The xrf Mystery

[bryantch]

As with all tools at the disposal of a gold buyer, an XRF is just one more. Although some may argue, an xrf is very accurate in the hands of an experienced operator...key word,"experienced". Results within 1/2% of a fire assay are expected.

With regards to detection of low gold concentrations, I know of no reason to expect a lesser degree of accuracy. Xrf machines are routinely used to detect low levels of heavy metals in contaminated soils and are also commonly used in mining operations to test gold and silver ores.

 

[nickvc]

.and how many in the UK are moving 30 kilos of fine gold a day ?? ie £1£million a day.and how many are actually refining it.....None of the guys I know No one sells it for 100% to the real refiners ie non-UK .and what about errors,silver or brass in the gold, gold price drops? and how many staff would you really have to have to handle 30 kilos a day of pure ie 80 kilos of 9ct a day?

And I also dont understand how you can get an accurate test from a drilling of a bar ie as opposed to cupellation of the whole bar?[/quote]

Their are several that I personally know of that move that sort of volume or close to it and in volume you can get very very good terms and one of the largest refiners in the world is UK based Johnson Matthey and several have depots to accumulate metals including Englehard and Metalor. The errors due to xrf testing are few and most will pay by carat so if its over 21 k they pay 21 or 20 k and most buyers of volume cover the metals they buy as they buy so no risk there. Very few staff are needed as the bigger buyers have large collectors and one has deals worldwide and not just jewellery scrap is traded remember but every type of precious metal scrap available and tradable.

With taking samples it really comes down to melting and mixing properly and most professional melters take dip samples but if you just do odd bars a drilling sample taken top and bottom at opposite ends of the bar will normally give you a fair result and if you know what you melted your rarely shocked, in other words if you really grade your scrap as you buy you should know exactly what assay to expect.

 

[benign01]

OK, yes you are right here, but how many 98% payers in Birmingham are doing JM volumes?

 

[nickvc]

As I said I know of two personally who probably move 30 kilos a day and maybe more and in those quantities and with the deals they will have with the big refiners 1 -1.5% is not not out of the question as a margin and plus the bonus metals, so yes they make good money believe me they don't need to cheat and it's up to you to give them no reason to by getting assays simple quick and cheap here.

 

[benign01]

Understood, the error must be on my part

 

[nickvc]

I'm sorry if my reply seemed to be blaming you it wasn't my intention.
To develop trust with your buyer it really helps if you know what exactly your trading which is why so many of the US based members refine to as pure as they can get their product but here in the UK assays are cheap and quick and allows you to know that you are been treated fairly by your buyer, not all lower results are due to the buyer but sometimes are due to errors on our own part and an assay allows the truth to be evident as to where the errors are.

 

[benign01]

I didn't think you were blaming me. I melt and sell around £1 million per quarter of gold. I know what I melt, I know its cost and I know what I sell it for. I have dealt with all the big "refiners in the JQ" and I know who does what. There are honest ones and there are less than honest ones. However recently I have had some very odd results on low-gold content bars and hence I return to the beginning of this topic
The maths I have used to calculate the gold yield must be incorrect but the fact is I have had inconsistent results on the last few low-gold bars I have produced and it doesnt add up. I know what I put in, I know what the first few results yielded so why are the results now dropping. Indeed in one case the refiner zapped the bar, got 7.8%, redid it for no apparent reason and got 5.6% and then had an assay done and it came back 5%. The bar, based on previous results should have been 8% gold.

But as you say an assay will sort it out prior to selling, so problem solved

 

[nickvc]

From your last post I can now understand your worries. To be fair an xrf reading from a gun that's been set for carat scrap buying means your bars at around 8% will not in my opinion get a decent result. If your source of material is the same each time and is identical I too would be extremely dubious of lowering results and if as I said the material is identical I would have individual pieces assayed as well to check it is exactly the same material been melted unless your paying on results.
If you leave a gap to be taken then maybe you will be hence as I said get assays and if you want to really be certain your been had get dip samples but don't say anything until you get the results from the buyer....could be interesting.
Ps if you need some glass rods to do this let me know and I'll see if I can scounge a few for free for you.

 

[jimdoc]

I can't see how you could melt gold filled items into a bar in a way to get a consistent reading.
You are better off just refining it first.

Jim

 

[nickvc]

It might not be gold filled he's melting but if he's using a purchased furnace it's almost certain to be an induction furnace and even if it's not with good temperature,fluxing and stirring he should have little problem achieving a good mix on every melt.
Some don't want to refine as it takes time and a lot more labour, here in the UK as I have said many times assays are cheap and quick so it can be worth just melting assaying and trading.

