The Cool-Water Rule Lost the Blind Tasting to a Boil

The Cool-Water Rule Lost the Blind Tasting

Tea shops will tell you that silver needle wants cool water. Between 70 and 80°C (about 158 to 176°F), because hotter water "can lead to bitterness and astringency, which masks the tea's natural sweetness."² The buds are called delicate, and the advice follows from the word.

A consumer panel was given white tea brewed at five temperatures and asked which it preferred. The recommended 70°C came last.

What Was Actually Tested

The work was done in Granada, and published in Food Chemistry in 2018¹. Pérez-Burillo and colleagues bought 34 commercial white teas from 16 local tea shops, 13 in bags and 21 loose. They brewed every one of them at every combination of five temperatures and five times: 60, 70, 80, 90 and 98°C (about 140, 158, 176, 194 and 208°F), held for 3, 5, 7, 10 and 15 minutes. Two grams of leaf to 150 mL of water throughout.

Then they measured the cup twice over. Once by chemistry: individual catechins, caffeine, antioxidant capacity by three separate assays. Once by people: a consumer preference panel, and a trained descriptive panel to name what the consumers were responding to.

Chemistry alone would have simply crowned the boil. Only the panel found where it stopped paying.

What Heat and Time Do to the Leaf

Below 80°C, time did almost nothing. Only a 15-minute steep pulled significantly more catechins and caffeine out of the leaf at those temperatures, which is to say the cool-water cup is a cup that has largely declined to extract.

Above 80°C the behaviour changes. Gallic acid, epigallocatechin gallate and caffeine rise almost exponentially once the steep passes 10 minutes. Antioxidant capacity measured by ABTS climbs steadily with both heat and time. The FRAP assay barely moves at all until 98°C.

98°C held the most of everything measured, and kept climbing the longer the tea sat in it.

The Panel Preferred the Boil

Given the same five-minute steep at each of the five temperatures, the panel rated each sample on its own. The shares that liked it:

70°C, the retail recommendation for this tea: 13%. 80°C, the other half of that recommendation: 14%, statistically indistinguishable from the first. 90°C, the temperature the trade reserves for green tea: 27%. And 98°C, a full rolling boil, hotter than any retail guide will suggest for silver needle: 43%.

That 43% is the largest share of any one temperature tested. It is not a majority, and it should not be reported as one. Fifty-seven percent of the panel liked something else, or liked nothing much. What can be said is narrower and still fatal to the convention: the water the trade specifically warns against was the best-liked of the five, and the water it recommends was the worst.

The Half That Corrects It

Then the same panel was given 98°C water at different times, and the result reverses direction.

Seven minutes: 50%. Ten minutes: 24%. Fifteen minutes: 25%.

The trained panel explains it. At seven minutes the brew read as measurably less bitter and less astringent, and its aroma was carried by floral, fruity and green notes, comparable to good green teas from Japan and Korea. At fifteen minutes, the same leaf and the same boiling water, the descriptors are brown, seaweed, spinach.

That is the harsh, over-extracted cup the trade warned about, arriving on schedule from the wrong cause: not the heat, the clock.

One Caveat for Loose-Leaf Drinkers

The bagged teas in the study produced almost double the antioxidant capacity of the loose-leaf teas brewed identically. The likely reason is particle size: cut leaf presents more surface, and water reaches more of it. Worth entering in the register, for a readership that mostly buys whole leaf and assumes the whole leaf gives more.

The Corrected Verdict

The convention was right that white tea can be ruined by carelessness with heat and time. It named the wrong one of the two.

Cool water does not protect the bud. It mostly prevents extraction, and the panel could taste the absence. What protects the cup is the clock. Seven minutes at a boil was preferred twice as often as fifteen minutes at the same boil, by the same people, on the same tea.

Our brewing reference sets out the temperatures we hold to by grade, and this study does not overturn it. What it does is move the point of attention. Withering and drying were left alone already; the last thing left to get wrong is how long the leaf sits in the water.