The Unmarked Scrap Parcel: Alloying, Sampling and the 750 Line
A regular customer delivers 213 g of mixed scrap in a sealed bag: about 120 g of 18 kt yellow chain stamped 750, fourteen 14 kt rings, six white-gold items of unknown white alloy, roughly 40 g of solder snippets trimmed from bench work, and three unmounted pieces carrying no stamp at all. He wants a single 18 kt yellow ring, 17.3 mm internal diameter with a 2.4 mm wide shank, plus the leftover metal returned as 18 kt sheet, all within eight days. The bench has an XRF gun, a touchstone and acid set, a rolling mill, and a borrowed melting furnace whose graphite crucible last held 14 kt yellow. Your colleague proposes melting the whole bag with 18 kt master alloy, "because the XRF says it averages about 705", and casting the ring from the result. There is no time or budget for a full fire assay before the deadline, and the customer has already told friends he will collect on Friday week.
1. Explain why melting the entire parcel into a single 18 kt alloy is unsafe, referring to the specific contents of the bag and to how the XRF figure was produced.
Read the worked answer
The 705 figure is a surface-biased reading of mixed items, not a statement about the gold actually present.
- Solder snippets (~40 g, almost 19% of the bag). Jewellery solders are chosen to flow below the parent alloy, so they are deliberately gold-poor and carry zinc, copper, sometimes cadmium or tin. At 750 they are not 18 kt at all and would drag the whole lot down while adding embrittling elements.
- Six unknown white-gold items. If they are nickel-white, nickel forms hard, brittle phases and will grey a yellow alloy and promote cracking on rolling; if palladium-white, the palladium is wasted and unbalances the colour.
- Three unstamped pieces. XRF reads only the surface, so a heavy plated brass or low-karat item reads high and contributes almost nothing.
- The crucible. A graphite crucible that last held 14 kt carries residual metal in its pores; it must be glazed or replaced, or the melt is contaminated before it starts.
- Colour and workability. Reaching 750 by adding fine gold alone leaves the base-metal balance accidental, so the colour, casting range and rolling behaviour are uncontrolled.
Correct answer in practice: sort and verify, sell or send the unknowns and solders for refining, and build the ring from verified material plus fresh fine gold and fresh 18 kt master alloy.
2. Using the colleague's figure of 705, calculate how much fine gold would be needed to bring 213 g to exactly 750. Then give two reasons why this number cannot be used as a production instruction.
Read the worked answer
Calculation.
Fine gold present = 213 g × 0.705 = 150.17 g. Let x = fine gold added.
(150.17 + x) / (213 + x) = 0.750 150.17 + x = 159.75 + 0.75x 0.25x = 9.58 x ≈ 38.3 g of fine gold
Check: 188.5 g fine in 251.3 g total = 0.750. ✔
Why it is not a production instruction.
- The denominator is unknown. 38.3 g of fine gold produces 750 only if the whole 213 g really is at 705. Remove the 40 g of solder and the plated unknowns and both the numerator and denominator change; the true figure could be several grams out in either direction, which is the difference between passing and failing assay.
- Karat is not only gold content. Adding fine gold gives an alloy of 750 gold and whatever base metals happened to be in the scrap. Correct 18 kt yellow needs a controlled copper/silver (and often a little zinc) ratio for colour and for a workable casting and rolling range. Fine gold alone cannot fix a zinc-heavy or nickel-bearing base.
A defensible workflow within eight days: sort, file a fresh window on each item and XRF that, magnet-test, quarantine solder and unknown white gold, melt only verified 750 plus fine gold plus fresh master alloy in a clean crucible, cast, finish, and send for assay before delivery.
3. Set out the workflow you would actually run in eight days, and state what you would tell the customer about the value of his scrap.
Read the worked answer
Day 1 – sorting and verification (no melting). Weigh and photograph everything. File or scrape a fresh window on every item so XRF reads metal, not plating or tarnish. Magnet test to flag nickel-white and steel findings. Touchstone and acid as a cross-check on the pieces the XRF is unsure about. Sort into: confirmed 750; confirmed lower karat; white gold; solder and unknown/unstamped → refining only.
Day 2 – melt and alloy. Clean or replace the crucible and glaze it. Melt the confirmed 750 with fine gold and fresh 18 kt master alloy to a calculated 750, flux, stir with a carbon rod, pour a button, roll to sheet. Record every weight.
Day 3–6 – make the ring. Bend, solder, size to 17.3 mm internal diameter, check with a ring gauge rather than calipers, file, emery, polish.
Day 7 – assay and hallmark. Submit to the assay office. Do not stamp 750 yourself on the strength of a calculation.
Day 8 – deliver with a written note giving the metal weight received, the weight returned, the refining charge, and the clear statement that scrap value is not karat value: solder, plated and unknown items are paid as refining feedstock, not as 18 kt, and the XRF average is not a purchase price. Get the customer's signature on that before starting work.