Research/Method

Fire assay and PhotonAssay

Fire assay, PhotonAssay, aqua regia, four-acid and BLEG compared by detection limit, sample size and 2026 lab price, plus lab accreditation and QA/QC.

Updated November 202613 min read

Fire assay is the standard lab method for gold in drill core, soil and rock chip samples, and it's a total decomposition technique (SGS, ALS). The lab fuses a 30 or 50 g charge of pulverized sample with a lead oxide flux above 1,000°C, the lead collects the gold into a button, and in cupellation a porous cupel absorbs the lead as lead oxide and leaves a precious-metal bead (ALS). The lab then weighs the gold or dissolves the bead and measures it by AAS or ICP (ALS). PhotonAssay measures gold in about 500 g of crushed sample with high-energy X-rays in about two minutes, against Chrysos's figure of about 3 to 4 hours for fire assay, and leaves the sample intact (ALS, Chrysos). Its larger sample reduces sampling error on coarse gold (Tremblay and others, 2019), and fire assay reaches lower detection limits, 0.005 ppm with an AAS finish against 0.03 ppm for PhotonAssay at ALS (ALS fire assay, ALS PhotonAssay).

Fire assay PhotonAssay
Sample 30 or 50 g charge of pulverized pulp About 500 g crushed to 2 to 3 mm, in a sealed jar
Gold range at ALS 0.005 to 10 ppm (AAS), 0.05 to 10,000 ppm (gravimetric) 0.03 to 350 ppm
Analysis time (Chrysos) About 3 to 4 hours About 2 minutes
Sample after analysis Charge consumed, pulp kept for repeat assays Whole sample kept for reanalysis
Limits on the method High sulphur or nickel can reduce the charge to as little as 5 g Uranium, thorium and barium raise the detection limit

Lead-collection fire assay

Most Australian labs crush a drill sample to 15 mm and then to 2 to 3 mm, split 1 to 3 kg and pulverize it until about 90 percent passes 75 µm, because a charge of 50 g or less needs that fine a grind to represent the sample (Tremblay and others, 2019). Gold is malleable, so it flattens and can smear onto the pulverizing equipment instead of breaking up (Intertek, Tremblay and others, 2019).

The lab mixes the charge with a flux of lead oxide and silica, plus borax, soda ash and other reagents, and fuses it above 1,000°C to break down the rock (ALS). The lead oxide reduces to lead, and the lead collects the gold, platinum and palladium (Intertek). The lab then heats the lead button in a porous cupel in an oxidizing atmosphere, the cupel absorbs the lead oxide, and a doré bead, or prill, of precious metal remains (ALS).

For an AAS finish the lab dissolves the bead in aqua regia and aspirates the solution into an acetylene flame, and for an ICP-OES finish the lab measures the same solution in a plasma (SGS). For ore-grade samples, labs use a gravimetric finish and weigh the gold (Standards Council of Canada).

Detection limits and sample weights

Codes and ranges are from the ALS fire assay, screen fire assay and PhotonAssay pages, checked November 2026.

ALS code Sample Method Gold range
Au-ICP21, Au-ICP22 30 g, 50 g ICP-AES 0.001 to 10 ppm
Au-AA23, Au-AA24 30 g, 50 g AAS 0.005 to 10 ppm
Au-AA25, Au-AA26 30 g, 50 g AAS 0.01 to 100 ppm
Au-GRA21, Au-GRA22 30 g, 50 g Gravimetric 0.05 to 10,000 ppm
Au-SCR21 1 kg screened at 106 µm, duplicate 30 g assays of the undersize Screen fire assay 0.05 to 100,000 ppm
Au-PA01 500 g crushed PhotonAssay 0.03 to 350 ppm

Au-AA23 reports gold up to 10 ppm, so a higher result needs an ore-grade method such as Au-AA25 or Au-GRA21 (ALS).

Screen fire assay and coarse gold

Coarse gold causes sampling error before the assay starts (ALS). A USGS study calculated that a sample needs an expected 20 gold particles for its result to fall within about 50 percent of the true grade 95 percent of the time, and that below five particles the chance of a sample with no gold at all rises sharply (USGS Professional Paper 625-C). Repeat assays of coarse-gold samples vary more than repeats of fine-gold samples (Tremblay and others, 2019), and geologists call this variability the nugget effect (Dominy and others, 2003). Dominy and co-authors describe lode-gold systems as often having a high to extreme nugget effect and gold particles coarser than 100 µm (Dominy and others, 2003).

