Cutoff grades. Everyone gets hung up about cutoff grades, as if they're voodoo. The basic text is Lane's Economic Definition of Ore (1968), but Lane was a boffin and it's a tough read. However. The first thing that needs to be understood is what the limiting factor of the mine is. It is mine limited, mill limited or market limited?
An open pit can mine as much dirt as you like, but the mill has a capacity of 1.5Mtpa, so it's mill limited. An underground mine can fit 1Mtpa up the decline to feed a 1.5Mtpa mill. It's mine limited. A crappy iron ore junior could sell 8Mtpa of 58% high-phos fines in 2012, but now can't sell a cracker. It's market limited.
Mill limited is generally the case. Market limited is difficult to analyse and I won't go into it here. Nor will I go into the mathematics, but, when analyzing cutoff grades, fixed costs are added to the limiting factor. So, for the open cut, the mill operating cost is $25/t plus $8/t fixed costs. Rule of thumb - each department in a mine costs at least $1Mpa: geology, mining, maintenance, admin, survey, safety & enviro etc. This has the effect of expanding the pit in an optimization to find more ore to pay the mill fixed cost. Seems counterintuitive, but that's the way it works.
Cutoff grades are time dependent. Each tonne of ore fed into a mill in a mill-limited or mined in a mine-limited operation potentially displaces a tonne of higher grade material that could produce higher margins at the same operating cost. This is why ore loss and dilution is so important (rule of thumb - 5% for each in a well managed, bog-standard pit, mostly much higher underground). Ye feed in a tonne of waste and it costs ya $33, for no return.
Having understood the limiting factor, there are then three time-dependent operating cutoff grades applied to the operation:
1. Incremental cutoff grade - mill operating cost/(grade x recovery x metal price). I won't go into it, but the mining cost cancels out and is irrelevant - if ya don't mine, ya don't go no ore. This is used to determine how much of the Resource may be economic. The best strategy is to stockpile this ore, if higher grade material is available and the dirt don't deteriorate with time, reducing recoveries through (usually) oxidation of sulphides.
2. Operating cutoff grade - (mill operating cost + mine operating cost + site overheads)/(grade x recovery x metal price). This is used in mine planning for monthly, quarterly, annual plans, whatever.
3. Full cutoff - (mill operating cost + mine operating cost + site overheads + corporate overheads + finance costs)/(grade x recovery x metal price). This is used to maximize the margins early on in the life of the mine to maximize NPV and repay capital.
From these cutoffs, the engineer can derive a schedule and cash-flow model. Full cutoff early on in the life of the mine, operating cutoff when things are steady state - although a full cutoff is introduced from time to time (say every five years) to pay for sustaining capital. At the end of the mine's life, when the Resource has been mined and the mill paid off and the Directors are backslapping and high-fiving, the stockpiled marginal gear can be fed through the mill, hopefully producing sufficient revenue to pay for rehab and closure costs.
Time dependent, kiddies. The Competent Person for Reserves MUST take a reasonable view on future commodity prices and exchange rates. If you see a cobalt "Reserve" that uses a cobalt price significantly greater than the consensus future price at the time of reporting (and there's plenty of places you can get this from), then run or start shorting.
A lot of this stuff can be automated by various software packages, which takes a lot of the labour out of it. Ultimately I prefer to see a spreadsheet schedule, modified by a competent engineer - I simply don't trust the software to react strategically to an abstract and constantly changing situation. I'm not the world's best scheduler (I'm crap at it), but I know what I'm looking for when a flunky bangs up a schedule or Linem Pockets Mining NL releases the results of its feasibility study.
Incidentally - there's no such thing as a Definitive or Bankable Feasibility Study. These are archaic and in my view misleading terms. The JORC Code (Clause 40ff) defines Scoping Study (implied accuracy of the parameters and underlying Inferred Resource +/- 35 - 50%); Pre-Feasibility Study (implied accuracy of the parameters and underlying Indicated Resource +/- 25 - 35%); and Feasibility Study (implied accuracy of the parameters and underlying Measured/Indicated Resource +/- 15 - 25%).
It be a statistical distribution. It be engineering tolerance. So never, ever accept a headline value - it's actually a preferred value within a range. IRR is a far better indicator of risk in a competently prepared cash-flow model than NPV; provided that everyone has worked appropriately, particularly the Resource geo. Error is cumulative.
