Assays drive every exploration decision. Here is how standards, blanks and duplicates prove the numbers can be trusted, and how to run a practical QA/QC programme.
Every exploration decision rests on assay results: where to drill next, whether a target is worth more work, and in time whether there is a resource at all. Quality assurance and quality control (QA/QC) is how you prove those numbers are right. It rests on three kinds of control samples: standards, blanks and duplicates.
Why QA/QC matters
Errors creep in at every step. Samples can be swapped or mislabelled, material can carry over from one sample to the next during preparation, and a laboratory can drift high or low. Without control samples, nobody can tell whether a good result is real.
Investors, partners and reviewers expect documented QA/QC, and international reporting codes such as JORC and NI 43-101 require the quality of the data behind public reports to be explained. Good QA/QC is cheap compared with redrilling or losing confidence in a whole dataset.
Standards: is the laboratory accurate?
Standards, also called certified reference materials (CRMs), are prepared material with a known grade, certified through testing by several laboratories. You insert them into the sample stream, and the laboratory's result shows whether its assays are accurate.
Choose standards that match the grades and rock type you expect, including one near your cut-off grade and one at higher grade. Each certified value comes with a standard deviation. A common approach treats a result outside two standard deviations as a warning and outside three as a failure that needs investigation.
Blanks: is there contamination?
Blanks are material with no meaningful grade. If a blank comes back with metal in it, something has contaminated the samples.
A coarse blank goes through the full preparation process of crushing and pulverising, so it tests for contamination during preparation. A pulp blank is already fine powder and tests the analysis itself. Blanks are most useful straight after samples you expect to be high grade, where carry-over is most likely.
Duplicates: is the result repeatable?
Duplicates are second samples taken from the same interval, and the difference between the pair shows how repeatable your results are.
- Field duplicates are a second split taken at the rig or core saw. They show the precision of sampling itself.
- Coarse duplicates are a second split after crushing. They test sample preparation.
- Pulp duplicates are a second portion of the final powder. They test the laboratory's precision.
Results are usually reviewed on scatter plots of original against duplicate, or as the relative difference between each pair. Coarse-grained or nuggety mineralisation, such as some gold, naturally shows wider differences.
How often to insert control samples
There is no single rule, and the competent person or project geologist sets the rates for each programme. Many programmes insert roughly one of each type of control sample in every 20 samples, which adds up to around 10 to 15 percent of all samples, with extra blanks after likely high grades. Early-stage programmes sometimes use lower rates, and resource drilling often uses higher ones.
Setting up a QA/QC programme in five steps
- Choose standards that cover your expected grades and match your rock type, and order enough for the whole programme.
- Source a blank material that is known to be barren, and test it before you rely on it.
- Decide the insertion rates and the positions of control samples in the numbering sequence, and write them into the sampling procedure.
- Agree turnaround times and reporting formats with the laboratory before the first batch is dispatched.
- Set the pass and fail rules in advance, and name the person who reviews results and signs off each batch.
Reviewing the results
Control samples only help if someone looks at them quickly.
- Plot each standard over time, with lines at two and three standard deviations
- Chart blank results against a threshold, usually a small multiple of the detection limit
- Plot duplicates and track the relative difference between pairs
- Investigate failures as soon as they appear, and ask the laboratory to re-assay the affected batch where needed
- Keep a log of every failure, what caused it and what was done
A failure is not a disaster. An unexplained failure, found months later, is.
What a QA/QC report should contain
A short QA/QC report at the end of each programme, or each phase, keeps the evidence together: the control samples used and their insertion rates, charts for every standard, blank and duplicate type, a list of failures with their causes and actions, any re-assays, and a conclusion on whether the data is fit for its intended use. Reviewers and future teams will read this report before they trust your numbers.
Sample handling and chain of custody
QA/QC starts before the laboratory. Clear sample numbering, checked sample sheets, secure bags, and a recorded chain of custody from the rig or core yard to the laboratory prevent many of the errors that control samples would otherwise catch. A written drill core sampling procedure keeps everyone doing the same thing, the same way.
Common mistakes
- Using standards that do not match the grades or rock type being tested
- Placing control samples in obvious positions the laboratory can recognise
- Collecting QA/QC results but only reviewing them at the end of the programme
- Mixing up original and duplicate samples on the sample sheets
- Losing the link between control samples and the batches they belong to
How Takween supports QA/QC
Our sample management and QA/QC team sets up protocols, inserts and monitors standards, blanks and duplicates, and reports exceptions while the programme is running. Because we also run the drilling and manage the exploration data, the samples, the paperwork and the database stay in step. Contact us to discuss QA/QC for your programme.
