Pit Optimization Software for Junior Miners

Pit optimization is where a block model turns into a question about money. Which blocks pay to dig, which stay in the ground, and what shape the pit takes once you account for revenue factors and stripping ratios. At pre-feasibility, getting that shell roughly right matters more than getting it perfect.

The problem for a junior is that most of the tooling assumes you are not one. Deswik, Vulcan, and the other enterprise suites rarely publish pricing, and when you ask, the number is built for a major. If your software budget is under $5K, the brochure tour does not help you.

Strip the jargon and the job is simple. You hand the software a block model: a 3D grid where every block carries an estimated tonnage, grade, density, and rough dollar value. It works out which blocks are worth digging once you account for what it costs to mine them, haul them, and process them, and what the metal sells for. Blocks that pay their way are ore. Blocks that do not are waste you may still have to move to reach the ore underneath.

The constraint that makes this hard is the wall. You cannot drop a vertical shaft around a single rich block. The pit walls have to lay back at a safe slope angle, which means every tonne of ore near the bottom drags a cone of waste above it. The optimiser's job is to find the shape of hole that captures the most net value while respecting those slope angles. That shape is the pit shell, and the line it draws between ore and waste drives every cost that follows it. A block on the wrong side of that line is a truckload sent to the wrong place, multiplied across the life of the pit.

Under the hood most tools run some version of the Lerchs-Grossmann algorithm or a pseudoflow method. You do not need the maths, but you should know the name, because "what method does your optimiser use" is a fair question to put to any vendor or consultant, and the ones worth hiring answer it without flinching.

Run the optimiser at a few different metal prices (revenue factors) and you get a set of nested shells, smallest at the lowest price, largest at the highest. Those nested shells are the raw material for pushbacks, the phases you mine in sequence. Two things are worth being clear on. The optimiser gives you a shape, not a schedule; sequencing the dig over time is a separate step. And your strip ratio falls straight out of the shell as the ratio of waste to ore. Change the price assumption and the whole picture moves. That sensitivity is the point. At pre-feasibility you are not chasing the perfect pit, you are learning how much the answer wobbles when the inputs do.

There is no single market here. There is a spread, and where you land on it has more to do with your budget and who is on staff than with the size of your orebody.

At the free end sit open-source optimisation libraries (pyomo, PuLP, OR-Tools), plus the academic codes and public datasets built around them. They will generate a pit shell if you have someone comfortable in Python who is willing to own the methodology. No licence fee, full control, and you can read every line of what it is doing. The flip side: no support line to call, no validation behind it, and the maintenance lands on one person's desk. When that person leaves, the capability usually walks out with them.

The middle is lighter commercial tools and academic-licensed software. Some have a proper interface, so you are not writing code to get an answer. They cost less than the big suites and can be plenty for testing economics. The trap here is licensing. Academic and research licences frequently forbid commercial or JORC use, and that restriction is easy to skim past until you try to put the output into a reportable study and find you cannot.

At the top are the enterprise suites (Deswik, Vulcan, Datamine, Micromine and the like) that run geology through optimisation into scheduling in one validated chain, with training, support, and an audit trail behind every number. That integration and that paper trail are what you are paying for. The pricing is not published because it is built around a major's budget, and for a junior that number is usually the whole problem.

One thing the tiers do not tell you: a more expensive tool is not a more correct one for the job in front of you. The enterprise suite and the Python script can land on the same shell from the same inputs. What separates them is support, defensibility, and how much of the work you carry yourself, not the answer they hand back.

A handful of questions sort the options faster than any feature list, and they are the same ones I would ask before signing anything.

Is your block model final, or still moving? If the geologists are still re-estimating, you are buying a tool to run again and again, which changes the sums on a per-study consultant fee.

Do you need pushbacks and phase design, or just a shell to test whether the economics hold up? A lot of juniors over-buy here, paying for scheduling capability they will not touch for a year.

Is the output for an internal go/no-go, or for something you will have to defend in a JORC or NI 43-101 report? Reportable work raises the bar on methodology and reproducibility, and a black box will not survive a competent person's questions.

Does it export to the formats your geological model and scheduler already speak? If moving data between tools means hand-editing files every time, you pay for that on every cycle.

And the one people forget: who runs this after the consultant goes home? If nobody on staff can reproduce the result next quarter, you do not own the capability, you rent it. Match the tool to the decision actually in front of you, not the mine you are hoping to build in five years.

The warnings are consistent across every tier, and none of them get cheaper to find out about later.

A tool that cannot export to a standard format. Lock-in is a cost even when the licence looks cheap.

An optimiser whose pit-shell method nobody can explain. If the consultant or vendor cannot tell you how the shell was generated, you cannot defend it, and a black box is worthless the moment someone competent starts asking questions.

A licence that quietly rules out the use you bought it for. Read the commercial and JORC restrictions before you commit, not after the study is half written.

A consultant who will not hand over the model file and the parameters. If you cannot re-run the optimisation yourself, you cannot reproduce the result, and a number you cannot reproduce is a number you cannot stand behind in front of a board or a regulator.

Buying the full enterprise suite for a single shell you will run twice. Over-buying is not the safe option, it is just the expensive one, and the capability you never touch does not make the answer any better.

Browse Mine Planning & Design and Geology & Resource Modelling for the products in this space.