Fermenting to cellar temperature instead of a thermostat
There is no temperature-controlled fridge in this setup and after four years I have stopped wanting one. The cellar under the garage keeps its own calendar, and brewing to that calendar has taught me more about yeast than a thermostat ever would have.
| Brewed | 30 March 2025 |
|---|---|
| Written up | 14 August 2025 |
| Into the fermenter | 20 litres |
| Mash held at | 67 °C for 60 min |
| Cellar on the day | 8 °C |
| Original gravity | 1.044 |
| Final gravity | 1.010 |
| Strength | 4.5 % by volume |
FG 1.010OG 1.044
Read with a hydrometer at twenty degrees, at the kitchen table, before anything got washed.
01 / 04
A year of Sunday readings
The cellar is a concrete box under the back of the garage, two metres down, reached by a hatch and a ladder. I have written the temperature off the same thermometer on the same wall every Sunday since the autumn of 2021, which adds up to a few hundred readings and is easily the most useful thing on this site. February sits between seven and nine degrees. April climbs to eleven, June to fifteen, and the peak lands in the second half of August at seventeen or eighteen.
The lag is the interesting part. Concrete and soil hold heat and give it back slowly, so the cellar is coldest about a month after the coldest week above ground and warmest a month after the hottest. Knowing that, I can lay out a year of brewing in March: cold, clean ferments from November through April, ordinary ales across the shoulder months, and the warm farmhouse batches in the last weeks of summer, when nothing else down there would behave.
Off the card
- February: seven to nine degrees, a month behind the frost
- April: ten to twelve, rising about a degree a fortnight
- June: fourteen to fifteen, the most forgiving stretch of the year
- Late August: seventeen to eighteen, three weeks after the heat
02 / 04
Choosing the beer to suit the month
Matching beer to month removes most of the arguments before they start. A clean ale yeast wants somewhere around eighteen to twenty degrees and will make apple and solvent flavours if you push it past twenty-four. In the cellar it never gets the opportunity, because the cellar cannot reach twenty-four. What it can do is sit at eleven degrees in April, which is too cold for that yeast and will leave it stalled half-finished with a mouthful of unfermented sugar still in the beer.
April and May batches therefore come up the ladder and live in the flat, in the bottom of a wardrobe against an interior wall, which holds nineteen degrees within a degree either way. The cellar takes the summer beers, when the flat is too warm, and the winter ferments, when everything upstairs is by the radiators. Three locations and a thermometer in each has turned out more flexible than one controlled fridge, and it cost me a shelf.
03 / 04
What the yeast does when it drifts
A ferment drifting a few degrees is not automatically a fault, and I spent a whole year treating it as one. What matters is when the drift happens. The first forty-eight hours, while the yeast is multiplying and laying down most of the flavour compounds, are the hours to hold steady: a swing of three degrees there shows up in the glass as banana, or as a hot, thin edge behind the malt. The same swing on day six is very nearly invisible.
The practical consequence is that I pitch and then leave the fermenter alone somewhere I know will not move much for two days, and only afterwards worry about where it ends up. A bucket of water standing around the fermenter buys enough thermal mass to flatten the daily cycle in the flat, which swings four degrees between afternoon and dawn. Twenty litres of beer inside twenty litres of water moves slowly, and slow is the whole point of the arrangement.
04 / 04
The two weeks after the bubbling stops
An airlock stops bubbling long before the beer is finished, and reading that silence as done cost me two crates of over-carbonated bottles in the first year. Fermentation ends when the gravity stops falling, which is a measurement rather than a sound. I take a reading on day seven and another on day nine, and if the two match to within a point the beer is finished. If they do not match, it waits, and I go and work out why it is sulking.
The fortnight after that is where the beer improves without any help from me. The yeast takes back some of what it made, the harshness comes off the bitterness, and the haze drops out as the cellar cools it. A batch bottled on day eight and a batch bottled on day sixteen out of the same fermenter are recognisably different beers — I have run that comparison twice — and the later one won both times. Patience is the cheapest ingredient in the cupboard.