Well, I think whatever the cut off grade you will order standards from a range of grades low, medium and high grade that is a pretty standard way of doing it.

For what type of project do you need your standard? Exploration or an existing operation! For our mine, we use two standards:

One for the production cut off
One with a low grade

I copy paste information given by a colleague when we did our standard, it may help you to take your decision:

"Project-specific standards used on advanced drilling projects should be of the same geologic matrix as the mineralized rock being drilled and analyzed. Such standards are routinely made from assay coarse rejects from previously drilled holes on the project. On advanced projects where resource delineation is in progress, normal practice is to prepare three separate standards with grades representative of the resource’s expected cutoff grade, average grade, and high-grade.

For low-grade deposits where the grade of a significant portion of the resource is near the cutoff grade, a standard at the expected cutoff grade may be the most important standard. For such deposits, assay bias on the high or low side can result in a significant portion of the mineralized material being incorrectly classified as ore or waste, resulting in incorrect decisions to proceed or not to proceed with a project."

Thank you very much for your contributions. Exploration projects on both advanced stage and near mine.

The standards (usually pulps) are designed to test the accuracy at the laboratory. Cut off grade is an important consideration but equally important is the accuracy of higher grade samples particularly for precious metals.

I usually recommend that you have at least six standards available (two near cut-off, two medium grades and two high grades). The idea of paired similar grade samples is it is difficult for the laboratory to identify the certified values when there is some overlap in the standard precision.

More important though understands the precision at the cut-off grade. As cut-off grade dictates ore/waste decisions you need to have enough duplicates to understand the likely error in using individual data for these decisions. If you don't routinely collect duplicates then aim to resample around the cut-off grade where possible.

The purpose of blanks is to test for possible contamination between samples during sample preparation (usually the crush and grind stage). Where possible they should be inserted deliberately in ore zones and ideally at the base of the ore zone to test for 'lab-smearing' downhole (particularly in percussion drilling. The main issue here is that downhole smearing can leadto overstating the mineralization in drillholes.

In percussion drilling, physical determination of the mineralized zone or interview needs dedicated and experienced geologists to do that, but in an establishment where QA/QC procedure is such that pre-determined numbers are allocated for blanks as a routine procedure, how do you change that?

Again, in early stage of exploration drilling where about 70% of the total samples submitted is within the waste, is it necessary to insert Blanks every 20 samples?
Comments are highly welcome.

The best way I think, is for each batch insert the blank sample just besides the sample that are known to return anomalous values, remember that blank samples are aimed to test contamination in the preparation stage. The worse way to perform a QA/QC procedure is inserting control samples systematically. This task should be done very carefully and by experienced geologist. Same for field duplicates, I found a lot samples returning values below detection limit for both original and duplicate, so these samples become useless and are a wasting of money.

Most laboratories, in Australia at least, have more pulverizes than crushers so if you put the blank immediately after a suspected high grade sample the blank is likely to be crushed in the same crusher as the high grade sample but not pulverized in the same mill as the high grade.

The close to 70% waste drilling for explo projects is an average from many projects. Candidly if it could be proven beyond reasonable doubts that the outcome of any drill hole pierced into the ground would be waste, it would not be drilled. There are a few instances where surface geology will expose waste contact (outcrops, soil color). At this stage, we are bound by industry practice to do the systematic thing since you will also have no solid basis for practicing vice.

You remarked that blanks are mainly to test for lab smearing downhole.

For advanced or development projects, when ore contacts and geology have been physically established, a blank in waste is not required when the "blank" is blank in all the analytes (<detection limit). Otherwise (like quartzite blank for BIF deposits), it should be inserted systematically. In the former instance, experienced geos should do this. It is advisable to control the rationale of any procedure in place and to challenge it if you deem reasonable. Some may just be copy-paste of similar projects of the company, which may helpful at inception of project but need to be fine tuned to the project with respect to the current understanding.

If you are an advanced exploration project and are well recognized geologic units, or is an operating mine, the general idea with the Reference is:

Materials Reference materials built by deposit (ideal)
Grades are in cut-off grade - average grade - high grade. The high and medium grade can be obtained from a statistical analysis of data and use the percentile 50 and 75.
Represent the geology (e.g., oxide - sulfide - Hypogene)
To generate reference materials, you can follow this recommendation and the budget available.

To build your batch, it is normal to use controls every 20 samples plus 5 controls:

Control 1: Reference Material to control accuracy of the analysis
Control 2: Blank, to control contamination
Duplicate 1: Duplicate field to control sampling
Duplicate 2: Duplicate preparation, to control mechanical preparation
Duplicate 3: Duplicate pulp, to control chemical laboratory

The comments from a few days ago are pretty accurate. I would add the observations that the QAQC program purpose is two-fold:

To monitor the batch results of assays coming back from the assay lab to deal with any problematic results rather immediately (hence the probable need for some level of semi-systematic standards/blanks submissions);
To monitor long term trends and accuracy from the assay lab(s).

