International Journal of Sustainable Development and

Planning
Vol., No., Month, Year, pp. **-**
Journal homepage: http://iieta.org/journals/ijsdp

Sustainable Transportation Planning for Djebel Onk's Open-Pit Phosphate Mine in Algeria:
Using Queuing Theory and Cost-Emission Analysis
Mokhtar Debbouz1 , Zoubir Aoulmi1,2 , Chamseddine Fehdi3 , Moussa Attia1,2,*
1
Institute of Mines, Echahid Cheikh Larbi Tebessi University, Tebessa 12002, Algeria
2
Environment Laboratory, Institute of Mines, Echahid Cheikh Larbi Tebessi University, Tebessa 12002, Algeria
3
Laboratory of Water and Environment, Faculty of Exact Sciences and Natural and Life Sciences, Echahid Cheikh Larbi
Tebessi University, Tebessa 12002, Algeria

Corresponding Author Email: [email protected]

Copyright: ©2024 The authors. This article is published by IIETA and is licensed under the CC BY 4.0 license
(http://creativecommons.org/licenses/by/4.0/).

https://doi.org/10.18280/ijsdp.xxxxxx ABSTRACT

Received: Efficient mineral transportation is critical for sustainable open-cast mining operations.
Revised: Queuing theory offers a practical approach to optimizing truck-shovel systems and address
Accepted: truck waiting times at loading and unloading sites. This research evaluates the excavator-truck
Available online: system at an Algerian open-pit phosphate mine using an M/M/1 queuing model. The model
reveals relationships between truck fleet size and queue length, waiting time, shovel
utilization, and overall production. Moreover, loading and transportation costs are analysed to
Keywords: determine the optimal truck fleet size that minimizes costs and emissions. The Match factor
Truck-shovel, Haulage system, Queuing further evaluates fleet compatibility for sustainable planning. This systems analysis provides
theory, Production rate, phosphate, insights into achieving efficient, low-emission truck-shovel operations through optimized fleet
Match factor. sizes, reduced waiting times, and cost-emission optimization. The integrated queuing model
and planning techniques presented can guide sustainable planning of open-pit mining
transportation systems. Focusing on efficiency, costs, and emissions allows strategic
optimization for both economic and environmental sustainability.

1. INTRODUCTION results in higher loader utilization but lower truck utilization,

whereas under-tucking lowers shovel utilization and hence
The global economy is significantly impacted by the lowers production. For instance, excessive trucking reduces
mining sector. The five primary stages of mining process are truck productivity while increasing shovel production until
exploration, development, exploitation, reclamation, and the service rate reaches its ideal level [7].
prospecting. Among the techniques employed in surface Queuing theory stands out as a commonly used method in
mining are dredging, mountaintop removal, open-pit, open- optimizing shovel-truck production, primarily for its capacity
cast, and strip mining. Interestingly, over 60% of total surface to handle randomly increasing demands [8]. Its effectiveness
output comes from open-pit operations [1]. Loading, loaded lies in offering enhanced estimations for wait times within a
transit, dump maneuvering, dumping, and returning to the haulage system. Compared to simulation techniques, queuing
loader are the steps involved in open-pit mining. The choices theory boasts faster computation times and simpler
about when and how to complete these tasks are made by formulation, expediting the transmission of crucial
mining engineers or planners [2]. information for more efficient operations [9]. In queuing
The primary duties in mining transportation involve theory, two primary service channel types are recognized: the
loading and hauling, commonly performed through the use of single-server channel and the multi-server channel. Within a
a truck and shovel technique [3]. This system is precisely single-server channel, trucks are directed to a lone loader
tailored to meet the company's production goals, both in the within the system [10].
short and long term. The design considers the size of the fleet Koenigsberg (1960) pioneered the utilization of queuing
and the loaders in order to obtain the optimal equipment theory within mining operations, specifically modeling
combination [4]. Equipment selection is an optimization conventional and mechanized room-and-pillar mining setups
technique that identifies the fleet size with the lowest total to establish a closed-loop queuing system. This system
haulage cost and maximum equipment utilization [5]. accounted for a finite number of mine faces, assuming
At loading and unloading locations, there are occasionally exponential service time distributions. Despite simulation's
wait durations. The waiting periods at these stations raise the prevalence in the 1960s, queuing theory gained popularity
unit cost per material carried since they lower operating due to the substantial computer memory and CPU (central
capacity [6]. It is clear that when there is over- or under- processing unit) demands of simulations at that time.
trucking, waiting times happen. Overtucking the system Consequently, analytical modeling methods like queuing
theory, requiring minimal or no computational resources, queue length. This model's applicability extends to any
emerged as a viable alternative to computer simulation haulage system provided the data on truck arrival times and
models [11]. shovel service times conform to an exponential distribution
Karshenas (1989) adapted the queuing approach for [16].
equipment selection, which was further investigated by The primary objective of this study is to improve decision-
Sabha [12]. Their aim was to create a model for determining making by optimizing the truck loader system using queuing
the optimal truck fleet size by employing a more precise theory. The specific aims of this research encompass:
productivity estimate, thereby minimizing the expenses • Assessing the efficiency of the shovel-truck system.
related to idle equipment. El-Moslmani (2002) introduced a • Investigating the effects of varying the truck fleet
queuing model, termed FLSELECTOR, that factored in size on parameters such as queue length, waiting time, shovel
server utilization in computing haulage system production. utilization, production, and operating costs.
This computer-based model facilitated the selection of the
optimal fleet size based on cost minimization and 2. METHODOLOGY
maximizing production output [13].
Krause and Musingwini demonstrated that a modified 2.1 Queuing theory
version of the "Machine Repair Model," an example of a
finite source queuing model, could effectively estimate fleet The basic sequence delineated in the queuing models is as
size and produce accurate results. They tailored the equations follows [17]: Trucks requiring loading services are
of this model to suit loading and hauling scenarios, adjusting consistently generated by an input source. Upon arrival, these
the average time trucks spent waiting for repair to represent trucks join a queue if no loaded truck is currently present. The
the average queuing time at dumpsites or loading sites [14, initial truck in the queue is selected for service, undergoing
15]. In a separate study, Hai (2016) utilized a single-channel the requisite loading operation, after which the truck departs
queuing model to establish correlations between varying from the queuing system (Figure 1).
truck numbers, shovel utilization, system production, and