JORC Table 1 asks whether sample sizes suit the grain size of the material, and it asks for more explanation where coarse gold creates sampling problems (JORC Code 2012).

A screen fire assay takes a larger sample, typically 1 kg, screens it at 106 µm (sometimes 75 µm), fire assays the whole oversize fraction and runs two charges of the undersize (ALS). The report gives both undersize assays and their mean, the oversize assay, the weight of each fraction and a weighted-average total. The oversize result shows how much of the gold is coarse, and ALS says screen fire assay data also give an early indication of the proportion of gravity-recoverable gold (ALS). Actlabs screens 500 g or 1 kg at 149 µm (Actlabs 2026 schedule), and Intertek assays the screen cloth with the coarse fraction (Intertek).

PhotonAssay

In a PhotonAssay unit from Chrysos Corporation, an 8.5 MeV linear accelerator sends X-rays through a sealed jar of about 500 g of sample. The X-rays excite gold-197 nuclei into a metastable state with a half-life of 7.73 seconds, each excited nucleus emits a 279 keV gamma ray as it decays, and germanium detectors count those gamma rays (Tremblay and others, 2019, Chrysos).

ALS lists Au-PA01 at 0.03 to 350 ppm on 500 g of crushed sample (ALS), Intertek gives 0.02 ppm (3-sigma) for jars of 400 to 650 g (Intertek), and Chrysos gives 0.015 to 0.030 ppm on typical gold ores and up to 35,000 ppm without dilution (Chrysos). Uranium, thorium and barium raise the detection limit (ALS), and Intertek calls PhotonAssay better suited to resource development and grade control samples (Intertek).

In a 2019 AusIMM conference paper by CSIRO, Chrysos and MinAnalytical authors, PhotonAssay results on about 1,200 samples from 10 Australian drill sites, crushed to 2 to 3 mm, generally correlated with fire assay at an adjusted R² of 0.97 or better, and two of the sites needed pulverizing for a strong correlation (Tremblay and others, 2019). The MinAnalytical lab concluded in its ISO/NATA validation that PhotonAssay results on 3 mm material are normally consistent and repeatable, with pulverizing needed for some ore types, and interchangeable with fire assay above 0.06 ppm, twice the detection limit (Tremblay and others, 2019).

On about 1,450 samples from an Australian mine with very coarse gold, the two methods agreed overall (R² of 0.953), but about 10 percent of samples differed by more than 25 percent. Screen fire assays of the 62 most different samples agreed much better with PhotonAssay than with the 25 g fire assay (Tremblay and others, 2019).

JORC and NI 43-101 reporting

JORC Table 1 asks the Competent Person to report the nature, quality and appropriateness of the assay method and whether it's partial or total (JORC Code 2012). NI 43-101 requires the type of analytical procedure, the sample size and the name, location and relationship to the issuer of each lab (NI 43-101, section 3.3), and a technical report adds the lab's certification and the qualified person's opinion of the procedures (Form 43-101F1, Item 11).

Assay methods

Labs finish a fire assay by AAS, ICP-OES, ICP-MS or weighing, and they dissolve multi-element samples in aqua regia (a partial digestion) or four acids (near total). The prices are Actlabs' 2026 US-dollar rates for the analysis alone (Actlabs 2026 schedule, checked November 2026).

