MAD BREWS
DAY 00 SG 1.090 THE ARCHIVE
THE DIARY, SINCE AUGUST 2022

Every batch here was actually logged.

Twenty-four batches are public. The first was started on 6 August 2022; the most recent came off its fruit in April 2026. These are diary entries, not recipes tidied up afterwards.

Each row below is one vessel. What went into it, how much water, the yeast if I wrote one down, the day it started, and every hydrometer reading I actually took. Nothing has been rounded into shape, and nothing has been added to make a row look complete. It is fruit wine nearly all the way down — bananas, pineapple, grapes, blueberries, mandarins, grapefruit, mango, pomegranate, plum — with one mead and one cider in it.

The curve on each plate is drawn only through readings that exist. Two readings means two points and a straight line between them, because I do not know what happened in between and I will not draw a guess. Some of these batches have three readings spread across six months. That is the entire record, and it is more than I kept for most of them.

Several rows have no yeast strain against them at all. The ones that do name biovin RC 212, Lalvin Bourgovin RC 212 or Lalvin 71B, and nothing else. Where the book is silent, this page is silent — so if you want to change a quantity rather than copy one, the calculators will do the arithmetic, and the yeast guide will tell you what a strain can actually survive.

Two rows are marked terminated: №4 and №2, both banana, both from the first day in the book. They are here on purpose, with the reading that went the wrong way still on the plate. A batch that failed with a cause written under it is worth more than a photograph of a full bottle.

Glass demijohn under airlock, cross-section 10 L HEADSPACE FIG. 3 — DEMIJOHN
PLATE BDEMIJOHNDRAWING
NOTHING IS FERMENTING RIGHT NOW

The demijohns are empty. №33 came off the pineapple and went onto the garage floor to clear, and that is the newest thing in the book. When something new starts it will appear here with its first reading, on the day the reading is taken, and not a day before. There is no live batch on this site because there is no live batch in the garage.

DAY 98 SG 0.998 THIRTEEN ROWS

Thirteen batches, and the two that never made alcohol.

Newest first. Fruit wine almost the whole way down, one mead, one cider, and at the bottom the two demijohns from 6 August 2022 that did not work — kept in place, with their readings, rather than quietly dropped.

0.9801.0001.0201.0401.0601.0801.100135.16.111.312.213.214.119252729BATCH № · START GRAVITY DOWN TO WHERE IT ENDED
PLATE EEVERY BATCH24 BATCHES2 CLIMBED

ALL 24 PUBLISHED BATCHES · 13 STOPPED EXACTLY ON 1.000 · 3 WENT FULLY DRY · 2 CLIMBED INSTEAD OF FALLING

0.9911.0181.0451.0721.099D0D49D98
PLATE ABATCH №33PINEAPPLE WINE1.090 → 0.99898 D12.1 %ABV

The newest thing in the book and the biggest vessel in it: 9KG of pineapple in 25L, left sitting on the fruit the whole way rather than strained off early. It is still clearing.

WHAT WENT IN
  • 9KG pineapple
  • 25L water
  • biovin RC 212
  • started 24 December 2025
THE READINGS
  • D0 — 1.090
  • D98 — 0.998

Two readings, ninety-eight days apart. The line between them on the plate is straight because that is the honest shape of what I know.

FROM THE DIARY

“It spent the whole time on the pineapple, unstrained. There is a bitterness to it. Put on the garage floor — it is clearing fairly fast. After it clears I will have to warm it and work out what to add, or how to knock that sharpness back. The smell so far is very good.”

YEASTbiovin RC 212
OG → FG1.0900.998
VOLUME25L water, 9KG fruit
TIME98D
STRENGTH12.1%ABV
WHERE IT ISAging. Clearing on the garage floor.
0.9961.0221.0471.0731.099D0D51D102
PLATE ABATCH №29BANANA MEAD1.092 → 1.010102 D12.1 %ABV

The only mead in the archive, and it is a banana one. Ten kilos of bananas into twenty-seven litres with Lalvin 71B — by a wide margin the largest banana batch I have run, and the one that worked. The two banana batches that did not are at the bottom of this page; two more from that same first day are still aging.

