Removing carrageenan from an acidified milk drink is harder than removing it from a neutral one, because at pH 3.8–4.5 casein is already close to its isoelectric point and wants to come out of solution. Carrageenan was doing two jobs at once: keeping the protein particles apart, and keeping anything insoluble in suspension. Replace it with one material and you will usually fail at one of the two. The working approach separates the jobs — a protective colloid for the protein, a very low dose of a suspending agent for the particles — and it is the second job that low acyl gellan is used for here, at 0.020–0.050%.
If you want the wider picture of which substitute replaces which function across all applications, start with carrageenan-free stabiliser systems.
What carrageenan is doing
| Job | Mechanism | Consequence if unprotected |
|---|---|---|
| Protein protection | Electrostatic complex with casein, keeping particles mutually repulsive | Sandiness, then sediment, then a clear whey layer on top |
| Suspension | Weak network that holds particles in place | Cocoa, pulp or mineral settles into a hard layer |
| Body | Viscosity contribution | Thin mouthfeel in a product that already reads as light |
The first row is the one that ruins reformulations. It is also the one where carrageenan is genuinely good, because kappa carrageenan forms a specific complex with casein that most substitutes do not replicate. Anything that replaces carrageenan here has to be judged on whether it stops the protein aggregating, not on whether the drink looks homogeneous on day one.
Candidates, and which job each one covers
| Substitute | Protein protection | Suspension | Working dose | Notes |
|---|---|---|---|---|
| HM pectin | Yes — the industrial standard | Weak | higher than gellan several fold | Needs a narrow solids and pH window |
| Soluble soy polysaccharide | Yes | Weak | higher than HM pectin | Very low viscosity, works at low pH |
| CMC | Partial | Weak | higher than HM pectin | Label is increasingly a problem |
| Low acyl gellan | No | Yes, strongly | 0.020–0.050% | Used for suspension, not protection |
| Combination, pectin + gellan | Pectin does it | Gellan does it | see below | How most clean-label drinks are built |
Be blunt about the gellan row: low acyl gellan does not protect acidified casein. If you swap carrageenan for gellan alone, the drink will be physically stable and will turn sandy. The replacement that works pairs a protective colloid — usually HM pectin or soluble soy polysaccharide — with a small amount of gellan to handle suspension. This is why the gellan line on the label is so short: it is doing one job at a very low dose.
The dose arithmetic
Everything below is a property of the formulation. Prices are yours to supply.
| Line | Carrageenan system | Pectin + gellan system |
|---|---|---|
| Protective colloid | kappa carrageenan, one material | HM pectin or soluble soy polysaccharide |
| Suspending agent | included in the above | low acyl gellan 0.020–0.050% |
| Suspension dose comparison | higher total | roughly one tenth for the suspension function |
| Total stabiliser powder | single figure | two figures, add them |
The useful observation is the suspension row: where carrageenan was carrying both jobs at one dose, gellan delivers the suspension part at roughly a tenth of that weight. Whether the combined system is heavier or lighter overall depends on how much protective colloid your pH actually requires, and that is a measurement rather than a rule. Run both routes against your own two prices.
Three cases where you should not switch
- Very low pH, below about 3.6. Below this the protein is close enough to its isoelectric point that even HM pectin struggles, and adding another variable tends to make things worse rather than better.
- Products with high native calcium. added minerals shift both the protein behaviour and the gellan set. If the drink is fortified, expect to re-develop rather than substitute.
- Anything already running on HM pectin successfully. If you have no carrageenan in it, this page is not for you — there is nothing to remove.
Migration in four steps
- Measure where you are. pH, titratable acidity, calcium content and particle size distribution before you change anything. The pH tells you how much protective colloid you need; calcium tells you how the gellan will set.
- Fix the protein first, alone. Qualify the protective colloid by itself — no suspending agent — until the drink survives three weeks without sediment or sandiness. If it fails here, no amount of gellan will fix it.
- Add gellan last, at the bottom of the range. Start at 0.020% for suspension and go up only as far as the particles demand. More than that gives a drink with a perceptible set.
- Run the real thermal process. If the product is pasteurised or UHT treated, test after the heat step, not before — gellan hydration and the protein behaviour both change through it.
Failure modes
| Symptom | Likely cause | First check |
|---|---|---|
| Sandy mouthfeel after a week | Protein aggregating — protection insufficient | Protective colloid dose, not gellan |
| Hard sediment at the base | Suspension insufficient | Gellan dose, then particle size |
| Clear layer on top (whey separation) | Protein network collapsed | pH and protective colloid together |
| Drink sets slightly in the pack | Gellan above the useful range | Drop to 0.020% |
| Grainy immediately after filling | Incomplete hydration | Dispersion and hydration temperature |
| Sediment appears only after transport | Vibration-sensitive weak network | Network strength and suspension dose |
Tools
- Dosage calculator — includes acidified milk drinks and drinking yoghurt at 0.020–0.050%.
- Formula pre-check — flags the pH and ion combinations that fail first.
- Troubleshooting tree — works back from the symptom to the likely cause.
Frequently asked
Can gellan replace carrageenan in acidified milk drinks on its own?
No. Gellan suspends; it does not keep acidified casein apart. That is the protective colloid's job, and it is why the standard answer combines HM pectin or soluble soy polysaccharide with a small amount of gellan.
Why does my drink look fine and then go sandy?
Because protein aggregation takes time. Sandiness appearing in the second or third week is a protection failure, not a suspension failure, and the fix is in the protective colloid rather than in the gellan dose.
How does UHT treatment affect it?
Both materials have to survive the heat step, and the protein behaviour changes through it. Anything validated cold and then put through UHT without retesting should be assumed unvalidated.
Is it still a clean-label position?
Generally yes, and better than CMC. Check the labelling rules for your destination market rather than assuming — requirements differ.