Swapping an oil means recalculating the lye
Maker & Craft Published October 10, 2026
The palm oil runs out with the pot already warming, and there is olive oil in the cupboard. Three hundred grams for three hundred grams, everything else left as written.
The soap that comes out of that mould is lye-heavy. Not because olive oil is harsh, but because it needs less lye than palm does, and the recipe is still carrying palm’s share.
Every oil has its own number
Saponification value is the quantity of lye required to turn one gram of a given fat into soap. It follows from the fatty acid makeup, so it is a property of the oil rather than a preference.
Coconut sits near 0.183 grams of sodium hydroxide per gram of oil. Olive is around 0.135. Castor is about 0.129, shea butter about 0.128. The spread between the extremes is roughly a third, and a recipe is simply the sum of those individual requirements.
What the error looks like
Take a 1,000 g batch and swap a few hundred grams of one oil for another with a lower value, without touching the lye. The lye requirement for that batch can shift by something like 19 g.
Against a typical lye quantity for that size of batch, 19 g is not a rounding error. It is the difference between the 5% superfat you intended and a bar that is effectively lye-heavy by a similar margin in the other direction, which is a bar that can irritate skin and that has to be rebatched or thrown away.
Run it the other way, substituting toward a higher-value oil, and you get a mild but soft, oily bar that never cures properly. Less hazardous, equally wasted.
The pairs that catch people out
Palm oil and palm kernel oil are the classic trap. Similar names, same fruit, different parts of it, and substantially different saponification values. They are not interchangeable in any sense and the labels do nothing to warn you.
Coconut oil sold at 76 degree and 92 degree melt points behaves differently in the pot but takes the same lye, so that one is safe. Lard and tallow are close to each other and not close to the vegetable oils. And anything sold as a blend has no single reliable value at all, which is a good reason to avoid blends in a recipe you intend to repeat.
Recalculate every time, even for small changes
The habit worth building is to treat the recipe as a set of oils and let the calculator produce the lye, rather than treating the lye as a fixed part of the recipe.
Put the new oil list into the soap lye calculator, with your superfat, and take the number it gives. It is the same thirty seconds whether you changed one oil or five, and it removes the entire category of error.
Keep the old sheet too. Comparing the two lye figures tells you immediately how big a change you just made, which is useful information about the bar you are about to get, not only about safety.
Percentages are not a shortcut
Recipes written as percentages invite the assumption that keeping the percentages constant keeps the chemistry constant. It does not, because the percentages describe the oils and the lye follows from which oils they are.
Swapping 30% olive for 30% coconut keeps the recipe looking identical on paper and changes the lye requirement substantially. The percentage column stays the same, the saponification arithmetic underneath it does not.
Work it out before anything is melted
The practical order matters. Decide the oils, run the numbers, weigh the lye, then start melting.
Discovering mid-batch that the oil you planned on has run out, and improvising with the pot already warm, is how substitutions happen without recalculation. Oils keep. A batch does not, once the lye is mixed, so the pressure at that point is all in the wrong direction.
Questions people ask
How different are the values really?
Coconut oil needs about 0.183 g of sodium hydroxide per gram of oil, olive about 0.135, castor about 0.129. So a gram of coconut takes roughly a third more lye than a gram of olive, which is a large gap when it is applied across a whole batch.
Is palm and palm kernel oil the same thing?
No, and this is the most dangerous near-identical pair in soap making. They come from different parts of the same fruit and have substantially different saponification values, so one standing in for the other throws the batch off badly.
What about butters?
Shea, cocoa and mango all sit near the lower end, around 0.128 to 0.137, so swapping one butter for another is a smaller change than swapping a butter for coconut. It still changes the number and still wants rerunning.
Can I just add the new oil at trace instead?
Only if it is inside your superfat allowance. Oil added at trace is still oil in the pot, and if it pushes the total past what the lye was calculated for, the result is a softer, oilier bar rather than a neutral addition.