Blueberries and honey, with an optional cinnamon, vanilla and oak trio at the aging stage that is where the name comes from. Ethan's blueberry batch was made with canned pie filling; what is written here is the frozen-fruit version, which is a different build and says so throughout.
Makes. 5 gallons of must, with the fruit on top of that rather than counted in it. What reaches a bottle is less — see the note on what “5 gallons” means.
Vessel. Open-top bucket of about 7.9 gallons, or two 6.5 gallon bucketsSame arithmetic as the cherry: 5 gallons of must under 15 lbs of berries is close to 7 gallons of contents, and blueberry skins make a thick cap.
Yeast. Lalvin 71B, one 5 g sachet. Lallemand's data sheet puts it at 20–40 g/hL, which over 5 US gallons is 3.8–7.6 g, so one sachet is about 26 g/hL and the two often suggested come to roughly 53 g/hL — past the top of the label rate. Ferment between 15 and 30 °C (59–86 °F); alcohol tolerance is about 14% v/v. Lalvin ICV D47 is the usual alternative on its own terms: 20–40 g/hL, 15–28 °C (59–82 °F), tolerance up to 15% v/v.
Water. Spring water if you can get it, otherwise whatever you have. Berries and honey go in first, then top up to the 5 gallon must line — the must line, not the rim, and not counting the fruit.
Scaling. All of these scale down. A single gallon bottle of fresh apple juice with some of it swapped out for honey will get you a small batch, no special kit required.
Not substitutions. Each one is a different drink, with its own sugar, its own volume and its own vessel.
15 lbs of thawed frozen blueberries in a bag. Complete, costed and reproducible.
Canned blueberry pie filling in place of the fruit. It works, and it brings three problems the frozen-fruit route does not have.
Wash everything the must will touch, then sanitise it with a brewing sanitiser at the concentration and contact time on its own bottle — bucket, lid, airlock, spoon, hydrometer, trial jar, bag and racking tubing. Sanitiser is not an ingredient and none of it goes into the must.
Thaw the blueberries and put them in the sanitised straining bag, then settle the bag in the bucket. Warm the honey only enough to pour it, dissolve it into a few gallons of warm water, and pour that over the fruit. Stir until nothing is sitting on the bottom.
Top up to the 5 gallon must line, then stir hard — honey sinks, and an unmixed must gives a reading that is true of the top inch only. Draw a sample, read it, note the temperature, and write both down. Every later reading is compared against this one, so it is the only reading that cannot be retaken.
By the sheet for the product in your hand. 71B without a protector: ten times its weight of clean 35–40 °C water, gentle stir, 20 minutes, then bring the slurry to within 5–10 °C of the must with a little must at a time and inoculate, inside 45 minutes total. With Go-Ferm Protect Evolution, use that sheet instead — 43 °C water and the yeast into 37 °C.
Estimate About half an hour, and never more than the 45 minutes the sheet allows
Hold the temperature inside the range for the yeast you pitched — 15–30 °C for 71B, 15–28 °C for D47 — and nearer the middle than the top. Feed on your nutrient's own schedule. Blueberry skins raise a firm cap: push it under once or twice a day with a sanitised spoon, both because a dry cap grows mould and because the colour is in the skins and only comes out while they are wet.
Estimate Active fermentation commonly runs a couple of weeks. The cap and the hydrometer decide, not the estimate
Lift the bag once the vigorous phase has died down and let it drip rather than wringing it. Rack off the gross lees into a vessel the mead fills, under an airlock. Blueberries throw a lot of sediment, so expect to want a second racking later.
Estimate Fruit usually comes out somewhere around the two-week mark
These go in at aging, not into the ferment, and they do not finish together. Cinnamon is usually the fastest and can go too far in a few days; oak cubes are slow; vanilla sits somewhere between. Taste every few days from a sanitised thief, and take each one out when it reaches the strength you want without waiting for the others — which is far easier if each is in its own small mesh bag rather than loose in the carboy. Nothing comes back out once it has gone too far.
Estimate Cinnamon is often days, vanilla and oak weeks. Taste, do not count
Bulk age under airlock with as little headspace as you can manage, cool and dark. When you think it is done: hydrometer reading, another two or three days later, both corrected for temperature. Equal readings mean the fermentation is over. Equal readings well above where the yeast should have landed mean a stall, not a finish, and bottling that makes bombs. Neither a quiet airlock nor a clear mead is evidence of anything.
Estimate No estimate. This step is decided by a measurement
Bulk age under an airlock with minimal headspace, adding and removing the optional spices and oak by taste along the way, and bottle only once repeated gravity readings agree. Several months is normal; the figure is an estimate about flavour and is never what makes it safe to bottle.
Fermentation produces carbon dioxide, and a sealed or blocked vessel is a pressure vessel — the Valentine's Brew blocked its own blow-off tube. These worked for the brewer, which is not the same as working with a different vessel, yeast and kitchen.
Manufacturer data sheets and published references, each one opened and read. The technical figures on this page come from them. The amounts of honey and fruit do not.
