Gold-Filled Scrap Refining: From Marking to Measured Recover

Ami Gur

A.G. METALS · RECYCLING CASE STUDY

A 7.25 kg batch of gold-filled scrap offers a practical comparison between nominal gold content, laboratory assay and the final recovered metal.

By Ami Gur | A.G. Metals Ltd.

We received 7,250 grams of gold-filled scrap marked 1/20 14K for refining. The project gave us an opportunity to compare the gold content indicated by the marking with the actual assay and the result of refining.

Photo 1. The incoming batch of gold-filled jewelry scrap, weighing 7,250 grams.

What gold content did we expect?

For this batch, the nominal calculation uses a 14K gold-alloy layer representing 1/20, or 5%, of the metal weight. Using 585 fineness for the 14K layer:

5% × 58.5% = 2.925% fine gold
7,250 g × 2.925% = 212.06 g of gold

This is a theoretical estimate based on the marking, rather than a measurement of the received scrap.

From scrap to laboratory assay

We melted the scrap and cast it into ingots, then sent a sample for fire assay. The reported gold content was 2.8%, corresponding to 203.00 grams of gold when applied to the incoming weight of 7,250 grams.

That is approximately 9.06 grams below the nominal estimate. A surface measurement of the gold-filled material returned 55.3% gold, compared with the 58.5% used in the nominal calculation.

As a numerical comparison, a 5% layer containing 55.3% gold would give an overall gold content of 2.765%, close to the 2.8% assay result. However, the surface reading alone does not establish the average composition or weight fraction of the layer throughout the batch. It is a possible clue, not proof of the cause of the difference.

Another possible contributor is the presence of small non-gold components, such as steel springs inside necklace clasps. These add to the incoming weight without contributing gold and may therefore slightly reduce the average gold content of the batch compared with a calculation that treats the entire weight as gold-filled material. Their contribution was not measured in this case.

Refining and the final result

The ingots were granulated, followed by dissolution and gold refining. The final recovered metal weighed 200.39 grams.

Photo 2. The recovered gold button on the balance: 200.39 grams. The displayed weight does not establish fineness.

Basis of comparison	Gold content	Calculated gold / recovered weight
Nominal marking, using 585 fineness	2.925%	212.06 g
Reported fire assay	2.8%	203.00 g
Final recovered button	Final fineness not provided	200.39 g

The final button weight was 2.61 grams below the assay-based estimate, equivalent to approximately 98.7% of that estimate. This is a weight comparison, not a fully verified fine-gold recovery rate: the final fineness and a complete process mass balance were not available for this case study.

Some gold may have remained in the process solutions or residues, but this was not confirmed. Further testing would be needed before assigning the difference to a specific cause.

A closer look at the gold-filled layer

Photo 3. Gold-filled strips after partial acid attack on the base metal, showing the remaining thin outer gold-alloy layer.

The strips illustrate the layered construction of gold-filled material. As the copper-based core is attacked, the thin outer gold-alloy layer becomes visible. The remaining layer is a gold alloy, rather than pure gold.

The practical lesson

In this batch, the nominal calculation indicated approximately 212 grams of gold, the assay indicated 203 grams, and the final recovered button weighed 200.39 grams.

These figures answer different questions. The marking provides an initial estimate; the assay measures the sampled material; and the final weighing records the recovered product. Keeping those stages separate makes a refining result easier to understand and evaluate.