Comment

Fire assay ยท Process guide

Cupellation Process: How Cupelling Gold Works in 6 Steps

Cupellation separates gold and silver from lead: a porous cupel absorbs about 98.5% of the molten lead oxide and leaves a precious-metal bead (USGS, 1977). It is the last furnace stage of a gold fire assay.

Authored by Farmonaut Research & Content Team
Reviewed and verified by Farmonaut Mining & Mineral Exploration Team
Updated 5 November 2026 ยท 11 min read
Sentinel-2 Satellite View Of Laurion Silver-Lead District, Attica, Greece
Sentinel-2 true-colour composite of Laurion silver-lead district, Attica, Greece (37.72ยฐN, 24.05ยฐE). Contains modified Copernicus Sentinel data, 2022 to 2026, processed in Google Earth Engine.

Key takeaways

  • A lead button is oxidized in a bone-ash cupel, which absorbs the lead oxide and leaves a bead (USGS, 1977).
  • A published USGS procedure runs about 820-920ยฐC after a brief 1,000ยฐC start; gold with a silver collector starts at 820ยฐC and finishes at 880ยฐC.
  • Heat drives loss: one 1905 USGS furnace series lost 0.5-3.89% of gold across five positions. That shows direction and is not a modern lab figure.
  • One lab’s 2026 list prices the five gold options compared here from C$23.00 to C$88.00 per sample (Actlabs).
  • Cupellation is the largest source of lead fume in fire assay, so hoods and monitoring come first (British Columbia, 2016).
In this article
  1. What is cupellation?
  2. The cupellation process in 6 steps
  3. What temperature is used for cupellation?
  4. How much gold is lost in cupellation?
  5. Convert a bead weight to a gold grade
  6. Fire assay finishes and lab prices compared
  7. Is cupellation sustainable? 5 ways to reduce its impact
  8. Cupellation of silver and the history of the method
  9. What cupellation means for your project decisions
  10. Frequently asked questions

Cupellation is a high-temperature refining step that separates gold and silver from lead. A lead button holding the precious metals is melted in a porous cup called a cupel, the lead turns to lead oxide, and the cup soaks that oxide up. A small bead of precious metal stays on top. The USGS fire assay manual (1977) reports that about 98.5% of the lead oxide goes into the cupel and about 1.5% is lost as vapour.

Today the cupellation process matters most in the assay laboratory, because it is the last furnace stage of a gold fire assay. A consultant or licence owner reading a lab certificate is reading a number that passed through this step. This guide covers the procedure, the temperatures, where gold is lost, what an assay costs, and the lead hazard that decides whether a lab runs it cleanly.

What is cupellation?

The word comes from the cupel, a shallow cup usually made of compressed bone ash. Anglo American describes cupels as typically made of bone ash or calcium phosphate. The USGS manual gives the most useful mental picture: the cupel acts like a membrane that lets molten lead oxide through and holds back metallic lead and the noble metals. (Noble metals are those that resist oxidation, such as gold, silver and the platinum group.)

Three materials do the work:

  • Lead acts as the collector. Gold and silver dissolve readily in molten lead, so lead gathers them out of the melted rock.
  • Litharge is the common name for lead oxide (PbO). It forms when air hits the molten lead, and it is what the cupel absorbs.
  • The cupel is the sponge. By the end, the saturated part of a used bone-ash cupel is usually about half bone ash and half litharge, according to the USGS manual.

Cupellation does not extract gold from raw rock by itself. In fire assay, a melting stage called fusion comes first and moves the precious metals from the rock into the lead.

How gold moves from rock to refined metal, step by step

โ–ถ How gold moves from rock to refined metal, step by step ยท 5:27

The cupellation process in 6 steps

This sequence follows the USGS manual and the Nevada Bureau of Mines and Geology fire assay explainer. Commercial labs set their own procedures, so treat the settings as one published example.

