Logistic and Operation Management Research (LOMR), Vol. 2 No. 1 (2023) https://doi.org/10.31098/lomr.v2i1.

1529
Research Paper

Optimizing Coal Hauling Operations: A Strategic Plan for Increased

Productivity at PT Borneo

Galuh Joko Pamungkas*1, Nur Budi Mulyono1

1Institut Teknologi Bandung, Indonesia

Received: January 10, 2023 Revised: May 3, 2023 Accepted: May 12, 2023 Online: May 29, 2023
Abstract

The purpose of this research is to determine the key operational variables that impact productivity in PT
BORNEO and identify optimal targets to improve productivity. A mixed methodology of qualitative and
quantitative approaches, supported by process observation, was used to collect data. The study primarily uses
existing quantitative data, but interviews may be conducted to capture constraints in the implementation plan
process. GPS data is also collected via an API to obtain information about truck movements, which can be divided
into various activities to identify cycle time and analyze system performance. The study found that reducing
loading time and minimizing random stoppages can significantly improve productivity, with average loading
time being the most important factor. Stoppage before rest time and stoppage in loaded condition and at major
prayer time had little impact on overall productivity. Future research could investigate the impact of other
variables such as equipment maintenance and driver training on productivity.

Keywords Productivity, Operational variables, Optimization

INTRODUCTION
Coal mining has been a major economic driver for many countries around the world,
including Indonesia (Armis & Kanegae, 2020). PT BORNEO, a subsidiary of a local conglomerate
group in Indonesia, is one of the country's top five coal producers. With an annual permitted
production of 36 million tons of coal, PT BORNEO aims to become one of the largest coal mines in
Indonesia. As with most mining companies, PT BORNEO follows a standard business process that
starts with land clearing and ends with shipment to customers.
However, coal mining has been associated with environmental concerns, including air
pollution, land degradation, and water pollution (Kalisz et al., 2022). Mining companies are often
required to meet stringent environmental regulations and to ensure that their operations do not
harm the environment. Additionally, the volatility of commodity prices, depleted reserves, rising
costs, and changes in government regulations have forced mining companies to reassess their
strategies to improve efficiency and flexibility to reap the opportunities.
In 2021, the world faced two major events that had a significant impact on the coal market:
the ongoing COVID-19 pandemic and the war between Russia and Ukraine (Adolfsen et al., 2022).
The pandemic caused disruptions in global supply chains, leading to shortages in some countries
and an increase in demand for certain goods. One of the goods that experienced a surge in demand
was thermal coal, which is used to generate electricity. At the same time, the conflict between
Russia and Ukraine disrupted natural gas supplies to Europe, which caused many countries to turn
to coal as an alternative source of energy (Brodny & Tutak, 2022). This led to a surge in demand
for coal, which caused the price of thermal coal to soar. Indonesia, as one of the world's largest
exporters of thermal coal, was no exception.
Although there are global efforts to reduce coal consumption and promote renewable
energy, the idea that coal is still necessary due to the lack of economic alternatives and lower
efficiency of renewable energy is a common misconception. Renewable energy sources such as

Copyright Holder: This Article is Licensed Under:

© Galuh & Nur. (2023)
Corresponding author’s email: [email protected]
 Log Op Man Res.

solar, wind, and hydro have already reached economical scales and have been replacing fossil fuels
in many parts of the world. Furthermore, coal is essential for the production process of steel and
cement, but its use can be reduced by transitioning to alternative processes that rely on renewable
energy. Finally, many countries have significant domestic reserves of coal, which are traditionally
used for domestic purposes.
Previous studies have examined the impact of improving the quality of coal on productivity.
Das and Sengupta (2004), found that productivity can be significantly increased by improving the
quality of coal. Similarly, (Leung & Scheding, 2015) found that the quality of coal has a significant
impact on productivity. Another study by (Zhang et al., 2011) examined the effects of coal-resource
integration in developing countries and regions. The analysis showed that such integration projects
can effectively improve mining technologies, collect capital, and encourage international
cooperation and exchange. These findings provide important insights into the factors that can
impact productivity and the potential benefits of coal-resource integration projects in developing
countries and regions.
Based on the literature review, while previous studies have highlighted the positive impact
of improving coal quality on productivity and the potential benefits of coal-resource integration
projects in developing countries, there appears to be a research gap in the specific context of
Indonesia. Specifically, there is a lack of research on the potential for improving coal quality in
Indonesia, and the impact that such improvements could have on productivity in the country's coal
mining sector. This thesis will explore the business process of PT BORNEO, a coal mining company
in Indonesia, and analyze the company's strategies to improve its efficiency and flexibility in
meeting the demand for coal. The study will also examine the challenges faced by PT BORNEO in
terms of meeting environmental regulations and ensuring sustainability. The research aims to
contribute to the understanding of the coal mining industry in Indonesia and provide insights into
the strategies that mining companies can adopt to overcome the challenges they face.

