By Carly Leonida, European Editor
The main challenges faced in mine planning remain relatively unchanged from those of 20-30 years ago. The question still stands: how can we produce more metal from assets in a way that maximizes operational and financial efficiency over both the short and long term?
During this time, the tools available to mine planners have reached new levels of capability in terms of computing power and speed and, as one would expect, the accuracy of plans has improved accordingly. However, at the same time, the level of uncertainty that mining operations face at every stage of their lives has also risen; from exploiting increasingly lower grade, more complex mineral deposits, to rapidly changing operational environments, rising production costs, and unstable markets that face numerous challenges including, most recently, a pandemic-induced recession.
“Turbulence in the market in terms of revenues and ever-increasing operating costs creates lots of instability. It is difficult to make decisions about key items in mine planning and design, such as the size of ultimate pit and scale of operation,” said Anoush Ebrahimi, principal mining engineer at SRK’s office in Vancouver, Canada. “In our view, the biggest challenge is to plan for flexibility. Challenges could range from increased difficulty raising capital, to increased scrutiny on environmental and social impacts leading to changes in designs.
“Although modern mine planning tools allow planners to analyze more data, they must be able to focus in on practical solutions at all levels of the mine planning process and provide input on which plan is most effective.”
To accomplish this, SRK relies heavily on practical experience in combination with high-level economic analysis to ensure that all necessary factors are taken into account.
“High-level economic analyses are included at every stage of project to provide guidance on how design changes affect the operation. By doing this, we help to lower risks and enable our clients make strategic decisions on time,” Ebrahimi added.
It is worth noting that the fundamentals of planning are the same for both open-pit and underground operations — plans are created in sequences with economical assessments at each step. However, for underground design, information with higher resolution is required, which is time consuming and generally more expensive and, for the purpose of this article, it will focus mainly on open-pit operations.
Some analyses that SRK has found beneficial to its clients in planning open-pit operations include simulation of mining progress and equipment; pre-concentration of ore, prior to processing, and probability of failure geotechnical analyses on all pit designs, including interim phases. The team has also been working to develop techniques to better understand potential mining dilution, which usually is a source of reconciliation issues in mining operations. This helps to increase the level of accuracy for ore selection.
A Tailored Approach
SRK recently helped several mines to develop and apply sophisticated geo-metallurgical models in their designs and production schedules, resulting in higher profitability. Another project saw the company develop a novel approach to mine design that incorporated complex geotechnical design criteria into a model for a complex coal deposit.
Justin Smith, senior mining engineer with SRK Consulting’s team in Reno, Nevada, elaborated: “Traditionally, design work in complex coal would be done in sections, however, this new methodology allowed for 3D modelling of the pit. Several trade-offs were then conducted in much less time than it would have taken to complete a single sectional design.”
The rise in autonomous and remote-operated mining equipment is also challenging the norm in pit design.
“In 2019, we helped two major mines in Canada to review and model the impacts of new technologies on mine design,” Ebrahimi told E&MJ. “New technologies such as autonomous trucks are changing the way we operate mines. This should be reflected in our designs. We helped the operations to explore the items that may need to be changed in mine design and assessed the impact of new technology implementation.”
Another key focus reflects the enhanced scrutiny mining companies are under with regard to their environmental impact, particularly in tailings and waste management. Mines are often now designed with their end-of-life in mind to lower closure risks and costs wherever possible.
“The key to ensuring that a mine plan is profitable while minimizing the environmental impact and maintaining a social license is to look at all aspects of the mine plan and optimize it as a system, rather than just focusing on each individual component and piecing them together at the end,” said Smith “This could be referred to as a holistic approach to mine planning. For environmental impacts, SRK has a concept in planning called ‘design for closure.’
“At the beginning of the project planning stages, we include team members from all disciplines to review and challenge ideas. Armed with this wide range of information, our mine planners are able to develop a site-wide plan that considers how all factors interact with each other. Potential solutions and alternatives can be assessed, and compared to determine their effectiveness, practicality, and effects on profitability of the operation.”
The Right Tools for the Job
Digital mines require digital tools to plan their operations, and the software available to mine planners today is far more powerful and sophisticated than predecessor versions. These packages can handle more iterations, scenarios and complete a higher level of analysis in a shorter time period. Items such as cycle times, volumetric calculations, dilution factors, and densities can be included with much higher granularity than was previously possible, and the simulation of results can also add significant value by decreasing risk for mining operations.
The accuracy of mine plans has risen in line with the amount of data available, but Smith pointed out that while data can bring opportunities to optimize, the sheer quantity can also create a challenge.
“Today, we are collecting, and analyzing, more data than has ever been available in the past,” he said. “While this helps us to understand the unique challenges facing each mine, it can be quite overwhelming and difficult to manage. Being able to process huge amounts of data and sort out what is important is critical to successful mine planning today.”