WHAT WENT IN
  • 10KG bananas
  • 27L water
  • Lalvin 71B
  • started 24 September 2025
THE READINGS
  • D0 — 1.092
  • D30 — 1.000
  • D102 — 1.010, at bottling

The day-30 reading is lower than the day-102 one. Both are in the book, so both are on the plate; I am not going to publish only the tidier of the two.

FROM THE DIARY

“It did not clear all the way, but it is OK.”

YEASTLalvin 71B
OG → FG1.0921.010
VOLUME27L water, 10KG fruit
TIME102D
STRENGTH12.1%ABV
WHERE IT ISBottled, not fully clear.
0.9961.0101.0251.0401.054D0D72D144
PLATE ABATCH №26PLUM WINE1.050 → 1.000144 D6.6 %ABV

Plum, five litres, one of a run of small batches started in spring 2025 and treated the same way as the rest of them. It racked clean and then disagreed with itself in the book.

WHAT WENT IN
  • plums — weight not written down
  • 5L water
  • no yeast strain recorded
  • started 10 April 2025
THE READINGS
  • D0 — 1.050
  • D30 — 1.000
  • D144 — 1.000, unchanged

Dry inside a month, then a hundred and fourteen days of the hydrometer saying exactly the same thing.

FROM THE DIARY, AT RACKING

“It racked off nicely. Everything is clean, the smell is not much good at all.”

AND AT TASTING

“You can taste the plum — but I could not tell you clearly what is going on with it. The smell is normal. The taste is thin. It lacks tannin, it lacks depth. You can tell it is not strong at all. I would want it a bit sweeter.”

One note says the smell is poor, the other says it is normal. They were written months apart and I have left both of them standing.

YEASTNot written down
OG → FG1.0501.000
VOLUME5L water; fruit weight not recorded
TIME144D
STRENGTH6.6%ABV
WHERE IT ISBottled.
0.9961.0091.0221.0361.049D0D88D175
PLATE ABATCH №25POMEGRANATE WINE1.042 → 1.000175 D5.5 %ABV

Pomegranate, five litres, started the same day as the mango and carrying the happiest note in the whole archive — followed, six months later, by exactly the same complaint as everything else from that spring.

WHAT WENT IN
  • pomegranates — weight not written down
  • 5L water
  • no yeast strain recorded
  • started 10 March 2025
THE READINGS
  • D0 — 1.042
  • D31 — 1.000
  • D175 — 1.000, unchanged

It starts at 1.042, which is a low start, and it finishes at 5.5%ABV, which is what a low start pays out.

FROM THE DIARY, AT RACKING

“Everything is great. It cleared beautifully, the taste is good, the smell is good.”

AND AT TASTING

“The pomegranate is there. The smell is cosmic. The taste is thin. It lacks tannin, it lacks depth. You can tell it is not strong at all.”

YEASTNot written down
OG → FG1.0421.000
VOLUME5L water; fruit weight not recorded
TIME175D
STRENGTH5.5%ABV
WHERE IT ISBottled. Cleared well.
0.9961.0101.0251.0401.054D0D88D175
PLATE ABATCH №24MANGO WINE1.050 → 1.000175 D6.6 %ABV

The best-smelling thing I have made, by my own note, and still thin in the glass. Same five litres and the same spring week as the pomegranate, and the same short list of things done to it.

WHAT WENT IN
  • mangoes — weight not written down
  • 5L water
  • no yeast strain recorded
  • started 10 March 2025
THE READINGS
  • D0 — 1.050
  • D31 — 1.000
  • D175 — 1.000, unchanged
FROM THE DIARY, AT RACKING

“A great deal of lees left behind, and only 50% yield.”

AND AT TASTING

“The mango is there — the best I have ever tried to make. The smell is cosmic. The taste is thin. It lacks tannin, it lacks depth. You can tell it is not strong at all.”

Those two sentences belong together: the best aroma in the archive, and half the liquid I started with still sitting in the solids.