LITERATURE REVIEW
Coal Hauling Process
An efficient coal hauling process is crucial to meet the production targets set by mining
companies. The coal mining process has several stages, starting from the upstream activities of
digging coal and processing it until it reaches the downstream stage of marketing or delivering the
coal to customers. One of the most critical stages in the mining process is the coal hauling activity,
which plays a vital role in maintaining a stable supply of coal products for marketing or delivery to
customers. Therefore, it is necessary to evaluate the effectiveness of the coal hauling process and
identify ways to optimize it.
To achieve optimal hauling productivity, the track flow should be designed to remain
smooth, minimizing waiting times and queues. Efficient hauling processes rely on the seamless
movement of vehicles, requiring the reduction of any obstacles along the route to enhance
operational effectiveness. In this regard, PT BORNEO uses GPS-based tracking technology to
monitor all activities and time-related events in the hauling process. Riyandi and Wibowo (2020)
support this method, emphasizing the importance of hauling activity in the mining industry. By
improving the coal hauling process, mining companies can achieve efficient and effective
production to meet their targets.
Several studies have explored efforts to optimize the coal supply chain by sequencing
business processes to align with shipping schedules. In the mining industry, the coal hauling
process plays a crucial role, and various strategies have been proposed to enhance its efficiency.
However, limited research has specifically addressed coal hauling. Therefore, this study focuses on
the coal hauling process from ROM (Run of Mine) to the port at PT BORNEO, aiming to identify the

64
 Log Op Man Res.

challenges faced by the company and provide recommendations for optimizing the process.
This study's significance lies in its contribution to the professional world in the mining
industry and other industries that use closed-loop hauling systems. It is believed that this research
will lead to organizational process improvement and business sustainability for PT BORNEO in the
long term. This research will help identify the factors that affect the efficiency of the coal hauling
process and provide recommendations for optimizing the process. Ultimately, this study seeks to
contribute to the development of a more efficient and effective coal-hauling process in the mining
industry.

Hauling Truck Productivity

The productivity of hauling trucks is crucial for the efficient operation of mining industries.
Several factors can affect the productivity of trucks, including the type and characteristics of soil,
payload in terms of weight or volume, road grades, distance, and haul return, the efficiency and
condition of equipment and workplace, and the loader's capacity/productivity. To measure truck
productivity, a formula is used, which involves calculating efficiency and cycle time. Cycle time
comprises several components, including loading time, loaded travel time, dump positioning time,
dumping time, empty travel time, and spotting time.
According to a study by Thompson and Visser (2006), the selection of truck size and the
number of trucks required to move material depends on several factors, including the type of
material, haul distance, and frequency of hauls. In addition, the study suggests that the productivity
of hauling trucks can be improved by optimizing the haulage route and maintaining the equipment
properly.
Similarly, a study by Chaowasakoo et al. (2017) on the performance of mining equipment
recommends using a fleet management system that can help optimize the routing and scheduling
of trucks, reducing cycle times and idle time for loaders. The study also suggests using predictive
maintenance techniques and training operators to improve the efficiency of equipment and reduce
downtime.
Moreover, a study by Arteaga et al. (2018) on the selection of the optimal fleet size for an
open-pit mine recommends considering the trade-off between the capital cost of equipment and
the operational cost of running the fleet. The study also suggests using simulation models to
evaluate the impact of different factors, such as maintenance policies and road conditions, on the
productivity of hauling trucks.
Another study by Bozorgebrahimi (2004) on the impact of road design on the productivity
of mining trucks highlights the importance of considering road design factors, such as grades,
curves, and width, in improving truck productivity. The study suggests that optimizing road design
can reduce cycle times, fuel consumption, and maintenance costs, resulting in overall productivity
gains for mining operations.
In summary, the productivity of hauling trucks in mining operations depends on several
factors, including equipment efficiency, road conditions, and maintenance policies. Various studies
recommend using fleet management systems, optimizing haulage routes, and maintaining
equipment to improve truck productivity. Additionally, the selection of the optimal fleet size and
the consideration of road design factors can also contribute to the overall productivity of mining
operations.

RESEARCH METHOD
Root Cause Analysis
Cause Mapping & 5 Why Method
Cause mapping is a technique that provides a visual representation of why a particular

65
 Log Op Man Res.

event or issue occurred. It creates a connection between individual causes and effects to unveil the
system of causes behind an event. The technique is widely used in various studies as a root cause
analysis tool. A cause map always starts with a deviation from the ideal or target state and is built
by investigating the problem and then backing into the causes by asking why questions. A cause
map is built from left to right, and the correlation between causes and effects is created through
arrows that read "was caused by" (Putri & Susanto, 2017). The technique asks questions beginning
with "Why did this event happen?" and continues until the cause(s) are identified. The Cause
Mapping Method also includes the question, "What was required to produce this effect?" to build a
more detailed cause map that provides a complete representation of the issue.

Control Chart
A control chart is a graphical representation of how a process changes over time. Data are
plotted in time series and a central line is drawn to indicate the average, while upper and lower
lines are used to indicate the control limits. These lines are determined from historical data. By
comparing current data to these lines, one can conclude whether the process variation is consistent
or unpredictable (Flores et al., 2020). Control charts are used in various industries to control
processes by identifying and correcting problems as they occur, predicting outcomes from a
process, analyzing patterns from special causes or common causes, and determining whether a
quality improvement project should aim to make fundamental changes. There are many types of
control charts available that can be used, depending on the nature of the process and data. The basic
procedure for building a control chart involves choosing the appropriate type, selecting a suitable
period for data collection, collecting and analyzing data, and looking for out-of-control signals.