  1. Fuse the sample. Crushed rock is melted with litharge and other fluxes (materials that help the rock melt). The lead sinks and the waste rock floats off as slag, a glassy by-product.
  2. Shape the lead button. After cooling, the slag is hammered away and the lead is shaped into a button.
  3. Preheat the cupels. The cupel is matched to the button: the manual uses 3.8 cm cupels for buttons up to 32 g and 4.4 cm cupels for buttons of 32 to 48 g.
  4. Open the button. The button goes on the hot cupel at about 1,000ยฐC until the lead is molten and red throughout. An assayer calls this being open.
  5. Drive off the lead. The heat is then cut quickly and raised in stages as the button shrinks, while air oxidizes the lead and the cupel absorbs the oxide.
  6. Weigh or analyze the bead. The bead is cooled and either weighed or dissolved for instrument analysis. If it holds both metals, nitric acid dissolves the silver and leaves the gold, a step called parting.

What temperature is used for cupellation?

Most of the USGS procedure runs between about 820ยฐC and 920ยฐC, after a brief start at 1,000ยฐC. For platinum-group work with a gold collector, the manual drops to 840ยฐC once the buttons open, then climbs through 860ยฐC and 880ยฐC to finish at 900ยฐC. For gold determined with a silver collector, the same routine starts at 820ยฐC and finishes at 880ยฐC.

Chart

Furnace settings in one USGS cupellation procedure

ยฐC, in order, platinum-group work with a gold collector

0ยฐC300ยฐC600ยฐC900ยฐC1,200ยฐC1,000ยฐCOpen840ยฐCFirst hold860ยฐCSecond hold880ยฐCThird hold900ยฐCFinishSource: USGS Bulletin 1445, A Manual on Fire Assaying, 1977.
0ยฐC300ยฐC600ยฐC900ยฐC1,200ยฐC1,000ยฐCOpen840ยฐCFirst hold860ยฐCSecond hold880ยฐCThird hold900ยฐCFinishSource: USGS Bulletin 1445, A Manual on Fire Assaying, 1977.
The heat falls fast after the button opens, then climbs in stages to the finish.

The reason for the fast drop is stated plainly in the manual: lower the heat as quickly as possible to avoid losing noble metals into the cupels. The finish can need 910ยฐC to 920ยฐC on a cold day with a down draft, which is a good reminder that these are working settings and depend on the furnace and room.

How much gold is lost in cupellation?

Some gold is always lost, and heat is the main driver. The clearest published measurements are old but still instructive. USGS Bulletin 253 (1905) placed rows of cupels from the front to the back of a furnace chamber, using pure mint gold and silver with 25 g of lead. In the series plotted below, gold loss rose from 0.5% to 3.89% and silver loss from 1.7% to 7.98% across five positions. The authors found that a shift of just one cupel width in the furnace had a large effect, and that the loss of pure gold was almost all absorption into the cupel. When gold and silver were cupelled together, as charted below, a related series lost roughly 30% of it as vapour and 70% to the cupel.

Chart

Gold and silver loss rose with furnace position

% of metal lost, cupels placed front to back, 25 g lead, proof gold and silver

0.0%2.0%4.0%6.0%8.0%12345Silver 8.0%Gold 3.9%Source: USGS Bulletin 253, Table IX, Series A, 1905.
0.0%2.0%4.0%6.0%8.0%12345GoldSilver8.0%3.9%Source: USGS Bulletin 253, Table IX, Series A, 1905.
Each step toward the back of the furnace raised the loss; this is one early-1900s test series.

Those percentages describe one furnace, more than a century ago, with about 30 mg of gold and 90 mg of silver in each test. They are not an expected loss for a modern lab, and they should not be applied to a certificate. They show the direction of the risk.

