The best remaining orebodies are deep, and depth changes the economics of mining in ways that do not yield to incremental effort. As an operation descends from a few hundred meters toward three kilometers, in-situ stress climbs several-fold and rock temperatures roughly double, seismicity and rockburst risk rise with the stress, ground support grows more expensive, ventilation and cooling come to dominate energy use, and the time workers spend simply traveling to and from the face erodes the hours available to produce. None of these is a detail to be engineered away one at a time. Together they ask a harder question: can a deep orebody be mined at the scale and safety a modern operation requires at all?
We took that question in its full form, asking how to mine three million tonnes per year, at three kilometers depth, in a safe and scalable way, and treated it as a design problem to be solved rather than a forecast to be hedged. The approach we used is the one we bring to any hard technical challenge, and it rests on three reinforcing moves: frame the challenge, re-imagine the system, and enable delivery.
Frame the challenge: if a problem looks infeasible, it has not been defined well enough
The first move is the one most often skipped. Faced with a daunting problem, teams reach for technologies before they have understood what they are actually up against, and so they optimize inside the wrong solution space. We do the opposite. We set a deliberately bold ambition, treat the constraints as design drivers rather than reasons to stop, and align the whole business on what success would have to mean before any solution is on the table.
Framing well means separating two kinds of constraint that are usually tangled together. Hard constraints are the ones physics and geology impose: stress, heat, thermodynamics. Soft constraints are the ones the industry has imposed on itself: standard methods, legacy layouts, organizational habits, the way it has always been done. Much of what an operation experiences as an immovable limit turns out, on inspection, to be a design choice wearing the costume of a law of nature. Pulling the two apart is what makes a problem that looked closed start to open.
"Many perceived limits at depth were design choices, not laws of nature. Separating the hard constraints from the inherited ones is what turns an infeasible problem into a solvable one."
Re-imagine the system: question the orthodoxies, then design end to end
With the problem framed, we challenge the truths the industry takes for granted. Is high stress only an enemy, or could it be made to do useful work? We then ask whether heat and ventilation are first-order problems or consequences of choices made elsewhere in the design, whether a known method is being forced to fit when the real question is what a design would have to look like to make the orebody mineable, and whether a deep mine must stay small or could change the equation by going deliberately large. Taken seriously, each of these questions points toward a different class of solution.
To answer them we go to the people who have pushed the boundaries, in mining and well beyond it. On this challenge we identified more than two hundred experts, interviewed over forty-five of them in depth, and convened twenty-six across a series of working forums. That dialogue, structured through value analysis and clustering, turned a broad challenge into one hundred and ninety-six focusing questions, then into one hundred and seventy-four ideas, then into forty-three clusters, and finally into thirty-one concepts worth developing further. The discipline of that funnel matters: it lets a team explore widely without converging too early on a single answer that feels safe but may be wrong.
The concepts themselves organized around the orthodoxies we had questioned. Some sought to change the stress, making a deep mine behave more like a shallow one. Some went deliberately small, using novel mechanisation and design to shrink exposure and work around the stress. Others went deliberately large, using new design, destressing, and sequencing to remove the consequence of stress rather than fight it. Across all of them, one finding held: value came from designing the whole system together, the mining methods, the innovations, and the people, rather than from any single technology added to an unchanged design.
Enable and deliver: technology as enabler, with ownership and gates
A study that ends at a slide of clever concepts has not finished the job. The third move is to build the path from today to the future: using technology as an enabler rather than the solution, sequencing concepts into roadmaps, integrating the people and processes that will have to carry them, and establishing clear ownership and decision gates so that promising ideas are de-risked in a deliberate order rather than all at once. The concepts are deliberately kept optional, each offering more than one technological pathway, so that the operation preserves choice as uncertainties resolve. From this work, early mining-method changes have begun to move into practice while further development continues.
What deep mining demonstrates is general, and it is as much about process as about rock. Take the time to deconstruct a problem before reaching for solutions, and be honest about which constraints are physical and which are inherited. Invest early effort in understanding root causes rather than scouting technologies, so the work is not optimizing within the wrong frame. Anchor the thinking in deep dialogue with world-leading experts, drawn from global extremes and adjacent industries, and balanced by the internal specialists who know the site's realities and trade-offs. Design systems, not standalone technologies. And constrain the thinking deliberately while using scenarios to expand the envelope, exploring multiple futures to test robustness rather than converging too early on a single optimal answer.
Mining at depth represents an opportunity, if operations are able to overcome the challenges that worsen with every additional meter. The work so far suggests those challenges are more tractable than they first appear, but only for teams willing to question what they thought was fixed, and to build the system rather than buy the parts.