Three days into a late-October batch I lifted the airlock to top up the water and got a face full of struck match. By the next morning it had moved on to something closer to a boiled egg forgotten in a gym bag. I knew the word for it from forum threads, and I did exactly what those threads told me to do: I stirred a heaping teaspoon of yeast nutrient into the carboy and waited for the hero moment. The smell got worse for two days before it got better, and I spent that week blaming the yeast strain. The strain was fine. The teaspoon was the problem.
Why apple juice is a hard place to be a yeast
Yeast needs nitrogen to build the proteins it uses to eat sugar, and the form it can actually use — free amino acids plus ammonium, measured together as yeast assimilable nitrogen, or YAN — is scarce in apples compared with grape must or a grain wort. A malt wort arrives with nitrogen built into it. Fresh-pressed juice frequently does not, which is why a cider ferment can look healthy in the airlock while the culture underneath is quietly running short.
When nitrogen runs low mid-ferment, the yeast starts stripping sulfur out of the sulfur-bearing amino acids it can reach, and hydrogen sulfide comes off as a byproduct. Our noses catch that molecule at a couple of parts per billion, which is an absurdly small quantity, so a batch can carry a fault you can smell across the room and still be nowhere near a real problem chemically. That sensitivity is the whole reason this shows up in home batches so often: you are detecting a trace, not a disaster.
The nitrogen curve is not a straight line
Here is the part that cost me a batch. In a 2020 study out of Cornell, Song and colleagues fermented cider at three supplemented nitrogen levels and measured what came off. The low treatment sat at 86 mg N/L, the intermediate at 208, the high at 433. The intermediate one — the sensible-looking middle — produced the most hydrogen sulfide of the three, roughly 2.3 times what the low treatment gave. The highest nitrogen level produced the least.
So the reflex I had, tossing a moderate scoop at a nitrogen-poor juice, lands in the worst zone on that curve: enough nitrogen to wake the yeast up and push it hard, not enough to keep it fed while it works. One study on one juice with one strain is not a universal law, and I would not read it as an instruction to dose every carboy to 433. What it did change for me is that "add a bit more nutrient and see" stopped being a free move.
Timing beats total
Yeast takes up ammonium nitrogen eagerly in the first half of a ferment and barely bothers with it near the end. Dumping the whole dose at pitch gives the culture a burst it cannot store, and dumping it at the end leaves unused nitrogen sitting in a nearly-finished cider where the only organisms interested in it are the ones you did not invite. Splitting it does more for a batch than raising it.
My schedule now is half the planned nutrient at pitch and half when the gravity has dropped about a third of the way to dry, which in a room-temperature carboy is usually somewhere around day three.
Three things I check before I open the nutrient tub
- Temperature. A ferment running hot is a ferment running stressed, and stressed yeast makes more of everything you did not order. If the carboy is warm to the back of your hand, move it before you add anything.
- The strain. Sulfide production varies enormously between strains, and some wine and champagne yeasts are selected and sold specifically for low hydrogen sulfide output. If you keep getting the same fault from the same packet, the packet is a suspect — the notes on choosing a cider yeast lay out how the families differ.
- How much sulfite went in. Yeast can reduce sulfur dioxide to hydrogen sulfide, so an enthusiastic Campden dose is one of the ways a clean juice starts smelling like eggs on day three. That is another argument for dosing off a pH reading rather than off the gallon count, which I learned the expensive way and wrote up in the sulfiting notes.
The smell ladder, and the clock you are racing
Hydrogen sulfide does not sit still. Left in contact with oxygen and time, it reacts into mercaptans, which read as burnt rubber, struck flint, or cooked onion, and those in turn oxidize into disulfides that come across as garlic or stale drain. Each rung down that ladder is harder to remove than the one above it, and the last one is essentially permanent at the home scale.
That is the single most useful thing to know about this fault. Rotten egg is the easy version. If your carboy smells like eggs today, you are holding a problem with a straightforward fix and a deadline. If it smells like burnt rubber and old garlic four months later, the window closed while you were waiting to see whether aging would sort it out.
Splash racking, and the day it stops being safe
Hydrogen sulfide is volatile, and moving the cider through air carries a lot of it off. During an active ferment that is a cheap fix: rack into a clean vessel with the tubing left high so the cider falls and breaks up, or simply stir the surface vigorously for a minute or two. The carbon dioxide still coming off the batch protects it, so the oxygen you introduce gets consumed rather than parked in the cider.
Once fermentation is finished, that protection is gone, and the same splash that rescued a young batch will flatten a finished one and set it browning. After dry, transfers should be quiet and full-to-the-neck, the way the racking gear notes describe. So the honest advice is: smell your ferment daily while it is working, because the free fix has an expiry date.
The copper test, run in a glass and not in the carboy
Copper binds sulfides, which is why the old trick of stirring cider with a clean copper pipe works at all. What it also does is tell you which rung of the ladder you are on, and that is the reason to bother.
- Pull three small glasses of the suspect cider. Leave one alone as your reference.
- Drop a piece of clean, unlacquered copper into the second, swirl it for thirty seconds, and take it out. Do not leave it sitting.
- Smell all three with a few seconds between them. If the treated glass has gone clean, the fault is still hydrogen sulfide and you caught it in time. If it barely moves, you are downstream into mercaptans and copper is not going to rescue the batch.
- Only then decide whether to treat the whole volume, and scale the contact from what the glass told you rather than from a number somebody posted.
The trial takes five minutes, and it has twice stopped me treating five gallons for a fault a splash rack would have cleared on its own.
What press day looks like now
I write the sulfite dose and the pH on the tag before the yeast goes in, pitch a strain I know is quiet, and put half the nutrient in at pitch with the other half marked on the calendar for day three. I smell the airlock every morning for the first week, which takes four seconds and is the only monitoring this fault actually needs. If eggs turn up, the batch gets a splash rack that day, not that weekend.
Two seasons of that and I have not lost a carboy to sulfur since. The change was not a new product on the shelf. It was moving from one large hopeful addition to two small planned ones, and treating a smell on day three as something to act on rather than something to wait out.