Abstract
The steepness of the slopes of an open pit mine has a substantial influence on the financial return of the mine. The paper proposes a novel design methodology where overall steeper pitwalls are employed without compromising the safety of the mine. In current design practice, pitwall profiles are often planar in crosssection within each rock layer; i.e., the profile inclination across each layer tends to be constant. Here instead, a new geotechnical software, OptimalSlope, is employed to determine optimal pitwall profiles of depth varying inclination. OptimalSlope seeks the solution of a mathematical optimization problem where the overall steepness of the pitwall, from crest to toe, is maximized for an assigned lithology, geotechnical properties, and factor of safety (FoS). Bench geometries (bench height, face inclination, minimum berm width) are imposed in the optimization as constraints which bind the maximum local inclination of the sought optimal profile together with any other constraints such as geological discontinuities that may influence slope failure. The obtained optimal profiles are always steeper than their planar counterparts (i.e., the planar profiles exhibiting the same FoS) up to 8° depending on rock type and severity of constraints on local inclinations. The design of a copper mine is first carried out employing planar pitwalls, secondly adopting the optimal pitwall profiles determined by OptimalSlope. The adoption of optimal slope profiles leads to a 34% higher net present value and reductions of carbon footprint and energy consumption of 0.17 Mt CO_{2} eq and 82.5 million MJ respectively due to a 15% reduction of rockwaste volume.
1 Introduction
In the last four decades, a clear trend of open pit mines being excavated at increasing depths is apparent (Fig. 1a). As existing mines deepen owing to the increased efficiency of mining equipment and improved exploration techniques and technology, the ore bodies left to be exploited are everywhere deeper (Fig. 1b). Between 1930 and 2000, the depth of the average discovery in Australia, Canada, and the USA increased from surface outcropping to 295 m [1]. As a consequence, ensuring pitwalls as steep as possible has grown in importance since the deeper a mine, the higher the effect of pitwall steepness on the amount of rockwaste to be excavated and therefore mine profitability [2].
Also, there is an increasing trend to excavate pits of significant depth in weak rocks such as saprolites [4]. In weak rocks, the steepness of the pitwalls is lower than the one for pits excavated in competent rock to prevent slope failures. This in turn implies more waste rock is produced. Therefore, any gain in the overall slope angle (OSA) of pitwalls is more important than ever for the economic profitability of open pit mines.
Anecdotal evidence of the fact that a slope profile nonlinear in crosssection, i.e., a profile whose inclination varies with depth, is better than a linear one, i.e., a planar profile, was reported as far back as 1890 [5]. In fact, Newman [5] observed that cuttings of concave shape excavated in homogeneous clay layers tended to be more stable than planar ones with the same OSA which in turn are more stable than cuttings of convex shape (Fig. 2a). Many decades later, Hoek and Bray in chapter 12 of the second edition of Rock slope engineering [6] analyze the stability of some concave circular slopes in crosssection. Assuming the slopes to be excavated in homogeneous rock and using the Mohr–Coulomb (MC) failure criterion to characterize its strength, i.e., cohesion, c, and internal friction angle, φ, they found a higher stability number (which is a dimensionless index capturing the mechanical stability of a slope that will be introduced in Section 3) for circular profiles than for their planar counterparts, i.e., the planar slopes with the same OSA which share the same toe and crest points. They also mention that circular shapes had been considered by [7] for the design of the pitwalls of an iron pit mine in Canada. However, Hoek and Bray strongly caution against the assumption of uniform slope since it completely disregards the fact that the strength of geomaterials tends to vary with depth, being usually lower in the upper part of the excavated slope due to wellknown geological processes (e.g., soil deposition, weathering). The first systematic theoretical study on the mechanical properties of concave slope profiles for geomaterials exhibiting some cohesion, applicable to all rocks and clayey soils, appeared in [8]. In that study, the superior stability of logarithmic spiral (logspiral) profiles, with the logspiral being featured by a radius of curvature increasing with the depth of excavation (Fig. 2b), was systematically proven over their planar counterparts having assumed the slopes to be uniform and with strength characterized by c and ϕ. By employing the limit analysis upper bound method, Utili and Nova [8] first systematically determined the optimal logspiral shape, i.e., the shape associated with the highest stability number for several prescribed OSAs, second compared the optimal logspiral slopes to their planar counterparts. They show that logspiral profiles exhibit higher FoS than their planar counterparts for any value of c, and φ considered with the highest gain for inclinations midway \(\varphi\) and the vertical line, i.e., for \(\mathrm{OSA}\sim \frac{\pi }{4}+\frac{\varphi }{2}\). Since then, other researchers [9,10,11] have independently investigated the stability of concave profiles excavated in a uniform c–φ geomaterial, employing different methods, namely the slip line method [9], limit equilibrium methods [10], and the finite element method for the assessment of slope stability. They all reached the same conclusion concerning the superior stability of nonlinear concave profiles. However, a key limitation of these studies is the assumption of a specific shape, either a circle [10] or a logspiral [8] or a curve stemming from the slipline field theory and the associated characteristic equations [9], so that the shape claimed to be optimal is found as the shape associated to the highest stability number among curves belonging to a very restricted family. It is obvious that these profiles are instead suboptimal and the shape of the truly optimal profile cannot be inferred from the aforementioned studies. In fact, in the case of a profile to be excavated in a uniform c–φ geomaterial, the optimal shape calculated by OptimalSlope [12] turns out to be partly concave and partly convex (see Fig. 2c) so significantly different from the purely concave shapes considered in [8,9,10,11]. Another perhaps even more important limitation resides in the assumption of uniform slope present in all the aforementioned methods that prevent the application of these findings to real open pit mines which are typically featured by complex lithologies involving multiple rock formations of different mechanical strengths and various geological discontinuities.