Four practical points follow for anyone judging assay work:

  • Silver protects gold. The USGS manual lists this as one reason for adding silver to the charge before fusion: it reduces gold losses during cupellation and makes tiny amounts of gold easy to move from the cupel.
  • Rich samples are the most sensitive. Bulletin 253 warns that small temperature differences have a marked effect on gold yield, especially with rich ore.
  • Used cupels can be re-assayed. The USGS manual describes fusing the saturated part of a cupel to retrieve absorbed metal when a low-grade sample had been used up and the collector had been left out.
  • Blanks and standards catch drift. Ask for the lab’s blanks, certified reference materials and duplicates. They reveal problems that a single grade cannot.

Convert a bead weight to a gold grade

Grade is the amount of metal per tonne of rock, usually in grams per tonne (g/t). A gravimetric finish divides the weighed bead by the sample weight. A 0.06 mg bead from a 30 g charge means 2 g/t, because 1 mg per gram equals 1,000 g/t. The calculator below does the sum, adds a loss correction you choose, and shows the value of the metal in one tonne at your chosen price. That value is contained metal, before dilution, mining, processing and refining costs.

Calculator

What gold grade does my bead weight mean?

Enter your own bead weight, sample weight, loss allowance and price.

mg

g

%

US$/troy oz
2.00 g/tgold grade after your loss allowance
  • Grade before the allowance (g/t): 2.00
  • Contained gold per tonne (troy oz): 0.06
  • Contained value per tonne (US$): $257

Example values only, not a benchmark or a forecast. The loss correction is your own assumption; use your lab’s figure. The gold price is an input: example, use today’s price. 31.1035 g = 1 troy oz. Contained metal is not recoverable metal.

Fire assay finishes and lab prices compared

Cupellation sits inside most gold fire assay packages. The options differ in how the bead is finished. In the Actlabs Canadian fee schedule (February 2026), one lab’s published list shows the trade-offs. Prices are per sample in Canadian dollars, with sample preparation listed separately, and other labs and volumes will differ.

Chart

One lab’s list price per gold sample

C$ per sample, Actlabs Canadian schedule, February 2026

C$0.00C$25.00C$50.00C$75.00C$100.00Metallic screen, 500 gC$88.00ICP-MS, 30 gC$32.00Gravimetric, 30 gC$29.50PhotonAssay, 500 gC$27.25Atomic absorption, 30 gC$23.00Source: Actlabs, Geochemistry Schedule of Services, Canadian, 2026.
C$0.00C$25.00C$50.00C$75.00C$100.00Metallic screen, 500 gC$88.00ICP-MS, 30 gC$32.00Gravimetric, 30 gC$29.50PhotonAssay, 500 gC$27.25Atomic absorption, 30 gC$23.00Source: Actlabs, Geochemistry Schedule of Services, Canadian,2026.
Metallic screen work costs the most here. PhotonAssay takes a 500 g sample at a lower price than the 30 g ICP-MS and gravimetric options, and does not use cupellation.
Method (Actlabs code) Sample weight Reporting range List price (C$) Best for
Fire assay with atomic absorption (1A2) 30 g 5 to 5,000 ppb 23.00 Routine exploration samples
Fire assay with ICP-MS (1A2-ICPMS) 30 g 0.5 to 30,000 ppb 32.00 Very low gold levels
Fire assay, gravimetric (1A3-30) 30 g 0.03 to 10,000 g/t 29.50 Higher-grade ore
Metallic screen fire assay (1A4) 500 g 0.03 to 10,000 g/t 88.00 Coarse gold that spreads unevenly
PhotonAssay (no cupellation) 500 g 0.03 to 350 ppm 27.25 Larger sample without lead flux

Ppb means parts per billion, and 1,000 ppb equals 1 g/t. Actlabs also notes that a 50 g charge may not give the best recovery and that it may cut the sample weight for a proper fusion. For guidance on picking a provider, see our guides to choosing a gold assayer and assay services for mining.

Is cupellation sustainable? 5 ways to reduce its impact

Cupellation does not make gold production sustainable on its own. It is a lead-based process that needs careful control, and its benefit lies in giving reliable numbers from a small sample. These five practices are where a lab can actually lower its risk and waste.

