipment production? In statistical modeling, estimation of the relationship among variables is done with help of regres-sion analysis method. The result obtained in a process is denoted by the dependent variable, whereas factors leading to result are termed as predictor variables. In mine
Uncertainty analysis of production in open pit mines – operational parameter regression… 151
production, output in terms of tonnage by each equipment can be termed as result or dependent variable. Availability, performance, utilization and their factors can be con-sidered as predictor variables. Effect of each factor on the dependable variable can be evaluated through the regression model. However, the equipment output is affected by all the factors simultaneously. The multiple regressions will yield results of comparing all factors effect on the dependable variable.
The methodology is represented in flow chart in Figure 2.
Fig. 2. Methodology followed for data collection and analysis
CASE STUDY
DATA COLLECTION
For developing a new production uncertainty model, a case study was conducted in a large Swedish open pit mine. The data was collected for 30 months from January 2013 to June 2015. Mine operates with temperatures varying from minus 30°C to 25°C. During the period of peak winter the snowfall reached 1.21 meters, varying within the range of 0.5 to 1 meter during the whole winter season. These conditions cause harsh effects not only on ore but also on equipment operation and operators skills. Haulage system in this mine composed of 31 trucks with two different capaci-ties and six shovels. The mine is operated for 24 hours with 7 days per week. The ap-
Determine data analysismethod
Determine factors affectingequipment output
Categorize factors in internaland external category
Collect data aboutselected factors
MINESTAR©
Combined effect of measured factors&correlation of each factors on
equipment production can be analyzed
Further root cause analysis usingstatistical method
Optimization and improvementsuggestions
NO
YES
Amol A. LANKE, Behzad GHODARATI, Seyed Hadi HOSEINIE 152
plied data for modeling was extracted from “Minestar©” software. The data is recorded output from sensors installed on equipment. For trucks data related to its destination (a specific crusher or dump site), assigned shovel and its rated capacities acts as an input and are recorded for each loading and unloading cycle. This data is complicated since trucks are assigned to different areas in the mine in real-time and keep changing their destination. The shovel input includes its identity, its source destination (extrac-tion site in mine) and assigned crusher. A common output recorded for both equip-ment is production in terms of tonnage. Shovel data includes sensor data for each of its activity, which include its availability time (working duration, standby duration) and utilization (idle time, queuing time etc.). Similarly, for trucks data which is useful for evaluating availability and utilization is recorded. The collected data was in a big size which was a mixture of automatic and manual entries. Data related to factors such as availability, utilization and performance were calculated for each day based on availa-ble raw data. Within 21 working hours, data for trucks fleet and shovel fleet is consol-idated. Rated capacity of a single shovel is 3840 tons per hour, with six shovels; and nominal 21 hours non-stop operation, the maximum rated capacity is 483840 tons/day. This is considered as theoretical maximum rated capacity for a shovel. In similar man-ner maximum rated capacity for trucks is calculated. Based on these calculation opti-mization report and interpretation of results can be drawn.
Conversion from raw information, application of consistency and frequency matching was done by developing and applying for a specific computer program with spreadsheet software. Internal factors which affect the output by equipment are repre-sented with by data. Completing these stages will lead to the formulation of a model which can represent the uncertainty.
DATA ANALYSIS
Variation or uncertainty in output is caused by three main factors availability, utili-zation and capacities of equipment involved. In order to determine the uncertainty in production and its correlation with these factors, multiple regression analysis was car-ried out. From the raw data preliminary analysis shows that the availability of whole mine fleet varies from 1 hour to 277 hours (Fig. 3). The mean available hours are 117 hours with standard deviation of 80 hours. The mean time for the utilization of the whole fleet is 114 hours per day (Fig. 4). The utilization time varies between 1 hour to a maximum of 245 hours. This is associated with standard deviation of 66 hours. The performance graph shows that payload achieved has been 167000 tons per day at its maximum value (Fig.5). Mean production per day is 97000 tons with standard devia-tion of 30000 tones.
