Shepherds Warning

About This Beer

A Belgian red on the Gladfield red malts, Aurora, Redback and Shepherds Delight, over an all-Pilsner base, no candi syrup. Brewed in Lower Hutt on 27 August 2026.

Those malts normally sit behind a wall of hops in a red IPA. Here they lead, with a yeast chosen to get out of the way.

SafAle BE-256 is POF-negative: no clove, no pepper, just the banana and pear of isoamyl acetate. The dark fruit here comes off the Shepherds Delight rather than the yeast, so nothing the yeast makes competes with it. The catch is that it also removes a Belgian ale’s usual source of spice, so the spice gets built instead. Crystal Rye for pepper and nutty caramel, Celeia for the noble herbal character.

Mandarina Bavaria is the third note, a Cascade daughter that went mandarin where its parent went grapefruit, near enough the dried Curaçao peel Belgian brewers reach for. Luminosa underneath it for candied peel and dark berry.

Red, not brown

No chocolate malt, no roasted barley, no black malt. Every colour-contributing grain is a red-hue Maillard malt, and Gladfield build two of them, Redback and Redback Wheat, to add colour without haze.

Redback Wheat is the quiet one. Huskless, so colour without the harshness husk brings at roast colours, and its low modification lifts body and head retention. A quarter kilo doing three jobs is what lets Shepherds Delight stay down at 1.8%.

The numbers

Grain milled at a friend’s place rather than crushed at the shop, and visibly finer: more starch exposed, more complete conversion. With a 6.8 kg grist and 23 litres to put it into, a strong wort was the expected result.

Bitterness-to-gravity of 0.50, most of it from the whirlpool rather than the kettle, so it should land soft rather than sharp.

Seventy minutes of boil, held hard and rolling throughout: 31 litres down to 25 or 26, near five litres an hour on this rig.

Eighteen SRM, thirty-six IBU, 1.073. A strong Belgian red.

The ferment

Pitched at three in the afternoon at 24°C into the 30 litre FermZilla, down to the 21°C start within two hours.

A full day before anything showed, then krausen by seven the next evening and 1.062 by late that night. Eighteen points overnight after that, and a peak of 27.9°C by nine the following morning. Target back to 26.5°C to hold it under twenty-seven, and five psi on the valve.

Most of the work was done inside three days and the curve was flat by the fourth. Another four days warm on the yeast after that, and it finished at 1.012 for eight percent, two points off where it was expected to land.

Then down slowly and on its own. The cooling was switched off and the setpoint walked down by hand instead, twenty degrees, then seventeen, then thirteen, then nine, four days of twenty-three litres coasting rather than being pulled. A crash to three and a half overnight after that, chitosan in the following morning, then a day cold before it moved.

One note on the trace below. It ends six points lower than the beer did, and the beer did not do it. A floating hydrometer stops telling the truth once it is a long way from the temperature it was calibrated at, so the number that counts came off glass at room temperature with the gas knocked out of the sample.

Kegged

Closed transfer under gas on 10 September, fourteen days from pitch. Nineteen litres run full into the keg and the trub left where it was, with a litre and a half drawn off for tasting either side of it. No sulphur anywhere, which was the thing most worth checking on a beer held under pressure from day two.

Three or four weeks from here is a floor rather than a target. At this strength it is the alcohol that needs the time, folding the heat back into the malt, and a beer this size usually keeps getting better well past a month.

The hops pull the other way. A hundred and fifty grams went in at whirlpool, and that aroma fades over exactly the months the malt and the alcohol improve across. It was never meant to arrive as hop character, and it has not. Mandarina was picked to do the job dried Curaçao peel does in a Belgian, which is to join the fruit rather than stand over it. So the question at four weeks is whether it grows as the alcohol settles, or quietly goes.