Due to the current economic climate, significant increases in the cost items most used by mine investors, such as fuel, electricity, explosives, and transportation, are greatly raising production costs and putting economic strain on companies (Emiroğlu, 2019). According to a report by PricewaterhouseCoopers (PwC), reducing surface mining costs and increasing production volume alone are not sufficient for making the operation more efficient. According to the information provided, even if companies increase production volumes and reduce costs, they may still complain about low productivity values. The report concludes that productivity in mines depends on many factors, the most important of which are choosing the right equipment under the right production and economic conditions (Mining Turkey Magazine, 2014). At this point, in open-pits – before transitioning to full automation – the optimum number of trucks and loaders required for overburden removal and production should be determined during the investment period. Because calculating the optimum number of trucks-loaders will reduce the cycle time and consumable costs such as fuel, machinery/equipment, and materials in the investment and operation periods of open-pits. In this way, open-pits will avoid significant additional investment costs that could arise later due to an incorrect loader-truck count assessment.

The technical aspect of the selection process is likely the primary determinant of the numerous feasible loading and transportation system alternatives. Considering the differences between capital cost, operating cost, and the estimated lifetime of the equipment, a cost comparison may be necessary to evaluate the total cost per unit of production. Beyond traditional engineering economics analyses, another additional step is computer simulation of the various systems under consideration. Simulation selection ensures the correct choice of alternatives and also helps identify potential changes in system design that could result in the best alternatives (Başçetin, 2007). The inclusion of Industry 4.0 in the processes of all mining companies and the use of simulations and other digital technologies can save mining companies from losses and production shortages. Making predictions for the future will help increase profitability in mines, improve safety, and optimize their operations and processes (Karikari and Askari-Nasab, 2023). According to the analysis by Dzakpata et al. (2016) on the most efficient loader-truck matching orientation for cost optimization, 45% of the loader's working time is spent identifying trucks. This demonstrates the inherent losses in truck-loader dispatch and management in mines. Improvements in the truck dispatch system can achieve much higher available time and usage time for the truck fleet. (Erçelebi and Kırmanlı, 2000) examined the surface mining equipment selection techniques found in the literature. They classified these techniques as classical, optimization research, and artificial intelligence techniques. Among the optimization research techniques, the most commonly used methods are linear and integer programming, simulation, and queuing theory. Artificial intelligence methods have become quite popular in recent years, and expert systems, knowledge-based decision systems, and genetic algorithm techniques have been frequently applied in mining equipment selection. According to (Faraji, 2013) and (Chaowasakoo, 2017), the main shortcomings in the models examined in the literature are: a) insufficient consideration of the waiting time and queues of excavators and loaders, b) simplification of the models and the consideration of a limited amount of detail in the models, c) neglecting the stochastic nature of the truck and excavator transportation system, and d) development of a model based on a homogeneous fleet (trucks and excavators). The developed software overcomes the difficulties in selecting appropriate equipment, which is crucial for surface mining operations, and provides significant savings (Kesimal and Başçetin, 2000; Erçelebi and Başçetin, 2009; Kırmanlı and Erçelebi, 2009).

With these developments, multi-criteria decision-making (MCDM) methods have become useful for a wide variety of applications in the mining sector for ~ 20 years. Their role in the decision-making process, where multiple conflicting criteria are considered under both precise and uncertain conditions, is of great importance (Sitorus et al., 2016). Today, due to the very high capital cost of surface mining equipment, any error in the selection of quantity, type, and capacity of equipment can have an irreparable impact on the net present value of the mining project. Equipment selection is a complex, MCDM problem. For example, (Bazzazi et al., 2011), proposed a new MCDM method that considers both tangible and intangible factors in the mining equipment selection problem, enabling the selection of the most suitable equipment that satisfies the decision-maker. The results showed a significant reduction in computational time consumption and good accuracy compared to traditional methods such as the fuzzy Analytical Hierarchy Process (AHP) method. (Rakhmangulov et al., 2024) ranked the four groups of criteria (technical, technological, environmental, economic, and organizational) according to their importance using the MCDM method when considering the complex truck selection criteria. (Alpay and Iphar, 2018) selected the most suitable hydraulic excavator for a magnesite mine using fuzzy TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) and VIKOR (Multi-Criteria Compromise Ranking), which are fuzzy MCDM methods.