Uncertainty analysis of production in open pit mines – operational parameter regression… 153
Fig. 3. Overall fleet availability with frequency and cumulative %
Fig. 4 . Overall fleet utilization percentage and cumulative %
Fig. 5. Overall system payload achievement
Amol A. LANKE, Behzad GHODARATI, Seyed Hadi HOSEINIE 154
OVERALL ANALYSIS OF DAILY PRODUCTION OF MINE
Fleet analysis was started with considering the effect of shovel and trucks’ fleet to-gether along with the actual payload. For selecting the significant factors and the ac-ceptable relationship between the studied parameters, two metrics were chosen: 1) percentage of variation explained by parameters and 2) P-value. At the system level, regression analysis reveals that 43.4% variation in achieved payload is explained with three variables namely available time, operating time and maximum theoretical capac-ity for the overall system. P-value (<0.001) also suggests that there is a strong correla-tion between these factors and the payload achieved as shown in Figure 6(a). Based on analysis, presented model in Equation (4) presents the daily production in overall sys-tem level:
��������� ��� � ����� � �� � ���� � ����������� (4)
This equation suggests that achieved payload for the entire fleet of trucks and shovels is dependent upon availability, utilization, and maximum theoretical capaci-ties. Achieved payload is more sensitive to changes in maximum capacity followed by availability.
a)
b)
Fig. 6. Statistical analysis of important factors: a) variation percentage explained by model presented; b) bach factor regressed over the other variables
The interaction between factors shows that with increasing of mean availability and utilization mean, mean of daily production will increase. As shown in Figure 7, interaction plot between the maximum capacity of both trucks and shovels show that mean daily production will increase with increasing of availability. The analysis shows that at a certain point, an increase of available capacity will not help to increase the daily production. Considering the effects of entire equipment parameters on achieved production, uncertainties are caused by changes in overall availability.
Uncertainty analysis of production in open pit mines – operational parameter regression… 155
Fig. 7. Analysis of daily production (change in payload when availability,
utilization and capacity changes)
This means that it is essential to delve into system at component level. In current scenario trucks and shovel fleet are components. With further analysis with trucks and shovel separately, it will be clear to see which elements affect their output most. This will also give insight into what changes in availability and utilization should be done to achieve maximum possible daily production.
PRODUCTION ANALYSIS OF SHOVEL FLEET
As discussed earlier, overall daily production of any open pit mine is resulted by interaction and combination of shovel fleet and truck fleet production. In shovel fleet perspective, number of available shovels and their nominal capacity are dynamic and uncertain parameters which deeply affect the daily production of the whole mine.
Based on the available data, regression analysis of shovel fleet production reveals that all variables (availability, utilization and maximum capacity) have a significant effect on fleet production with P-value less than 0.1. Equation (5) shows the relation-ship between all mentioned factors and shovel fleet production.
� !"��#�""��$�� ��� � ���%& ' %���()*+,- ' .����/)*+,- ' .�.%���01,�-��234�- (5)
This model states that availability is the most correlated factor for shovels to achieve production. It also represents that once the ideal actual capacities are consid-
Mea
n Da
ily P
rodu
ctio
n (t
one)
Available Time × Maximum Capacity
Availability (%)
0
40000
80000
120000
160000
200000
0 20 40 60 80 100
Amol A. LANKE, Behzad GHODARATI, Seyed Hadi HOSEINIE 156
ered, utilization of shovels becomes a more impactful factor in achieving production goal. Figure 8 shows that available time causes more impact, followed by the ideal actual capacities and then the utilization time. However, when operating time is re-gressed over other factors involved, it reaches R2 value of 95% as compared to 93% of R2 value for availability. The significance of availability i.e. reliability of equipment is the key factors that must be focused.
Fig. 8. Impact of factors on shovel fleet output when regressed individually
and when regressed over all other parameters
Number of working shovels per day is another important factor for achieving high-er availability. Considering the availability and actual maximum capacities, this gives us insight into the optimal number of shovels that can be used for the operation. When the model is analyzed for current condition with objective of maximizing the payload, it is seen that use of 4 to 5 shovels daily would be ideal as seen in Figure 9. This would not only impact shovel capacities but also increases the overall or production of the shovel fleet.