What this recipe depends on you getting right
“Finished” is a hydrometer reading, not a quiet airlock
An airlock goes quiet for reasons that have nothing to do with sugar — a lid that is not seated, a cold night, a cap of fruit lying over the surface — and a mead can be brilliantly clear with sugar still in it. Clarity is not a measurement either. The hydrometer is the only thing that answers the question. The American Homebrewers Association's rule is a reading, then another, and if they are the same the ferment is done; on a mead, which works slowly, leave two or three days between them rather than one, and correct for temperature before comparing. A reading that has stopped moving is still not automatically a finish, because a stall looks the same on the scale — it sits well above where the yeast should have landed and stays there. Telling them apart means knowing roughly where your yeast can get to, which is the alcohol tolerance on its data sheet and is repeated in the yeast line further up this page. A must built past that tolerance is EXPECTED to stop short, and that is a ceiling rather than a fault. A must built well under it that stops anyway is the one to worry about.
Every time on this page is an estimate
None of the durations here is a schedule and no date decides anything. How long a mead takes moves with temperature, pitch rate, nutrient and how much fruit is sitting in the vessel, and two people following this to the letter will not finish in the same month. What ends a step is the condition described in it. What allows bottling is repeated stable gravity readings — never the calendar, never clarity, and never a silent airlock.
What “5 gallons” means here, and what you will actually bottle
Five gallons is the target volume of MUST — the honey and the liquid, measured before whole fruit goes in. Fruit is not part of that number and adds to it: 15 lbs of fruit takes up roughly the volume of its own weight in water, about 1.8 gallons, so a 5 gallon must under 15 lbs of fruit is close to 7 gallons of contents before fermentation lifts a cap above it. That is arithmetic on the weight rather than a measurement — frozen berries pack with air gaps and take up less until they thaw and break down. What reaches a bottle is less than 5 gallons in every case: the fruit holds liquid that leaves with the bag, every racking leaves some behind on the lees, and every hydrometer sample is another cup gone. Build for 5 gallons of must and expect to bottle under it. If 5 gallons in the bottle is the point, scale the whole recipe up rather than squeezing the bag harder.
Measure the fermenter rather than trusting its name
The other meads here ferment 5 gallons in an open-top bucket of about 6.5 gallons, and that air gap is what keeps them out of trouble. That vessel will not take this batch. Five gallons of must plus roughly 1.8 gallons of fruit is about 6.8 gallons of contents, which is over the rim of a 6.5 gallon bucket before there is any cap or foam at all. Either use a larger primary — 7.9 gallon fermenting buckets are a standard size — or split the batch across two 6.5 gallon buckets and combine them at the first racking. A vessel sold as a “6 gallon bucket” is named for what it is meant to hold rather than for where its rim is, so fill yours from a measuring jug once and write the real number on the side. The same caution applies going the other way: these scale down, but the FRUIT has to scale down further than the honey does. Craft a Brew puts a 1 gallon carboy at 1 to 2 lbs of fruit per gallon rather than 3, precisely because it has no room for a cap.
Fermaid O and Fermaid K are different products with different schedules
Both are yeast nutrients and neither one follows the other's dosing. Lallemand's sheet for Fermaid O gives 2 × 20 g/hL — 20 g/hL at the start of alcoholic fermentation and 20 g/hL at a quarter to a third of the way through — which is about 3.8 g per addition in 5 gallons, 7.6 g in total. Fermaid K is 25 to 50 g/hL with a stated maximum of 50 g/hL, split at the end of the lag phase and again at one-third sugar depletion: about 4.7 g to 9.5 g in total for 5 gallons. Either one is suspended in ten times its weight of water or must before it goes in. Fermaid K's sheet addresses honey by name, and what it recommends for a poor fermentation medium is heavier than its general rate — 4 lbs per 1000 gallons of Fermaid K with 2 lbs per 1000 gallons of yeast hulls and 6 lbs per 1000 gallons of DAP, fed incrementally from pitching onward, which is about 9.1 g, 4.5 g and 13.6 g in a 5 gallon batch. Mead-specific feeding protocols circulated in the hobby go higher still, on the argument that honey carries almost no assimilable nitrogen of its own; that is a deliberate departure from the label rather than the label, and the figures here are the manufacturer's. Pick one product and follow that product's own sheet.
Backsweetening: stabilise first, and sorbate alone will not do it
If it finishes drier than you want, the honey does not simply go into the bottle. Virginia Tech's enology extension is blunt about what potassium sorbate is: it “does not kill yeasts but, if properly employed, inhibits their growth”. It cannot stop a fermentation that is still running, which means it is only meaningful once repeated hydrometer readings say the ferment is genuinely over — the two halves of this page people most often skip are the same half. It also does not work alone: lactic acid bacteria growing in the presence of sorbic acid produce 2-ethoxyhexa-3,5-diene, the “geranium” taint, and that source puts the free sulphur dioxide needed at bottling to prevent it at approximately 30 mg/L. The rate given is 150 to 300 mg/L of sorbic acid, and potassium sorbate weighs 1.34 times the sorbic acid level being aimed at — 201 to 402 mg/L, or roughly 3.8 g to 7.6 g in 5 gallons. How far up that range you need to go moves with pH, alcohol and free SO2, and the same source advises against sorbate altogether in anything intended for long aging. Stabilise, wait, then sweeten a measured sample and taste it before treating the batch.