1. Keep the heat as low as the work allows

The USGS sources agree that excess heat sends more metal into the cupel. Staged temperatures protect the result, and that can reduce the need for repeat assays when results look doubtful.

2. Capture the fume at the furnace

The British Columbia fire assay lead guideline (2016) says cupellation produces the largest volume of contaminated fume in the process. It recommends an enclosed hood at the cupellation furnace door, with a shelf where cupels can cool while fumes clear. Moving hot cupels to a separate bench is discouraged because the fumes escape into the room on the way.

3. Treat lead waste as lead waste

The same guideline requires dust from the exhaust filters to be stored in sealed, labelled containers as hazardous lead waste and collected by qualified people wearing protective equipment. Rules vary, so check how your jurisdiction classes used cupels and slag as well.

4. Measure air and blood lead

In the United States, the OSHA lead standard sets an action level of 30 ยตg/mยณ and an exposure limit of 50 ยตg/mยณ, both averaged over 8 hours. A NIOSH evaluation of a Montana assay operation measured 170 ยตg/mยณ for an assayist. In an Alaska investigation of 2010-2011, 12 of 15 employees with known duties had elevated blood lead at least once, and 10 of the 12 reported working in the fire assay room. Four employees’ children had blood lead of 5 ยตg/dL or more, and the investigators concluded that workers could have brought lead home.

5. Use a method without lead where it fits

PhotonAssay uses high-energy X-rays instead of a furnace. Ledoux & Company, a lab offering assay services, describes it as non-destructive and says it avoids flux and lead, while still calling fire assay the standard when accuracy is critical. If you move a project to a new method, run paired samples against fire assay first.

The main environmental impacts of mining, explained in plain words

โ–ถ The main environmental impacts of mining, explained in plain words ยท 2:31

Cupellation of silver and the history of the method

Silver cupellation goes back thousands of years. The USGS manual traces fire assaying to finds at Troy II (about 2600 BC) and concludes that the method began in Asia Minor in the first half of the third millennium BC. It also records that Roman refiners could bring silver in lead down to 0.01%, or even 0.002% in some cases. Archaeologists still debate how early the method was applied to lead ores: Wood and colleagues in Internet Archaeology argue that lead ores were not used as a silver source in their own right until the mid-first millennium BC. They place that later use at sites such as Laurion in Greece, shown at the top of this page.

Georgius Agricola described the assay in 1556 in De Re Metallica, and the Nevada Bureau of Mines and Geology notes that the technique still uses the same principles. Silver also behaves differently in the furnace. Bulletin 253 lost two to three times as much silver as gold, in percentage terms, in the series charted above, and it notes that silver losses from too much heat were already well known.

What cupellation means for your project decisions

For a consultant, licence owner or investor, the main question is whether the assay data can be trusted. A short review covers most of it:

  • Method code. Check the finish and sample weight on the certificate, and whether its range suits the grade you expect.
  • Coarse gold. A 30 g charge can miss or overstate nuggety gold. Compare results with a metallic screen or a larger-sample method.
  • Quality control. Look for blanks, standards and duplicates in each batch.
  • Ore type. Sulfide-rich or unusual samples may need a changed charge. Our note on oxidized gold ore explains why the ore type affects processing, and our guide to types of gold deposits covers host rocks.
  • Lab safety. For your own or a contracted site lab, ask about furnace hoods and blood lead monitoring.

An assay measures only the samples you took, so where you sample matters too. Satellite imagery can help choose those locations by mapping surface alteration and structure, but it does not replace assays or drilling. Farmonaut offers satellite-based mineral detection for this early screening; you can start with the mining query form or contact the team.

What satellites can and cannot show about gold ground

โ–ถ What satellites can and cannot show about gold ground ยท 2:00

Frequently asked questions

What is cupellation in simple terms?