To see how utilization affects production levels by shovel fleet, correlations be-tween reasons which could lead to loss of utilization are evaluated. For this analysis idle time, delay time and downtime were considered. Analysis of the components re-veals that shovel fleet is more sensitive to idle time compared to delay and downtime. In order to increase the production by a reduction in shovel idle time, one of the prom-inent area of research is related to trucks and shovel system. Requirement of optimum number of trucks or match factor has potential to save cost for the system (Ercelebi
Uncertainty analysis of production in open pit mines – operational parameter regression… 157
and Bascetin, 2009; Alarie and Gamache, 2002). A comprehensive strategy which reduces the shovel idle times based on increasing utilization of truck has been dis-cussed by Alarie and Gamache (2002). Operational planning, efficiency of blasting operation, impact of ageing machinery, optimum truck dispatching time are important reasons for idle time for shovel operation (Mohammadi et al., 2013; Rai et al., 2000; Mohammadi et al., 2016; Patnayak et al., 2008).
Fig. 9. Optimal number of shovels required for the current level of production
TRUCK FLEET ANALYSIS
In initial truck evolution, P-value indicates a significant relation between utiliza-tion, availability, performance and achieved payload. The analysis shows that 47.6% of the variation in carried payload by trucks is explained by the regression model. The regression model for truck fleet productions is presented by Equation (6).
5!"����$������#��� 6�#�""� � ����% ' �����37428 ' .���37428 (6)
The achieved model shows that availability and utilization of trucks are signifi-cantly important in reaching the production goal. As shown in the model increase in 1% utilization will increase the payload by 1.24%. This leads to attention for analysis of utilization and its elements. As lower truck utilization translates to higher idle times for shovels and overall delay in production process causing loss of ore tonnage. The analysis shows that it is possible to achieve 50% utilization with 25 hours of idle time for the whole fleet per day. The mean of utilization varies with changes in idle and delay time. Based on the data analysis, with an increase in idle times and decrease in delay times to minimum hours it is possible to achieve 70% mean utilization.
To pinpoint the reason for uncertainty in production through the availability of trucks, further analysis is carried out. During this model building between payloads achieved and trucks downtime, and total operating hours are evaluated. Uncertainty or variation of achieved payload is explained 40% by downtime and total operating hours.
Avai
labi
lity
(%)
Maximum number of available shovels1 32 4 5 6
Amol A. LANKE, Behzad GHODARATI, Seyed Hadi HOSEINIE 158
DISCUSSION
In open pit mining, uncertainties related to equipment and operations are prime reason for loss of production. In mining literature there is lack of method for directly relating uncertainty with production by equipment. With this research attempt had been made to fill such a gap. The aim of study was to exemplify quantification of un-certainties with production output by fleet of mining equipment. Root cause analysis for production loss through this quantification was another objective of this study. To demonstrate this method, 2 years and 6 months of data was obtained from mining or-ganization and analyzed. Based on requirement of study and statistical principle it was observed that multi-regression analysis modeling is suitable methodology. Using this methods analysis was carried out considering two equipment fleet together (system) and further for fleet of shovel and truck separately.
CONCLUSION
During the period of observation it was seen that daily production is sensitive to availability more than current maximum theoretical capacities at overall system level. With detailed analysis of shovel fleet it was seen that number of shovels and their idle time are dominant factors to achieve target production. With shovels’ analysis it is observed that 4 to 5 number of shovels is optimal size of fleet to achieve current level of production and operating of less than full number of shovels for operation could be potentially economical. This will reduce the downtime for fleet of shovel thus causing increase in utilization of shovels leading to further increase in output.
Model for truck fleet showed that utilization followed by availability is main im-portant criteria that must be focused on increase of production. The model for truck fleet lacks ideal actual or ideal maximum capacities. When truck fleets utilization was analyzed it was seen that combined standby time is limiting factor compared to down-time. Truck fleet idle time i.e. waiting for the ore or waiting at crusher is one of root causes that cause hindrance to increased production level. To increase the in mine output from current levels, capacities of trucks and shovels can be exploited further. The analysis shows that current configuration of both equipment fleets is able to re-spond to higher demand of production. It is possible to achieve current level of pro-duction of ore with 4 to 5 shovels instead of 6 shovels. The truck fleet capacities are more than adequate for achieving required performance. Economical optimization thus can be achieved by reduction in number of equipment.
ACKNOWLEDGEMENTS
The authors would like to acknowledge CAMM project for funding this research. We would also like to acknowledge mining organization and personnel for their help in providing valuable data.
Uncertainty analysis of production in open pit mines – operational parameter regression… 159
REFERENCES
ABDEL SABOUR S., DIMITRAKOPOULOS R., KUMRAL M., 2008. Mine design selection under uncertainty, Mining Technology 117,2, 53–64.
