Drawing a sample without spoiling the batch

A wine thief is a tube — glass, stainless or a rigid food-grade polymer — long enough to reach the depth of interest. The simplest kind is open at both ends: it is lowered, fills, and is closed with a thumb or a stopper over the top so that atmospheric pressure holds the column in while it is lifted out. Better ones have a valve at the tip that opens on contact with the bottom or is operated from the top, which lets a sample be drawn from a chosen depth rather than from wherever the tube happened to fill.

A cask thief or valinch is the same idea sized to pass through a bung hole, and it is the only way into a closed cask that does not involve opening it up. At demijohn scale a long pipette or a large syringe with a length of tubing does the same job and takes a much smaller sample, which matters more than it sounds.

A sample valve on a tank is the arrangement to prefer wherever it exists, because nothing enters the vessel at all. It has one important characteristic: it is a dead leg, so the cider standing in it between uses is not representative and is very often not clean. A sample valve is flushed to waste before the sample is taken, and it is cleaned and sanitised by hand rather than by any clean-in-place cycle.

A vessel of cider is not one liquid. A fermenting tank has yeast in suspension throughout and more of it lower down; a settled tank has a gradient of solids and a lees bed; a tank that has just been dosed or blended has whatever was added sitting where it was put until something mixes it. A sample from a bottom valve and a sample from the top of the same tank can differ enough to change a decision.

The practical rules follow directly. Take from the same place every time if the readings are to be compared, and write down where. Flush a valve first so the sample is cider from the tank and not cider from the pipe. After an addition or a blend, circulate or wait before sampling, because a sample taken too early describes the top of the tank rather than the batch. Before bottling — the moment when the number really matters — mix the vessel properly and sample from the middle.

For gravity specifically, consistency of place beats accuracy of place. A series of readings all taken from the same valve at the same depth tells the truth about the trend even if every one of them is a little off; a series taken from wherever was convenient tells nothing at all.

Going in, everything that enters the cider is a route for organisms: the thief, the tubing, the pipette, a hand, a sleeve. All of it is cleaned and sanitised before it goes in, and then allowed to drain, because sanitiser carried into a small vessel is both a dilution and, at the wrong concentration, a taint. On a demijohn this is not a trivial fraction.

Going out is the rule that gets broken. Cider that has been in a trial jar with a hydrometer in it, in a glass on a bench, or in a jug for ten minutes has met the room and everything in it, and returning it to the batch to save the volume puts all of that back into a vessel that was previously protected. The volume saved is small; the population introduced is not. Sampled cider is drunk, tasted or discarded — never poured back.

The third route is the one nobody counts: the air that enters to replace what left. Drawing a litre out of a sealed vessel draws a litre of air in behind it, and doing that weekly for six months on a maturing cider is a slow oxidation with a known cause. This is one of several reasons to take fewer, better samples.

At demijohn scale the arithmetic is stark. A trial jar takes a substantial fraction of a small batch, and taking one every few days through a ferment removes a real volume and leaves a real headspace. This is the strongest practical argument for the refractometer, which answers the pre-fermentation questions on two drops, and for reserving the hydrometer for the readings that actually drive a decision: the start, the point where the ferment appears to have finished, and the confirmation a few days later that it really has.

The other economy is to plan the order of measurements so that one sample serves several. Take the temperature first, in the sample vessel. Read the gravity next, in the trial jar, since the hydrometer needs the volume and adds nothing to it. Then use the same cider for pH, and then for acidity if it is being titrated, and taste last of all — because a probe that has been in a titration or a finger that has been in the glass has no business anywhere upstream.

A fermenting sample also needs degassing before a gravity or an acidity reading. Dissolved carbon dioxide raises the apparent gravity, sticks bubbles to a hydrometer stem and is itself an acid that will be titrated along with the malic. Swirling the sample vigorously or letting it stand is enough for most purposes and takes a minute.