The extra data generated, for example, by more sophisticated geo-metallurgical models, higher resolution dilution calculations and more complex operating cost models, means that the number of variables that must be considered have also increased substantially.
“Mine planning tools are, for the most part, well equipped nowadays to handle this complexity,” Smith said. “So, as the tools get better and more powerful, we are considering a much larger problem and, in some instances, it may take more time for preparation and analysis than in the past. The key reason being more complicated inputs and the demand for higher accuracy.”
This is why flexibility in planning is so key. New startups must often stage their development to reduce preliminary capital costs, and fewer greenfield projects mean many operations must make costly adjustments to mine in previously disturbed areas. Lower grades are also driving companies to reduce costs, and significantly longer permitting timelines increase both the initial costs and the risk that operations may not be approved following large expenditures.
“Software advances have also increased the speed at which trade-off studies can be conducted, allowing companies to test several scenarios before committing to a final mine plan,” noted Smith.
Advances in Software
Philippe Lebleu, principal mining engineer and open-pit manager at AMC Consultants in Vancouver, recently presented a webinar on the capabilities of modern mine planning software.
“Recent advances in open-pit strategic mine planning software enable the optimization of intricate mining problems associated with a complex and vast array of parameters and constraints,” he said in the follow-up paper.
“These programs simultaneously optimize the mining sequence, cut-off grade selection, amount of mining equipment needed and capital expenditure to maximize the net present value (NPV) of a project or deliver on corporate goals.”
Lebleu explained that, typically, a mine plan is produced based on fixed cut-off grades and strives to achieve a primary goal, such as a target mill feed, while maintaining a smooth total material movement to simulate the operation of a predetermined mining fleet. The mine plan is then used to calculate equipment hours and numbers.
“The latter form the input into cost models that calculate mining costs and overall project value. The process is repeated with different assumptions for total material movement, sequencing or cut-off grades,” he explained. “Once the mining engineer creating the plan is satisfied that an adequate number of scenarios have been evaluated, the best outcome is selected and used for the rest of the process such as waste dump and stockpile design.”
However, this approach can be time-consuming and sub-optimal from a value perspective — especially when the operation being planned is complex. It is also heavily reliant on the mining engineer’s understanding of the deposit and their experience — something that is becoming more difficult as experienced engineers retire and skills shortages hit home.
“That is why the use of advanced mine planning software is becoming more widespread among mining companies and consultancies,” Lebleu said. “Mixed-integer linear programming processes and advanced algorithms allow rapid evaluation of complex problems and help engineers and management make educated decisions regarding the best mine development or optimization strategy to adopt.”
Lebleu explained that the advantage of using advanced mine planning software lies in its ability to achieve multiple targets while respecting a variety of constraints. For example, by looking ahead to ensure that the choice of mine development made in the first year of production does not jeopardize its ability to achieve targets in subsequent years and achieve optimum value.
“Improving a project’s NPV can be achieved by applying variable cut-off grades by scheduling period. This is accomplished by defining grade bins based on the spatial distribution of the ore and grade tonnage curves,” he said. “Advanced mine planning software can define a ‘high-grading’ strategy that brings high-grade material to the processing facility earlier in the mine life while balancing the total material mined, and therefore the mining costs to do so. Following a high-grading strategy can typically result in a 15% higher NPV compared to directly processing run-of-mine ore.”
The main issue with high grading is that mining more material requires additional equipment. The impact of that cost can be evaluated within the software by accounting for equipment hours as a variable in the model. Assigning a capital cost to a larger fleet allows the software to gauge whether and when to increase production capacity to optimize the project’s NPV. Capital expenditure decisions are not limited to the mining fleet and can extend to evaluating the merits and best timing for increasing processing plant capacity or putting capital toward developing a new pit.
Lebleu concluded that, despite the advances in mine planning software, mining practicality still needs to be central to a development strategy.
“It is in the engineer’s best interest to produce multiple scenarios to demonstrate the incremental value generated and help tell the story behind the decision-making process,” he said.
Developing New Features
In order to delve a little deeper into software development, E&MJ turned to Sandeep Sandhu, general manager for North America at software specialist RPMGlobal.
“Not that long ago, generating and analyzing a single fairly basic scenario could be a challenging task. But now, users can evaluate dozens of alternatives and really start to understand the impact and sensitivity to key drivers,” he said.
“Furthermore, software has shifted from highly customized, siloed, desktop applications to a more commercial off-the-shelf offering that supports product integration and data sharing. Planning software is producing more data than previously possible, and without being able to expose planning data to analytical systems, effectively comparing options becomes much more difficult.”
Sandhu said that a common misconception is that software needs to be customized and scripted for a site.