Figure 1. Single server queue system

The subsequent analytical equations were formulated to +

represent the practical loading and hauling procedures Lq = N − (1 − P0 ) (3)
through the application of the queuing model (M/M/1), which 
was employed for the truck-shovel system at the Open Pit
Mine [18]. Probable number of trucks in system
Probability that the system operating without any trucks.
LS = Lq + 1 − P(0) (4)
1
P (0) =
N  N!  
n
 (1) Expected waiting time for trucks in the queue, measured in

n =0 
   minutes.
 ( N − n)!    
Lq
λ Average number of trucks washed per period. Wq = (5)
µ is the Service rate  ( N − L)
Probability of “n” trucks in system
Expected waiting time of trucks in system, (minutes)
n
N! 
( N − n ) !   
P (n) = P (0) (2) 1
WS = Wq + (6)

Probable number of trucks in the queue
 The utilization of the shovel, denoted as S , and the utilization of the trucks, denoted as tr , are provided in
Equations 7 and 8, respectively [8]. Ctruck: Cost of operating truck and shovel per hour,
Cshovel : Cost of operating shovel per hour.

 S = 1 − P0 (7) 2.2 Method to find a mismatch of equipment

Match factor (MF) is the most commonly used method to

Wq measure efficiency and productivity of an equipment fleet. It
tr = (8) is defined as the ratio of truck arrival rate to shovel service.
Wq −Turck Cycle time
Match factor is commonly used to measure the compatibility
among trucks and shovels in terms of fleet size, truck cycle
Hourly production of the shovel Qn and shovel loading times [19].

Qn = N    M tr (9) The calculations made to get the match factor for trucks
are given

Average Truck cycle time

Capacity of truck in ton. Match Number = (13)
N: number of trucks Average Shovel Load cycle time
Truck operating cost, (DA/hr.)
Qn = N    M tr (14)
N .Cturck
Ctr = (10)
Qn 3. CASE STUDY

Shovel operating cost, (DA/hr.) The research area, located in the Djebel Onk mining basin,
is positioned in the Tebessa region, approximately 600 km
southeast of Algiers and just 20 km from the Algerian-
C shovel
Csh = (11) Tunisian borders (Fig. 2). In this region, attention is directed
Qn towards the Kef Essenoun deposit among the five identified
deposits, which is presently undergoing extraction of
Total cost (cost of use of trucks + Shovel operating cost), phosphate raw material. As of 2019, the annual production
(DA/hr.) from this site was estimated to be around 1.2 million metric
tons, with the overall resource of phosphate raw material
approximated at 2.2 billion metric tons [20].
CT = Cshovel + N .Ctr (12)