YEASTNot written down
OG → FG1.0501.000
VOLUME5L water; fruit weight not recorded
YIELD50% — the rest stayed in the fruit
TIME175D
STRENGTH6.6%ABV
0.9910.9991.0081.0161.024D0D47D94
PLATE ABATCH №14STRAWBERRY CIDER1.020 → 1.00094 D2.9 %ABV

Twenty kilos of strawberries into fifteen litres with Lalvin 71B, split across two vessels. Half the fruit went in at the start and half went in later; this is that half-and-half one. Its twin №14.1 took the rest of the strawberries plus about a kilo of pineapple, with five litres pulled off purely as a test.

WHAT WENT IN
  • 20KG strawberries, half at the start and half later
  • 15L water
  • Lalvin 71B
  • started 28 March 2023
THE READINGS
  • D0 — 1.020
  • D27 — 0.998
  • D94 — 1.000, at bottling

1.020 is a cider start rather than a wine start, and 2.9%ABV is what came out of it.

FROM THE DIARY

“The second half of the strawberries went in. 50% in the primary and the other 50% now. One bottle left completely dry — hard cider, just to taste it.”

TASTING, BOTH HALVES

14: “The taste is lovely when it is sweet.” №14.1: “Unreal when it is sweet — actual strawberry jam.”

HOW №14.1 ACTUALLY ENDED

Both of those notes were written at bottling. №14.1 did not keep: the taste turned, and the whole of it went down the sink rather than into a glass. The ferment itself was never the problem — 1.020 to 0.998 in 27 days is a clean cider, and the gravity behaved all the way through. A good ferment and a good drink are not the same thing, and this is the batch that taught that.

YEASTLalvin 71B
OG → FG1.0201.000
VOLUME15L water, 20KG fruit
TIME94D
STRENGTH2.9%ABV
SPLIT14 and №14.1 out of one primary
0.9861.0141.0421.0711.099D0D48D95
PLATE ABATCH №13.1BLUEBERRY WINE1.090 → 1.00095 D13.1 %ABV

Eight kilos of blueberries in eight and a half litres, fermented as one lot and then split into two secondaries so that a question could actually be answered. This half got about 50G of American oak. The other half is the next row down.

WHAT WENT IN
  • 8KG blueberries
  • 8.5L water
  • about 50G American oak, in the secondary
  • no yeast strain recorded
  • started 22 February 2023
THE READINGS
  • D0 — 1.090
  • D21 — 0.990
  • D95 — 1.000, at bottling
FROM THE DIARY

“Clearing. It has about 11–12 degrees in it. Five litres left in the secondary. Not sweetened. About 50G of American oak added on top.”

My own note guesses 11–12%. The readings give 13.1%ABV. The readings are what is on the plate.

AND AT TASTING

“The oak is very present — it overpowers every other flavour. We will see how it ages. Not much so far.”

YEASTNot written down
OG → FG1.0901.000
VOLUME8.5L water, 8KG fruit
VARIABLEAmerican oak, about 50G
TIME95D
STRENGTH13.1%ABV
0.9861.0141.0421.0711.099D0D48D95
PLATE ABATCH №13.2BLUEBERRY WINE1.090 → 1.00095 D13.1 %ABV

The other half of the same must. No oak; 400G of frozen shop-bought blueberries instead. One primary, two secondaries, one thing changed between them — this is the only genuine side-by-side in the archive, and it is the reason I can say what the oak did rather than guess at it.

WHAT WENT IN
  • 8KG blueberries
  • 8.5L water
  • 400G frozen shop-bought blueberries, in the secondary
  • no yeast strain recorded
  • started 22 February 2023
THE READINGS
  • D0 — 1.090
  • D21 — 0.990
  • D95 — 1.000, at bottling

The two plates are identical because it was one ferment. Everything that separates these two rows happened in the secondary, after the split.

FROM THE DIARY

“Clearing. About 11–12 degrees in it. Five litres left in the secondary. Not sweetened. 400G of frozen shop-bought blueberries added.”