It is heating a lead button in a porous cup so that the lead turns to oxide and soaks into the cup. Gold and silver do not oxidize in the same way, so they remain as a small bead.

What is the purpose of cupellation in fire assay?

It removes the lead collector and leaves a bead of precious metal that can be weighed or analyzed by an instrument.

What is a cupel made of?

Usually compressed bone ash, as in the USGS manual. Anglo American also lists calcium phosphate.

What happens to the lead during cupellation?

It oxidizes to molten litharge. The USGS manual reports about 98.5% absorbed by the cupel and 1.5% volatilized.

What is the difference between cupellation and parting?

Cupellation removes lead and leaves a gold-silver bead. Parting comes afterward and uses nitric acid to dissolve the silver, leaving gold.

What is the difference between cupellation and smelting?

Smelting or fusion melts rock with fluxes to produce metal and slag. Cupellation is a later oxidizing stage that refines the lead carrying the precious metals.

Why is silver added before cupelling gold?

The USGS manual says silver reduces gold losses during cupellation, makes trace gold easier to transfer, and gives a silver-gold bead that can be parted.

Is cupellation dangerous?

It can be if fume is not controlled, because it releases lead. Labs need exhaust hoods, hygiene rules, monitoring and trained staff. It is not a home or field procedure, so send samples to an accredited lab instead.

Can cupellation recover gold from tailings or ash?

Cupellation only refines metal already collected in lead. Lab assays can measure gold in such materials, but recovery at scale needs its own tested process, and the cost must be weighed against the measured grade.

Is cupellation still used today?

Yes. Commercial labs such as Actlabs still list fire assay methods in their current fee schedules.

This page is for information only. It is not investment, legal or safety advice. Fire assay involves toxic lead and must be done only in equipped laboratories under local rules.

Authored by: Farmonaut Research & Content Team

Reviewed and verified by: Farmonaut Mining & Mineral Exploration Team

Reviewed November 2026 against USGS Bulletins 1445 and 253, the British Columbia fire assay lead guideline, OSHA 1910.1025, NIOSH HETA 89-213 and the Actlabs 2026 Canadian schedule.

Farmonaut mining

Choose where to sample before you pay for assays

Map alteration and structure across your project area from satellite data, then decide where field samples should go.

Farmonaut Mining Pro 3D Target View (Sample, Synthetic Data)

Mining Pro sample view, synthetic data

Sources

  1. USGS Bulletin 1445, A Manual on Fire Assaying (1977): pubs.usgs.gov
  2. USGS Bulletin 253, Methods of Assay of Gold Telluride Ores (1905): pubs.usgs.gov
  3. Nevada Bureau of Mines and Geology, Chemistry and Fire: nbmg.unr.edu
  4. Anglo American, Assay definition: angloamerican.com
  5. Actlabs, Geochemistry Schedule of Services, Canadian (February 2026): actlabs.com
  6. British Columbia Ministry of Energy and Mines, Reducing Inorganic Lead Exposure in Fire Assay Laboratories (2016): www2.gov.bc.ca
  7. OSHA, 29 CFR 1910.1025 Lead: osha.gov
  8. NIOSH, Health Hazard Evaluation HETA 89-213 (1992), via OSTI: osti.gov
  9. Elevated Blood Lead Levels Among Fire Assay Workers and Their Children in Alaska, 2010-2011 (Public Health Reports): pmc.ncbi.nlm.nih.gov
  10. Wood et al., Sending Laurion Back to the Future, Internet Archaeology 56: intarch.ac.uk
  11. Ledoux & Company, Photon Assay vs Fire Assay: ledouxandcompany.com



Discussion

Questions, corrections or field experience? Share them below. First-time comments are reviewed before they appear, and your email is never published.

No comments yet. Be the first to add a question or your experience.

Leave a comment

Farmonaut Farmonaut Trusted by 200,000+ users and 100+ businesses 200,000+ users trust us Get started