Old Assay 1929 vs Today

[Placer Paul]

If a copper/silver/gold ore was assayed back in 1929 would they have the means back then to give an accurate assay of the copper and silver if they were in sulfide form but the gold was in native/free mill form? Or if the gold, copper and silver were all in sulfide form?

Would today's methods be more accurate?

Reason I'm asking is that I had some of the ore assayed and when I dropped the samples off I talked with the assayer about the ore, and he noted that the minerals were sulfides and he suggested the fire assay. So I don't think they also assayed for native/free mill gold as well or did they just do their tests for sulfide minerals. If the gold is in the sulfides only would this test have given accurate results for all of them?

Is there a separate test I should have asked them to do to make sure accurate numbers of gold were calculated?

My concern is that the old government archive assay info from 1929 show gold values in the ore from this mine ranged from 2 g/t to 20 g/t. Yet on these assays I had done don't show gold values anywhere near those numbers. And one of the samples was specifically taken from the ore they had blocked out and were shipping out when they shut down.

 

[Placer Paul]

I also had a Qemscan mineral analysis done on the ore as well on one of the same samples the fire assay was done on and the results on that were higher values of the copper and the silver than that of the fire assay. The Qemscan mineral composition analysis results were 8.9% copper and 140 gr silver while that same sample ground mixed and split and fire assayed results were 5% copper and 96 gr silver.
Is that normal for the fire assay to not match the Qemscan analysis? Those numbers are pretty far apart.

 

[Deano]

You are not working with splits of the samples from 1929, you are working with new samples from the same formation.

It is impossible to compare these sample results as they come from totally different samples.

The only thing you can say about the results is that they both show similar numbers when the difference in sample collection is allowed for.

Usually, but not always, the older samples were collected with more care and are more representative of the orebody.

You would still like to know who took the old sample and why it was taken, was it for advertising the grades to a prospective buyer.

If you take the most cynical approach you are usually right.

You may also have had someone high grading the samples before assaying was done.

The usual approach to sampling by untrained or poorly trained people is to take the prettiest looking rocks and ignore all else.

Just because the assay was performed by a government assayer does not mean that the samples were truly representative of the orebody.

Properly performed fire assay will show all values for precious metals no matter where the metals are located in a sample.

The older assays were usually either custom fluxed depending on observation of the ore sample or standard fluxed if they came from a location where the ore was well known to assayers.

These days there are few assayers who have the knowledge to custom flux, the labs are trying to get a lot of samples through cheaply so most have just two flux types, sulphide and oxide. Tough if your samples do not fit neatly into either camp.

Your assay report sheets show good repeatability on the fire assays, this is what you would expect from a certified lab.

Note that the lab is using 30 gram samples for the fire assays but there are large variations in results from different samples.

This is why you do statistical analysis on the results to get results for the sample average with a high confidence interval.

Your results do not show the sample weight for the Qemscan test, hard to compare results with fire assay if you do not have the weights.

It is hard to know from your explanation of procedure whether the samples used for fire assay and Qemscan were from splits of a milled ore bulk sample or from splits of unmilled ore which was then milled for analysis.

The former procedure is how it should be done, the latter procedure is automatically subject to sample variation.

Deano

 

[Placer Paul]

Thank you for the reply Deano.