Scatter Plot / Scatter Diagram

A scatter plot is a diagram that graphs pairs of numerical variables, with one variable on
each axis to examine the relationship between the two variables. If the variables are correlated, the
points will fall along a line or curve. Scatter plots are used when we want to compare pairs of data
when the dependent variable might have multiple values for each of the independent variables
when we want to find correlations between two variables, identify potential root causes of
problems, or test an autocorrelation. Creating a scatter plot involves identifying the variables,
determining the type of correlation between the variables, plotting the points, drawing a line of best
fit, and interpreting the results (Elearn, 2021).

Hauling Truck Productivity Calculation

According to the Mining Equipment and Material Handling book by Mousavi & Naghdi
(2013), to measure truck productivity, we can use this formula:

P = E x 60 / C
with:
P = productivity (measurement unit (BCM)/hr)
E = Efficiency (%)
C = Cycle Time Truck (minutes)

The cycle time was calculated from the time trucks started loading at the loading point until
they came back to the loading point and were ready to be loaded again. Commonly the cycle time
components can be broken down into:

C = Cycle Time Truck = a + b + c + d + e + f

66
 Log Op Man Res.

a. Loading time: time for the truck to be fully loaded starting after the truck finished spotting.
b. Loaded travel time: time for truck to move from loading point to dumping point
c. Dump positioning time: time for the truck to position itself on the dumping point
d. Dumping time: time for the truck to empty all the cargo in the vessel and put the vessel again
on the truck.
e. Empty travel time: time for truck to move from dumping point to loading point
f. Spotting time: time for the truck to position itself near the excavator/loader and ready to be
loaded

This cycle time calculation will lead to truck number requirement calculation for moving the
material nonstop from the loading point to the dumping point in such a way that the
loader/excavator will have minimum idle time.

Matching Factors
Matching factor is an approach commonly used in mining to measure the matching degree
between loader equipment and hauling equipment in every load and haul activity. The matching
factor value lies between 0 and 1, whereas 1 means that the loader and hauler work at 100%
capacity, and there is no idle time either for the loader or hauler.
There are some variations of the matching factor formula, but all compare loader capacity
and hauler capacity. The matching factors that will be used in this research are:

MF = 𝑁𝑡 𝑥 𝐶𝑙 ; where,
𝑁𝑙 𝑥 𝐶𝑡

MF = Matching Factor
Nt = number of trucks or haulers
Nl = number of diggers or loaders
Cl = cycle time of diggers or loaders
Ct = cycle time of trucks or haulers

Based on the MF’s formula, there are three possibilities of value from the matching factor:
a. MF <1 means the hauler’s capacity is assigned at 100% capacity while the loader works under
100% capacity. This causes the loader to wait for trucks to arrive.
b. MF = 1 means loader and hauler working at 100% of their capacity. There is no waiting time
for the loader and there is no queue time for the hauler.
c. MF >1 means the hauler’s capacity is assigned under 100% and the loader works at 100%
capacity. This will lead to a situation where trucks will form a queue line at the loading point.

Matching factors are indirectly affected by other factors such as material type, depth of cut,
swing angle, working conditions, truck capacity, bucket capacity, operator skills, and equipment
condition.

Data Collection Method

For this study, a mixed methodology of qualitative and quantitative approaches was used
to collect data, which was further supported by process observation. The study aims to address
business problems and provide proposed solutions that are suitable for the current conditions,
needs, and constraints of PT BORNEO. The research will primarily focus on evaluating the
underlying causes of consistently unachieved RTP (ROM to PORT) productivity.
67
 Log Op Man Res.

The first phase of the study involved identifying the problems that need to be addressed by
PT BORNEO. This will be followed by the second phase, which will delve deeper into the problem
and produce possible business problem scenarios. The root causes of the business problem will be
identified during this phase. The third phase will determine the most feasible solution and
implementation plan.
This study primarily uses quantitative data that already exists but has not been used to
generate meaningful insights and solutions. Interviews may be conducted to capture any
constraints in the implementation plan process. GPS data is also collected through IT-enabled
means via an API, as it consists of millions of data rows. Truck movements are essential to obtain
information and trends about transportation processes, including any congestion or obstacles that
occur in the system. These movements can be divided into several activities, and by knowing each
of these activities, the cycle time will be identified and the expected system performance can be
analyzed.
Lastly, the data collected and used by the company to evaluate performance in daily or shift
aggregates will be utilized. Using this data, the researchers can identify variables that may be the
root cause of issues in the hauling process. Through this mixed methodology, the study aims to
provide feasible solutions to improve the company's productivity and performance.