It may seem obvious to some, but it's not uncommon for the first of those to be neglected.

For the second point, choice of standards is somewhat dependent on the size of the drilling program that you are running. The need is for enough results for a particular standard to give you some statistically valid results in order to determine the accuracy/trend issues. i.e. if you are submitting a large quantity of samples over a period of time, use a full spread of standards. If you are submitting a relatively small quantity of samples over a short period of time limit your choice of standards to those that you feel are critical and carefully consider the submission rates, maybe putting in a few more than you would otherwise. No point having a 12 standards if you only have 2-5 results for each of them better to have 2 good standards with 12-30 results (although I'd still like to have more results! This is just an example!).

Your ability to preferentially insert SRMs in and around mineralised material is entirely dependent on your ability to visually identify the mineralisation. And I can't answer that for you.

However, I would note that from a resource estimation perspective, it is equally necessary to know that the "waste" samples produce "waste" grade assays--and e.g. that the assay lab hasn't had a bad day calibrating their AAS machine inadvertently producing spurious results or that the prep lab hasn't had some contamination issues. It happens.

I often hear the same cost-savings justification regarding selective sampling of diamond core holes which is another discussion altogether.

The best QAQC system that I have seen tends to send samples off to a preparation lab (possibly including coarse blanks); pulps are returned to client where they are randomized and renumbered and standards/blanks randomly inserted (in similar sachets to the pulps) and then sent off to a different assay lab. Not a simple, straightforward process. And not the cheapest. But at least the standards/blanks should then be totally blind to the assay lab which is sort of the ideal objective in the submission process. The definite weakness in the process is the re-handling and the client’s ability to manage multiple sets of sample numbers for the same sample and to assess the results on the lab certificates quickly. If your geo's, database person(s) and system are not up to scratch, don't do it. In other words, in some cases, simple is better.

While on the subject of blanks and standards it is probably worth mentioning 'spiking' samples with a known concentration of an element that is not of economic interest to the deposit you are testing. The idea here is that if you are doing multi-element analysis (say XRF or ICP) it is relatively inexpensive to analyze an extra analyte (for example let's say tin in a nickel laterite suite). You then add a say some tin oxide powder to the standards and the blanks and monitor tin concentration. This spiking has the advantage of being able to monitor whether the tin is being smeared from the blank to following samples. The spike method can also help with the common problem of standards being incorrectly identified during insertion. The idea here is to have different tin concentrations or perhaps different spikes elements in each standard.

I agree with you, it seems most people religiously collect duplicates every 20 meters, but if your ore zone is 10m wide and you need to drill a hole of 150m to get there why take duplicates of some 120m of <0.01 material every 20 meters? Rather take 4 duplicates (if not more) from the actual ore zone (if you can pick it of course) plus a good waste zone and maybe put one duplicate in the start of the hole if you have to. You are testing the precision of the stuff with the gold (or other goodies) in it! Precision plots throw out everything below x15 the LOD anyway waste of money.

Similar with the standards and blanks to a degree, and I agree, it is nice to test sampling issues in general, which the standards and blanks will pick up straight away (i.e. sample mock-ups). So therefore nice to put in standards throughout the hole but don’t see the need for systematic intervals really, let’s not make it too easy for the lab (even though of course they can pick it anyway). Blanks in immediate waste zones i can see but not 60m outside an ore zone, really what is the point. Put 3 blanks in a 10m ore zone and just after it, which is where you want to make sure no smearing occurs. Indeed not only the first of your points is often neglected, both are and people seem to be putting them in because the SOP tells them to and that where it stops.

Pre-allocating standards blanks and duplicates? If you are on a true exploration hole, make sure you have a geo that can pick alteration. If you get alteration/sulphides, crank up your duplicates and blanks, if it’s fresh as a daisy, then turn the tap to low frequency. If you don’t get anything in your hole, make sure you have at least a few in any way to indeed make sure the lab's machine doesn’t have an off day.

Its great to be able to get pulps back from a prep lab and then change the numbering sequence and resubmit but lab problems like those will show with your 5% blind pulp repeats (including the original standards and following pretty much the same path) should tell you whether your lab is doing a good job and cost you 20 times less effort to organise. Otherwise: good luck with the sample number muckups.

I am so much happy for the responses. This shows that sample is the most important product as far as exploration is concerned; hence its integrity cannot be compliance. I hope the discussion will not end here. Indeed the discussions have brought up a lot of issues and have painstakingly gone through all your comments.