The search for the optimal shape of a slope profile is a difficult problem of topological optimization since slopes develop very significant irrecoverable (plastic) deformations which cause significant stress redistribution to occur before reaching failure, but the theory of topological optimization developed so far in engineering deals predominantly with elastic media [13]. Neglecting the onset of plastic deformations in a slope, i.e., assuming a purely elastic behavior, is not a viable option since this would severely underestimate the slope resistance to failure hence the slope factor of safety. And the application of plasticity theory to topological optimization is only in its infancy [14]. Therefore current algorithms in the literature for topological optimization are not viable to search for the optimal profile of slopes. The code OptimalSlope exploits the fact that slope failures occur either as a rotational mechanism (a planar failure being a particular type of rotational failure with an infinite radius of curvature) or mechanisms whose kinematics is dictated by the presence of discontinuities (e.g., the interface between two rock layers, a fault, joints, beddings). For a homogeneous slope in a c–φ geomaterial, the limit analysis upper bound method allows finding the critical rotational mechanism simply by determining the minimum of an analytical objective function without requiring any discretization of the slope into finite elements [15]. The function is obtained by imposing the energy balance between the external work done by the mass of the candidate failure mechanism and the energy dissipated along its failure surface. The equation has been extended to find the critical mechanism for piecewise linear slope profiles in a uniform layer and then to the case of layered slopes [12]. Also, the formulation has been extended to slopes in rocks obeying the generalized Hoek–Brown (GHB) failure criterion [16, 17]. The minimum of the function stemming from the energy balance equation and therefore the critical mechanism is found by OptimalSlope [12]. Since the time the GHB criterion was introduced, a strong consensus has gathered in the rock mechanics community that it is a better criterion than MC to describe the strength of rock masses [18, 19]. But for highly weathered rocks and residual soils MC is better [4], so there are several cases of mines where the MC criterion is employed to fit the rock strength data of some layers while the GHB criterion for some other layers. OptimalSlope can deal with these situations too, i.e., some rock layers being characterized by the MC criterion while others by the GHB criterion, since the GHB parameters of any layer are converted by OptimalSlope into equivalent c–φ parameters, employing the method of Renani and Martin [20].
In the excavation of an open pit, several rock layers of different strengths are usually encountered. OptimalSlope can find the optimal profile for any specified lithological sequence (any number of layers can be specified as input with each layer strength characterized by either MC or GHB parameters) without unduly restricting the search to any predefined family of shapes. The optimal slope profile is found as the solution of a mathematical optimization problem where the OSA of the slope, i.e., the inclination from slope crest to toe, is maximized for an assigned stratigraphy, rock strength properties, and a prescribed FoS. Geometric requirements stemming from bench sizes (bench height, face inclination, and minimum berm width) are imposed as constraints binding the maximum local inclination of the sought optimal profile together with any other geometric constraint, e.g., constraints to prevent the occurrence of local failure mechanisms due to geological discontinuities such as faults and joints (see Section 3).
In the next section of the paper, the case study considered to demonstrate the financial and environmental benefits stemming from the adoption of optimal profiles for mine pitwalls is described. Then, in Section 3, the methodology employed is provided followed by results in Sections 4 and 5. In the methodology section, it will be illustrated how the proprietary code OptimalSlope [21] works and interacts with the mining software packages employed to perform strategic pit design.
2 Case Study
The block model of the copper deposit employed as case study has been provided by a mining company collaborating with the Delphos Mine Planning Laboratory at the University of Chile (Fig. 3). Due to a nondisclosure agreement (NDA), its name and location cannot be revealed. All the parameters needed to perform the pit design were taken from [22] except for the mining cost adjustment factor (MCAF), which was estimated for this study. The characteristics of the block model are as follows: cubic blocks of 10 m × 10 m × 10 m, block model dimensions of 2250 m × 2250 m × 360 m, 647,446 blocks in total. From the site lithology and geotechnical properties, two roughly uniform pit sectors were identified in [22] (see Fig. 4). In each pit sector, one representative crosssection was assumed to design the pitwall profile.
The rock geotechnical parameters are reported in Table 1. The rock strength is characterized by the GHB failure criterion [16]. The uniaxial compression strength (UCS), the Geological Strength Index (GSI), m_{i}, and the disturbance factor (D) are all provided in the table. The disturbance factor is to accounts for rock weakening due to blasting and stress relaxation. We acknowledge that assuming D = 1 throughout the rock mass is overly conservative since in modern mines blasting is well controlled. Also, the depth of the open pit is such that the amount of stress relaxation is unlikely to cause significant damage. However, we opted to retain the assumption of D = 1 to be consistent with the set of input data from [22], thus accepting a significant degree of conservativism on the value of this parameter.
The values adopted for the economic parameters and metallurgical recovery together with the discount rate assumed were taken from [22] and are listed in Table 2. With regard to the capital costs, in [22] only the cost for the processing plant is accounted for. Here, instead, we wanted to account for all the typical costs needed in an open pit mine project: they are provided in Table 3.
3 Methodology
The design of pitwalls in open pit mines is an iterative process involving a multidisciplinary team of geologists, geotechnical engineers, and mining engineers [1]. The design process requires iterative steps between the teams [23]. Typically, several boreholes are drilled as part of a site investigation and laboratory tests are performed on the core samples retrieved to characterize the mechanical strength of the geomaterials encountered and the key lithological units. Then, a preliminary simplified design is performed and a pit crest contour is drawn. The pit is then split into sectors in order to design the pitwalls [23]. It is wise to divide the mine into sectors small enough so that a 2D vertical crosssection representative of the lithology of the sector can be determined for each sector. Then for each sector crosssection a pitwall profile needs to be designed as steep as possible and at the same time to satisfy a prescribed FoS against slope failure. The pitwall profiles are then prescribed as input together with the relevant economic and metallurgical data for the mine into a pit optimizer software package to calculate the optimal Ultimate Pit Limit and pushbacks. These are typically based on the LerchGrossman algorithm [24] or the more recent pseudoflow [25]. Both algorithms require to identify precedences between blocks of the block model lying within the pit boundary. How block precedences are built for pitwalls with depth varying inclinations is well described in [26, 27].
