You add the fragrance, give it one more pulse, and the batter turns into something you didn't order — thick, or lumpy, or split, or a solid block around the blender. It is the most common way a batch that was going perfectly stops going perfectly, and it is nearly always the fragrance.
What isn't common is anyone distinguishing between the ways it happens. "It seized" gets used for all of them, and that matters, because they have different causes, different rescues, and different prognoses.
Four failures that look alike in the pot
| What went wrong | What you see | What's actually happening | Can you save it? |
|---|---|---|---|
| Acceleration | The batter thickens far faster than the recipe should — light trace to pudding in seconds, sometimes while you are still pouring the fragrance in. | Something in the fragrance is speeding the reaction. It is still soap, and still going right; it is just going right too quickly to work with. | Yes. Stop blending, stir by hand, and get it into the mold. The bars are fine — only the design is lost. |
| Ricing | Hard grains suspended in batter that is otherwise still liquid. Looks curdled, like scrambled egg or wet couscous. | Not a saponification failure at all — something in the fragrance is coming out of solution and setting into lumps with the harder fats. | Usually. Keep blending: the grains often break down and the batter comes back smooth. That costs you working time, so expect a thicker pour. |
| Seizing | The pot goes from liquid to a solid mass, in seconds, and cannot be stirred at all. | Acceleration taken to its conclusion. Some fragrances do it every time in cold process, and a stearic-heavy recipe will do it without any help. | Not as a poured loaf. Press it into the mold for a rustic bar, or finish it as hot process in a slow cooker. It is still good soap. |
| Separation | Oil pooling on top, or the batter splitting into layers after it looked fine. | The emulsion has broken — the fragrance has acted as a solvent and pushed the oil and lye solution back apart. | Often. Blend it back together properly. If it will not hold, hot process it. Do not pour a split batch and hope. |
| Discoloration | The loaf turns tan, brown or coffee-colored over hours to days, sometimes in patches or rings. | Vanillin in the fragrance oxidizing in an alkaline environment. Predictable, harmless, and nothing to do with the batter behaving. | Nothing to save. Ask the supplier for the vanillin percentage before you buy, and design the color around it. |
Ricing is the one most often misread. Grains in liquid batter is ricing and comes from the fragrance; a uniform thickening with no fragrance in the pot yet is false trace, which comes from cold hard fats and is covered in the trace guide.
Why no label can warn you
The obvious question is which fragrances misbehave, and the honest answer is that the bottle cannot tell you and the supplier can only half tell you. A fragrance oil is a blend of dozens of aroma chemicals, and the ones that accelerate a batch are not the ones you can smell — spice and floral notes have the worst reputation, but plenty of florals behave impeccably and plenty of clean linens do not.
Worse, the behavior isn't a property of the fragrance alone. The same bottle in a different recipe, at a different temperature, at a different water content, will do something else. When a supplier says a fragrance "accelerates slightly", that is a true report about their test batch, which is not your recipe. The published lists of well-behaved fragrances are worth reading and are not a guarantee.
So the useful skill isn't memorizing which bottles are trouble. It's knowing how to find out cheaply, and how to soap something difficult once you know.
Find out for 2 grams
Wholesale Supplies Plus publishes the sensible version of this: make a plain batch, split it into 2 oz (56.7 g) portions, stir one fragrance into each, and watch every portion at five points — as it's mixed, as it's poured, five minutes later, at unmolding, and at four weeks. What nobody says is how little that costs.
| Test batch | NaOH | Water | Batter | Fragrances you can test |
|---|---|---|---|---|
| 150 g oils | 21.6 g | 44 g | 215 g | 3 |
| 300 g oils | 43.1 g | 88 g | 431 g | 7 |
| 450 g oils | 64.7 g | 131 g | 646 g | 11 |
The classic bar at 5% superfat and a 33% lye solution, split into 56.7 g portions. Computed by this site's engine — the lye calculator will size any other base recipe.
A 56.7 g portion contains about 39.5 g of oils, so at a 5% load it needs 2.0 g of fragrance. A 30 mL sample bottle holds roughly 28.5 g — enough for 14 tests. And one test portion is 4.1% of a 2.5 lb loaf. So the real choice isn't between testing and not testing; it's between spending two grams to find out and spending a whole loaf to find out. Fragrance density taken as 0.95 g/mL, which varies a little by blend.
How to soap a fragrance you don't trust
Assume the worst and give yourself room. In rough order of how much they buy you:
- Add it at emulsion, not at trace. The single biggest one. The moment the batter is uniformly opaque, you have soap that will not separate — stir the fragrance in there and you have spent none of your working time yet.
- Stir it in by hand. Put the stick blender down before the fragrance goes in. A whisk or spatula will incorporate it perfectly well, and a blender pulse on top of an accelerating fragrance is what turns a fast batter into a solid one.
- Soap cool. Both parts at the low end of the workable range. The reaction feeds on its own heat, so starting cooler buys minutes — though not below the point where hard fats set, which the trace guide covers.
- Use full water. A 25–28% lye solution is a thinner batter with more room to move. It costs you a longer cure, which the water discounting guide quantifies.
- Pick a slow recipe. High-oleic, low in palmitic and stearic. The trace guide ranks the common blends by how much working time each gives you — a difficult fragrance in a castile-leaning recipe is manageable, and the same fragrance in a butter-heavy one is not.
- Simplify the plan. One color, a single pour, a flat top. Design ambition is what turns an accelerating fragrance from an inconvenience into a ruined loaf.
- Pre-mix it with a little of your oils before adding, which sometimes helps a fragrance that rices disperse instead.
Some fragrances accelerate in cold process every single time, and no technique fixes them. That isn't the end of the fragrance — it's a signal to change method. In hot process the soap is already saponified when the fragrance goes in, so there is no reaction left for it to accelerate: it cannot seize what is already soap. The bars are rustic rather than smooth, and the scent survives better besides, since it never sits in hot caustic batter. Stearic-heavy shave soaps are made this way for exactly the same reason.
When it happens anyway
Gloves and goggles stay on — a seized pot is still raw soap and still caustic. Then work out which of the four you have and act accordingly: keep blending for ricing and separation, stop blending for acceleration, and reach for the slow cooker if it has genuinely seized. Do not pour a batch that has split; do not add water to loosen a thick one, which changes the recipe and produces a soft bar that never quite recovers. The trace guide covers the same rescues from the other direction, where the cause was the blender rather than the bottle.
And write down what happened. A fragrance that seized once will seize again, and the note you make today is what stops you buying a liter of it.
SoapmakingToolbox.com · fragrance test log
Fragrance test log
One row per fragrance, 56.7 g of plain batter each. Mark each column: ✓ behaved · A accelerated · R riced · S seized · P separated · D discolored.
| Fragrance / supplier | % | At mix | At pour | +5 min | Unmold | 4 weeks |
|---|---|---|---|---|---|---|
Base recipe: ______________________________ · Oils: ________ g · Lye concentration: ______% · Soaped at: ______ °C · Date: ____________
The short version
- Four failures, not one. Ricing and separation want more blending; acceleration and seizing want less.
- The label can't warn you, and neither can the supplier's notes, because behavior depends on your recipe and your temperature.
- Test for 2.0 g. A 300 g batch screens 7 fragrances.
- Add fragrance at emulsion and stir it by hand. Two habits that cost nothing and prevent most of this.
- A fragrance that always seizes isn't unusable — it's hot process.