The discipline of the number
Most plants run on unaudited numbers.
We settle disputed numbers, explain drifting ones, and verify the ones a decision rests on.
We settle disputed numbers, explain drifting ones, and verify the ones a decision rests on.
Every figure in your board pack came from a sample. Almost none of them carry a stated uncertainty.
Is the equipment capable of taking an unbiased sample?
How much uncertainty does the reported number actually carry?
Is there a systematic error, proven rather than asserted?
Sampling is one of the largest sources of error in any reported figure and the least audited. It is where we work.
Whether the numbers an operation runs on are good enough for the decisions taken on them.
Mining and processing operations are measured obsessively and audited hardly at all. A plant produces thousands of numbers a day, and almost every one is used with more confidence than its derivation supports.
The consequences are not dramatic, which is why they persist. A figure carrying four points of uncertainty does not fail. It quietly misinforms a settlement, a trial and a board pack, and every one of those decisions is taken in good faith on a number nobody has audited.
To make the numbers behind mining decisions defensible, by measuring how good they actually are and saying so plainly.
An industry where no capital decision, settlement or public report rests on a number whose uncertainty has never been quantified.
Thirty minutes, no charge. We establish what is being measured and what it is used for.
Exactly what will be done, what it produces, and what it costs.
Measurement on site, and a report that includes the working.
Your plant reports a number. We tell you how far that number is from what is actually there.
Sampling is the one part of an operation where a fault leaves no trace. The samples keep arriving on schedule and they look entirely normal, while the error they carry travels into every figure downstream.
What you get is not a repair. It is a clear statement of how reliable your numbers are, and a verdict on whether they are good enough for the decisions being taken on them.
Each answers one question and each can be bought alone.
Is the equipment capable of taking an unbiased sample?
We measure the equipment that takes your samples and check it against the rules that govern correct sampling. Most systems that fail, fail here, and most have never been checked.
How much uncertainty does the reported number actually carry?
We follow a sample from the moment it is taken to the moment it is analysed, and work out how much error each step adds. The step that matters most is rarely the one people expect.
Is there a systematic error, proven rather than asserted?
We test your sampling against a reference taken by hand and analyse the difference statistically. It is the only method that settles the question, which is why it is worth doing properly once.
We audit any number we can independently take a correct reference of.
Where we cannot verify a parameter reliably, we say so rather than sell you an opinion.