Figure 2. Satellite Image (Google Earth Map) Depicting the Position of the Phosphate-Ore mine
 Table 1. displays the input data, with the arrival and 4. ANALYSIS AND RESULTS
service rates being derived from experimental data
The haulage system in this study employed a queuing
DATA Values model with a service parameter of 1, adhering to the (M/M/1)
Arrival rate (λ) 2.66 trucks/hour format. Following the computation, the model produced
Service rate (μ) 22.47 trucks/hour various outcomes, including waiting time for truck loading
Number of shovels (s ) 1 (Wq), server utilization, shovel production, and related costs.
Traffic of shovels (ρ = λ/ μ) 0.1183 These precise values are succinctly presented in Table 2.
Cost of operating shovel , Cshovel 250 $ (33600 DA)
Cost of operating trucks; Ctruck 180 $ (24192)
Shovel average loading time (min.) 2.67
Truck average cycle time (min.) 22.47
Truck capacity 55 ton
Loading capacity 6 m3

Table 2. Queuing model outputs

Truck Lq Ls Wq Ws Shovel Production Cost of Cost of Total Cost, Truck

fleet (trucks) (trucks) (min) (min) utilization Qn, ton/h loading, hauling DA/h utilisation
size ηs Csh/Qn KCtr/Qn ηt
2 0.022 0.232 0.283 2.95 0.2094 258.730 129.865 187.006 316.871 0.9876
3 0.072 0.382 0.618 3.29 0.3100 383.056 87.716 189.466 277.182 0.9732
4 0.155 0.562 1.025 3.71 0.4070 502.936 66.808 192.406 259.214 0.9564
5 0.281 0.780 1.537 4.27 0.4995 617.328 54.428 195.941 250.369 0.9360
6 0.458 1.044 2.192 5.00 0.5866 724.997 46.345 200.210 246.555 0.9111
7 0.697 1.364 3.038 5.95 0.6672 824.543 40.750 205.379 246.129 0.8809

8 1.009 1.749 4.118 7.14 0.7400 914.479 36.742 211.635 248.377 0.8451

9 1.402 2.206 5.441 8.56 0.8043 993.976 33.804 219.048 252.851 0.8051
10 1.896 2.754 7.057 10.25 0.8578 1060.153 31.694 228.193 259.887 0.7610
11 2.483 3.384 8.854 12.069 0.902 1114.2056 30.156 238.836 268.992 0.7173
12 3.165 4.100 10.775 13.959 0.935 1155.7676 29.072 251.179 280.250 0.675

Graphs illustrating the impact of variations in fleet size on utilization, leading to increased production. Conversely, the
queue length, waiting time for trucks in the queue, shovel addition of five more trucks to the system resulted in a nearly
utilization, and costs can be generated and observed in 10% decrease in truck utilization.
Figures 3 through 7. Analyzing the relationship depicted in Figure 7 between the
The graphs reveal an increasing trend in the correlation number of trucks and operational costs, it becomes apparent
between the number of trucks and various outputs from the that loading costs and hauling costs are in direct opposition.
queuing model, including queue length, waiting time in Increasing the number of trucks leads to a reduction in
queue, shovel utilization, and production. With an expansion loading costs but an elevation in hauling costs. The total cost
in fleet size, there is a corresponding rise in the number of is derived by combining the hauling and loading cost curves.
trucks awaiting service at the shovel, leading to an increase This cost is minimized at the curve's lowest point, indicating
in queue time. Interestingly, a greater number of dispatched the optimal number of trucks in the queue system. Referring
trucks results in a more efficient performance of the shovel. to Figure 7, the optimized position is at 7 trucks. In other
As illustrated in Figure 6, adding two more trucks to the words, deploying 7 trucks for 1 shovel in this scenario
system resulted in an approximately 20% rise in shovel ensures the system operates at the most cost-effective level.

Figure 3. Trucks in queue length, and in system Figure 4. Relationship between waiting time and
number of truck
 2. Match Factor < 1.0:
- Indicates a form of mismatch.
- Excess loading capacity is observed within the system.
- While loader efficiency remains near 100%, transporter
efficiency experiences a reduction.

3. Match Factor > 1.0:

- Signals a different type of mismatch.
- Shovel efficiency reaches 100%.
- However, hauler efficiency declines in this scenario.

In essence, it's apparent that the match factor serves as a

pivotal parameter in evaluating the system's efficiency. A
match factor of 1.0 symbolizes an ideal equilibrium between
Figure 5. Relationship between production and number the loader and hauler, whereas deviations from this value
of trucks highlight imbalances leading to either surplus loader capacity
or compromised hauler efficiency.

