Guides

Baumé, Brix, SG and API — what your hydrometer is actually telling you

A drum says '20° Baumé,' a tank says '66° Baumé,' a crude assay says '39.6° API,' a refractometer says '24° Brix.' Four numbers, four industries, and two of them aren't even the same formula. Here's what each one means and how to convert between them.

By D. Shin · Published September 3, 2026

Pick up an unlabeled glass hydrometer off a shelf in an old plant and it might read straight in specific gravity, or it might carry a scale marked only "°Bé" with no further explanation. None of this is arbitrary. Every scale on this page was built by hand, by someone solving a specific 18th- or 19th-century measurement problem for a trade that had no easy access to a lab balance — and every one of them converts cleanly to the specific gravity your calculator already understands, once you know which formula belongs to which scale.

Start with the one that isn't a trick: specific gravity

Specific gravity (SG) is just a ratio: the density of your liquid divided by the density of water under some reference condition. It's unitless, which is the whole point — a hydrometer doesn't need to know grams or millilitres, it only needs to float higher or lower depending on how the liquid compares to water. The catch, and it trips up more conversions than the algebra ever does, is that "water" isn't one number. Chemistry references often use 20°C water against 20°C sample (written 20°/20°), older definitions use water at its density maximum near 4°C, and the petroleum industry insists on 60°F against 60°F. Mixing reference temperatures across a conversion introduces an error before you've touched a formula.

SG shows up everywhere in the trades once you start looking. A charged lead-acid battery's electrolyte reads roughly 1.265 SG; run it down and that same electrolyte drifts toward 1.12–1.15 as sulfuric acid is consumed at the plates — which is exactly why a battery hydrometer is calibrated in SG rather than volts. Every scale below is SG run through a different arithmetic, invented so someone reading a floating glass rod on a shop floor didn't have to do the division themselves.

Baumé: one name, two unrelated formulas

Antoine Baumé, a French pharmacist, built his scale in 1768 for tradespeople handling acids, brines and syrups who needed a fast field reading, not a lab result. The complication he left behind is that he actually built two scales, calibrated from opposite directions, and they don't share a datum:

Notice the datum mismatch: on the heavy scale, plain water reads 0°. On the light scale, plain water reads 10°. That single fact is the fastest way to identify an unmarked Baumé hydrometer — drop it in a glass of water and read where it settles. Near zero, it's a heavy-scale instrument, meant for acids, brines and syrups. Near ten, it's light-scale, meant for oils and alcohols.

Worked example: 20° Baumé muriatic acid

"Muriatic acid" — technical-grade HCl sold for pool chemistry and concrete etching — has been graded by Baumé strength for well over a century, and 20° Baumé is the standard commodity number. Plug it into the heavy formula:

SG = 145 ⁄ (145 − 20) = 145 ⁄ 125 = 1.160

That specific gravity corresponds to roughly 31.45% HCl by weight — the figure that actually appears on most 20° Baumé muriatic acid spec sheets, and it's routinely rounded up to "32% muriatic" in casual trade talk for the same product.

Worked example: 66° Baumé sulfuric acid

Baumé's own scale was partly calibrated around concentrated sulfuric acid, and "66° Baumé oil of vitriol" became the historical benchmark for commercial concentrated H₂SO₄ — the number is still printed on tank cars and spec sheets today.

SG = 145 ⁄ (145 − 66) = 145 ⁄ 79 = 1.835

That works out to roughly 93% H₂SO₄ by weight — the standard commercial-grade concentration that's actually sold under the "66° Baumé" name, sitting on the flat top of the density curve where a fair range of concentrations all read within a hair of the same specific gravity.

The caustic trade runs the identical formula for the same reason: "50° Baumé caustic soda," the classic commercial grade of roughly 50% NaOH, is just SG 1.526 run through the same heavy-scale arithmetic — see the NaOH concentration calculator if that's the acid or base you're actually working with.

Worked example: the light scale, and where it still shows up

Pure ethanol at 20°C has a specific gravity of about 0.789 — lighter than water, so it's the light-scale formula this time:

°Bé = 140 ⁄ 0.789 − 130 ≈ 47.4°

Before Gay-Lussac's directly-in-percent alcohol scale and, later, API gravity took over their respective industries, light Baumé was the general-purpose trade tool for anything lighter than water — vegetable oils, syrups, and some spirit assessment. A distilled spirit at drinking strength is much closer to water than to pure ethanol, so on this same scale it reads far lower — in the high teens rather than the high forties. Don't try to interpolate that number by eye, though: ethanol and water don't mix in a straight line. Half a litre of each makes about 965 mL, not a litre, and the density curve bends the same way (see why alcohol plus water is less than the sum of its parts). The real figure at any given proof comes out of a measured density table, not a linear guess.

Brix: Baumé's sugar-industry descendant

The Brix scale descends from the same lineage — Karl Balling's sucrose hydrometer work refined by Adolf Brix in the mid-1800s — but it solves a narrower problem: for a pure sucrose-and-water solution, degrees Brix reads directly as percent sucrose by weight. 24° Brix means roughly 24 g of sucrose per 100 g of solution.