 

[benign01]

Thanks for your offer, I have a few already though. You are right about getting each bar assayed beforehand though. Its just that the refiner I use doesn't rate the Assay office's results,which i find slightly odd. They prefer to use the company mentioned earlier,

 

[benign01]

I can't see how you could melt gold filled items into a bar in a way to get a consistent reading.
You are better off just refining it first.

Jim

Gold filled stuff doesnt give 8%. The stuff I have is now more like 1-2% No its a mixture of 9 carat scrap and gold-filed

 

[qst42know]

Is it out of the question to do at least a nitric leach to up grade this a bit?

As it was explained to me 9k responds reasonably well to nitric as does gold filled.

 

[Harold_V]

it's almost certain to be an induction furnace

If that be the case, stirring is a non-issue. The eddy currents developed in induction furnaces provide stirring that borders on being violent.

Harold

 

[nickvc]

I didn't phrase that as well as I could have Harold, thank you for putting into plain language what I was trying to say, that without an induction furnace good stirring would be highly recommended but with an induction furnace this would be done by the furnace itself.

If this material was mine I'd melt the gold filled as is and have an assay done on it and if the terms are low on this material simply grain it and add it in known quantities to karat melts to keep the charges down.

 

[benign01]

Stirring! Yes I stir and I stir and I stir, but only maybe twice for 30 seconds or so over the whole melt period! Can i ask Is that sufficient? Clockwise15 seconds then counter -clockwise. for the same period, two or three times prior to pouring. I don't have a fancy induction machine and nothing stirs if I don't stir it Done many melts!

 

[nickvc]

With melting it's as much an art as a science, I find that the melts take on a very liquid look, almost like staring into a pond when the metals are all molten...that's the time to stir and feel what's occurring in the melt if you feel solids you need more time and more heat if on the other hand you feel the whole swirl then it's done but a good melt requires attention to detail and as I said any lumps in your melt means it isn't right and will require more stirring or heat...eventually you just know when it's time to pour the bar and if it's right it's like pouring water from a jug,smooth and silky and no lumps and the surface is clear and bright with heat rings been generated across the surface...I have always enjoyed watching molten metal it's almost alive and highly dangerous if not treated with thought and care.

 

[DarkspARCS]

here's another thing to consider:

The annalist was astounded by the amount of sulfer in the sample lol, which prompted him to reveal a lil device flaw the Xsort has, which is that the elements Iron and sulfer have the unfortunate affect of masking the actual ppm count for gold and pgm.

 

[NobleMetalWorks]

First, if you do not correctly melt (stirring, mixing, observing, correcting), and pour your bars incorrectly, the heavier gold will try to move towards the bottom. A melt has to be one homogenous mass, meaning no dark spots and having a nice uniform color before you pour into a mold. But if you take a mass that has not been melted correctly, and pour it into a mold, in the center lets say, the gold that isn't melted correctly with the rest of the metal will try to travel to the bottom and dependent upon how fast the bar becomes solid, it might or might not make it. So all along the outside of the bar might read lower than the inside. The top might read differently than the bottom. When the top and bottom are drilled and the shavings hit with an XRF gun, you might be getting a of 5 shaving amounts from low content areas, and one shaving reading from the sweet spot in the middle of the bar. And how can you accuse them of ripping you off if they are just testing your bar?

I don't understand the concern over being able to flip a switch on an XRF to read differently. It wouldn't even take that much. When I owned a grocery store I had to constantly watch for people ripping me off using all kinds of different methods. Unfortunately, because I would like to believe everyone is good, my experience has been that if you give someone an opportunity to steal from you, eventually they will. A common way to rip off the store you work for as a cashier is to do what is called sweethearting, selling things for less than their worth, to friends and family. I have seen a lot of sweethearting and some pretty clever ways to do it. I had a 19 year old girl who tapped a barcode under her bracelet and would simply cover the barcode with her hand, and then move her bracelet to scan the barcode on her wrist. Forget about the calibration, if someone wants to rip you off it just takes a small piece of metal, nothing more. By scanning one time, and then a second time, they can get the reading of the actual content, and then force a second reading with another small sample you don't see, and get the reading they want not too far off from your own. This is just a suggestion of how it could be done.

Also, don't forget XRF scanners operate on software, anything that has software can be altered, no switch required.

Third party fire assay is the way to go.

Scott

 

[Dan Dement]

I refer to my original question and add another
Does an xrf analyser test differently if the gold content is relatively low eg 2%... What I mean is does the accuracy vary with lower amounts?

Any takers?

The answer is yes. Also, you can set up the XRF to read correctly on .999 and be off on 14kt. It's a matter of having to adjust the standards. Sometimes on 14kt Gold, you can have 3 to 9 or more different alloys. I have my fourth XRF machine on order.

Dan Dement