Abstract In short-term mine planning, mining scheduling is generally defined by designing dig-lines, allocated on benches. The mined ore will be sent to stockpiles, homogenization piles, or a concentration plant. The process to design dig-lines is usually done manually, whereby multiple simultaneous mining fronts are time-consuming and labour-intensive. The manual design of dig-lines tends to produce high variability of the grades throughout certain periods. Due to the limited time to manually multiple test dig-line design alternatives in short term planning, it is impossible to ensure production under stationary mean grades and variance. This article proposes an alternative to design short-term dig-lines, through an optimization process that joins and sequences the blocks in the block model over weeks or months, ensuring low variability of grades among periods. The methodology proposed generates multiple random paths starting at seed-points representing the locations and numbers of shovels previously selected by the mine planner. It tests multiple polygons representing a set of first dig-lines, comparing them with others, and keeping the dig-lines of low variability closer to a specific ore grade probability distribution, discarding the rest of the iterations. The process is repeated for the next dig-line. The block grades' probability distribution of all iterations is compared to a reference-grade histogram, and the iterations with the grade histogram more adherent are selected. Union-find and genetic algorithms were used to optimize the dig-lines aiming at the possible stationary grade distribution. The mean and variance of the reference model are 2.13% and 0.64%2, respectively. The mean for the automated draw dig-lines is closer to these values than the ones manually drawn. The method ensures more constant quality and quantity of ore production along a period planned, matching a target grade probability distribution. The methodology is illustrated using SiO2 values at a major iron ore mine.

Abstract Knowledge of the physical properties of rock masses is fundamental for the economics and safety of mining projects. The determination of these properties in rock samples in the laboratory requires time, expensive equipment and qualified personnel, which considerably increases the information's cost. Indirect methods were developed to obtain properties related to rock masses, which have been shown to be a viable alternative to traditional procedures. The determination of the compressional mechanical wave velocity (Vp) and subsequent correlation with lithological mechanical properties are indirectly obtained. This study’s objective was to obtain correlations between Vp and the resistance to uniaxial compression, UCS (Unconfined Compressive Strength), as well as the density and porosity of the siltstone and sandstone lithologies present in the coalfield of Candiota, located in the southern region of Rio Grande do Sul, Brazil. The Vp records were obtained in laboratory samples, using ultrasonic velocity sensors, and in-situ by geophysical well logging (directly in boreholes). The results indicate the possibility of using Vp to determine the physical parameters of the investigated lithologies. In the specific case of the correlations between Vp and Unconfined Compressive Strength, determination coefficients R2 above 0.70 were obtained, indicating sufficiently high reliability for using this information (e.g. in roof support projects). The correlation between Vp and density was also high.

Abstract Non-linear geostatistical methods are known to deal appropriately with the geological and geometrical complexity of gold deposits. This article reports the results related to an investigation to improve the gold content estimate based on restricted ore modeling to honor the structural aspects that control the mineralization. The grade domains are defined by using structural measurements to guide the indicator kriging (IK) estimator. Relevant grade intervals are chosen as indicators. Kriging the indicators provides a measure of the grade uncertainty at the sample support. The probability indicator modeling relies on thresholding the estimates which are represented by cumulative distribution functions (cdf) at the unsampled locations. The implicit concept of probability means that the chance of an estimated node belonging to a given grade domain is as big as the estimated IK value. The geological consistency of IK models requires a proper definition of some key parameters: The probability thresholds and indicator variogram models must honor the structural features and stationarity conditions of grade intervals. The geological representativeness of these models depends heavily on thresholding the estimates. For instance, extremely permissive estimates may produce overrepresented ore domains. The decision of the optimal indicator probability for defining the ore boundaries is made by iterative comparison. Several thresholds were applied to kriged maps and the results reconciled to the most sampled areas until achieving reasonable geological adherence. The mineralization continuity often varies according to local structural features and so dynamic anisotropy is used to control the variogram direction and search ellipse to consider the significant scale trend and small-scale fold geometries. A case study based on a real gold deposit dataset was performed and the method was discussed. The IK models can define precisely the mineralization bounds in the most detailed areas. However, the results presented some limitations on reproducing the geological expectation in regions of wide drilling spacing. The lack of information in some areas led to an excessive number of small sub-zones. The method allows a faster and efficient modeling of structurally complex geometries and provides an uncertainty assessment which may be useful to support exploratory and short-term decisions.