36% of publications on this topic deal solely with the problem of mining method selection. The remaining publications relate to areas such as ore enrichment, mining geology, post-mining activities, and health, safety, and environment (Sitorus et al., 2016). Furthermore, while numerous publications exist in the literature on optimizing loader-truck cycle times and increasing operational efficiency, there is a noticeable lack of studies optimizing loader-truck selection with the aim of reducing investment and operating costs and increasing efficiency using the MCDM method. Considering this deficiency, this study aims to optimize loader-truck cycle times and select the most efficient loader-truck with minimum cost to achieve this.

This study aimed to demonstrate the critical role of truck-loader dispatch in transporting overburden material from 5 different loading points in the mine to 6 different dumping sites (Fig. 1) throughout the mine's lifecycle. In particular, the truck's carrying capacity, one of the most important parameters affected by cycle time, was considered a key parameter in this study. The methodological steps are summarized in Fig. 2:

Figure 1. Truck-excavator match (Representative image).

Figure 2. Summary of the study's workflow.

Although calculations and analysis of machinery and equipment investments and other data could reveal suitable options, this is currently a rather laborious and primitive approach. Human errors and mistakes can occur during analyzes. Therefore, the Weighted Sum Model was applied in the MCDM analysis in this study. In this method, decision options were determined. The options used in the model were: low unit truck cost (non-beneficial), high daily transport capacity of the truck (beneficial), low number of trucks and excavators (non-beneficial), and low investment amount (non-beneficial). Then, the weight of each decision option was determined, a decision table was created in the study, and the value of each match was summed as a score, and the highest value - that is, the closest to 1 - was selected as the decision. From this, the best decision matches were identified. These matches were compared with previous data, i.e., whether they have high investment costs, the number of machines and equipment, the amount of work they can do daily, and their unit costs.

CAT brand mining equipment, specifically 770 series trucks and hydraulic excavators of models 365, 374, 385, 6015, and 6018, were used in the match process. These 25 truck-excavator matches are presented in Table 1. The capacities (transport and load) of each machine, fuel consumption per km and per hour, and excavator cycle time were taken from catalog values. The data obtained in this study are presented in Table 2. In this table, the data is as follows: 1) The columns for "Total number of trucks required" and "Total number of loaders required" are calculated values. 2) The columns for loader bucket capacity, loader cycle time, truck capacity, fuel consumption by the truck, and fuel consumption by the loader represent data taken from the machines' catalogs.

Table 1. Truck-excavator matches.

Table 2. Data obtained from calculations (indicated by H) and catalog data (indicated by K).

Cycle times and truck transport capacities, central to this study, are among the most important parameters to be examined to operate an efficient transportation system. In mining operations, it is critical planning that excavators do not remain idle or transport trucks do not have to wait in line to take loads from excavators. Therefore, careful evaluation of machine cycle times and synchronized truck movements without waiting is necessary. As mentioned above, there are 5 different areas (Areas A, B, C, D, and E) and 6 different dumping sites in the mining area. The distances between these areas and dumping sites are presented in Table 3.

In surface mining preparatory work, overburden removal operations constitute a significant portion of the costs (Özdemir, 2023b). The relationship between the unit truck overburden removal cost and transport capacity is shown in Fig. 3. This demonstrates the relationship between increasing the amount of work done per unit time and decreasing the unit cost. The purchase prices of the trucks and excavators used in this study are presented in Table 4 in USD ($). Based on this data, the investment cost and the unit operating costs of the truck-excavator for each match, according to the previously determined number of machines, were calculated and presented in Table 5. The truck unit cost shown in Table 5 is the operating cost and includes only diesel fuel consumption. Other costs are included in the overburden removal cost. In the calculation of the overburden removal cost; The total work volume (218,000,000 m³), ​​including all items such as explosives, materials, labor, machinery maintenance and repair, general production and management costs, was divided by the overburden volume, and the unit overburden costs (given in the last column) were calculated. Examining the unit costs of the trucks, even with different matches, the unit costs are essentially the same. The important point here is that these unit costs were not derived from transportation based on distances to the dumping sites, but rather from a fixed average daily round-trip distance. When optimization was performed by considering cycle times that included the distance to each dumping site, it was observed that these costs decreased by ~ 25%. This result can contribute to a more realistic economic evaluation of the project.