Data Analysis
Considering the quality of sensors and the possibility of noise in raw data and report data,
data analysis begins with data preparation and cleaning, outlier detection, and clean data isolation.
The next step is to test the assumptions that need to be met for statistical analysis. After the data is
ready, the analysis will be done using statistical tools such as the Goodness of fit test, one sample t-
test, paired sample t-test, spearman correlation, statical power measurement, run test, and
Wilcoxon test.
Each component describes the average cycle time of the truck in each segment. In the next
step, researchers will correlate each variable that potentially affects normal cycle time. Any
significant variables will be isolated for the verification step. From this last step, it can be concluded
which variables truly contribute to the unachieved performance. The final stage of this study
formulate proposed solutions to improve performance and tools to detect and predict performance
deviations from normal conditions or to detect changes from normal conditions in operations so
that performance can be maintained in optimal conditions.

FINDINGS
In this analysis, the first step is to translate the actual process into an analysis model that
represents the system. The tracking system is designed to use geofences at every 500 meters. With
the decent GPS quality, there are many unstable GPS traces in the data. To obtain more meaningful
information from the data, an aggregate model that removes unstable GPS while retaining the GPS
data's significance and capturing relevant information must be created.
In the analytical model, there are two routes in the system. One is built, owned, and
maintained by PT BORNEO (blue line), and the other is built, owned, and maintained by another
company (green line). The cost of using the green route is much higher than the blue PT BORNEO-
owned route, limiting movement using the green route in terms of cost savings and capacity. The
road also only supports from ROM EST to KM12 and from KM12 to PORT-T, with no other entry or
exit points. In the analytical model, there are six ROMs and two PORTs that can be used to transfer
coal commodities from ROM to PORT. POOL is a term used by transport companies as a place for
their operational crews to carry out operational activities such as shift changes, rest time, minor
repairs, etc. Some of them combine their POOL with the workshop, where they can carry out major

68
 Log Op Man Res.

maintenance or other activities that require trucks to be unloaded.

Figure 1. Hauling Road Network

The analytical model divides the road with major intersections where the main road splits
or merges (denoted by S) (Figure 1). There are two roads in the system. One was built, owned, and
maintained by PT BORNEO (blue line), and the other was built, owned, and maintained by another
company (green line). The cost to use green roads is much more expensive than the Blue Line road
owned by PT BORNEO so movement using the Green Line road is limited in terms of saving cost
and capacity.
The distance of the main road varies between one (1) and thirteen (13) kilometers. Some
minor sections from ROM are symbolized by R, while minor sections that connect the Pool with the
main road are symbolized by W. The distances of R and W are very short, all less than 3 km. To
determine the root causes of business problems for the transport company, researchers will employ
root cause mapping and the 5-why method. These methods help identify multiple possible
underlying causes. Each of these root causes must be verified either by observation or by collecting
data that justifies and confirms the causal relationship with the business problem. One root cause
may cause several problems, but it does not mean that the root cause is the main cause. Each
potential root cause should be treated the same until proven otherwise.

69
 Log Op Man Res.

From the problem tree mapping below (Figure 2), several potential root causes could be
the main cause of low transporter productivity due to their possible systematic nature in the
business. Factors that affect all trucks in the system will not be considered as root causes because
when that happens, all trucks will stop, and there will be no production at the appropriate time,
such as flooding, social issues, etc.

Figure 2. Mapping Diagram for Low RTP Productivity

The empty and loaded cycles are built by adding all main road sections for different
conditions (empty and loaded). The queue at the ROM will be approached by counting trucks at
each ROM minus the ROM's output. ROM activities will be measured by direct observation and will
be used as verification for ROM output calculations. The queue at the PORT uses the same approach
method as the ROM. The loaded and empty cycles will be affected by many factors, and these factors
will be represented by standard variations in the loaded and empty cycles. All time used by trucks
from the loaded cycle and return to ROM activities will be the truck cycle time (Table 1).

Table 1. Average Time per Segment (minutes)

Distance Shift 1 Shift 1 Shift 2 Shift 2
Segment
km Empty Loaded Empty Loaded
S1 5 48.7 39.6 51.6 44.1

70
 Log Op Man Res.

Distance Shift 1 Shift 1 Shift 2 Shift 2

Segment
km Empty Loaded Empty Loaded
S1A 4 21.6 50.7 14.6 59.5
S2 3 3.8 4.1 4.2 4.4
S3 4 4.7 5.7 5.1 6.3
S4 5 6.3 8.2 6.9 9.1
S5 3 4.5 4.5 4.7 4.9
S6 12 9.8 29.6 10.1 26.1
S7 19 17.0 24.3 17.9 25.9
S8 13 7.5 3.5 7.5 3.7
R1 0 0.0 0.0 0.0 0.0
R2 2.5 4.0 4.9 4.3 5.7
R3 1 1.3 1.6 1.5 2.2
R4 1 2.4 3.8 2.6 4.1
R5 1.5 2.0 3.4 2.2 3.9
R6 2.5 3.3 5.5 3.5 6.0

ROM and Port Activity

The findings of the study showed that the normal ROM activity time for loading coal onto
trucks was found to be an average of 218 seconds. This was inferred using 1 sample t-test
hypothesis testing, which did not provide enough evidence to accept H1 (≠218 seconds) with a p-
value of >5% (Figure 3). On the other hand, the normal PORT activity time for unloading coal from
trucks and delivering it to the stockpile or transferring it to barges was found to be an average of
92 seconds. Again, this was inferred using 1 sample t-test hypothesis testing, which did not provide
enough evidence to accept H1 (≠92 seconds) with a p-value of >5%.