ALARIE S., GAMACHE M., 2002. Overview of solution strategies used in truck dispatching systems for open pit mines, International Journal of Surface Mining, Reclamation and Environment 16,1, 59–76.
BARABADI A. (2011), Production performance analysis: Reliability, maintainability and operational conditions. University of Stavanger,Norway, Phd. Thesis .
BARABADI A., BARABADY, J. MARKESET T., 2011. A methodology for through- put capacity anal-ysis of a production facility considering environment condition, Reliability Engineering & System Safety 96,12, 1637–1646.
DEHGHANI H., ATAEE-POUR M., 2012. Determination of the effect of operating cost uncertainty on mining project evaluation, Resources Policy 37,1, 109–117.
DHILLON B. S. (2008), Mining equipment reliability, maintainability safety, Springer Science & Busi-ness Media.
DIMITRAKOPOULOS R.G., SABOUR S.A.A., 2007. Evaluating mine plans un- der uncertainty: Can the real options make a difference? Resources Policy 32,3, 116–125.
EKIPMAN A., BİR S.İ., SİSTEMİ K.D., PROSESİ A.H., 2003. A decision support system for optimal equipment selection in open pit mining: analytical hierarchy process. Istanbul University, Mining Engineering Department, 16(2), 1-11.
ERCELEBI S., BASCETIN A., 2009. Optimization of shovel-truck system for surface mining, Journal of The Southern African Institute of Mining and Metallurgy 109,7, 433–439.
GHODRATI B., 2005. Reliability and operating environment based spare parts planning, Lulea, Univer-sity of Technology, Ph.D Thesis.
GUSTAFSON A., 2011. Automation of load haul dump machines, Lulea University of Technology, Ph.D.Thesis.
HAIDAR A., NAOUM S., HOWES R., TAH J., 1999. Genetic algorithms application and testing for equipment selection, Journal of Construction Engineering and Management 125,1, 32–38.
KAZAKIDIS V.N. (2001), Operating risk: Planning for flexible mining systems, University of British Columbia, PhD thesis.
LANKE A.A., HOSEINIE S.H., GHODRATI B., 2016. Mine production index (MPI)-extension of OEE for bottleneck detection in mining. International Journal of Mining Science and Technology, 26,5, 753-760.
MOHAMMADI M., RAI P., ORAEE K. KUMAR M., 2013. Analysis of availability and utilization of dragline for enhancement of productivity in surface mines-a case study, in: Proceedings of the 23rd World Mining Congress, Montreal.
MOHAMMADI M., RAI P., SINGH U., SINGH S., 2016. Investigation of cycle time segments of drag-line operation in surface coal mine: A statistical approach, Geotechnical and Geological Engineering. 1–10.
PATNAYAK S., TANNANT D., PARSONS I., DEL VALLE V., WONG J., 2008. Operator and dipper tooth influence on electric shovel performance during oil sands mining, International Journal of Min-ing, Reclamation and Environment 22,2, 120– 145.
RAI P., TRIVEDI R., NATH R., 2000. Cycle time and idle time analysis of draglines for increased productivity-a case study, Indian Journal of Engineering and Materials Sciences 7,2, 77–81.
Amol A. LANKE, Behzad GHODARATI, Seyed Hadi HOSEINIE 160
RAMAZAN S., DIMITRAKOPOULOS R., 2013. Production scheduling with uncertain supply: a new solution to the open pit mining problem, Optimization and Engineering 14,2, 361–380.
SAMANTA B., SARKAR B., MUKHERJEE S., 2002. Selection of opencast mining equipment by a multi-criteria decision-making process, Mining Technology 111,2, 136–142.
SAMANTA B., SARKAR B., MUKHERJEE S., 2004. Reliability modelling and performance analyses of an LHD system in mining, Journal of the South African Institute of Mining and Metallurgy 104,1, 1–8.
VERNON R., 1984. Uncertainty in the resource industries: the special role of state- owned enterprises, in: Risk and the Political Economy of Resource Development, Springer, 207–223.
WALKER W.E., HARREMOËS P., ROTMANS J., VAN DER SLUIJS J. P., VAN ASSELT M. B., JANSSEN, P., KRAYER VON KRAUSS M. P., 2003. Defining uncertainty: a conceptual basis for uncertainty management in model-based decision support. Integrated assessment, 4(1), 5-17.