The iterative procedure we followed to calculate the UPL is illustrated in Fig. 5. The procedure is the same irrespective of the shape of the pitwall profile adopted, i.e., planar or optimal profile: at the beginning, an initial pit depth (H_{0}) was assumed as equal to the total height of the block model minus the air blocks so H_{UP L S1,0} = H_{UP L S2,0} = 370 m with H_{UP L S1,0}, and H_{UP L S2,0} indicating the pitwall height in sector S1 and S2 respectively. Then, we calculated the representative pitwall profiles for the specified pit depth in each sector: in the case of planar pitwalls employing limit equilibrium method (LEM) analyses by Slide2 [28] whereas for optimal pitwalls employing OptimalSlope (see Section 3.1). Then, we assigned the pitwall profiles into the pit optimizer (Geovia Whittle 4.7.3) and ran it to produce the UPL (the steps entailed are described in Section 3.2). The detail of the procedure followed to integrate OptimalSlope with Geovia is presented in Fig. 6. In the figure, it is also illustrated how to integrate OptimalSlope with three other major commercial mining software packages, namely Datamine, Maptek, and Hexagon Mining. Note that conceptually the process is independent of the software package employed, but the names of the software modules to manipulate the block model, their scripting languages, and some routines within the pit optimizers may differ. For this reason, we verified that the design procedure here employed was applicable in all the software packages mentioned in Fig. 6, namely Geovia Surpac [29] and Whittle [30], Datamine Studio OP [31] and Studio NPVS [32], Maptek Vulcan Open Pit Mine Planning [33] and Hexagon MinePlan3D [34] and Project evaluator [35], to ensure the pit optimization procedure herein described works correctly with each software package. The UPL depths obtained as output of the strategic pit optimization, H_{UP L S1,1} for sector S1, and H_{UP L S2,1} for sector S2 were then compared to the values prescribed as input in the slope design process, H_{UP L S1,0} and H_{UP L S2,0}, respectively. Since they turned out to be different, a second iteration was performed where H_{UP L S1,1} and H_{UP L S2,1} were assumed as input for a second slope design process followed by a second run of the pit optimizer. The iterations were then stopped when the UPL depths obtained as output of the pit optimizer ended up being equal to the values prescribed as input in the slope design process for each pit sector. The iterations required to reach convergence are reported in Table 5.
3.1 Pitwall Design
Open pit mines are increasingly excavated in complex lithologies where usually different failure mechanisms (e.g., bench failure, interramp failure, overall slope shear failure, failures involving faults and discontinuities) turn out to be dominant in different sectors of the pit so that they all need to be analyzed. In the design of pitwalls, we followed the standard practice of starting with the design of benches and then moved to the overall pitwall profiles [36, 37].
To determine the maximum inclination of each face bench we performed limit equilibrium method MorgensternPrice analyses using the Rocscience program Slide2 [28] to satisfy the prescribed factor of safety (in this case FoS = 1.1, see Table 4). In general, the amount of backbreak and the effective bench face angle are controlled by the joints and faults intersecting each bench. Here, joints were not considered purely due to the lack of this type of data for the mine analyzed. If data on joints and faults were available, software packages such as SWedge [38] or Frac_Rock [39 ] would allow determining the maximum inclination of each face bench.
The height of the benches adopted for the whole mine is 10 m [22]. Consistently with the set of input data from [22], we computed the minimum berm width, b_{w}, using the equation proposed by Call [40] known as modified Ritchie’s criteria, which has been demonstrated to be effective in field tests in several benched mine slopes [41 ]:
We acknowledge that another equation, b_{w}(m) = 3.5(m) + 0.17 ∗ H_{bench}, was proposed in [41 ] but it exhibits the undesirable property of being less conservative [42] so we preferred to use Eq. (1). Berms have multiple objectives that need to be met when establishing a berm width , the most important ones being rockfall retention, capture of debris from bench excavation and operations of the chosen fleet of excavators. With regard to rockfall retention, berm widths are not designed for 100% retention of rockfalls since this would lead to unviably large widths but instead to shelter most of the rockfall. Eq. (1) leads to 70–85% retention of rockfall volumes [18]. Such a percentage may be less than what a mining company would wish for. For this reason, in the last two decades, several rock fall types of research studies were performed utilizing numerical simulations considering potential trajectories on slope designs using either twodimensional lumpedmass impact models (2DLM), e.g., the Rocscience code RocFall [43], and/or threedimensional rigidbody impact models (3DRB) such as the code “Trajec3D” [44] . These numerical models use coefficients of restitution to characterize the amount of energy lost due to the inelastic deformations stemming from the collision of a rock fragment bouncing over a berm or bench. Unfortunately, the input parameters are vastly different for 2DLM. Also 3DRB and they are seldom calibrated with any sitespecific rock fall case studies or field test data at the mine feasibility stage, and may remain uncalibrated throughout the operating life of the mine [45] . In conclusion, given the amount of information available at the feasibility stage of mine design , we believe Eq. (1) is adequate to establish minimum berm widths especially given the low bench heights of our case [45] . Finally, note that as pointed out in Read and Stacey [18, see Sect. 10.2.1.2], Eq. (1) is still largely used by practitioners in North and South America. Another requirement for the chosen berm widths is the ability to capture most debris falling from the bench faces above due to unstable wedges. The adequacy of the chosen berm widths can be verified via either computer modeling [39] or analytical equations [46] . In [47] , a recent extensive review of the bench design methodologies currently employed by the open pit mine industry is provided.
After the computation of the bench geometries, we computed the geotechnically optimal slope profiles for the crosssections of the two pit sectors of the mine (Fig. 4) using the proprietary code OptimalSlope [21]. The code requires bench height, bench face inclination, minimum berm width, and road width as input from the user (see Fig. 7) since these geometric data will act as constraints in the search for the optimal profile. Any pitwall profile is defined in OptimalSlope by a discrete set of points in the vertical plane : see the (x_{i}, z_{i}) coordinates in Fig. 8, with z_{i} being values specified according to the bench height (Δz = bench height) input by the user while x_{i} represents the unknown variables to be determined. The search for the optimal profile is constrained to feasible profiles (which lie within the red and blue bounds of Fig. 8). A profile is feasible if \(\frac{{z}_{i} {z}_{i1}}{{x}_{i} {x}_{i1}}\le \mathrm{tan}{\alpha }_{i\mathrm{max}}\) for every i, i.e., the inclination of each segment of the profile is capped to \({\alpha }_{i\mathrm{max}}\). The \({\alpha }_{i\mathrm{max}}\) values are determined by the code before the optimization algorithm is called on the basis of bench height, bench face inclination, and minimum berm width provided by the user (see Fig. 8b). In case a ramp needs to be included as part of the pitwall profile, a lower \({\alpha }_{i\mathrm{max}}\) value is imposed for the profile segment corresponding to the vertical position of the ramp.
The optimal pitwall profile is defined as the overall steepest safe profile, i.e., OSA = OSA_{max}, with OSA being the inclination over the horizontal of the line joining the pitwall toe to the crest (see Fig. 8). OSA_{max} is determined by OptimalSlope iteratively (see Fig. 9). The main algorithm finds the optimal pitwall shape for an assigned OSA and geometric constraints (\({\alpha }_{i\mathrm{max}}\) values). An initial guess OSA value is first determined for the specified pitwall height, the geotechnical properties of all the layers, and the specified FoS on the basis of a database of stability charts based on [63] builtin OptimalSlope. In the case of rock strength described by GHB, the conversion of the HB parameters to MC was performed via Eq. (14) in [20]. Note that a few equations for the conversion of the GHB parameters into MC for rock slope stability analyses have been proposed in the literature with [48, 49] being the most prominent. All these relationships depend on the range of confining stress experienced by the slope. Hence, the main challenge in finding equivalent strength criteria is selecting the appropriate range of confinement. However, contrary to what the previous relationships suggest, the appropriate range of confinement was shown in [20] to be insensitive to rock mass strength and instead primarily controlled by slope geometry and only the equation proposed in [20] reflects such a dependency. The FoS_{i} associated with the optimal profile found at the ith iteration is then compared to the target FoS_{target}: if it is higher a steeper OSA is prescribed at the next iteration; if it is lower a less steep OSA is prescribed. The termination criterion is specified in terms of the percentage difference between FoS_{target} and FoS_{i}. For the pitwalls of the case study considered, four iterations were enough to obtain a FoS less than 1% different from FoS_{target}.
