FG 1.020 is only a problem if your OG doesn't support it. From a 1.060 OG, 1.020 is 66.7% apparent attenuation — noticeably below the 73-82% most ale yeast strains are rated for, which means this batch is more likely stalled than simply finished. From a 1.040 OG, 1.020 would be 50% attenuation, which is stalled for essentially any yeast. From a 1.090 OG imperial stout, 1.020 is 77.8% attenuation — a completely normal, healthy finish. The number alone means nothing without the OG next to it; check your own pair against attenuation directly.
Rule out "still fermenting" before you rule out "stalled"
Fermentation slowing doesn't mean it's finished — a strong ale yeast can still be grinding through the last few gravity points well past the point where visible airlock activity has all but stopped. Take two readings 2-3 days apart. If they're identical, it's genuinely done or genuinely stuck; if the gravity is still dropping even slowly, give it more time before doing anything else. This single check eliminates the most common false alarm.
If the gravity really has stopped moving
A handful of causes account for most true stalls, roughly in order of likelihood:
- Fermentation temperature dropped too low. Yeast slows dramatically below its rated range, and a fermenter sitting in a cool room or a temperature swing overnight is enough to stall a strain that would otherwise finish cleanly. Check the strain's published range on the yeast reference against where the fermenter actually sat, not just where the room thermostat was set.
- Underpitched or old yeast. A pitch too small for the batch's gravity, or yeast that's past its viable date, runs out of healthy cell mass before the fermentable sugar is gone.
- Insufficient yeast nutrient, especially on higher-gravity worts or worts made from a large proportion of adjunct sugar with little of the nitrogen and micronutrients yeast need to keep working through a long fermentation.
- A genuinely low-attenuating yeast strain paired with a recipe that assumed higher attenuation. Check the manufacturer's published range for your exact strain — see what attenuation should I expect from my yeast — some English and Belgian strains simply don't attenuate as far as a clean American ale strain, and that's by design, not a fault.
- Mash temperature ran high, producing a wort with a higher proportion of unfermentable dextrins than the recipe intended — this is a mash-day cause showing up at bottling, and there's no fixing it after the fact for that batch; it's worth checking against what strike water temperature you were actually targeting for the next one.
Rouse it, warm it, then wait
Rouse the yeast gently (a swirl, not a vigorous shake that risks oxidation), raise the fermentation temperature a few degrees if it's been running cool, and wait another 3-5 days before reading again. If a second rousing and a warm-up genuinely produce no further gravity drop across a week, the batch has attenuated as far as it's going to — package it at that gravity rather than waiting indefinitely, and note the strain and pitch rate for next time. A beer that finishes a bit sweeter than planned is still good beer; one that's primed for a CO2 target assuming a lower FG than it actually reached is the real risk — see what happens if I use too much priming sugar before bottling a batch you're not fully sure has stopped fermenting.