“We have had a lot of success deploying software, which requires no customization to get running, but merely configuration for a specific operation,” he explained. “You no longer need to spend months to get new software up and running — it can be as little as a week and you no longer need a consultant on site full time providing support.”
Another misconception Sandhu and his team often hear is that there is an “optimal plan.”
“In reality, anything that is optimal from a value standpoint is probably unachievable from an operations standpoint,” he said. “There really needs to be a focus on finding maximum value within an achievable framework.”
RPMGlobal’s XPAC Solutions are built for planning according to different commodities and mining methods. The last few releases have included new features to enhance modelling of haulage within open-pit mines.
“With haulage costs being a major portion of mining costs in pits, these enhancements have allowed for more accurate haulage calculations within the solutions,” Sandhu said. “We have also expanded the range of design capabilities available in all our solutions.”
In November 2019, RPMGlobal unveiled several upgrades for its XECUTE short-term scheduling solution to help expand planning across the mine value chain.
“The market reaction to the new release has been very positive,” Sandhu enthused. “With more clients looking to move to remote operating strategies, and more companies pushing digital strategy agendas harder, XECUTE’s offering as an enterprise mine planning application is filling a gap that other planning products cannot provide. Since the latest release in November, we have made our first sales of XECUTE into Indonesia and Russia and have gained further interest in Western Australia (Australia) and North America.”
Sandhu said one of the most exciting aspects is the enthusiasm with which the industry is embracing new technologies.
“Mining has traditionally been a bit slow to the mark with these sorts of shifts but there is a large amount of activity focused on new technology,” he told E&MJ. “Mining companies are now looking at things like better automation of tasks that have traditionally been very manual across the pit-to-port value chain.
“The other big trend is the accessibility of the new software coming out to market. In the past, mining software was typically reserved for use by those with a high degree of skill in the software. However, with new technology, we are seeing the barrier to entry with new software being reduced.”
Connecting the Dots
The team at the COSMO Stochastic Mine Planning Laboratory, based out of McGill University in Canada, is also looking at ways to optimize planning across the mineral value chain.
“Every few decades, we see a big focus on and new technologies for mine planning, production and scheduling for optimization,” Roussos Dimitrakopoulos, director of the COSMO Lab, told E&MJ. “The current challenge is how to harness new digital technologies to produce more from mining assets that we currently have and those we discover in the future. This brings up the topic of connecting all parts of the mineral value chain, or mining complexes as we like to call them.
“We are developing ways to connect and simultaneously optimize all of the components within a mining complex — from mines to products — to produce more metal, more efficiently and to capitalize on synergies between those components. We are also investigating new ways to treat mine waste.”
Stochastic mine planning aims to take into account and generate models that describe the uncertainty of the materials in the ground, as well as uncertainty in operating environment and in the markets.
Dimitrakopoulos explained: “So we start, rather than by saying we have one possible scenario of what’s in the ground, that we have several, to reflect the potential variability of the materials as well as quantify their uncertainty. When you understand what you are extracting, where it could flow and where it could end up, you create a different understanding of the supply of materials that comes from the whole setup.”
“Conventional mine planning methods, although they are very advanced methods, don’t allow for multiple scenarios of what might be in the ground. If you take mine planning software that is on the market, it generally uses what we call deterministic optimization — one input and one output — whereas stochastic optimization allows for multiple inputs to build up a framework that informs and simultaneously optimizes short and long-term mine asset planning. That kind of framework can accommodate uncertainty around what is in the ground and also in the markets, commodity prices, equipment availability… The stochastic framework has been shown to produce more from mining assets than conventional methods.”
One of COSMO’s current research projects is focused on harnessing the vast quantities of data generated by mining equipment and processes, as well as advanced software to create what the team call a “self-learning mining complex.”
“Nowadays, sensors are applied to everything and they generate huge amounts of data. We are looking at ways to capitalize on that information and use it to automatically update models with information on what is in the ground to better inform weekly, daily and hourly planning activities,” said Dimitrakopoulos.
The project uses artificial intelligence to create self-learning, deep neural networks that adapt production schedules in real time. Dimitrakopoulos believes we are only about five years or so away from the self-learning mining complex being a reality.
The COSMO team is also working to understand how stochastic mine planning techniques can be used to optimize the production of mine waste, with a view to better rehabilitating operations at the end-of-life, or even creating new streams of income through the repurposing of what were previously considered waste materials.
“We need to switch our thinking in this area,” Dimitrakopoulos stressed. “Why can’t we take the same technologies and techniques that we use to optimize the production of metals and apply them to engineer waste materials and ensure optimal outcomes in mine rehabilitation as well?
“Understanding how materials flow through mining complexes, and the characteristics of materials that we are sending to the processing plant will play a huge part in that. Metals are not the only products that a mine can produce. We should expand the tools and technologies that we have and apply them to tailings and waste as well.”