In the case of a homogeneous slope, i.e., a single uniform rock/soil layer, the optimal profile shape is found by the main algorithm for a prescribed input OSA as the profile associated with the largest stability factor defined as \({N}_{\mathrm{s}}=\frac{\gamma H}{c}\) in the case of the MC criterion or \({N}_{\mathrm{s}}=\frac{\gamma H}{{\sigma }_{ci}}\) in the case of the GHB criterion. \({N}_{\mathrm{s}}\) is a wellknown single scalar parameter that was introduced by Taylor [50, 51] in drawing dimensionless stability charts and is the appropriate metric to compare the performance of different slope profiles [21]. In the case of a layered (nonhomogeneous) slope, OptimalSlope determines the optimal profile as the profile associated with the maximum of an equivalent \({N}_{\mathrm{s}}\) which is calculated as a physically based weighted average of the γ, H, c or \({\sigma }_{ci}\) parameters of all the slope layers. Note that in a layered slope a failure mechanism may go through several layers of vastly different strength so the length of the failure surface (a curve in 2D) in each layer may affect significantly the overall resistance (i.e., the amount of energy dissipated in the limit analysis energy balance equation). Therefore, OptimalSlope calculates the energy dissipated along the failure surface of each candidate mechanism considered based on the actual length of the failure curve present in each layer ensuring the \({N}_{\mathrm{s}}\) calculated for each mechanism considered is a true reflection of the mechanism stability factor (see Fig. 10). Crucially, to determine the stability factor of any candidate profile shape, all the possible failure mechanisms are considered by OptimalSlope including any failure surface daylighting everywhere above the slope toe (see Fig. 10).
3.2 Pit Optimization
To assign the pitwall profiles to the pit optimizer Geovia Whittle, we split the block model into “zones” (according to the Whittle terminology) using Geovia Surpac and assigned a slope inclination to each “zone.” The number of zones to be employed depends on the shape of the pitwall profile (see Fig. 6). Then, to compute the ultimate pit limit and pushbacks, first, we ran Whittle to produce the discounted bestcase scenario curve in the pitbypit graph (see Fig. 15). Next, we employed the socalled Milawa NPV algorithm to generate a specified case scenario curve for an initial set of pushbacks, chosen in correspondence of sharp increases exhibited by the bestcase scenario curve. Then, we recomputed the specified case scenario curve a few times exploring the choice of different pitshells as pushbacks nearby the ones initially selected to make the specified case curve as close to the bestcase scenario as possible. Having found the combination of pushbacks that maximizes the NPV of the UPL, we plotted the annual ore production scheduling graph (Fig. 16) and changed the selection of pushbacks if needed to ensure an annual tonnage input to the processing plant as uniform as possible over the mine lifetime. Finally, we chose the UPL in correspondence of the highest point of the plateau exhibited by the specified case curve (Fig. 15). We are aware practitioners may make slightly different choices: for instance, they may decide to pick a pit shell different from the one associated with the peak of the specified case scenario as UPL for various considerations (to maximize the amount of ore extracted or of reserve amount or due to operational reasons) and may not be willing to iterate for the selection of pushbacks and UPL. Moreover, design practices vary between practitioners due to different company objectives and designer experiences. Nevertheless, the key objective of the design exercise performed here is to run a consistent and meaningful comparison between the traditional design based on planar pitwalls and the one here proposed based on geotechnically optimal pitwalls in order to quantify in a rigorous way the financial and environmental gains due to the adoption of geotechnically optimal profiles . To this end, it is logical to adopt the same procedure for the selection of UPL and pushbacks irrespective of the pitwall profile shapes. We also believe that even if design objectives different from the one used here were to be adopted (e.g., compromising some NPV to maximize reserve conservation) and even if the procedure for the determination of UPL and pushbacks was different from the one adopted here (e.g., prioritizing payback time over NPV maximization, using bench discounting), the adoption of optimal profiles for the pitwalls will always bring financial and environmental gains since it leads to a reduced amount of waste rock excavated whatever the design procedure and corporate priorities .
4 Results
The geometric features of the ultimate pit limits (H_{UPL} and OSA) obtained as the result of each iteration of the design process (Fig. 5a) are reported in Table 5. A total of six iterations was performed for the traditional design based on planar pitwalls, while for the design based on optimal pitwalls, two iterations were enough to reach convergence.
Note that to have a consistent and fair comparison between the two designs, we adopted a design procedure envisaging iterations between the geotechnically based design of the pitwall and the pit optimization carried out by Whittle (Fig. 5). However, practitioners may perform fewer iterations or not at all due to time constraints accepting a less optimal pit design. Nevertheless, in the case of no iteration, i.e., looking at the results obtained after iteration 1 in Table 5, we can state that the adoption of optimal pitwalls provides financial gains of the same order of magnitude.
4.1 Pitwall Profiles
The pitwall profiles obtained as a result of the pit design process are plotted in Fig. 11a for pit sector 1 and Fig. 12a for pit sector 2. The factor of safety values of all the pitwall profiles were verified by performing a LEM analysis with the MorgensternPrice method, which is a rigorous LEM method where all equations of equilibrium are imposed on all slices [52], in Slide2 [28] with noncircular failure surface and other default settings. Preliminary analyses were carried out to make sure the results are independent of the number of slices adopted and of the domain boundaries chosen. The pitwall profiles employed in the Slide2 analyses are reported in Figs. 11b and 12b for the pit design adopting planar profiles and in Figs. 11c and 12c for the pit design adopting optimal profiles together with their FoS. In all the cases, the FoS found is less than 1% from the target value of 1.30 (Table 4). For the optimal pitwall profiles, given the importance of an independent verification of their FoS, we performed additional stability analyses by an explicit finite difference method with shear strength reduction technique (FDMSSR) employing FLAC3D 7.0 [53] having assigned a unit length in the out of plane direction. In the FLAC analyses, the GHB criterion was employed together with the “model factorofsafety” command and default values for the settings affecting the convergence criterion (specified in terms of unbalanced nodal forces) and the detection of slope failure. Details of the algorithm employed by FLAC to calculate the FoS are provided in the FLAC manual [53]: in summary, the standard procedure adopted