How do you fix a wort that missed its gravity?
One idea does all the work: gravity points × volume doesn't change in the boil. Boiling drives off water, not sugar, so 1.042 in 7.5 gallons is 315 "point-gallons" whether you boil it down to 6 or top it up to 8. Every fix is a rearrangement of that product:
points × volume = constant · too strong → add water · too weak → boil longer, or add points (extract)
| Fix | Example | What else it changes |
|---|---|---|
| Dilute | 1.060 in 5 gal → 1.050: add 1 gal of water, ending at 6 gal | IBUs and color drop by the same 17%; body a shade lighter. Free, instant, more beer. |
| Boil longer | 1.042 in 7.5 gal → 1.048: boil off 0.9 gal — about 45 min more at 1.25 gal/hr | Concentrates hops and color too (a bit more bitter, a bit darker), costs propane and an hour, and you end with less beer. |
| Add extract | 1.044 in 5.5 gal → 1.050: 0.75 lb (340 g) of DME | Keeps the volume and the hop schedule; adds a little unfermentable body and pale color. Costs a bag of DME you should have in the cupboard. |
Default US profile boil-off rate (1.25 gal/hr); DME at 44 PPG. From this page's engine — the calculator uses your profile's rate.
Wrong gravity and wrong volume: do the volume first
The common brew-day case isn't "gravity is off" — it's "I have 1.042 in 21 liters and the recipe wants 1.050 in 19". Fixing the gravity alone (boil to 1.050) would leave you at 17.6 L; fixing the volume alone (boil to 19 L) leaves you at 1.046. So hit the volume with the boil, then hit the gravity with extract: boil off 2.0 L (about 27 minutes at 4.5 L/hr), land at 1.046, and stir in 185 g of DME for the last 3.6 points. Enter a target volume in the calculator and it works this out for you — including the opposite case, where the wort is too strong and too plentiful, and the answer is simply "dilute, don't boil".
Take a gravity reading as soon as the kettle is full — cooled, corrected — and compare it with the pre-boil gravity your equipment profile predicts (Total Water and Expected OG both show it). Every fix is cheaper before the boil: extract dissolves and sanitizes itself, extra boiling costs only time, and dilution water gets boiled with everything else. After the boil you can still add boiled-and-cooled water or a DME syrup you've boiled separately, but you're now handling cooled wort, and the hop utilization is already set.
Which fix should you choose?
Too strong: dilute. It's free, it makes more beer, and the beer you diluted was over-bittered by the same ratio, so the IBUs come back to plan too. Only boil less if you haven't started yet. Too weak by a couple of points: extract, if you have it — the volume and the hop additions stay exactly as planned. If you don't, boil longer and accept slightly more bitterness and a smaller batch; you can add the extra hops' worth of water back later, but you can't un-boil. Too weak by ten points: ask why before you fix it (a stuck sparge, a mis-measured volume, a bad crush) — then extract, because that much extra boiling turns a pale ale amber.
Your hydrometer is good to about 0.001, and one gravity point is about 0.13% ABV. A 1.048 wort that "should" have been 1.050 is the same beer with a rounding error, and correcting it will introduce more variation than it removes. Fix misses of three or more points, or anything that also broke your volume; log the small ones in your efficiency and move on. The pre-boil number in your profile is a target, not a promise.
Why gravity rises during the boil — and by how much
Nothing about the sugar changes; the water leaves. On the default US profile the kettle boils off 1.2 gal in an hour from 7.5 gal, so a 1.042 pre-boil wort finishes around 1.050 without any help. That is the whole reason "pre-boil gravity" and "OG" are different numbers, and why a longer boil is a legitimate gravity tool: each extra 1.25 gal evaporated is worth roughly a point and a half on a 6-gallon batch. It's also why kettles differ — a wide, shallow pot on a big burner can boil off twice what a tall narrow one does — and why the rate in your profile should be a measurement, not a guess.