Unlike Baumé, there's no clean algebraic bridge from Brix to SG. Real sugar solutions pack together the same way real acid and caustic solutions do — non-linearly, through hydrogen bonding between sucrose and water — so the relationship is empirical, built from actual measurements and published as reference tables (the ICUMSA international sucrose tables and AOAC methods are the standard sources), the same honest approach this site's own acid and caustic density tables use rather than forcing one formula across the whole range. As a rough landmark from those published tables: 10° Brix sits around SG 1.040, 20° Brix around SG 1.081.

Three closely related scales get used almost interchangeably across the beverage industry, and it's worth knowing which is which: Brix for wine, juice and honey; Plato (°P) for brewing wort — the number a brewer means by "original gravity" is usually quoted in Plato; and Balling, Brix's older predecessor, still printed on a few legacy hydrometers but functionally obsolete now. They agree to within a small fraction of a degree for most practical work, but a certificate of analysis will specify which one was used.

A standard winemaking rule of thumb converts a Brix reading into an estimate of potential alcohol: multiply by roughly 0.55. Grape must reading 24° Brix at harvest suggests a potential strength around 24 × 0.55 ≈ 13.2% ABV — a planning number, not a promise, since actual yeast efficiency varies. Where a must comes in too sweet for the target style, winemakers sometimes cut it with water to bring the Brix down before fermentation; once you accept that Brix behaves like a mass-percent figure, that's an ordinary solution dilution problem, C₁V₁ = C₂V₂, same as diluting any other stock solution.

API gravity: petroleum built its own scale on purpose

By the early 1920s the American Petroleum Institute had a real problem: the light Baumé hydrometers in circulation across the US oil industry had drifted out of consistency with each other, glass batch to glass batch, and crude oil was already being bought and sold by the reading. In 1921 API defined its own scale, close to light Baumé in shape but re-anchored to fix the drift:

°API = 141.5 ⁄ SG − 131.5  ·  SG = 141.5 ⁄ (°API + 131.5)

Reference basis is 60°F against 60°F water (ASTM D287/D1298) — not 20°C like most of the rest of this page, which is the single most common unit-basis mistake when someone pulls an SG from a chemistry table into a petroleum calculation. Notice water still comes out to exactly 10.0° API, the same anchor point light Baumé used — API is best understood as light Baumé's more carefully calibrated grandchild, not an unrelated invention.

The scale runs backwards from what feels intuitive: higher API gravity means lighter, more valuable crude. The US Energy Information Administration's rough classification bands are light crude above 31.1° API, medium 22.3–31.1°, heavy 10–22.3°, and extra-heavy below 10°. West Texas Intermediate, the US benchmark grade, trades around 39.6° API:

SG = 141.5 ⁄ (39.6 + 131.5) = 141.5 ⁄ 171.1 ≈ 0.827 — comfortably light crude.

At the other end, Alberta oil-sands bitumen is commonly quoted around 8° API — SG ≈ 141.5 ⁄ 139.5 ≈ 1.01, meaning it's barely lighter than water and won't flow through a pipeline on its own. That's why it ships as "dilbit": bitumen cut with a lighter condensate purely to bring the blend's API gravity, and therefore its viscosity, up to something a pipeline can move.

Reading the instrument without guessing which scale it is

• Confirm heavy or light before you calculate anything. Running the heavy formula on a light liquid (or the reverse) doesn't throw an error — it just hands you a confident, wrong number. The water-dunk test above (reads ≈0° vs ≈10°) settles it in ten seconds if the scale isn't labeled.
• Match the reference temperature to the scale, not to your process. Most Baumé and Brix work is defined at 20°C; API gravity is defined at 60°F. A hot sample reads lighter than it really is on every one of these scales — proper correction tables exist (ASTM D1250 for petroleum) rather than eyeballing an adjustment.
• Read the meniscus at eye level, with the hydrometer floating freely and not touching the container's wall or bottom. Looking down at an angle is a small, systematic error that's easy to eliminate and easy to forget.
• Treat Brix and Baumé readings on real (non-pure) solutions as process-control numbers, not lab-certified concentrations — every dissolved thing in the liquid nudges a density-based reading, which is exactly why titration remains the reference method whenever a figure needs to be certified rather than just monitored.

Quick reference: the same specific gravity, four ways

Because every scale on this page is pure algebra on SG, there's no ambiguity in building a side-by-side table — these are computed directly from the formulas above, not measured data:

*API gravity is arithmetically defined for any SG, but negative API numbers don't correspond to any real trade usage — the scale exists for petroleum products, which never approach these densities. They're included here only to show the formula is the same equation as light Baumé with different constants, not a different kind of math.

None of these four scales measures anything a plain density figure doesn't already capture. They exist because, for two and a half centuries, the person who needed the number was standing at a tank with a glass float and no calculator — and each trade built the arithmetic that put the reading they cared about directly on the stem. The float hasn't changed much since. The conversion back to something a modern spec sheet understands is the only part that's gotten easier.