Figure 3. The relationship between unit cost and truck capacity.

Table 4. The purchase costs of the machines.

Table 5. Unit costs.

The following data were used in the optimization: distances from loading areas to unloading areas (Table 3), cycle times of the trucks depending on these distances (Table 3), and cycle time calculation parameters for each truck (Loaded speed (km/h), unloaded speed (km/h), loading time (min), unloading time (min), capacity (m³)). The objective function is as follows:

Here, X_ij is the amount of load transported in m³, and C_ij is the unit cost. The constraints are as follows:

Although the calculations made so far could reveal a suitable option by examining the tables, this is currently a rather laborious and primitive method. Errors and mistakes stemming from human factors may occur during interpretation. Therefore, the Weighted Sum Model was applied in this study for MCDM analysis. In this method, decision options were determined. The options to be used in the model are as follows: a) Unit truck cost: Should be low (i.e., non-beneficial), b) Daily transport capacity of the truck: Should be high (i.e., beneficial), c) Number of trucks and excavators: Should be low (i.e., non-beneficial), d) Investment amount: Should be low (i.e., non-beneficial). In this method, those whose outcome we want to minimize are calculated as "non-beneficial" (minimum value / value), and those whose outcome we want to maximize are calculated as "beneficial" (value / maximum value). The data should first be normalized using linear normalization. In this case, non-beneficial and beneficial options should be transformed using the following formula: Non-beneficial = (Minimum value / value), and beneficial = (value / maximum value). Then, the weight of each decision option should be determined. The sum of the weights cannot exceed 1. In our model, these weights were chosen as follows: a) Unit truck cost: 0.35, b) Daily transport capacity of the truck: 0.30, c) Number of trucks and excavators: 0.1 each, d) Investment amount: 0.15. These values ​​are multiplied by each converted value and written in the Table 7.

Here, K_d is the calculated decision value, W_j is the determined weight value (between 0 and 1), and X_ij is the transported load. Then, the decision table (Table 7) was obtained in the study. Here, the value of each match is summed as a score on the right side, and the highest value, closest to 1, is selected as the decision. Examining the Table 7, we see that the best decision is the E₄ match with 0.95, and the second-best option is E₁. Comparing these with the previous data in a table, we obtain the data in Table 8. As can be seen, although the E₄ match has a higher investment cost compared to E₁, it has fewer machines and equipment, can do more work daily, and has a lower unit cost.

Table 7. Raw data and decision scores.

Table 8. Comparing the two highest scores.

This study analyzes the optimization of loader-truck combinations using the MCDM method to reduce cycle time and minimize operating costs. First, multiple loader and truck models were selected to create combinations. Then, the distances between loading and unloading areas were included in the calculation to determine the minimum investment and unit operating cost, and the minimum cycle time. The loader-truck combination that best supports the company's economic operation was then selected. Determining the optimum number of loader-truck allows for the optimal calculation of investment (such as purchase and depreciation costs) and operating costs (such as fuel, oil, tires, vehicle parts replacements, and maintenance costs) of the selected loaders and carriers during the investment period. This, along with other cost items, allows for the calculation of the ratio of loading and transportation costs to investment and operating costs in open-pits.

This article focuses on the relationship between cycle time and cost in loader-truck selection, providing a support tool for mining operators to make optimal decisions during or after the investment phase. Unlike previous literature, this article applies the Weighted Sum Model with MCDM and identifies the most cost-effective loader-truck selection as a decision-maker's choice. This ensures that potentially overlooked data remains mathematically present in the system, creating a support mechanism for rapid decision-making. This mechanism will allow mining operators to obtain alternative solutions by comparing parameters or modifying their weightings to find the most advantageous match. The results of this article are expected to improve cost estimation and increase equipment efficiency in mining operations by focusing on the lowest investment and operating costs in the equipment selection process. By using the MCDM method, mining operators can make quick decisions regarding the most suitable equipment match by evaluating multiple criteria simultaneously within the same operating process using a simple scoring method.

