What does “33% lye concentration” mean?
It describes the lye solution, not the recipe: a 33% solution is one-third sodium hydroxide and two-thirds water by weight. If your recipe needs 143.8 g of lye, a 33% solution means 292.0 g of water (143.8 ÷ 0.33 − 143.8), for 436 g of solution in the jug. Water : lye ratio says the same thing the other way round: 2 : 1 is two grams of water per gram of lye, which is a 33.3% solution. The two scale with the lye, so a recipe written in either makes the same bar whether you pour 500 g or 5 kg.
concentration = 1 ÷ (1 + ratio) and ratio = (1 − concentration) ÷ concentration. So 2 : 1 → 1 ÷ 3 = 33.3%; 1.5 : 1 → 40%; 1 : 1 → 50%. Grams of water = grams of lye × ratio.
Why “water as a percent of oils” is the odd one out
The third convention, SoapCalc's default of “38% water as percent of oils”, sets water from the oil weight instead of the lye weight. That sounds reasonable until you notice that different recipes need different amounts of lye for the same oils. A castile needs about 129 g of NaOH per kilo; a 100% coconut bar needs 147 g. Give both 380 g of water and you get two quite different solutions:
| Recipe (1,000 g oils, “38% water”) | NaOH | Water | Lye concentration | Water : lye |
|---|---|---|---|---|
| 100% olive (castile) | 128.7 g | 380 g | 25.3% | 2.95 : 1 |
| Classic bar: olive 35 / coconut 30 / palm 30 / castor 5 | 143.8 g | 380 g | 27.5% | 2.64 : 1 |
| 100% lard | 134.1 g | 380 g | 26.1% | 2.83 : 1 |
| 100% coconut, 20% superfat | 146.6 g | 380 g | 27.8% | 2.59 : 1 |
| Shave soap: stearic 50 / coconut 25 / shea 15 / castor 10 | 141.0 g | 380 g | 27.1% | 2.70 : 1 |
NaOH from each oil's SAP value at 5% superfat except where stated; water fixed at 380 g, 38% of the oil weight. The same “38%” spans 2.5 percentage points of lye concentration, from a slow, wet castile to a coconut bar that will move fast.
None of those bars is wrong, but the recipe writer didn't choose them; the arithmetic did. That is why this site (and most modern soapers) state water as lye concentration or ratio, and why the converter above asks for your lye and oil weights before it will call a “% of oils” figure exact: without them it assumes a typical bar at 0.14 g of lye per gram of oil and says so.
Should I use lye concentration or water : lye ratio?
Either. They are interchangeable, and the builder prints both. Ratio is the older habit on the forums (“2 : 1”, “1.5 : 1”); concentration is what the chemistry says and what this site's calculators default to. Pick the one you find easier to picture and keep a chart of the other by the scale:
SoapmakingToolbox.com · water : lye ↔ lye concentration
| Water : lye | Lye concentration | Water per 100 g lye |
|---|---|---|
| 1 : 1 | 50.0% | 100 g |
| 1.2 : 1 | 45.5% | 120 g |
| 1.5 : 1 | 40.0% | 150 g |
| 1.7 : 1 | 37.0% | 170 g |
| 1.8 : 1 | 35.7% | 180 g |
| 2 : 1 | 33.3% | 200 g |
| 2.2 : 1 | 31.3% | 220 g |
| 2.5 : 1 | 28.6% | 250 g |
| 2.7 : 1 | 27.0% | 270 g |
| 3 : 1 | 25.0% | 300 g |
concentration = 1 ÷ (1 + ratio). Below 1 : 1 (above 50%) the lye will not fully dissolve at room temperature.
Choosing a concentration
The converter translates between the conventions; it doesn't choose a strength for you. The lye concentration guide does: the same recipe at six concentrations with the water, shrinkage and bar weights worked out, what a steeper discount does to trace, gel and soda ash, and why "a 10% water discount" means little until you say discount from what. As a starting point, this site's calculators default to 33% (2 : 1) for everyday bars, and 25–28% gives a first batch or an intricate swirl more working time.
Reading an old recipe
If a recipe lists only oils, lye and water in grams, the converter can tell you what convention it was written in. Type the water as a ratio or a concentration (divide water by lye), add the lye and oil weights, and the third figure appears. A castile listed with 1,000 g oils, 129 g lye and 380 g water was written at 38% of oils; it is a 25.3% solution, and you can now rewrite it at 33% with 262 g of water: the same oils and lye, with less water to cure out.