Figure 3. T-test sample for observation of ROM data

However, some anomalies were observed during the study that caused the PORT output to
slightly decrease. Certain trucks were waiting for the dump body to be perfectly aligned with the
body frame before moving forward, which is an ideal practice but not the best practice for reducing
cycle time. To reduce cycle time, trucks usually move forward when the dump body is only halfway
into the body frame. Additionally, some hoppers dedicated to double trailer dump were only being
dumped by a single truck at a time, which presents an opportunity to further reduce cycle time in
coal hauling operations.

71
 Log Op Man Res.

Figure 4. T-test for observing PORT data

ROM and Port Queue

The study analyzed the factors affecting the productivity of the RTP in Indonesia,
specifically the impact of the number of hauling trucks on productivity. The study found that the
number of trucks had a strong positive correlation with productivity per shift. However, the
relationship was not linear, indicating that the trucks had reached the road limit, and additional
trucks would lead to lower output for one input unit. This phenomenon is known as diminishing
marginal returns. Based on the findings, it is suggested that the company should monitor truck
quantity closely and not add any more trucks into the system.

Figure 5. Average Queue at PORT

The study also examined the truck quantity per hour and found that there were significant
differences in truck quantity at certain times, such as during fatigue time (3 to 5 AM), post-change
shift time (6 to 8 AM and 6 to 8 PM), and near change shift time (3 PM to 6 PM). The method of
changing shifts resulted in drivers assembling at a certain point, which caused a loss of time. The
statistical test results showed that there were significant differences in truck quantity at specific
hours, which suggests that it is important to monitor and manage the number of trucks during those
hours to ensure optimal productivity.

72
 Log Op Man Res.

Figure 6. Daily Average Total Queues in ROM

In addition, the study found that there were other delays that caused a loss of time in
production activity, which were unidentified because each truck had unknown details. Therefore,
it is important to conduct further research to identify these delays and address them accordingly to
improve productivity. Overall, the study highlights the importance of monitoring and managing the
number of hauling trucks and identifying and addressing delays that cause loss of time to ensure
optimal productivity in the RTP in Indonesia.

Truck Quantity
The researchers found that the quantity of trucks used to transport coal to the port is
another factor that affects the productivity of the RTP. However, the movement of hauling trucks
can vary due to various causes, such as breakdowns, refueling, driver rest times, regular check-ups,
GPS intermittent signals, and other unforeseeable delays. Despite these delays, the number of
trucks that are effectively operating on the haul road should be kept constant or at least meet a
certain quantity in every operational hour.

Figure 7. Truck quantity and productivity relationship

Based on the visual tracking data, the researchers found visible trends where trucks tend
to exit the system during change shift time between 5.00 to 6.00 WITA and 17.00 to 18.00 WITA,
during prayer time especially between 15.00 to 16.00 WITA and between 2.00 to 3.00 WITA, and
near rest-time in 12.00 and 23.00 WITA. However, other delays cause loss of time in the production
activity which are unidentified because each truck has unknown details.
As a benchmark based on the planning process, the researchers estimated the number of
trucks per shift equivalent to the actual trucks that perform full-time production activity. They
estimated the quantity using a weighted average with time fraction in the data as the weight,
resulting in a relationship between truck quantity and productivity per shift as shown in Figure 8.

73
 Log Op Man Res.

Figure 8. Truck quantity compared with productivity.

The scatter plot in Figure 9. shows that there is a strong positive correlation between truck
quantity and productivity per shift. However, this relationship is not linear and can only be
explained by a quadratic or cubic equation. Linear regression can only explain around 55% of the
data, while quadratic and cubic equations can explain around 59% of the data. However, the cubic
fitting resulted in an increased p-value without a significant increase in R-square, suggesting that
the truck quantity is already reaching the road limit, and any additional trucks will result in lower
output for one input unit. This phenomenon is commonly known as diminishing marginal returns.
At the current capacity, it would be better if the company monitors the truck quantity closely and
does not add any more trucks to the system.

Figure 9. Truck quantity with productivity relationship

To further analyze the truck quantity trends, the researchers isolated the truck quantity in
each hourly time slot and compared it with the next time slot. Using 3 months of data, this study
found that there are significant differences in truck quantity during certain hours. For example,
during fatigue time from 3 to 5 AM, many drivers are slowing down or taking a rest along the road.
At 6 to 8 AM, which is the post-change shift time, many drivers are just starting to enter the system
gradually and do the production process. Another significant difference occurs during non-Friday
time from 3 PM to 6 PM, which is a narrow time combining short prayer time at 3 PM, followed by
near change shift time at 5 PM and 6 PM (Table 2). The method of changing shifts makes drivers
assemble at a certain point, resulting in a loss of time.