Abstract Most mining decisions are based on models estimated/simulated given the information obtained from samples. During the exploration stage, samples are commonly taken using diamond drill holes which are accurate and precise. These samples are considered hard data. In the production stage, new samples are added. These last are cheaper and more abundant than the drill hole samples, but imprecise and are here named as soft data. Usually hard and soft data are not sampled at the same locations, they form a heterotopic dataset. This article proposes a framework for geostatistical simulation with completely heterotopic soft data. The simulation proceeds in two steps. First, the variable of interest at the locations where soft data are available is simulated. The local conditional distributions built at these locations consider both hard and soft data and are obtained using simple cokriging with the intrinsic coregionalization model. Second, the variable of interest in the entire simulation grid using the original and previously simulated values at soft data locations is simulated. The results show that the information from soft data improved both the accuracy and precision of the simulated models. The proposed framework is illustrated by a case study with data obtained from an underground copper mine.

Abstract Knowing the quantity and the quality of products and tailings generated by a beneficiation plant, even before ore processing, can make the mining operations more sustainable, more profitable, and safer. To forecast these values, it is necessary to submit samples to batch tests which mimic the processing workflow used on an industrial scale. Then, the results need to be analysed with the aim of finding a statistical model able to comprehend how Run of Mine (ROM) characteristics impact the performance at the beneficiation. After developing a model, it is possible to apply it to blocks where the ROM characteristics are known, but the metallurgical information is not, making it possible to estimate these. With this goal, a geometallurgical model was developed with a neural network technique using 37 samples collected at two Brazilian gold mines. The Au and S grades in ROM, and the mine from where the sample was collected, were used as input variables. The model was able to forecast the following variables with a Pearson correlation coefficient on the cross validation test set equal to the value in parenthesis: mass (0.55) and metallurgical (0.54) recovery in the gravimetric concentrate, mass (0.80) and metallurgical (0.12) recovery in the flotation tailings, mass (0.77) and metallurgical (0.11) recovery in the leaching tailings, mass recovery (0.84) of gas sent to the sulphuric acid plant, and metallurgical recovery (0.65) in the leaching concentrate. The results obtained with neural networks were superior to the ones obtained when three alternative techniques were tested.

Abstract Evaluating mineral resources requires the prior delimitation of geologically homogeneous stationary domains. The knowledge about the ore genesis and geological processes involved are translated into three dimensional models, essential for planning the production and decision-making. The mineral industry usually considers grade uncertainty for resource evaluation; however, uncertainty related to the geological boundaries are often neglected. This uncertainty, related to the location of the boundary between distinct geological domains can be one of the major sources of uncertainty in a mineral project, and should be assessed due to its potential impact on the ore tonnage, and consequently, on enterprise profitability. This study aims at presenting three different methodologies capable of generating multiple geomodel realizations and thus, assessing uncertainty. A real dataset with high geological complexity is used to illustrate the methodology. The results are compared to a deterministic model used as a reference scenario.

Abstract The inclusion of grade uncertainty for multivariate mineral deposits is of great importance for the correct management of subsequent decisions involved in mining planning. Mapping grade uncertainties allows maximization of profit and resource extraction. In this article, the co-simulation turning band algorithm is applied with the aim of predicting multivariate grade uncertainties. Moreover, a probabilistic analysis in long term mining sequencing is proposed in order to select the best given grade scheduling uncertainty derived from the simulations. A case study in a phosphate mine shows that the correlation of co-simulated variables honors the original data and there is an improvement in the project by an increase in Net Present Value (NPV) planning considering grade uncertainties. A comparison is performed with the results derived from the selected schedule and the results using the model based on kriged grades.