Upon reviewing the results presented in Table 3 and

analyzing Figure 8, it becomes evident that the most
advantageous scenario emerges with the utilization of 8
trucks. This conclusion closely aligns with the outcomes
derived from the queuing method, as evidenced by the
closely matching cost values. This finding underscores the
significance of carefully considering the match factor in
optimizing transportation systems within the studied context.

Table 3. Efficiency and Match Factor

N (trucks) MF % Efficiency
2 0.23712256 23.7122558
3 0.35568384 35.5683837
Figure 6. Correlation between Shovel Utilization and Truck 4 0.47424512 47.4245115
Utilization in Relation to the Number of Trucks 5 0.59280639 59.2806394
6 0.71136767 71.1367673
7 0.82992895 82.9928952
8 0.94849023 94.8490231
9 1.06705151 93.294849
10 1.18561279 81.4387211
11 1.30417407 69.5825933
12 1.42273535 57.7264654

Figure 7. Relationship between Cost and number of trucks

From Figure 8, it's evident that the efficiency of the system

correlates directly with the match factor. A match factor of
precisely 1.0 signifies a perfect alignment, resulting in
optimal efficiency at 100% within the system. However, any
deviation from this value indicates a degree of mismatch
between the components. Here's a detailed breakdown of the
insights provided:

1. Perfect Match (Match Factor = 1.0):

- System efficiency achieves peak performance at 100%.
- Both the loader and hauler operate at their maximum Figure 8. Efficiency and Match Factor Analysis
efficiency levels.
5. CONCLUSION [3] Derbal, C., Djellali, A., Attia, M., Zaamouche, F., &
Belkhiri, A. (2023). Enhancing Sustainable Mining
In conclusion, the mining industry's focus on Practices Through Fracture-Informed Blasting
transportation and packaging systems, which collectively Strategies: A Case Study of the Aïn El Kebira Limestone
represent about 60 percent of production costs, underscores Quarry in Algeria. International Journal of Sustainable
the critical importance of optimizing these processes. This Development & Planning, 18(12).
study delves specifically into the shovel-truck system within https://doi.org/10.18280/ijsdp.181201
an open-cast phosphate mine located in eastern Algeria. By [4] Liu, S. Q., Lin, Z., Li, D., Li, X., Kozan, E., & Masoud,
employing the queuing model (M/M/1), we have M. (2022). Recent Research Agendas in Mining
significantly advanced our understanding of the Equipment Management: A Review. Mining, 2(4), 769-
transportation system's performance and identified optimal 790. https://doi.org/10.3390/mining2040043
solutions. [5] Burt, C. N., & Caccetta, L. (2018). Equipment selection
Practical experimentation conducted on-site provided for mining: with case studies (Vol. 150). Springer.
crucial input parameters, including arrival and service rates, https://doi.org/10.1007/978-3-319-76255-5
which were essential for modeling the system accurately. [6] Moghrani, R., Aoulmi, Z., & Attia, M. (2023). Hybrid
Through this method, we were able to obtain various outputs, RPI-MCDM Approach for FMEA: A Case Study on Belt
shedding light on critical system parameters such as queue Conveyor in Bir El Ater Mine, Algeria. Journal
length, waiting time, shovel utilization, and overall system Européen des Systèmes Automatisés, 56(3).
production. https://doi.org/10.18280/jesa.560314
A key finding of this study was the discernible correlations [7] Attia, M., Bechouat, M., Sedraoui, M., & Aoulmi, Z.
observed between these variables and the size of the truck (2022). An Optimal Linear Quadratic Regulator in
fleet. As the fleet size increased, the number of trucks waiting Closed Loop with Boost Converter for Current
for the shovel and the time spent within the system also Photovoltaic Application. European Journal of Electrical
increased. However, this led to noteworthy improvements in Engineering/Revue Internationale de Génie Electrique,
shovel utilization rates and overall production efficiency. 24(2). https://doi.org/10.18280/ejee.240204
The determination of the optimal number of trucks to be [8] Cundiff, J. S., Grisso, R. D., Resop, J. P., & Ignosh, J.
dispatched into the system within the queuing model was (2023). Load-Out and Hauling Cost Increase with
pivotal. This optimization was achieved by minimizing the Increasing Feedstock Production Area. AgriEngineering,
total operating cost, which was derived by graphically 5(4), 1694-1712.
summing the loading and hauling expenses. The point where https://doi.org/10.3390/agriengineering5040105
the total cost curve reaches its lowest represents the [9] Grozdanović, P., Gligorijević, A., Andrejić, M., Nikolić,
optimized number of trucks for this queuing system. M., & Kilibarda, M. (2023). A New Model for
Furthermore, the utilization of the Match factor allowed us Determining the Price of Product Distribution Based on
to assess the compatibility between trucks and shovels. By Fuzzy Logic. Logistics, 7(3), 62.
considering fleet size, truck cycle, and shovel loading times, https://doi.org/10.3390/logistics7030062
we gained valuable insights into the efficiency and synergy [10] Elijah, K., Joseph, G., Samuel, M., & Mauti, D. (2021).
between these components of the transportation system. Optimisation of shovel-truck haulage system in an open
In essence, this study not only provides practical insights pit using queuing approach. Arabian Journal of
into optimizing transportation systems within open-pit mines Geosciences, 14(11), 973.
but also underscores the importance of utilizing mathematical https://doi.org/10.1007/s12517-021-07365-z
models such as queuing theory to enhance operational [11] Rovetto, C., Cruz, E., Nuñez, I., Santana, K., Smolarz,
efficiency and ultimately drive sustainable practices within A., Rangel, J., & Cano, E. E. (2023). Minimizing
the mining industry. Intersection Waiting Time: Proposal of a Queue
Network Model Using Kendall&rsquo;s Notation in
ACKNOWLEDGMENT Panama City. Applied Sciences, 13(18), 10030.
https://doi.org/10.3390/app131810030
The authors wish to extend their heartfelt thanks to the [12] Galimulina, F. F., & Barsegyan, N. V. (2024).
management professionals, employees, and laborers at the Application of Mass Service Theory to Economic
Djebel Onk mine in Algeria. Their support in furnishing Systems Optimization Problems—A Review.
essential technical details and offering feedback on the data Mathematics, 12(3), 403.
and experimental work is greatly appreciated. https://doi.org/10.3390/math12030403
[13] Koenigsberg, E. (1960). Finite queues and cyclic queues.
REFERENCES Operations Research, 8(2), 246-253.
https://doi.org/10.1287/opre.8.2.246
[1] Welch, C., Barbour, S. L., & Hendry, M. J. (2021). The [14] Sabha, F. H. (2012). Modeling reclamation earthwork
geochemistry and hydrology of coal waste rock dumps: operations using special purpose simulation tool.
A systematic global review. Science of the Total https://doi.org/10.7939/R3D433
Environment, 795, 148798. [15] El-Moslmani, K. (2002). Fleet selection for earthmoving
https://doi.org/10.1016/j.scitotenv.2021.148798 operations using queueing method Concordia
[2] Al-Masri, M. (2022). An approach for Evaluating the University].
Full Truck and Full Bucket Loading Strategies in Open- [16] Krause, A. J. (2006). Shovel-truck cycle simulation
Pit Mining Using a Discrete Event Simulation and methods in surface mining University of the
Machine Learning. Witwatersrand.
https://doi.org/https://doi.org/10.7939/r3-1dcb-rx31
[17] Musingwini, C., Ali, M. M., & Dikgale, T. (2009). A
linear programming and stochastic analysis of mining
replacement rate for typical Bushveld Complex platinum
reef conventional mining under variable geological
losses. In Proceedings of the Eighteenth International
Symposium on Mine Planning and Equipment Selection
(MPES 2009) (pp. 551-565).
[18] Hai, D. V. (2016). Optimization of truck and shovel for
haulage system in the cao son mine, Viet Nam using
queuing theory Prince of Songkla University.
[19] Selinka, G., Franz, A., & Stolletz, R. (2016). Time-
dependent performance approximation of truck handling
operations at an air cargo terminal. Computers &
Operations Research, 65, 164-173.
https://doi.org/10.1016/j.cor.2014.06.005
[20] Czaplicki, J. M. (2008). Shovel-Truck Systems:
Modelling, Analysis and Calculations. CRC Press.
https://books.google.dz/books?id=IzDMBQAAQBAJ
[21] Burt, C. N., & Caccetta, L. (2007). Match factor for
heterogeneous truck and loader fleets. International
Journal of Mining, Reclamation and Environment, 21(4),
262-270. https://doi.org/10.1080/17480930701388606
Boumaza, B., Kechiched, R., & Chekushina, T. V.
(2021). Trace metal elements in phosphate rock wastes
from the Djebel Onk mining area (Tébessa, eastern
Algeria): A geochemical study and environmental
implications. Applied Geochemistry, 127, 104910.
https://doi.org/10.1016/j.apgeochem.2021.104910