Metal Recovery After Melting: Reconcile Every Gram

Quick answer

Calculate metal recovery after melting from a closed mass balance, not ingot weight alone. Reconcile accepted charge into cooled product, reusable returns, dross or skim, crucible and tool residue, spills, samples, and an unexplained difference. Use a suitable calibrated scale and never label the unexplained difference “burn-off” without evidence.

Use a closed mass-balance equation

For one identified lot:

Accepted charge = cooled product + reusable returns + process residue + equipment residue + spills + samples + unexplained difference

Then calculate:

Accounted recovery % = (cooled product + reusable returns + recoverable documented residue) ÷ accepted charge × 100

Whether a residue is “recoverable” depends on its metal content, contamination, handling, economics, and the receiving refiner’s rules. Do not count it at full metal mass merely because it feels heavy.

Use seven buckets

Bucket	Examples	Control
Accepted charge	Verified ingot, shot, sprues, traceable scrap	Weigh after rejecting stones, steel, coatings, dirt, and packaging
Cooled product	Ingot or accepted casting	Weigh at a consistent clean, cool state
Reusable returns	Gating, sprues, runners, clean buttons	Label alloy and lot; do not silently mix
Process residue	Dross, skim, flux glass, floor sweep	Collect separately; assay when value justifies it
Equipment residue	Metal retained in crucible, mold, tongs, skimmer	Inspect and assign to the lot
Spills and samples	Test coupon, assay sample, contained spill	Record destination and custody
Unexplained difference	Difference after all measured buckets	Investigate measurement and process; do not rename it

This equation separates material accounting from casting yield. A sprue can be fully accounted metal even though it is not an accepted casting. Conversely, a good-looking ingot can have poor accounting if residue and measurement are ignored.

Apparent loss may be measurement error

NIST guidance emphasizes calibration, traceability, repeatability, environmental effects, minimum sample quantity, and uncertainty in balance use. For a small precious-metal batch, a kitchen scale with coarse resolution may be unable to distinguish the process change you want to measure.

Record:

scale ID, capacity, resolution, and calibration/check date;
stable location and warm-up requirement;
tare container and method;
repeated readings;
minimum sample quantity appropriate to the scale;
whether hot, dirty, or moving material affected the reading; and
rounding rule.

Use the same controlled method before and after the heat. Do not weigh a warm ingot on one scale and a mixed cold charge on another, then report the difference to two decimal places.

Do not assume every residue is waste—or metal

In aluminum melting, oxidation creates dross and can entrap metallic aluminum. The Aluminum Association says some metal is typically lost through operations or oxidation, while DOE sources explain that dross quantity depends on furnace, charge quality, contaminants, temperature, and operating practice.

Those industrial ranges are not promises for a benchtop furnace. They show why a recovery percentage without feedstock and method is weak evidence.

For precious metals, visually dark residue may contain metal, flux, crucible material, oxides, dirt, or mixed contaminants. Do not use color or magnetism as a purity assay. Store valuable residues by lot and ask a qualified refiner what form and documentation it accepts.

Audit recovery claims on product pages

The current ToAuto TGF3000 page states that a test using five 20 g gold melts produced zero slag and more than 98% recovery. The public wording does not provide the raw before/after masses, scale uncertainty, alloy, starting material, residue handling, crucible tare, sampling, or independent verification.

Treat that figure as a dated manufacturer claim, not a universal performance guarantee. A publishable recovery test should include:

exact furnace and crucible revision;
identified metal and alloy;
charge mass and form;
scale method and uncertainty;
cold/warm start and electrical supply;
temperature and time record;
every output bucket;
photographs and raw readings;
repeated trials; and
calculation rules fixed before the test.

A result can then be compared with later runs from the same controlled process. It still should not be transferred to a different metal, charge, operator, or furnace.

Diagnose a falling recovery trend

If accounted recovery declines, investigate categories rather than guessing:

Pattern	Evidence to inspect
More dross or skim	Charge surface area, contamination, alloy, temperature, hold time, disturbance
More crucible residue	Crucible condition, flux history, incomplete pour, metal compatibility
Larger unexplained difference	Scale checks, tare errors, mixed lots, uncollected spills, missing samples
Stable recovery but lower accepted casting yield	Mold fill, gating, defects, trimming, rejection rules
Recovery changes after feedstock switch	Alloy identity, coatings, dirt, moisture, hollow or light-gauge material

Do not respond by adding an unsourced flux recipe or raising temperature. Both can change chemistry, fumes, crucible life, and residue. Use a qualified process review.

Frequently asked questions

What is a good metal recovery percentage after melting?

There is no universal target across metals, charges, furnaces, and residue rules. Establish a controlled baseline for one identified process, report uncertainty, and investigate changes. Do not copy an industrial range or seller claim.

Is the missing mass always oxidation?

No. It may be in dross, crucible residue, tools, spills, samples, rejected feed, tare error, scale uncertainty, or mixed lot handling. Oxidation is one mechanism, not a bookkeeping category.

Should I throw away dross from precious-metal melting?

Not automatically. Isolate and label it, avoid cross-contamination, and ask a qualified refiner whether analysis or recovery is economical. Do not assume all residue is valuable metal or all of it is waste.

Bottom line

Metal recovery after melting becomes useful when every gram has a defined bucket and the measurement system is fit for the batch. Close the balance first; diagnose chemistry second. ToAuto can turn its recovery claim into strong EEAT evidence by publishing raw, revision-matched, repeated test records with uncertainty and residue disposition.