Method Measures Sample Gold range Price (USD) Use it for
Fire assay, AAS finish Total gold 30 or 50 g 5 to 5,000 ppb $20.00 (30 g), $22.60 (50 g) Routine exploration gold
Fire assay, ICP-OES finish Total gold 30 g 2 to 30,000 ppb $21.10 Low-level gold
Fire assay, ICP-MS finish Total gold 30 g 0.5 to 30,000 ppb $27.85 The lowest gold limits
Fire assay, gravimetric finish Total gold 30 g 0.03 to 10,000 g/t $25.65 High grades
Metallic screen fire assay Total gold, coarse and fine fractions 500 g or 1 kg 0.03 to 10,000 g/t $76.55 (500 g), $85.70 (1 kg) Coarse gold
PhotonAssay Total gold, sample kept 500 g crushed 0.03 to 350 ppm $23.70 Coarse gold, resource and grade control
Aqua regia, ICP-MS Partial gold 30 g 0.5 to 2,000 ppb $21.75 Soils and grassroots exploration
Aqua regia, ICP-OES and ICP-MS Partial gold (semi-quantitative on 0.5 g) and a multi-element suite 0.5 g 0.5 to 10,000 ppb $25.25 Pathfinder elements in soil and rock
Four-acid, ICP-OES and ICP-MS Near-total base metals and trace elements Pulp Gold not reported $31.30 Base metals, pathfinders and lithogeochemistry
BLEG, ICP-MS Cyanide-extractable gold 1 kg 0.1 to 10,000 ppb $49.35 Stream sediments
Cyanide extraction, ICP-MS Cyanide-extractable gold 50 g 0.02 to 10,000 ppb $15.25 Recoverable gold for metallurgy
Nickel sulphide fire assay, ICP-MS Gold, platinum, palladium, rhodium, iridium and ruthenium 30 g 1 to 10,000 ppb $191.40 each for 3 or more PGE analysis
Bismuth fire assay (Bureau Veritas) Gold, with bismuth as the collector Not listed Similar to lead fire assay By quote Lead-free fire assay

Aqua regia (hydrochloric and nitric acids) leaches sulphides, some oxides and some silicates, and it barely attacks barite, zircon, monazite, chromite, garnet, ilmenite, rutile or cassiterite. Four-acid digestion adds perchloric and hydrofluoric acids and dissolves most minerals, but it can be incomplete for barite, rare earth minerals and Ti, Sn and W minerals (Actlabs 2026 schedule). Intertek recommends its 52-element aqua regia packages for exploration only and not for resource analyses, because silicate and refractory minerals stay largely undigested (Intertek).

Actlabs warns that nickel sulphide fire assay can read gold low and recommends its fire assay ICP-MS package for accurate gold (Actlabs), and Bureau Veritas says its bismuth fire assay has detection limits and accuracy similar to lead fire assay at comparable cost (Bureau Veritas).

Cost per sample

Add sample preparation to each analysis. Actlabs charges $12.40 to crush a rock or core sample of up to 7 kg to 80% passing 2 mm, split 250 g and pulverize it to 95% passing 105 µm, and $6.75 to crush and split 500 g for PhotonAssay, so a 30 g fire assay with an AAS finish comes to $32.40 a sample and PhotonAssay to $30.45 (Actlabs 2026 schedule, checked November 2026). Actlabs asks for two to three times the listed sample weight for quality control, and it may cut a fire assay charge to 15 g or less for a proper fusion.

Lab selection

CIM's guidelines ask for a reputable and preferably ISO-accredited lab qualified for the element, signed certificates with a statement of the methods, and proof from the geologist that any new method suits the task (CIM). ISO/IEC 17025 accreditation covers the method codes on the lab's scope, so check the scope for the code in the release. The Standards Council of Canada scope for SGS Burnaby lists its 30 and 50 g fire assays with AAS and gravimetric finishes (SGS scope), and Actlabs lists ISO/IEC 17025 for 7 of its 14 labs, including Ancaster, and sends readers to the SCC site for the accredited methods (Actlabs 2026 schedule).

QA/QC programs

CIM asks for blanks, certified reference materials (CRMs, or standards) and duplicates in every campaign, often enough for statistical confidence and across the full range of grades, plus regular check assays at a third-party lab (CIM).

Control Checks Example rate
CRM (standard) Accuracy against a certified grade 1 in 25 at Hot Chili, 1 in 20 standards or blanks at Ramelius
Blank Contamination from the previous sample 3 in 100 at Hot Chili, placed after mineralized samples
Field duplicate Sampling precision 1 in 25 at Hot Chili (RC field splits and core coarse-crush duplicates), 1 in 20 at Ramelius
Umpire assay Bias between labs Bureau Veritas checks of ALS results at Hot Chili in 2021 and 2023

The rates come from Hot Chili and Ramelius. OREAS, a CRM maker, suggests treating one result outside two standard deviations of the certified value as a warning, and several results outside 2SD or one outside 3SD as grounds for rejection, at the QC manager's discretion (OREAS 991 certificate).

