Bench methods
Heat stability, by a heated sample
Whether the wine will turn hazy when it warms up in a shop window or a car. Unstable protein comes out of solution with heat, so you hold a sample hot, cool it, and compare its turbidity with the same wine unheated. The rise in turbidity is the evidence, and you decide the verdict.
These are the standard bench procedures in plain words, not a copy of any one published protocol. If your reagent kit comes with its own procedure, follow the kit. It was written for the normality in that bottle. The end of each method explains which of its figures come from the chemistry and which are standard practice.
Wear eye protection and gloves for every one of these methods. Sodium hydroxide is caustic, iodine stains and irritates, and the acids and peroxide burn. The chromatography solvent is flammable, so use it in a hood, not on an open bench. Each reagent's safety data sheet is the authority on handling and storage. These steps don't replace it.
- Produces
- A verdict, stable or unstable, and the turbidity rise behind it, recorded as a stability trial on the lot.
- Takes
- Ten minutes of bench work, then six hours hot and a night cooling.
What you need
- Two clear bottles of the wine, filtered bright. You can't read a rise on a hazy sample.
- A water bath or oven that holds 80 °C (176 °F)
- A turbidity meter, or a dark card and a strong light for a visual read
- For a bentonite trial: a bentonite slurry and the doses to try
What this test is for
This is a TEST, not a treatment. Fining with bentonite is cellar work done to the wine, and it's recorded with finishing. This test tells you whether the wine needs it, and how much.
Run it
Measure the turbidity of the filtered wine and record it as the before figure.
The test reads the rise, so you need the starting figure on record. A wine that's already cloudy has no rise to read, and the test can't tell you anything about it.
Hold one sample at 80 °C (176 °F) for six hours, sealed so nothing evaporates.
This is the most common setup. Some wineries hold 90 °C (194 °F) for an hour, or 80 °C for thirty minutes, and each gives slightly different results. Whichever you use, run it the same way every time, and enter the temperature and hours on the trial so you can compare a result from March with one from October.
Cool it to room temperature overnight, or to 4 °C (39 °F) for a few hours and then to room temperature.
Measure the turbidity of the heated sample beside the unheated one and record it as the after figure.
The trial shows the rise and how it was worked out. A rise under about 2 NTU is commonly read as stable and a larger one as unstable. That's common practice, not a figure from any regulation, and the verdict you record is your call.
A bentonite trial, if it's unstable
Dose a set of samples across a range (a quarter, a half, three quarters and one gram per liter is a usual set), let them settle, filter each one bright, and heat-test each one the same way.
Use the smallest dose that passes. Bentonite strips flavor and volume as well as protein, so the trial is for finding the least that works, not for proving that a lot of it works.
Record the passing dose in the trial's response, and log the addition on the lot when you make it.
The trial is the reason for the dose. The addition records the dose that went in and the volume it went into. Without the trial beside it, no one can check the dose later.
Where it goes wrong
A stable result here tells you nothing about tartrates.
Heat and cold instability are different chemistry: protein versus potassium bitartrate, and each test only detects its own. A wine that passes this test can still drop crystals in a cold shop. That's what the cold test checks, and it's recorded as a separate trial.
Test the wine you will bottle.
Protein stability shifts with pH, alcohol and everything that happens between the test and bottling. A late acid addition or a blend can bring back a haze in a wine that passed in March. That's why each verdict is dated. An old verdict on a wine that has since changed tells you about a different wine.
Too much bentonite costs you wine, and nothing else records that loss.
Bentonite takes aroma and volume along with the protein, and the wine lost in its lees doesn't come back. Trialing a range of doses keeps you to the least that works. A round number picked without a trial is just the largest dose someone was willing to try.
Everything here is convention, not chemistry. A heat test has no equivalent weight behind it and nothing to calibrate. The hold, the cooling and the level a rise is judged against are common practice found across winery lab manuals, and Fermata records the verdict you reach using your own. The one thing it calculates is the rise, a simple subtraction, shown with its working.