74
 Log Op Man Res.

Table 2. Statistical test for Truck Quantity per Hour excludes Friday

hour_before hour_after statistics p_value result

0 1 7311.5 0.1907 Same
1 2 6681 0.2964 Same
2 3 6680 0.0566 Same
3 4 6453.5 0.0005 Not Same
4 5 5888.5 0.0000 Not Same
5 6 7713.5 0.4316 Same
6 7 4326.5 0.0000 Not Same
7 8 6485 0.0101 Not Same
8 9 7412 0.3927 Same
9 10 7022.5 0.4569 Same
10 11 6129 0.4742 Same
11 12 7832 0.2720 Same
12 13 7584 0.4273 Same
13 14 7248.5 0.4478 Same
14 15 6923 0.1253 Same
15 16 7218.5 0.0132 Not Same
16 17 6862 0.0315 Not Same
17 18 7415 0.0183 Not Same
18 19 5930.5 0.4919 Same
19 20 7027 0.0578 Same
20 21 5275 0.2952 Same
21 22 6810.5 0.4742 Same
22 23 6449.5 0.1644 Same
23 24 4454 0.0913 Same

The statistical test in Table 2 shows the comparison of truck quantity per hour between the
hours before and after. The test excluded Friday because Friday is a special day in Indonesia, and
there is a longer stoppage time. The statistical test shows that there are significant differences in
truck quantity during certain hours, such as during fatigue time from 4 to 5 AM and during post-
change shift time from 6 to 7 AM, and between 3 PM to 6 PM, except for the hour before 3 PM. This
analysis helps the researchers identify the hours with significant differences in truck quantity and
may be useful for optimizing the scheduling of trucks to improve the productivity of the RTP.

Truck Payload
The trucks that carry coal have different carrying capacities. Some trucks can carry up to
40 tons of coal, while others can only carry around 25 tons. The weight bridge reading data
distribution chart shows that the mean carrying capacity is around 30 tons (Figure 10). The
company's target for truck payload is being met, and there is no issue with the payload.

75
 Log Op Man Res.

Figure 10. Truck payload Aug – Nov 2022

Effect of Queues:
There is no relationship between queues at PORT and queues at ROM. Several ROMs are
used every day, and the trucks regulate themselves along the way to their destinations. The
distribution of the data tends to be not Normally distributed with an extreme tail, so Spearman
correlation is used to test the correlation between the queues at PORT and ROM. The correlation is
calculated every 15 minutes offset, up to 135 minutes, and there is no significant correlation
between queues at PORT and ROM at any given time offset.

Table 3. Correlation between Queues at PORT and ROM with Offset

PORT Capacity and Incoming Hauling Capacity

Trucks regulate themselves before the last segment to PORT (S6) by increasing their speed,
overtaking, or slowing down. Therefore, other segments cannot be used to monitor the queue
potential in PORT due to the high uncertainty and time needed to reach segment 6 road. The
distribution of truck quantity and PORT capacity is not normally distributed, and Spearman
correlation is used to measure the correlation. Incoming trucks in S6 are moderately positively
correlated with PORT capacity.

76
 Log Op Man Res.

Figure 14 Truck quantity vs PORT capacity

DISCUSSION

Table 4. Root Cause Analysis

*NRC = Not Root Cause, RC = Root Cause

No Sub Possible Root Cause Remark & Summary
Status*

1 Intersection The intersection has the same effect for all

trucks and it's relatively permanent, this is
not a root cause for small trucks as bigger NRC
trucks have no problems with it.

2 Road Maintenance Road maintenance will have the same

Road Activity effect for all trucks, this is not a root cause
NRC
for small trucks as bigger (DDT) trucks
have no problems with it.
3 Road Surface This also will have the same effect for all
trucks, this is not a root cause for small
NRC
trucks as bigger trucks have no problems
with it.
4 High Loading time Evidence shows that Loading time has
relatively stable performance. Some spike
occurs but not to the point of reducing NRC
productivity significantly.

5 ROM Activity High Maneuver time Evidence shows that in ROM, 1.5 or 2 side
loading methods are used, there are several
anomalies but the impact is balanced by
NRC
others that have lower maneuver time.
This is not a root cause.

77
 Log Op Man Res.

6 Unbalanced Truck When there are obstructions at hauling

movement roads or idles in ROM or PORT for some
time, there will be inefficient resource
RC
usage. This is a waste and the root cause of
the low productivity.

7 Loading Capacity< Hauling Queue at ROM rarely happens. High queue

capacity (>12 trucks) at ROM happens due to NRC
special cause). This is not a root cause

8 ROM Activity Loader breakdown Loader is critical and focused in operation.

If a loader will be unavailable for a long
time, the supervisors will adjust the fleet NRC
accordingly. This is also not a major root
cause
9 High dumping time By observation, this variable is relatively
stable but has some points to be NRC
highlighted as improvement potential.
10 High Maneuver time

11 Unbalanced Truck The same with item 6

movement NRC

12 PORT Activity Crushing Capacity< Hauling There is little capacity to re-route trucks in
capacity this activity. Evidence shows that hauling
capacity at the nearest segment (S6) has a
RC
medium to high correlation with a queue at
PORT-B. This is a root cause