AND AT TASTING

“Better than the one with the oak, but it has the bitterness blueberries always carry. Overall not bad though; back-sweetened, it should be good in a year or two.”

YEASTNot written down
OG → FG1.0901.000
VOLUME8.5L water, 8KG fruit
VARIABLE400G frozen blueberries instead of oak
TIME95D
STRENGTH13.1%ABV
0.9861.0101.0351.0601.084D0D73D146
PLATE ABATCH №12.1GRAPE WINE1.080 → 1.000146 D11.8 %ABV

Ten kilos of grapes into six litres, malolactic bacteria added before the first racking, then racked into glass in the garage with a little simple syrup. Its twin №12.2 came out of the same primary and got pea blossoms for colour, which turns out to be the useful part of the story.

WHAT WENT IN
  • 10KG grapes
  • 6L water
  • malolactic bacteria, before the first racking
  • a small amount of simple syrup
  • started 11 January 2023
THE READINGS
  • D0 — 1.080
  • D9 — 0.990
  • D146 — 1.000, at bottling

Nine days from 1.080 to 0.990. Grapes carry their own sugar, and it shows in how fast this one moved.

FROM THE DIARY

“Malolactic bacteria added before the first racking. Racked in the garage into glass bottles. A small amount of simple syrup added.”

THE TWO HALVES, TASTED

12.1: “Sweet with nothing added at all, even though they are dry.” №12.2, the one with the blossoms: “The blossom tea added a strange flavour that I am not really keen on. It is not unpleasant, but the plain grape ones are much nicer without it.”

YEASTNot written down
OG → FG1.0801.000
VOLUME6L water, 10KG fruit
TIME146D
STRENGTH11.8%ABV
TWIN12.2, same primary, plus pea blossoms
0.9761.0001.0251.0501.074D0D34D67
PLATE ABATCH №11.3MANDARIN WINE1.068 → 0.98067 D10.2 %ABV

Ten kilos of mandarins into ten litres, started on the last day but one of 2022, with peppermint tea leaves added. It finished at 0.980 — the driest finish anywhere in this archive, and lower than water.

WHAT WENT IN
  • 10KG mandarins
  • 10L water
  • peppermint tea leaves
  • no yeast strain recorded
  • started 30 December 2022
THE READINGS
  • D0 — 1.068
  • D9 — 0.990
  • D67 — 0.980, at bottling
FROM THE DIARY

“Peppermint tea leaves added.”

AND AT TASTING

“The peppermint is very present. Curious what it will be like in a year or so. Tastes decent right now.”

Bone dry, with a strong flavouring sitting on top of it. This is the row I most want to open again in a year or two, and the one I most regret not writing more about.

YEASTNot written down
OG → FG1.0680.980
VOLUME10L water, 10KG fruit
TIME67D
STRENGTH10.2%ABV
NOTEDriest finish in the archive at 0.980
0.9961.0181.0401.0621.084D0D3D5
PLATE ABATCH №5.1GRAPEFRUIT WINE1.080 → 1.0005 D10.5 %ABV

Grapefruit, October 2022, run in two vessels at once. This is the fifteen-litre one, with 100ML of sugar syrup; the twin №5.2 was a five-litre vessel with 50ML of syrup and malolactic bacteria. In the end they were blended together anyway.

WHAT WENT IN
  • 48KG grapefruit, as logged across the pair
  • 4L water
  • 100ML sugar syrup
  • no yeast strain recorded
  • started 7 October 2022
THE READINGS
  • D0 — 1.080
  • D5 — 1.000

Five days, two readings, and then nothing written down until the bottles. Whole citrus in a small volume of water moves quickly.

FROM THE DIARY

“Fifteen-litre vessel with 100ML of sugar syrup.”

AND AT TASTING

“Semi-dry. The bitterness is interesting, but bearable. Clarity I would call acceptable, though certainly not clear. I mixed the five-litre and the big vessel together and got twenty bottles.”