Abstract One of the basic factors in mine operational optimization is knowledge regarding mineral deposit features, which allows to predict its behavior. This could be achieved by conditional geostatistical simulation, which allows to evaluate deposit variability (uncertainty band) and its impacts on project economics. However, a large number of realizations could be computationally expensive when applied in a transfer function. The transfer function that was used in this study was the NPV net present value. Hence, there arises a necessity to reduce the number of realizations obtained by conditional geostatistical simulation in order to make the process more dynamic and yet maintain the uncertainty band. This study made use of machine-learning techniques, such as multidimensional scaling and hierarchical cluster analysis to reduce the number of realizations, based on the Euclidean distance between simulation grids. This approach was tested, generating 100 realizations by the sequential Gaussian simulation method in a database. Proving that similar uncertainty analysis results can be obtained from a smaller number of simulations previously selected by the methodology described in this study, when compared to all simulations.

Abstract Mine planning is dependent on the natural lithologic features and on the definition of their limits. The geological model is constantly updated during the life of the mine, based on all the information collected so far, plus the knowledge developed from the exploration stage up to the mine closure. As the mine progresses, the amount of available data increases, as well as the experience of the geological modeller and mine planner who deliver the short, medium, and long-term plans. This classical approach can benefit from the automation of the geological mapping on the mining faces and outcrops, improving the speed of repetitious work and avoiding exposure to intrinsic dangers like mining equipment, falling rocks, high wall proximity, among others. The use of photogrammetry to keep up with surface mining activities boarded in UAVs is a reality and the automated lithological classification using machine learning techniques is a low-cost evolution that might present accuracies above 90% of the contact zones and lithologies based on the automated dense point cloud classification when compared to the manual (or reality) classified model.

Abstract High anomalous grades are very common in gold deposits, whose presence requires careful treatment to prevent overestimation of metal content. Mineral resource analysts have worked on the estimation of several gold deposits, and none of the classical methods were able to avoid manual interventions, such as cutting high grades for local estimation or using more information beyond the data for the variogram inference. The Field Parametric Geostatistics (FPG) is presented as an alternative for the application of linear kriging methods to estimate highly skewed distributions, proposing a mathematical model which incorporates the grades and its representativeness into a new variable, reducing the influence of high grades without empirical manual interventions. In this article, the mathematical formulation of the FPG theory is presented, as well as its application in datasets with outliers and high skewed distributions: the Walker Lake dataset and the Amapari gold deposit. The results are compared to results obtained by the application of standard techniques, demonstrating that FPG is a feasible alternative to estimate local grades and local reserves for highly skewed variables.

Abstract The current developments with unmanned aerial vehicles (‘UAVs’) are revolutionizing many fields in civil applications, such as agriculture, environmental and visual inspections. The mining industry can also benefit from UAVs in many aspects, and the reconciliation through topographic control is an example. In comparison with traditional topography and maybe modern techniques such as laser scanning, aerial photogrammetry is cheaper, provides faster data acquisition and processing, and generates several high-quality products and impressive level of details in the outputs. However, despite the quality of the software currently available, there is an uncertainty intrinsic to the surfaces acquired by photogrammetry and this discrepancy needs to be assessed in order to validate the techniques applied. To understand the uncertainty, different surfaces were generated by UAVs for a small open pit quarry in southern Brazil. Well-established topographic surveying methodologies were used for geolocation support and comparison, namely the RTK (real-time kinetic) global navigation satellite system (GNSS) (here called conventional method) and laser scanning. The results showed consistency between the UAV surfaces with a few outliers in when vegetation, water and mobile objects during the flight missions. In comparison with the laser-scanned surface, the UAV results were less erratic surrounding the RTK points, showing that surfaces generated by photogrammetry can be a simpler and quicker alternative for mining reconciliation, presenting low uncertainty when compared to conventional methods.