13 Loader breakdown The same with item 8

NRC

14 cut-off road This will affect all trucks in the cycle. This
is not a root cause NRC

15 Material condition This is a given condition because the cost NRC

16 Material not uniform to purchase good materials is very
expensive and takes a long time. Mine must
Road NRC
have a long lifetime to make the purchase
congestion feasible.
17 Incident This will only affect all trucks in the cycles.
This is not a root cause NRC

18 Force Majeure If this happens, this will affect all trucks in

the cycles. This is not a root cause NRC

19 Low Pit Performance Pit performances are buffered by stocks in

ROM, so if any problems happen with Pit
performance, it will have little effect on NRC
hauling. This is not a root cause

20 ROM Weather This will affect all trucks in the cycle. This
is not a root cause NRC
Problems
21 Self-Combustion This will only increase the delay for one or
two loaders, this is not a root cause
because the combustion area will be NRC
isolated and the loading point will switch
accordingly.

78
 Log Op Man Res.

22 Incident The same with item 17 NRC

23 Weather This will affect all trucks in the cycle. This
is not a root cause NRC

24 No Driver This will reduce truck quantity and the

trucks will not operate from the beginning NRC
if there's no driver. This is not a root cause

25 Incident The same with item 17 NRC

26 Low Working Demo/Social/Community If this happens, this will affect all trucks in
Hour Problem the cycles. This is not a root cause NRC
Utilization
27 Other standby This will reduce trucks in the cycle and
affect truck quantity in the system. This
NRC
will relate to unbalanced truck
movement/cycles.
28 Breakdown This will reduce the truck quantity and
automatically will be replaced by a NRC
transporter. This is not a root cause

29 Low truck quantity This will reduce truck quantity, this is not
the root cause NRC

30 Low truck payload The truck payload was relatively stable and
most have considerably good payload. This NRC
Low
is not a root cause
productivity
31 per hour High truck cycle time This item analyzed in item no 1 - 10
NRC

32 Low loader total capacity The same with item 8

NRC

The cause mapping and root cause identification analysis presented above provide a
valuable framework for understanding the factors contributing to the low productivity in the coal
hauling system. By identifying two major root causes, namely unbalanced truck movement/cycles
and crushing capacity less than hauling capacity, the analysis highlights the areas that require
improvement to optimize the system.
A comparison of this analysis with previous research reveals similar root causes for low
productivity in mining operations. Studies have shown that unbalanced truck movement is a
significant factor affecting productivity. Similarly, a study by Rahman et al. (2020) on a coal mine in
Indonesia identified crusher downtime as a major contributor to prolonged truck waiting times.
To address the identified root causes, the strategies proposed above align with established
approaches for improving mining productivity. Regular maintenance and repair of equipment,
employee training and development, and the implementation of technology have been shown to
enhance equipment performance and reduce downtime. Similarly, continuous improvement
through regular evaluation and analysis of performance data, as well as collaboration with other
companies, has been found to optimize mining operations and improve productivity (Rahman et
al., 2020). Furthermore, safety remains a critical aspect of mining operations, with regular safety
audits and training programs playing a key role in ensuring compliance and fostering a strong
safety culture among employees.
In conclusion, the cause mapping and root cause identification analysis presented above
provide valuable insights for improving the efficiency and productivity of the coal hauling system.
The strategies proposed, including regular maintenance and repair of equipment, training, and

79
 Log Op Man Res.

development of employees, implementation of technology, continuous improvement, collaboration

with other companies, and a focus on safety, are consistent with prior research and can help PT
BORNEO optimize its coal hauling operations and increase profitability.

CONCLUSIONS
Based on the research findings, it can be concluded that several key factors significantly
impact the productivity of RTP in PT. BORNEO. The most important factor is the ROM activity
(average loading time), which has a direct correlation with productivity. Random stoppages along
the shift also have a significant impact on productivity, as they can cause queues and delay the
overall system. Additionally, the loaded condition during change shift is an important variable that
needs to be addressed to prevent pseudo-traffic jams and increase efficiency. While stoppage before
rest time, stoppage in loaded condition, and stoppage at major pray time (Ashar) can have some
impact on overall productivity, they are not considered priority areas for improvement due to their
relatively minor impact. Finally, by optimizing a combination of operational parameters, such as
reducing loading time, increasing loaded trucks in change shifts, and allowing for a small percentage
of truck stoppage along the shift, it is possible to increase productivity. These findings suggest that
targeted improvements to operational variables can significantly increase the efficiency of the RTP
system in PT. BORNEO.