YEASTNot written down
OG → FG1.0801.000
VOLUME4L water, 48KG fruit as logged
TIME5D
STRENGTH10.5%ABV
BOTTLED20 bottles, blended with №5.2
1.0561.0671.0781.0881.099D0D7D14
PLATE ABATCH №4BANANA WINE1.064 → 1.09014 DTERMINATED

One of four banana demijohns filled on 6 August 2022, the first day in the book. Two kilos of bananas in two and a half litres, pitched with beer yeast, and it stopped.

WHAT WENT IN
  • 2KG bananas
  • 2.5L water
  • beer yeast — the note names it, no strain is recorded
  • added sugar
  • started 6 August 2022
THE READINGS
  • D0 — 1.064
  • D14 — 1.090, higher than the start

The reading went the wrong way. There is no ABV printed for this row because it never fermented, and a negative percentage would be a lie dressed up as a measurement.

FROM THE DIARY

“It looks as though the beer yeast could not stand the alcohol concentration and that was the end of it. We will see whether it wakes up for a further ferment or drives itself to vinegar. Right now it is like a syrup with a sour edge. Not much. Really not much. Something to think about — perhaps it is worth trying this one without added sugar?”

YEASTBeer yeast, strain not recorded
OG → LAST READING1.0641.090
VOLUME2.5L water, 2KG fruit
TIME14D before it was written off
STRENGTHNone. It never made alcohol.
CAUSEBeer yeast in a wine must. It could not survive the alcohol and quit.
1.0061.0101.0151.0201.024D0D7D14
PLATE ABATCH №2BANANA WINE1.010 → 1.02014 DTERMINATED

Another of the four from that first afternoon. Extra sugar went in, and what came out was vinegar. It started at 1.010, which is barely above water, and read 1.020 a fortnight later.

WHAT WENT IN
  • 2KG bananas
  • 2.5L water
  • extra sugar
  • no yeast strain recorded
  • started 6 August 2022
THE READINGS
  • D0 — 1.010
  • D14 — 1.020, higher than the start

No ABV: it never made any. The plate is captioned TERMINATED for the same reason.

FROM THE DIARY

“Extra sugar added. Actual vinegar — or rather, just not nice. Super dry. Simply no good.”

1 and №3 came out of the same afternoon and the same fruit and are both still aging. Four vessels, one day, two failures — and the only reason I can say anything at all about why is that all four of them were written down.

YEASTNot written down
OG → LAST READING1.0101.020
VOLUME2.5L water, 2KG fruit
TIME14D
STRENGTHNone. It never made alcohol.
CAUSEIt went to vinegar. Sour, bone dry, undrinkable.
DAY 175 SG 1.000 THE LONG WAIT
READING THE ARCHIVE SIDEWAYS

What this archive keeps getting wrong.

Yield first. №24 says it plainly — a great deal of lees and only 50% yield — and №27, the banana, carrot and sea buckthorn one, says the same thing in different words: half of it stayed in the solids. That is what fruit wine does without a press. The fruit keeps a large share of the liquid and you rack a good deal less than you thought you had.

Then thinness. Three rows from spring 2025 — №26, №25 and №24 — end with almost the same sentence written three separate times: thin, lacks tannin, lacks depth, clearly not strong. Same five litres, same short list of interventions, same result. That is not three tasting notes. That is one problem recorded three times, and the calculators are where I go to start fixing it — a start of 1.042 was never going to give me depth.

And the one thing the archive did answer properly. №13.1 and №13.2 came out of a single blueberry ferment, split in two: fifty grams of American oak in one, 400G of frozen berries in the other. The oak overpowers everything else in the glass. I only know that because the split existed — one variable, two vessels, and both of them tasted on the same evening.

If you are starting out, start where the fruit does the work. №6.1 went from 1.072 to 1.000 in four days on grapes and a single litre of water, and the note about it is the shortest happy note in the book. Only three strains are named anywhere in this diary — biovin RC 212, Lalvin Bourgovin RC 212 and Lalvin 71B — so the yeast guide is about what to buy next, not about what I have already used.

Hydrometer floating in a test jar SG 1.062 AIR FIG. 2 — HYDROMETER
PLATE BHYDROMETERDRAWING