Butterfat vs Vegetable Fat in Gelato: What Changes
Table of contents
Two bases can carry the same 8 percent fat and behave nothing alike. Milkfat melts across a 70 degree window; refined coconut oil melts across about ten. That single difference reshapes body, melt speed, flavour release and what you are legally allowed to print on the tub.
Same job on the balance sheet. Two completely different crystallisation curves.
Two fats, two melting curves
Milkfat is not one fat. It is a mixture of several hundred distinct triglycerides built from more than 400 identified fatty acids, and that diversity is exactly why it behaves the way it does. Marshall, Goff and Hartel describe its melting behaviour in Ice Cream as a broad range running from roughly minus 40 degrees Celsius to plus 40, split into a high-melting fraction, a middle fraction and a low-melting fraction. There is no single melting point to quote, because at any temperature between those extremes some of the fat is solid and some is liquid.
Refined coconut oil is close to the opposite. USDA FoodData Central puts it at roughly 82 to 92 percent saturated fatty acids, dominated by lauric acid at around 45 to 48 percent and myristic acid at 16 to 18 percent. Those chains are similar in length, so they crystallise in a narrow band. The oil is a firm white solid at 20 degrees and a clear liquid by about 26. Palm kernel oil behaves the same way for the same reason.
That contrast is the whole article. Everything below follows from it.
What partial coalescence needs from the fat
Body in gelato does not come from fat sitting passively in the mix. It comes from partial coalescence: during churning, fat globules whose interiors are partly crystalline collide, pierce each other's membranes and clump into an irregular network. That network traps air cells, holds the structure upright in the pan and slows the melt.
The mechanism needs two things at once. It needs crystals, so a globule has something rigid to puncture with. It also needs liquid fat, because the liquid phase is what bridges two globules and glues the clump together. A globule that is entirely liquid deforms and bounces off. A globule that is entirely solid has nothing to bridge with.
Milkfat sits in that useful middle almost everywhere. At the 4 degree ageing temperature a large part of it is crystalline and the rest is still oil, which is why a cream base destabilises readily in the batch freezer. Lauric fats arrive at ageing temperature essentially fully crystallised, with very little liquid phase left to do the bridging. They can still build structure, but the process is less forgiving and leans much harder on the emulsifier to strip protein off the globule surface first.
Quick reference. Milkfat melts across roughly minus 40 to plus 40 degrees Celsius and is part solid, part liquid at ageing temperature. Refined coconut oil melts sharply near 24 to 26 degrees and is close to fully solid at 4 degrees. Same fat percentage, different destabilisation behaviour.
Figure 1. Milkfat is part solid and part liquid across the whole working range. The plant fats switch state in a narrow band.
What the label lets you call it
In the United States the answer is not a matter of taste. 21 CFR 135.110 defines ice cream as containing not less than 10 percent milkfat and not less than 20 percent total milk solids. Replace the milkfat with a vegetable fat and the product no longer meets that standard of identity. The relevant category becomes mellorine under 21 CFR 135.130, which permits fat of non-milk origin at not less than 6 percent fat and 2.7 percent protein, and which must be labelled as mellorine.
The European Union has no single compositional standard for ice cream, so national codes and trade practice do the work instead. What is fixed at EU level is disclosure: Regulation (EU) No 1169/2011 requires refined oils of vegetable origin to be declared in the ingredient list with their specific vegetable source named, so a base built on coconut or palm has to say so. If you sell in more than one market, decide the fat source before you print anything.
Flavour, and the part the balance sheet misses
Milkfat carries a flavour payload no plant fat reproduces. Short-chain fatty acids, butyric acid at C4 chief among them, are volatile at serving temperature and are a large part of what reads as dairy. Cream also brings MSNF with it, which a bottle of oil does not: swapping heavy cream out of a recipe removes both fat and around 5.5 percent milk solids per 100 g of cream, so protein has to be rebuilt with skim milk powder or the base turns thin.
Refined coconut oil is deliberately neutral, which is either the point or the problem. Virgin coconut oil brings lactones and an unmistakable coconut aroma that will dominate a delicate flavour. Cocoa butter sits somewhere else again: it melts sharply just below body temperature, which is why cocoa butter reads as clean rather than waxy. Fats that finish melting above 37 degrees, such as fully hydrogenated fractions and hard palm stearins, are the ones that coat the palate.
Melting range also sets when aroma is released. A fat that is still solid in the mouth holds its dissolved aroma compounds hostage.
Swapping one for the other in a real recipe
| Property | Milkfat (cream, butter) | Refined coconut oil | Cocoa butter |
|---|---|---|---|
| Saturated fraction | about 65 to 70 percent | about 82 to 92 percent | about 60 percent |
| Melting behaviour | broad, minus 40 to plus 40 C | sharp, near 24 to 26 C | sharp, near 34 to 38 C |
| State at 4 C ageing | part solid, part liquid | close to fully solid | fully solid |
| Carries MSNF | yes, about 5.5 g per 100 g cream | no | no |
| Flavour contribution | strong dairy note | neutral when refined | cocoa note |
| Emulsifier dependence | moderate | high | high |
| US label impact | ice cream permitted | mellorine, not ice cream | mellorine, not ice cream |
The practical swap looks like this. Take a white base at 8 percent fat built on whole milk plus cream, pull the cream, and you lose fat and milk solids together. Put the fat back with 25 to 30 g of coconut oil per kilo of mix, rebuild the solids with a plant protein or a starch, and raise the emulsifier a little, because the fat phase now needs help destabilising. Soy lecithin at the usual dose is the common starting point. Then re-check overrun, since a change in the fat network changes how much air the mix will hold and keep.
Expect the melt to change too. A sharply melting fat can make a scoop that reads firm in the pan and then collapses quickly on the plate, which is the same failure mode described in why gelato melts too fast.
When vegetable fat is the right answer
Three cases, and they are all legitimate. A vegan line has no other option, and the choice then is between coconut, cocoa butter and nut fats, which is covered in vegan gelato fat sources. Cost matters in industrial production, where lauric fats are cheaper and more stable in storage than cream. And some flavours genuinely want a neutral fat: a strong fruit or a delicate tea is easier to read without butyric acid in the background.
What does not work is treating the swap as a one-line substitution on a spreadsheet. Same percentage, different fat, and the homogenisation step, the emulsifier dose and the overrun all need re-checking before the recipe is finished.
Related Concepts
- Ideal fat percentage for gelato covers the 6 to 9 percent window both fats have to land in
- Coconut milk for vegan gelato is the whole-ingredient route to the same fat
- Cashew cream in gelato adds body alongside the fat
- Homogenisation in gelato explains why globule size sets the ceiling on either fat