Assays in a release

Run these five checks on a drill release before you buy:

  • Which lab ran the assays, and is it independent and accredited? NI 43-101 requires the lab's name, location and relationship to the issuer (NI 43-101), and the SGS Burnaby and Bureau Veritas Vancouver labs hold ISO/IEC 17025 accreditation for their gold fire assays (SGS scope, Bureau Veritas scope).
  • Which method, charge weight and finish? A release should name the method and lab code, such as a 50 g fire assay with AAS finish (Au-AA24) or PhotonAssay on 500 g (Au-PA01), plus the ore-grade method for high grades, and JORC asks whether the method is partial or total (JORC Code 2012).
  • Does the release report visible gold or erratic repeat assays? Both indicate coarse gold, and ALS assays coarse gold by screen fire assay or by PhotonAssay on a 500 g sample (ALS, ALS PhotonAssay).
  • What QA/QC did the company insert? JORC asks for standards, blanks, duplicates and external laboratory checks, and whether the results show acceptable accuracy and precision (JORC Code 2012).
  • Does one high-grade interval account for most of a long intercept? NI 43-101 requires disclosure of significantly higher-grade intervals within a lower-grade intersection (NI 43-101), and JORC requires a qualification wherever a release omits true widths (JORC Code 2012, JORC Code explained).

GeoTheories for prior work

Each GeoTheories project page lists the historic holes, intercepts and assays from earlier reports and company filings, so you compare a new intercept with what earlier operators drilled on the same ground before you buy, option or drill. You digitize assay tables and drill logs from PDF reports into layers.

Agents build out your thesis from the deposit model and then walk it through the geology, geophysics, geochem, satellite alteration, mine records, drill holes, filings and claims on that ground, in any country or mining district, including Yukon, the Athabasca Basin, Greenland, Argentina, Sweden and South Australia.

Solo is $299 a month or $2,499 a year, Team is $699 a month or $5,849 a year for up to 5 people, through self-serve checkout.

FAQ

What is the fire assay method for gold?

Fire assay fuses a 30 or 50 g charge of pulverized rock with a lead oxide flux above 1,000°C, collects the gold in lead, removes the lead by cupellation and measures the gold in the remaining bead by AAS, ICP or weighing (ALS).

Is PhotonAssay better than fire assay?

For coarse gold, yes, because it measures about 500 g against a 30 or 50 g fire assay charge, and each analysis takes about two minutes (ALS, Chrysos). For low-grade exploration samples, fire assay reaches lower limits, 0.001 ppm with an ICP-AES finish against 0.03 ppm, and Intertek calls PhotonAssay better suited to resource development and grade control (ALS, Intertek).

What is the alternative to fire assay?

PhotonAssay is the non-destructive alternative for total gold (Intertek). Aqua regia digestion is a cost-effective partial method for grassroots exploration, and cyanide leach and BLEG measure cyanide-extractable gold (Intertek).

How much does a fire assay cost?

Actlabs charges $20.00 for a 30 g gold fire assay with an AAS finish and $12.40 for standard rock or core preparation, so about $32.40 a sample, and $23.70 plus $6.75 preparation for PhotonAssay (Actlabs, 2026 USD schedule, checked November 2026).

How long do assay results take?

It depends on the batch size, the methods and the lab's workload, and labs such as Actlabs offer rush turnaround for a surcharge (Actlabs). The analysis itself takes about 3 to 4 hours for fire assay and about 2 minutes for PhotonAssay (Chrysos).

What does ISO/IEC 17025 accreditation mean for an assay lab?

It means an accreditation body such as the Standards Council of Canada has assessed the lab for the specific methods on its scope, such as SGS Burnaby's 30 and 50 g fire assays (SGS scope). CIM's guidelines ask for a preferably ISO-accredited lab (CIM), so check that the scope lists the method code in the release.

What is the detection limit of fire assay?

At ALS it's 0.005 ppm (5 ppb) for a 30 or 50 g charge with an AAS finish and 0.001 ppm with an ICP-AES finish, and the gravimetric finish starts at 0.05 ppm (ALS).

Is PhotonAssay accepted under JORC and NI 43-101?

Yes, when the Competent Person or qualified person judges it suits the deposit. Chrysos says PhotonAssay results appear routinely in resource and reserve statements under JORC and NI 43-101 (Chrysos), and both codes ask the release to state the assay method and the QA/QC, with NI 43-101 adding the lab's name, location and relationship to the issuer (JORC Code 2012, NI 43-101).