Although the method in this study is applied only to the overburden removal process, it can also be applied in studies that consider the effect of ore grade, a parameter not included in the overburden removal process. Furthermore, while this article does not examine all parameters and efficiencies in surface mining, as mentioned above, it is possible to examine mining processes as a whole by increasing the number of these parameters. The study matches trucks and excavators involved in transporting overburden material from 5 different loading points to 6 different dumping sites. However, this match does not consider equipment performance, breakdown/maintenance losses, the impact of climatic conditions, the effects of routes and paths, and the size of the material after blasting on the loader and truck, etc. Additionally, this article assumes that all truck and loader equipment was purchased as unused equipment at the beginning of the investment phase. In other words, a match was made without considering both fleet aging and the simultaneous shipment of truck and loader equipment of different ages to different routes.

This study analyzes the investment costs of loaders-trucks, the unit cost of overburden, and the unit costs per m³ of trucks and loaders used for overburden. Furthermore, while direct materials (such as diesel fuel), consumables, and machinery and equipment materials that affect the unit cost have been included in the calculations, costs such as depreciation and labor, which can be considered as fixed values, have not been considered.

In this study, an analysis was conducted by considering the age of the loader and truck equipment as fixed. On the other hand, using this model, all kinds of quantitative data that will affect the selection of loader-truck equipment can be determined as criteria. At this point, it is recommended that new studies be conducted to develop a score-based MCDM method by including more criteria in the evaluation.

As this study shows, it is possible to solve such problems without using complex algorithms and methods. Just as it is impossible to predict the results of uncertainties arising from human-based studies with 100% accuracy, it is also impossible to achieve 100% satisfactory results in sectors like mining where dozens of parameters need to work synchronously and harmoniously. However, with the super-fast development of artificial intelligence technology, it will be possible to achieve optimal results by reducing decision-making processes to very short timeframes and taking advantage of the space opened up for intervention. Therefore, considering the time value of money, it is recommended to prioritize such studies.

In the future, to ensure more efficient and synchronized operation, it will be beneficial to use alert programs that can monitor cycle times in real-time and ensure that cycle times remain within certain limits. In addition, dispatch systems that can optimally manage and control artificial intelligence technologies or machines within the open-pits may also be considered. Programs that can integrate these systems will contribute to decision-making mechanisms that create predictable scenarios.

Annex 1. Equipment Selection for the Overburden Process

Here, a CAT 775G (56-ton) loader has been selected. The loader will fill the truck at an average speed of 0.19 m³/s (= 0.19 m³/sec x 2.30 tons/m³ (tons/sec) x 25 seconds = 10.96 tons/cycle) with a bucket cycle time of 25 seconds. Considering different truck types, the loader selection is as follows: If the 56-ton 775G truck is selected, [56 tons / (10.96 tons / loader cycle time)] = 5.11. This means that with an average of 5 passes (5 x 10.96 tons = 54.80 tons), 97.88% of the truck's capacity can be loaded. Therefore, this truck was chosen. Calculating the required number of loaders also requires calculating the truck's cycle time.

5 pass x 25 sec (Loader bucket cycle time) = 125 sec. T= (54.5 + 125 + 289 + 199.71) sec = 668.21 sec. This value (668.215 sec) is the maximum cycle time for a truck in the mine site.

[56 tons x 90% (efficiency)] / 668.21 sec = 0.075 tons/sec = 271.53 tons/h = > 686.58 (tons/h) / 271.53 (tons/h) = 2.53. So, the capacity of one loader is equivalent to ~ 3 trucks.

The average number of active trucks required for overburden removal at the mine site = (Average hourly capacity of a loader / Average hourly capacity of a truck) => (686.58 tons/h) / (271.53 tons/h) = 2.45 (trucks/loader). This means that for 3 actively working loaders (2.45 trucks/loader), 3 loaders equal 7.36 trucks, or 8 trucks in terms of capacity for overburden.