by FDMSSR as in the case of MC geomaterials [54] is followed with the GHB criterion approximated locally by the MC criterion, \({\tau }_{\mathrm{max}}={\mathrm{tan}\Phi }_{\mathrm{loc}}+{c}_{\mathrm{loc}}\), where local cohesion, \({c}_{\mathrm{loc}}\), and angle of shearing resistance, \({\mathrm{tan}\Phi }_{\mathrm{loc}}\), are calculated from the local value of the minor principal stress and GHB parameters. In each FLAC analysis, \({c}_{\mathrm{loc}}\) and \({\mathrm{tan}\Phi }_{\mathrm{loc}}\) are decreased by dividing them by an increasing reduction factor until active slope failure is detected. The FoS is found as the reduction factor at the verge of slope collapse. Both associated and nonassociated with zero dilation analyses were run obtaining practically identical FoS values. The negligible influence of dilatancy on slope stability is good news since its value is not known for the rock types of this mine. The critical failure mechanisms identified by Slide2 and FLAC3D and the associated FoSs are reported in Figs. 11c and 12c. In both mine sectors, the FoSs found are less than 1% from the target value of 1.30. Note that FDMSSR is an entirely different method of slope stability analysis from LEM and from LA (which is employed by OptimalSlope) so the fact that the FoSs found by Slide2 and FLAC3D are so close to the ones determined by OptimalSlope gives confidence to geotechnical practitioners about the trustworthiness of the FoS of the pitwall profiles determined by OptimalSlope. In conclusion, the FoS values of the pitwall profiles found by OptimalSlope were independently verified by two most popular geotechnical software employed for the geotechnical verification of open pit mines, Slide2 and FLAC, confirming that the pitwall profiles determined by OptimalSlope are as safe as their planar counterparts.
Figures 11a and 12a allow visual comparison between the planar pitwall profile of a traditional design, demarcated by the orange line, and the optimal pitwall profile, demarcated by the blue line. Considering pit sector S1, the overall height of the profiles is the same. The optimal profile is clearly steeper than the planar one. The area colored in light green (Fig. 11a) highlights the difference between the two profiles: apart from the small difference in the central zone, the optimal profile requires significantly less rock excavation (see the bottom and top third parts). Considering now pit sector S2, the overall heights of the profiles are significantly different with the optimal one being 70 m less deep than the planar one (Fig. 12a). Once again, the optimal profile is steeper than the planar one. Comparing the two profiles, the volumes of rock excavated look similar albeit distributed differently in space. In the upper half of the pitwall, the optimal profile requires a larger excavation to extract mainly ore whereas the planar profile requires extraction of ore to a deeper depth.
4.2 3D Slope Stability Analysis
In Fig. 13a and b the 3D views of the ultimate pit limits obtained for the traditional design based on planar pitwalls and for the design based on optimal pitwalls respectively are plotted. It is apparent that in both cases, a multipit shell made by roughly two conical shapes, one distinctly deeper than the other one, is obtained. This is due to the different rock properties and orebody distributions between sectors S1 and S2 with each cone lying within a pit sector.
The FoS of the entire UPL was verified by performing a 3D FDM analysis with FLAC3D 7.0 [53]. Note that the numerical settings adopted in the 3D stability analysis were the same as those employed for the 2D analyses reported in Section 4.1 for the sake of consistency. The critical failure mechanism identified by the software and the FoS contour maps are plotted in Fig. 14. To check for any potential effect of mesh dependency, two analyses were run: one for a coarser (but still fine) mesh and another one for a finer mesh obtained by halving the element sizes of the first mesh. Throughout the entire mine, the minimum FoS found was 1.70 for the first analysis and 1.67 for the second analysis: such a small difference implies the mesh size adopted is sufficiently small so that the resulting FoS is not affected by mesh size for practical purposes. We can then assume convergence of the FoS from the FLAC3D analysis to 1.67, which is significantly higher than the FoS values obtained by the 2D FLAC analyses of Section 4.1, of ~ 1.30 and ~ 1.31 respectively. The main reason for this is due to arching effects that in the case of slope shapes concave in plan view enhance stability [6, 55]. In this case, the very pronounced planar concavity of the UPL acts to constrain movement. This is not taken into account by 2D analyses, that as a consequence are conservative.
4.3 Key Financial Indicators
In Table 6, the key output data for the two design cases are provided. The net present value (NPV) for the design based on optimal pitwalls is around 12 million USD higher than the NPV of the design based on planar pitwalls. Therefore, the adoption of the optimal profiles would lead to an increase of NPV of 34%. Such an increase is to be ascribed to a very substantial decrease of rockwaste volume, around 15%, from 23.7·10^{6} tonnes to 20.7·10^{6} tonnes, while the amount of ore extracted is similar. This implies a reduction of the Stripping Ratio from 0.40 to 0.35. Another metric measuring the financial return of a mine is the internal rate of return (IRR). Adoption of the optimal profiles leads to an IRR of 15.8% instead of the 13.9% obtained from the design with planar pitwalls. Since an IRR of 15% is considered by some companies as a threshold for the viability of a mining project, it can be said that for this mine the adoption of optimal pitwalls could make the difference in terms of the economic viability of the mine project.
4.4 Production Schedule
In Fig. 15, the pitbypit graph of the mine is plotted for both the traditional design based on planar pitwalls (Fig. 15a) and the design based on optimal pitwalls (Fig. 15b). The vertical black lines indicate the pitshells selected as pushbacks including the UPL (line most to the right). In both cases, the UPL has been selected as the pushback corresponding to the peak of the discounted specified case scenario curve (green curve). The fact that the specified case scenario curves exhibit a plateau implies the choice provides a robust NPV optimum. In Fig. 16, the amount of tonnage is plotted against the life of the mine in years. It can be seen that an almost uniform amount of ore is to be extracted year after year for both designs. A constant production over time is a highly desirable feature from a logistical point of view.
4.5 Environmental Indicators
Recently, several methods have been proposed in the literature to calculate the life cycle assessment for open pit mines. Here, we have computed the energy requested to mine both orebody and waste rock together with the associated carbon footprint for both design types (design with planar pitwalls and with optimal pitwalls) based on [56]. In the Appendix, all the equations employed for these calculations are provided. Note that energy consumption and carbon footprint are calculated for each mined block of the UPLs with block properties such as grade, mass, and distance to a reference point on the surface used to estimate the energy required to produce 1 tonne of mined ore. The energy requirement per tonne is then translated into carbon footprint by using characterization factors and includes scope 1, 2, and 3 emissions associated with drilling, blasting, extraction, loading, and hauling [57]. This approach has been employed in [58] to incorporate sustainability parameters directly into the ore body block model and include them into strategic pit optimization. The characterization factors were taken from [59] having assumed the electricity production mix for Chile where the mine is located. The results are provided in Table 7 below.