LIMITATION & FURTHER RESEARCH

Despite the valuable insights provided by this research, several limitations should be
considered. First, the study was conducted only in one company (PT BORNEO), and the findings
may not be generalizable to other mining companies because the hauling process in every company
is different in terms of road length, longer roads will have higher uncertainty and checkpoint
positions. Second, the study on ten operational variables (empty time, loaded time, ROM activity,
PORT activity, Queue PORT, Queue ROM, Truck quantity, PORT capacity, ROM capacity, portion of
loaded condition at change shift), and there may be other variables that also affect productivity.
Third, the study was conducted over a relatively short time, and the results may not be indicative
of long-term trends. Finally, the study relied on data provided by the company, and there may be
errors or inaccuracies in the data that could affect the findings.
To overcome the limitations of this study, further research could be conducted in several
areas. First, the study could be replicated in other mining companies to determine if the findings
are consistent across different organizations. Second, additional operational variables could be
included in future studies to provide a more comprehensive analysis of the factors that affect
productivity. Third, future research could be conducted over a longer period to determine if the
results are stable over time. In conclusion, forthcoming research endeavors could employ various
data sources to corroborate the results and mitigate the prospective influence of errors or
inaccuracies in the data. Furthermore, the utilization of machine learning techniques could be
explored for predicting queues and hauling performance.

REFERENCES
Adolfsen, J. F., Kuik, F., Schuler, T., & Lis, E. (2022). The impact of the war in Ukraine on euro area
energy markets. Economic Bulletin Boxes, 4.
Armis, R., & Kanegae, H. (2020). The attractiveness of a post-mining city as a tourist destination from
the perspective of visitors: A study of Sawahlunto old coal mining town in Indonesia. Asia-Pacific
Journal of Regional Science, 4(2). https://doi.org/10.1007/s41685-019-00137-4
Arteaga, F., Nehring, M., & Knights, P. (2018). The equipment utilisation versus mining rate trade-off in
open pit mining. International Journal of Mining, Reclamation and Environment, 32(7).

80
 Log Op Man Res.

https://doi.org/10.1080/17480930.2017.1306674
Bozorgebrahimi, E. (2004). The evaluation of haulage truck size effects on open pit mining. Mining
Engineering, March.
Brodny, J., & Tutak, M. (2022). Challenges of the Polish coal mining industry on its way to innovative
and sustainable development. Journal of Cleaner Production, 375.
https://doi.org/10.1016/j.jclepro.2022.134061
Chaowasakoo, P., Seppälä, H., Koivo, H., & Zhou, Q. (2017). Digitalization of mine operations: Scenarios
to benefit in real-time truck dispatching. International Journal of Mining Science and Technology,
27(2). https://doi.org/10.1016/j.ijmst.2017.01.007
Das, S., & Sengupta, R. (2004). Projection pursuit regression and disaggregate productivity effects: The
case of the Indian blast furnaces. Journal of Applied Econometrics, 19(3).
https://doi.org/10.1002/jae.756
Elearn. (2021). Scatter diagrams. In Quality and Operations Management.
https://doi.org/10.4324/9780080458793-13
Flores, M., Naya, S., Fernández-Casal, R., Zaragoza, S., Raña, P., & Tarrío-Saavedra, J. (2020).
Constructing a control chart using functional data. Mathematics, 8(1).
https://doi.org/10.3390/math8010058
Kalisz, S., Kibort, K., Mioduska, J., Lieder, M., & Małachowska, A. (2022). Waste management in the
mining industry of metals ores, coal, oil, and natural gas: A review. Journal of Environmental
Management, 304. https://doi.org/10.1016/j.jenvman.2021.114239
Leung, R., & Scheding, S. (2015). Automated coal seam detection using a modulated specific energy
measure in a monitor-while-drilling context. International Journal of Rock Mechanics and Mining
Sciences, 75. https://doi.org/10.1016/j.ijrmms.2014.10.012
Mousavi, R., & Naghdi, R. (2013). Time consumption and productivity analysis of timber trucking using
two kinds of trucks in northern Iran. Journal of Forest Science, 59(5).
https://doi.org/10.17221/10/2013-jfs
Putri, L. R., & Susanto, S. (2017). Lean hospital approach to identify critical waste in the outpatient
pharmacy installation of RSI PKU Muhammadiyah Pekajangan. Jurnal Medicoeticolegal Dan
Manajemen Rumah Sakit, 6(2). https://doi.org/10.18196/jmmr.6139
Rahman, Z. F., Mulia, S. A., Sugiharta, A. M. B., Susanti, L., & Tualeka, A. R. (2020). Coal dust and
acute respiratory infections in south Kalimantan PT ‘X’ coal mining workers. Indian Journal of
Forensic Medicine and Toxicology, 14(1), 444-
447. https://doi.org/10.37506/v14/i1/2020/ijfmt/192939
Riyandi, I., & Wibowo, M. B. (2020). Optimalisasi penggunaan alat berat dengan menggunakan unlock
system-fleet management system (FMS) dispatch modular: Studi kasus PT Bukit Makmur Mandiri
Utama job site Lati. Prosiding Temu Profesi Tahunan PERHAPI, 2020.
Thompson, R. J., & Visser, A. T. (2006). Selection and maintenance of mine haul road wearing course
materials. Institution of Mining and Metallurgy Transactions. Section A: Mining Technology,
115(4). https://doi.org/10.1179/174328606X155138
Zhang, J., Fu, M., Geng, Y., & Tao, J. (2011). Energy saving and emission reduction: A project of coal-
resource integration in Shanxi Province, China. Energy Policy, 39(6).
https://doi.org/10.1016/j.enpol.2011.03.026

81