Abstract Niobium metallurgical recovery measures how much of the metal content in the ore is separated in the concentrate after the mineral processing stages. This information can be obtained through laboratory tests with ore samples obtained during drilling. Thereunto, representative ore samples are subjected to tests mimicking the ore concentration processing flow, but these experiments are time consuming and costly. The main objective of this study was to develop a more efficient way to obtain the metallurgical recovery information from ore samples. Based on the development of chemometrical studies, the chemical components currently analyzed in the ore with correlation to the metallurgical recovery were identified. These correlated variables were used to build a nonlinear multivariate regression model to explain the response variable, i.e. metallurgical recovery. The Principal Component Analysis was used in this work to define which chemical variables contribute most to explain the metallurgical recovery phenomenon. The Second order regression equation (Response Surface) was the most suitable methodology to explain the metallurgical niobium recovery and was created by the interaction of the five most important chemical variables. After the exclusion of outliers, the linear regression coefficient between the metallurgical recovery calculated and the metallurgical recovery analyzed was 82.59%. The use of the second order regression equation contributes to reduce the amount of experimental analysis to assess the geometallurgical niobium ore response, promoting the reduction of costs for metallurgical characterization of the ore samples. The methodology proposed proved to be efficient, maintaining an adequate precision in the forecasted response.

Abstract Short-term mining planning typically relies on samples obtained from channels or less-accurate sampling methods. The results may include larger sampling errors than those derived from diamond drill hole core samples. The aim of this paper is to evaluate the impact of the sampling error on grade estimation and propose a method of correcting the imprecision and bias in the soft data. In addition, this paper evaluates the benefits of using soft data in mining planning. These concepts are illustrated via a gold mine case study, where two different data types are presented. The study used Au grades collected via diamond drilling (hard data) and channels (soft data). Four methodologies were considered for estimation of the Au grades of each block to be mined: ordinary kriging with hard and soft data pooled without considering differences in data quality; ordinary kriging with only hard data; standardized ordinary kriging with pooled hard and soft data; and standardized, ordinary cokriging. The results show that even biased samples collected using poor sampling protocols improve the estimates more than a limited number of precise and unbiased samples. A welldesigned estimation method corrects the biases embedded in the samples, mitigating their propagation to the block model.

Abstract Geostatistical simulation comprises a variety of techniques which can help on the decision-making process for uncertainties. They allow the uncertainty assessment of function responses (which depend on the simulated inputs) commonly through a non-linear relationship (net present value, interest tax return, geometallurgical ore recovery...). However, one of their limitations is that running the simulation can take considerable processing time to be executed in large deposits or large grids. Herein is presented an attempt to solve this problem in short-term modeling cases, via the use of Multiple Random Walk Simulation. This algorithm combines kriging with the simulation of independent random walks in order to generate simulated scenarios much faster than via traditional simulation algorithms. A case study is presented to illustrate the application of the method in an iron mine. The Multiple Random Walk Simulation models were properly built, respecting the reproduction of both histogram and variograms. Also, the speed-up was compared with standard methods of geostatistical simulation and there was a considerable speed gain with Multiple Random Walk Simulation (3.39 to 5.65 times faster than the others).

Abstract Prior to every geostatistical estimation or simulation study there is a need for delimiting the geologic domains of the deposit, which is traditionally done manually by a geomodeler in a laborious, time consuming and subjective process. For this reason, novel techniques referred to as implicit modelling have appeared. These techniques provide algorithms that replace the manual digitization process of the traditional methods by some form of automatic procedure. This paper covers a few well established implicit methods currently available with special attention to the signed distance function methodology. A case study based on a real dataset was performed and its applicability discussed. Although it did not replace an experienced geomodeler, the method proved to be capable in creating semi-automatic geological models from the sampling data, especially in the early stages of exploration.