It emerges that the adoption of optimal pitwalls leads to reductions of carbon footprint and energy consumption of 0.17 million tonnes CO_{2} eq and 82.5 million MJ respectively over the life of the mine. To provide some context, a reduction of 0.17 million tonnes CO_{2} eq is equivalent to the carbon sequestered by 2.8 million tree seedlings grown for 10 years and to the greenhouse gas emissions avoided by 35 wind turbines producing electricity for a year [60]. Both carbon footprint and energy usage savings are achieved by a significant reduction of rockwaste excavation, around 15% in volume, with the amount of orebody extracted being very similar in the two design options considered (see Table 6).
5 Conclusions
It is well known that the steepness of the pitwalls of an open pit mine bears a large influence on the volume of rockwaste to be extracted so that any increase of steepness leads to a better stripping ratio and therefore higher profitability and carbon footprint and energy use reductions. In current design practices, mines tend to be designed on the basis of planar pitwalls, i.e., with a constant interramp angle (IRA) or even overall slope angle (OSA), in each sector of the mine. However looking at the final geometry of any pit crosssection, this is anything but planar due to the need to accommodate for benches, stepouts, and roads; therefore, the assumption of constant IRA and/or OSA adopted at the design stage is a simplification which can and should be removed if a better design can be achieved as result. It seems only natural to wonder whether pitwalls of nonlinear shape could be used instead of planar ones. In the geotechnical literature, some specific shapes of nonlinear profiles have already been proven to be more stable than the planar one ([8,9,10], and [11]) for uniform cϕ slopes. The code OptimalSlope [12] determines optimal shapes for both uniform and nonuniform slopes with any number of rock layers, and for geomaterials whose strength is described by either the Mohr–Coulomb criterion or the generalized Hoek–Brown criterion. In this paper, we employed the geotechnically optimal pitwall profiles determined by OptimalSlope to systematically maximize the OSA of the pitwalls of a soon to be opened copper mine. The longterm schedule of the mine was carried out using Geovia Whittle first employing planar pitwalls, secondly adopting the optimal pitwall profiles determined by OptimalSlope.
The adoption of geotechnically optimal profiles led to a 34% higher net present value and very significant reductions of carbon footprint and energy consumption, 0.17 million tonnes CO_{2} eq and 82.5 million MJ, respectively, due to a 15% reduction of waste rock volume in comparison with the traditional design based on planar pitwalls. The stability of all the pitwall crosssections determined by OptimalSlope was independently checked by two most popular geotechnical software packages, namely Rocscience Slide2 to perform limit equilibrium method analyses with the MorgensternPrice method and FLAC3D to perform finite difference method analyses with shear strength reduction. The factor of safety values determined by Slide2 and FLAC turned out to be in very close agreement (less than 1% difference) to the ones determined by OptimalSlope. Also, a 3D stability analysis for the entire ultimate pit was carried out by FLAC3D verifying the resulting FoS was well above the specified acceptability criterion (FoS = 1.3 in this case) across the entire pit.
The case study considered here is of a relatively small pit in a mediumstrong rock. It can be expected that the increase of pitwall inclination will be more significant in weaker rocks. Also, we know that the larger the depth of a pit, the more significant the impact of the pitwall steepness is on the economics of the mine. Therefore, we believe the financial and environmental gains obtained for large and deep pits can be even higher. Finally, irrespective of the size of the open pit mine, because a planar slope is a particular case of a curved one, obtained by setting the radius of curvature to infinity along the entire slope, there is a theoretical argument to suggest that adopting planar pitwall profiles is a suboptimal choice for most rock masses and soil types apart from the case of frictional (zero cohesion) soils.
Data Availability
The block model is subject to an NDA from the mining company. The corresponding author can redirect any request to the mining company.
Code Availability
The input and outputs files of OptimalSlope, Whittle, and Slide2 can be made available upon request to the corresponding author.
References
Randolph M (2011) Current trends in mining. In: Darling P (ed) SME Mining Engineering Handbook. 3rd edn. Society for Mining Metallurgy and Exploration, pp 11–19
Hustrulid W, Kutcha M, Martin R (2013) Open pit mine planning and design. CRC Press
Brown ET (2004) Geomechanics: the critical engineering discipline for mass mining. In: Karzulovic A, Alfaro M (eds) Massmin 2004, Santiago, Chile, 2004. Chilean Engineering Institute, Santiago, Chile
Martin D, Stacey P (2018) Guidelines for open pit slope design in weak rocks. CSIRO Publishing
Newman J (1890) Earthwork slips and subsidences upon public works: their causes, prevention, and reparation. Spoon, London
Hoek E, Bray J (1977) Rock slope engineering, 2nd edn. The Institution of Mining and Metallurgy, London
Rana MH, Bullock WD (1969) The design of open pit mine slopes. Can Mining J 58–66
Utili S, Nova R (2007) On the optimal profile of a slope. Soils Found 47:717–729
Jeldes A, Drumm EC, Yolder DC (2015) Design of stable concave slopes for reduced sediment delivery. ASCE J Geotech Geoenviron Eng 142
Vahedifard F, Shahrokhabadi S, Leshchinsky D (2016) Optimal profile for concave slopes under static and seismic conditions. Can Geotech J 53(9):1522–1532
Vo T, Russell AR (2017) Stability charts for curvilinear slopes in unsaturated soils. Soils Found 57(4):543–556
Utili S (2016) OptimalSlope: software for the determination of optimal profiles for slopes and pitwalls. United States Copyright Office, registration: TXu 2254159.
Bendsoe M, Sigmund O (2004) Topology optimisation. 2nd edn. Springer, Berlin. https://doi.org/10.1063/1.3278595
Fin J, Borges LA, Fancello EA (2019) Structural topology optimization under limit analysis. Struct Multidiscip Optim 59(4):1355–1370. https://doi.org/10.1007/s001580182132y
Chen WF (1975) Limit analysis and soil plasticity. Elsevier, New York
Hoek E (1994) Strength of rock and rock masses. ISRM News J 2(2):4–16
Hoek E, Brown ET (1997) Practical estimates of rock mass strength. Int J Rock Mech Min Sci 34(8):1165–1186
Read J, Stacey P (2009) Guidelines for open pit slope design. CSIRO, Australia
Hoek E, Brown ET (2019) The HoekBrown failure criterion and GSI 2018 edition. J Rock Mechan Geotech Eng 11(3):445–463. https://doi.org/10.1016/j.jrmge.2018.08.001
Renani RH, Martin CD (2020) Slope stability analysis using equivalent MohrCoulomb and HoekBrown criteria. Rock Mech Rock Eng 53(1):13–21. https://doi.org/10.1007/s00603019018893
OptimalSlope (2021): Software for the determination of optimal profiles for slopes and pitwalls. User manual Version 1.
Parra A, Morales N, Vallejos J, Nguyen PMV (2017) Open pit mine planning considering geomechanical fundamentals. Int J Min Reclam Environ 32(4):221–238. https://doi.org/10.1080/17480930.2017.1278579
Stewart A, Hawley P, Rose N, Gilmore B (2004) Mining applications. In: Wyllie D, Mah C (eds) Rock slope engineering. 4th Edition edn. Spon press, New York, pp 357–376
Lerches H, Grossmann I (1965) Optimum design of open pit mines. Can Inst Mining Transact 68:17–24
Hochbaum D (2008) The pseudoflow algorithm: a new algorithm for the maximumflow problem. Oper Res 58(4):982–1009
Khalokakaie R, Dowd P, Fowell R (2000) LerchsGrossmann algorithm with variable slope angles. Transact Inst Mining Metallurgy Sect A 109:A77–A85
Khalokakaie R, Dowd P, Fowell R (2000) Incorporation of slope design into optimal open pit design algorithms. Transact Inst Mining Metallurgy Sect A 109:A70–A76
Rocscience (2021) Slide2
Geovia (2021) Surpac user manual. Geovia Dassault systemes
Geovia (2021) Whittle user manual. 4.7.3 edn., Geovia Dassault systemes
Datamine (2021) Studio OP. 2.8 edn. Datamine
Datamine (2021) Studio NPVS. 1.0.51.0 edn. Datamine
Maptek (2021) Vulcan open pit mine planning. 2020.1 edn. Maptek
Hexagon (2021) HxGN MinePlan engineering OP LTP Pro. Hexagon
Hexagon (2021) HxGN MinePlan project evaluator. Hexagon
Kliche CA (2011) Slope stability. In: Darling P (ed) SME Mining Engineering Handbook. 3rd edn. Society for Mining Metallurgy and Exploration, pp 495–525
Lorig L, Read J, Stacey P (2009) Slope design methods. In: Read J, Stacey P (eds) Guidelines for open pit slope design. CRC Press, pp 237–264
Rocscience (2020) SWedge user manual. https://www.rocscience.com/help/swedge/documentation
SRK (2016) Frac_Rock: programme for the analysis of discontinuous rock masses
Call RD (1992) Slope stability. In: Hartman HL (ed) SME Mining Engineering Handbook, vol 1. 2nd edn. Society for Mining Metallurgy and Exploration, Littleton, Colorado (USA) Ch 10.4
Ryan TM, Pryor PR (2001) Designing catch benches and interramp slopes. In: Hustrulid W, McCarter M, van Zyl D (eds) Slope stability in surface mining. Society for Mining Metallurgy Exploration, pp 27–38
Alejano LR, Pons B, Bastante FG, Alonso E, Stockhausen HW (2007) Slope geometry design as a means of controlling rockfalls in quarries. Int J Rock Mech Min Sci 44:903–921
Rocscience (2002) RocFall user manual. https://www.rocscience.com/help/rocfall/documentation
Basson FRP (2012), Rigid body dynamics for rock fall trajectory simulation, in Proceedings of the 46th US Rock Mechanics & Geomechanics Symposium 24–27 June 2012, American Rock Mechanics Association, Chicago
Bar N, Nicoll S, Pothitos F (2016) Rock fall trajectory field testing, model simulations and considerations for steep slope design in hard rock. First Asia Pacific Slope Stability in Mining (APSSIM) Conference, Brisbane, Australia
Gibson W, de Bruyn IA, Walker DJH (2006) Considerations in the optimisation of bench face angle and berm width geometries for open pit mines. Proc. South African Institute of Mining and Metallurgy Int Symp on Stability of rock slopes, Symp series S44 Stability of rock slopes in open pit mining and civil engineering situations, pp 557–579
Coetsee S (2020) An overview of bench design for cut slopes with an example of an advanced dataset assessment technique. Proc of the 2020 Int Symp on Slope Stability in Open Pit Mining and Civil Engineering, pp 731–748.
Hoek E, CarranzaTorres C, Corkum B (2002) HoekBrown failure criterion 2002 edition. In: Hammah R et al (eds) Proc 5^{th} North American Rock Mechanics Symp. University of Toronto, Toronto, pp 267–273
Aj Li, Merifield RS, Lyamin AV (2008) Stability charts for rock slopes based on the HoekBrown failure criterion. Int J Rock Mechan Mining Sci 45:689–700
Taylor D (1937) Stability of earth slopes. J Boston Soc of Civil Engineers 24:197–246
Taylor D (1948) Fundamentals of soil mechanics. Wiley, New York
Morgenstern N, Price V (1965) The analysis of the stability of general slip surfaces. Geotechnique 15(1):79–93
Itasca International Inc. (2021) FLAC 3D user manual
Dawson E, Roth W, Drescher A (1999) Slope stability analysis by strength reduction. Geotechnique 49(6):835–840
Lorig L, Varona P (2001) Practical slope stability analysis using finite difference codes. Slope Stability in Surface Mining, Colorado, 2001. Soc Mining, Metallurgy and Exploration, pp 115–124
Munoz J, Guzman RR, Botin JA (2014) Development of a methodology that integrates environmental and social attributes in the ore resource evaluation and mine planning. Int J Mining Mineral Eng 5:38–58. https://doi.org/10.1504/IJMME.2014.058918
Greenhouse Gas Protocol (2019) https://ghgprotocol.org/. Accessed 21/03/2021
Pell R, Tijsseling L, Palmer LW, Glass HJ, Yan X, Wall F, Zeng X, Li J (2019) Environmental optimisation of mine scheduling through life cycle assessment integration. Resour Conserv Recycl 142:267–276. https://doi.org/10.1016/j.resconrec.2018.11.022
Wernet G, Bauer C, Steubing B, Reinhard J, MorenoRuiz E, Weidema B (2016) The ecoinvent database version 3 (part I): overview and methodology. Int J Life Cycle Assess 21(9):1218–1230. https://doi.org/10.1007/s1136701610878
Greenhouse gas equivalencies calculator (2021) https://www.epa.gov/energy/greenhousegasequivalenciescalculator. Accessed 21/03/2021
Isheyskiy V, Sanchidrián JA (2020) Prospects of applying MWD technology for quality management of drilling and blasting operations at mining enterprises. Minerals 10:925. https://doi.org/10.3390/min10100925
Soofastaei A, Aminossadati SM, Arefi MM, Kizil MS (2016) Development of a multilayer perceptron artificial neural network model to determine haul trucks energy consumption. Int J Min Sci Technol 26:285–293. https://doi.org/10.1016/j.ijmst.2015.12.015
Tang GP, Zhao LH, Li L, Yang F (2015) Stability charts of slopes under typical conditions developed by upper bound limit analysis. Comput Geotech 65:233–240
Acknowledgements
The authors thank Mr A. Russo from SRK for reviewing Section 3.1 and Mr C. Zhang for calculating the Environmental Indicators (Section 4.5).
Funding
The PhD of the second author is funded by the UK Engineering Physical Sciences Research Council (EPSRC) Doctoral Training Account. Also, EU H2020 RISE Hercules grant (778360) is acknowledged as funding overseas stays of the first and second author useful for the project. Nelson Morales was funded by ANID/PIA Basal Project AFB180004 (Advanced Mining Technology Center, Universidad de Chile).
Author information
Authors and Affiliations
Contributions
S. Utili: ideation and conceptualization of the work, methodology, result analysis, writing of the paper, reviewing and editing. A. Agosti: productions of the results by running OptimalSlope and the pit optimizer software packages, methodology, result analysis, production of figures and tables, reviewing and editing. N. Morales: result analysis, reviewing and editing. C. Valderrama: performance of the FLAC analyses R. Pell: Section 4.4 result analysis. G. Albornoz reviewing and editing.
Corresponding author
Ethics declarations
Conflict of Interest
The authors declare no competing interests.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
The appendix presents the formulae employed to compute the energy consumption and the carbon footprint associated with drilling, blasting, loading, and hauling for all the ultimate pit limit blocks.
Our calculations were based on the formulae provided by [56], who were the first to provide a comprehensive set of equations for the calculation of energy consumption and GWP emissions for blocks of a block model. First, the energy consumption of each block is evaluated based on the processes affecting it and then its GHP is derived from the calculated energy consumption. Muñoz et al. [56] identify three main stages of energy consumption, namely mining, concentrating, and hydrometallurgical. The concentrating and hydrometallurgical stages have not been considered since we do not know the passing material at the feed nor at the product of the crusher. Also, no information about the processing method is available. However, because the amounts of orebody extracted for processing for the traditional pitwall design and for the design by optimal pitwalls are almost identical (see Table 6), we do not expect them to alter the differences of carbon footprint and energy consumption between the two designs considered in the paper. The energy consumption due to mining, \({E}_{\mathrm{M}}\), is made of four components [56]:
with \({E}_{\mathrm{drilling}}\), the energy consumption due to drilling, \({E}_{\mathrm{blasting}}\) due to blasting, \({E}_{\mathrm{loading}}\) due to loading of the mined material on dumper trucks, and \({E}_{\mathrm{hauling}}\) due to hauling to either processing plant or dump.
For drilling, the energy consumption is calculated as [56]:
where:

\(A = 116.9 {\mathrm{cm}}^{2}\) is the area of the drill hole;

\(L = 35 \mathrm{cm}\) is the charged length of the drill hole;

\(N = 10\) is the number of drill holes for each block;

\({E}_{\mathrm{v}} = 112148 \mathrm{J}/{\mathrm{cm}}^{3}\) is the drilling specific energy which depends on rock type and was estimated based on a unified classification system of rocks according to their drillability [61];

\({\eta }_{\mathrm{drill}} = 80\%\) is the assumed driller efficiency;

\({m}_{\mathrm{b}}\) is the mass of the block in tonnes.
The specific energy that an explosive can deliver when detonated is computed with the formula [56]:
where:

\(LF =6 {\mathrm{kg}}_{\mathrm{expl}}/\mathrm{t}\) is the load factor, defined as the amount of explosive per tonne of detonated rock;

\({E}_{\mathrm{expl}} = 3.81581 \mathrm{MJ}/{\mathrm{kg}}_{\mathrm{expl}}\) is the specific explosive energy [56] for ANFO, the type of explosive here employed.
The specific energy that a front loader consumes to load up the fractured material on the dumper can be computed with the expression:
where:

\({P}_{\mathrm{L}}= 0.18 \mathrm{MW}\) is the front loader power (having assumed a CAT 950 GC);

T = 45 s is the assumed average time to meet the loading capacity of the dumper using the front loader;

\({\eta }_{\mathrm{load}} = 70\%\) is the assumed front loader efficiency;

\({m}_{\mathrm{truck}} = 92.2 \mathrm{t}\) is the loading capacity of the dumper (having assumed a Komatsu HD 785–8), assuming the dumper fully loading at every trip.
The specific energy to haul a tonne of material from the pit to the processing plant or the waste dump can be computed with the formula [56]:
where:

S is the distance of the ith block from the processing plant or waste dump in kilometers which we have calculated for each block;

\(i =10 \%\) is the inclination of the ramp;

\({R}_{s}= 2 \%\) is the rolling resistance of the surface taken from [62];

\({R}_{i} =1\%\) is the assumed internal resistance of the dumper;

\({M}_{\mathrm{truck}} = 166 \mathrm{t}\) is the total mass of the loaded dumper (Komatsu HD 785–8).
To calculate the mine carbon footprint, the specific energy consumptions per tonne were translated into specific carbon footprints using characterization factors to include scope 1, 2, and 3 emissions associated with the mining activities by using the following equation [56]:
where the coefficients α, β, and δ are the coefficient of carbonization for diesel, electrical power matrix, and explosive taken from the Ecoinvent database [59]. Their values are:

α = 0.09159 \({\mathrm{t}}_{{\mathrm{CO}}_{2,\mathrm{eq}}}\)/MJ;

β = 0.5200 \({\mathrm{t}}_{{\mathrm{CO}}_{2,\mathrm{eq}}}\)/MJ;

δ = 2.270 \({\mathrm{t}}_{{\mathrm{CO}}_{2,\mathrm{eq}}}\)/MJ.
Finally, the total carbon dioxide emissions are computed as the multiplication of the specific carbon footprint by the tonnages of the ultimate pit limit.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Utili, S., Agosti, A., Morales, N. et al. Optimal Pitwall Shapes to Increase Financial Return and Decrease Carbon Footprint of Open Pit Mines. Mining, Metallurgy & Exploration 39, 335–355 (2022). https://doi.org/10.1007/s42461022005468
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s42461022005468